Description
A development finance institution (DFI) is an alternative financial institution which includes microfinance institutions, community development financial institution and revolving loan funds.
ABSTRACT
Title: AN INTEREST GROUP THEORY OF
FINANCIAL DEVELOPMENT
Nela N. Thomas Richardson, Ph.D., 2005
Directed By: Professor Peter Murrell, Department of
Economics
My work contributes to current explanations of the variance in financial
development across countries by considering the role of political and legal structure in
determining the effect of private interests on financial policy and illustrating the political
obstacles that policymakers face when reforming the financial system.
In Chapter 2, I present a political economy model to study the role of politics in
the process of financial development across emerging markets. The model concludes that
both special interest groups and political structure affect the level of credit market
development chosen in equilibrium. When policymakers are constrained by political
institutions that require democratic accountability, they are more likely to improve the
level of creditor rights enforcement in the financial system. Financial reform is also more
likely to occur in wealthy and highly productive economies. In contrast, the model shows
that openness to international capital inflows impedes financial development.
Furthermore elite special interest group members benefit more from financial repression
when wealth is unequally distributed; hence, income inequality provides a further
obstacle to financial reform in emerging markets.
Chapter 3 empirically investigates the role of political institutions in
implementing financial reform under three different levels of democratic accountability,
Free Countries, Partly Free Countries and Not Free Countries. I find that the institutional
details of the political system, as summarized by the number and cohesion of its veto
players – individuals whose consent is required for policy change – are weakly associated
with credit market development in Partly Free Countries.
Chapter 4 (co-authored with A. Knill) investigates the role of security laws on the
ability of firms to raise external finance by issuing capital. We find that securities laws
have disparate effects on capital issuance between small and large firms in G10 and
emerging market countries. Private enforcement of securities laws that codify existing
market arrangements is found to be a deterrent to capital issuance for small firms and
firms in emerging markets. Public enforcement of securities laws by government
regulation significantly increases the probability of issuance for emerging market firms.
AN INTEREST GROUP THEORY OF FINANCIAL DEVELOPMENT
By
Nela N. Thomas Richardson
Dissertation submitted to the Faculty of the Graduate School of the
University of Maryland, College Park, in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
2005
Advisory Committee:
Professor Peter Murrell, Chair
Professor Roger Betancourt
Professor Vojislov Maksimovic
Professor Carmen Reinhart
Associate Professor Peter Coughlin
© Copyright by
Nela N. Thomas Richardson
2005
ii
Dedication
I dedicate my dissertation to my son, Andrew Dubois Thomas Richardson, born during
the writing of this thesis. Drew, my wish for you is that your future is filled with
opportunities, abundance, music, happiness, and laughter and that you experience in your
life the joy that you have brought to mine.
iii
Acknowledgements
I am grateful to my advisor Peter Murrell for spending hours poring over my
theoretical model, for his expeditious and detailed feedback, and his guidance and
support. I also benefited greatly from the comments provided by Roger Betancourt,
Vojislov Maksimovic, David Smith, and John Rust. I thank seminar participants at the
University of Maryland, Syracuse University, the Financial Management Association
2003 Doctoral Workshop, and the Federal Reserve Board International Finance Division
for helpful comments. I also recognize the National Science Foundation for financially
supporting the first three years of my graduate education.
I wish to express my eternal gratitude and appreciation for my mother and role
model in all things, Mary Thomas, for instilling in me the tools necessary to complete
such an arduous task as a dissertation and encouraging me throughout this process.
It has been a long journey. My partner through all of it has been my husband and
soul mate, Christopher Richardson. I am profoundly thankful for his constant and
unwavering support, advice, love, encouragement, and most of all, patience.
iv
Table of Contents
ABSTRACT......................................................................................................................... i
Dedication........................................................................................................................... ii
Acknowledgements............................................................................................................ iii
Table of Contents............................................................................................................... iv
List of Tables ..................................................................................................................... vi
List of Figures ................................................................................................................... vii
Chapter 1: Introduction...................................................................................................... 1
Chapter 2: An Interest Group Theory of Credit Market Development.............................. 7
1. Introduction............................................................................................................. 7
2. A Simple Model of Financial Development ......................................................... 11
2.1 Setup ............................................................................................................. 11
2.2 Perfect Credit Markets: Case of 1 = ? .......................................................... 16
2.3 Complete Financial Repression: Case of 0 = ? .......................................... 19
2.4 Imperfect Markets: Case of ) 5 . 0 , 0 ( ? ? ....................................................... 19
3. Welfare.................................................................................................................. 25
3.1 Politically Organized Unconstrained Firms (The SIG) ................................ 26
3.2 Politically Unorganized Unconstrained Entrepreneurs................................. 28
3.3 Constrained Entrepreneurs............................................................................ 29
3.4 Poor Lenders ................................................................................................. 30
3.5 Aggregate Welfare........................................................................................ 31
4. Political Utility...................................................................................................... 32
5. Political Equilibrium............................................................................................. 37
6. Comparative Statics Under Imperfect Credit Markets.......................................... 42
6.1 Change in Productivity ................................................................................. 42
6.2 Percent Change in Wealth Per Capita........................................................... 44
6.3 Change in Financial Openness...................................................................... 45
6.4 Percent Change in Wealth Inequality ........................................................... 48
7. Concluding Remarks............................................................................................. 51
Chapter 3: An Empirical Analysis of Political Institutions and Credit Market
Development ..................................................................................................................... 55
1. Introduction........................................................................................................... 55
2. Veto Players under Different levels of Political Accountability .......................... 57
3. Data....................................................................................................................... 61
4. Empirical Strategy ................................................................................................ 66
5. Results................................................................................................................... 69
5A. Strategy 1: Clustering Standard Errors by Country..................................... 69
5B. Strategy 2: The Effect of Liberalization ...................................................... 71
5C. Strategy 3: Averaging Data in 5-year intervals............................................ 72
5D. Strategy 4: Ownership as the Dependent Variable ....................................... 73
6. Conclusion ............................................................................................................ 74
Chapter 4: Securities Laws, Firm Size, and Capital Issuance........................................ 102
1. Introduction......................................................................................................... 102
v
2. Methodology....................................................................................................... 105
3. Data..................................................................................................................... 109
4. Microeconomic Testing Strategy of LLS (2004)................................................ 116
5. Results................................................................................................................. 118
6. Robustness Check............................................................................................... 125
7. Conclusions......................................................................................................... 126
Appendices...................................................................................................................... 141
Appendix A: Derivatives of the Constrained Capital Equation................................. 141
Appendix B: Derivatives of the Firm Profit Equations with respect to .................. 142
Appendix C: Description of Variables, Chapter 3..................................................... 144
Appendix D: Description of Variable, Chapter 4 ...................................................... 146
Appendix E: Regression of Capital Issuance on Individual Sub Indexes of Public and
Private Enforcement.................................................................................................... 151
Bibliography ................................................................................................................... 152
vi
List of Tables
Table 2.1: Summary of Variables .................................................................................... 21
Table 2.2: The Parametric Model .................................................................................... 23
Table 2.3: Parameter Ranges ........................................................................................... 42
Table 3.1A : Free Country and Years .............................................................................. 76
Table 3.1B : Partially Free Country and Years................................................................. 77
Table 3.1C : Not Free Country and Years ........................................................................ 79
Table 3.2: Summary Statistics ......................................................................................... 80
Table 3.3: OLS Regression of the Basic and Openness Equations................................... 82
Table 3.4: OLS Regression of the Basic and Openness equations and Legal Origin....... 83
Table 3.5: IV Regression of the Basic Equation............................................................... 84
Table 3.6: IV Regression of the Openness Equation ........................................................ 85
Table 3.7: OLS Regression of the Basic and Openness Equations and Liberalization .... 87
Table 3.8: OLS Regressions of the Basic and Openness Equations, Legal Origin, and
Liberalization............................................................................................................ 88
Table 3.9: IV Regression of the Basic Equation and Liberalization................................. 89
Table 3.10: IV Regression of the Openness Equation and Liberalization........................ 91
Table 3.11: OLS Regression of the Basic and Openness Equations................................. 94
Table 3.12: OLS Regression of the Basic and Openness Equations and Legal Origin .... 95
Table 3.13: IV Regression of the Basic Equation............................................................. 96
Table 3.14: IV Regression of the Openness Equation ...................................................... 97
Table 3.15: OLS Estimation of Ownership...................................................................... 99
Table 3.16: OLS Estimation of Ownership and Legal Origin ....................................... 100
Table 3.17: IV Estimation of Ownership....................................................................... 101
Table 4.1: Security Issuance by Country....................................................................... 128
Table 4.2: Summary Statistics of Firm Specific Variables............................................ 130
Table 4.3: Summary Statistics of Macroeconomic Variables......................................... 130
Table 4.4: Firm Specific Correlations (with Capital Issuance Indicator) ...................... 131
Table 4.5 Macroeconomic Variable Correlations (with Capital Issuance Indicator) ..... 131
Table 4.6A: All Firms .................................................................................................... 132
Table 4.6B Small Firms .................................................................................................. 133
Table 4.6C: Large Firms ................................................................................................. 134
Table 4.7A: G10 Small Firms......................................................................................... 135
Table 4.7B G10 Large Firms .......................................................................................... 136
Table 4.8A: Small Emerging Market Firms .................................................................. 137
Table 4.8B: Large Emerging Market Firms.................................................................... 138
Table 4.9A: All Firms – Standard Errors Clustered by Country ................................... 139
Table 4.9B: All Firm Groupings – Standard Errors Clustered by Country ................... 140
vii
List of Figures
Figure 2.1: Graph of endogenously determined interest rates as a function of Creditor
Rights Enforcement. ................................................................................................. 24
Figure 2.2: Optimal Capital Investment for Decreasing Returns to Scale Production
Function as a Function of Creditor Rights Enforcement. ......................................... 25
Figure 2.3: Entrepreneurs by wealth endowment ............................................................ 26
Figure 2.4: Aggregate Welfare of the Special Interest Group as a Function of Creditor
Rights Enforcement .................................................................................................. 28
Figure 2.5: Social Welfare as a Function of Creditor Rights Enforcement ..................... 32
Figure 2.6: Sequence of events and decisions ................................................................. 37
Figure 2.7: Political Utility .............................................................................................. 40
Figure 2.8: Plot of points that make the policymaker just indifferent between high and
low financial development for different levels of productivity ranging from g = 4 to
g=5. ........................................................................................................................... 43
Figure 2.9: Plot of points that make the policymaker just indifferent between high and
low financial development for different levels of aggregate initial wealth as wealth
per capita varies. ....................................................................................................... 44
Figure 2.10: Plot of points that make the policymaker just indifferent between high and
low Financial Development for different degrees of capital openness..................... 46
Figure 2.11: Plot of points that make the policymaker just indifferent between high and
low financial development for different levels as wealth inequality varies............ 49
Figure 2.12: Plot of SIG welfare under different degrees if wealth inequality under
financial repression (=.1) and financial development (=.49) ................................ 50
1
Chapter 1: Introduction
The purpose of this dissertation is to examine the institutions that underpin
financial systems. Political and legal institutions function in tandem to shape financial
policy, in part, by determining the ability of those who benefit from financial repression
to block development. I find that legal institutions that fail to protect private property
rights generate a dichotomy between entrenched corporate interests and aggregate
welfare. For this reason, politicians are influenced by powerful, elite special interest
groups when forming financial policy.
My thesis contributes to current explanations of
the variance in financial development across countries by analyzing the effect of private
interests on financial policy and examining the institutional obstacles that policymakers
face when reforming the financial system.
The context of this dissertation is embedded in a large literature examining the
role of finance on growth. In a recent revitalization of this research, Levine (1997)
demonstrates the critical causal relationship between financial structure and economic
development by explaining the purpose of financial systems within society, how they
operate, and the mechanisms by which they affect and are affected by economic growth.
Shleifer and Vishny (1997) produce a second watershed paper under the finance and
development umbrella. Their article examines the different ways economies deal with
the problem of corporate governance: the set of laws and institutions created to ensure
that firms share their profits with the suppliers of capital. The authors focus on two
differentiating features of corporate governance systems across countries: ownership
concentration and the degree of legal protection of investors. Shareholder rights
determine the competency of the financial system to allocate society’s resources
2
efficiently. High ownership concentration is a market response to the agency costs that
plague financial systems with weak corporate governance institutions. A juxtaposition of
Shleifer and Vishny (1997) and Levine (1997) reveals that the effectiveness of a financial
system to generate economic growth is dependent on the institutional structure in which
the system is embedded.
La Porta et al. (1998) offers a rationale for the success of financial systems in
promoting growth known as the “law and finance” view of financial development. The
authors explore the contribution of a country’s legal origin in the formation of its
financial structure and its corporate governance institutions, finding that legal origin —
be it English common law, or French, German or Scandinavian civil law — partly
determines the quality of investor protection and the size of the stock market versus the
banking sector. The paper concludes that English common law systems generally have
the strongest investor protection enforcement, followed by Germany, Scandinavian, and
lastly, French civil systems. Beck et al. (2001) support these results, finding that legal
origin has a considerable influence on access to bank credit.
Though the law and finance view is the leading explanation for the variance in the
proficiency of financial systems across countries, the literature also recognizes a
relationship between political institutions and financial system development. Rajan and
Zingales (2002) analyze the importance of interest groups as opposed to legal origin or
culture in influencing financial development. They propose and test a theory that firms
are more willing to support financial liberalization in times of trade openness and
increased international competition. Firms are more resistant to financial development
when the economy is relatively more closed. The authors explain La Porta et al.’s (1998)
3
results by suggesting that in Civil law countries, like France, it is easier for governments
to implement new policies when swayed by interest groups to do so.
Biasis and Mariotti (2003) provide a theoretical model that implements the story
told by Rajan and Zingales. The authors show that soft bankruptcy laws, which are
indicative of low levels of financial development, may actually increase social welfare by
reducing the potential for inefficient liquidation caused by imperfect credit markets.
However, imperfect sanctions against default impose a collateral requirement that
prevents poor agents from accessing the credit market, thus, the gains to rich
entrepreneurs are bought at the expense of the poor. Within their model, the authors
point out that the “[a]gents with different initial resources typically have different
preferences towards the bankruptcy law. Hence different laws can be chosen in different
countries, reflecting the political influence of the different social classes, and possibly at
odds with social welfare.”
Other research in the political economy aspects of financial development include
Pagano and Volpin (2002), who survey the literature on corporate governance structures
by examining the ability of political economy methodology to analyze the economic
regulations and financial institutions that result from the balance of power between the
constituents of society. The main insights of the political economy approach is that it
explains international differences in financial policy by describing “which constituencies
are assuming a certain regulatory outcome, why they are currently dictating the rules, and
how and why the balance of power can shift against them.” From the political science
literature Haggard et al. (1993) give a detailed multi-country case study of the influence
of political institutions on financial repression and liberalization. The authors identify
4
several key factors that affect financial structure development, including macro instability
as a political liability, the existence of conservative central banks, new government
installation, and the presence of countervailing interest groups.
Other researchers, for example Denizer et al. (1998), have shown that through a
combination of powerful elites and low inter-party competition, interest groups have been
able to instigate and perpetuate implicit subsidies through government allocation of credit
to favored firms and industries. Though research on the politics and economics of
financial development is scarce, several papers suggest that the effects of political
economy on financial development deserve further investigation.
1
In addition, empirical studies, like those undertaken by Singh (1995) and Glen
and Pinto (1994), underscore the importance of policy reform on firms’ financial
behavior. Singh credits the intense intervention in financial markets by developing-
country governments for the tendency of large firms to rely primarily on new equity
issues to finance investments instead of the internal revenues or bank loans predicted by
theory. The experience of firms in developing countries – notoriously lacking in legal
rules, corporate governance institutions and legal enforcement – reveals that there must
be influences other than legal structure on the size of the stock market and the degree of
financial development. The paper by Glen and Pinto offers a qualitative framework to
analyze how the issues of financing cost, riskiness, disclosure and fear of loss of control
affect the firm’s capital structure decision. The authors connect government policies —
including capital controls, tax incentives and interest rate ceilings — to financing
decisions made by firms.
1
See Hellman et al. (1998); Hellman and Shankermann (2000); Krozner (1998, 1999); Faccio (2002);
Morck et al. (1998); Gordan and Lei (1999); Johnson and Titman (2002); and Laffont and Tirole (1991).
5
Booth et al. (1999) highlight differences in financing behavior between emerging
market and industrialized firms. The researchers find that firms in developing countries
are affected by the same sort of financial variables when making their capital structure
decisions as firms in industrialized countries. The crucial difference is that knowing the
country of an emerging market firm’s origin is as least as important in explaining its
capital decisions as the firm’s characteristics. This observation provides further
evidence, albeit indirect, that a country’s institutional factors, such as political
institutions, go a long way in accounting for firms’ financial decision-making.
Chapters 2 and 3 of my thesis explore the role of institutions in developing credit
markets. Credit market development is particularly critical in emerging markets because
firms receive the lion’s share of their investments funds from banks.
2
Chapter 2 presents
a general equilibrium model that analyzes the effect of political institutions on financial
intermediary development in an emerging market economy. The model finds that the
amount of political accountability imposed by the country’s political system affects the
level of credit market development that a policymaker chooses in equilibrium.
Chapter 3 explores whether there exists a connection between the institutional
details of the political system and intermediary development. I develop four empirical
strategies to analyze this relationship and determine the sensitivity of the influence of
political institutions on credit access to different legal regimes and levels of financial
openness. Though Chapter 2 identifies the importance of democratic accountability in
increasing accesses to credit, Chapter 3 finds that the detailed characteristics of the
2
For example, Rojas-Suarez and Wiesbrod (1996) study capital market development in seven Latin
American countries. The authors find that only in Chile is equity a substantial source of finance. For other
countries, firms are almost entirely dependent on bank loans to make investments.
6
political system play a weak role in credit development once political constraints on the
executive have been taken into account.
While most corporations receive much of their financing through bank loans,
beginning in the 1990’s large corporations in developing countries have relied heavily on
new issues to fund investments. Chapter 4 (coauthored with A. Knill) investigates the
role of securities laws on the ability of firms to raise external funds by issuing capital.
We find that securities laws have disparate effects on capital issuance between small and
large firms in G10 and emerging market countries.
7
Chapter 2: An Interest Group Theory of Credit Market Development
1. Introduction
Financial development, the capacity of the financial system to efficiently allocate
capital from saver to investor, enhances a country’s ability to generate economic growth.
3
Complicating this relationship between financial development and growth is the
acknowledgement that some economic agents benefit from dysfunctional financial
systems. The disparate effects of financial development on the economic prosperity of
different segments of society are part of the explanation for why financial systems vary
dramatically across countries.
The finance and growth literature has uncovered the role of economic institutions
in determining a country’s level of financial development and made the important
connection between legal institutions, property rights, and the enforcement of financial
contracts.
4
However, the role of institutions in financial development is not limited to
economic functions; political institutions affect the ability of those who benefit from
repression of financial markets to persuade their governments to restrict financial
development. In order to understand completely the role of institutions in explaining
differences in financial structures across countries it is necessary to consider both
channels of institutional influence.
3
See also Levine and King (1993), Levine (1997) Levine and Zervos (1998) for landmark studies on
financial development and growth.
4
La Porta et al. (1998) is the first to offer a rationale for the success of financial systems in promoting
growth known as the “law and finance” view of financial development. The authors explore the
contribution of a country’s legal origin – be it English Common Law or French and German Civil Law – in
the formation of its financial structure and its corporate governance institutions, finding that legal origin
partly determines the quality of investor protection.
8
The recent history of emerging markets illustrates the political obstacles faced by
governments that attempt to develop their financial system. In these countries, the
decades of the 1980s and 1990s were characterized, largely, by reform of the banking
system. With varying degrees of success, reforms focused on liberalizing financial prices
(interest rates) and reducing government-directed credit programs to favored corporations
and industries. Several studies present anecdotal evidence of how high capital costs,
increased firm competition, and reductions in government subsidies strengthen large
corporations’ political resistance to financial development.
5
To examine the role political institutions in achieving financial development in
emerging economies, I present a model in which a single policymaker sets financial
policy by determining the level of creditor rights enforcement. The enforcement of
creditor rights is a natural focus for examining government financial policy due to the
importance of banking regulation in emerging markets. Firms in emerging markets are
much more limited in their capacity to raise external finance than firms in developed
countries
6
. They do not have access to the range of securities and financial options
available in industrialized nations. Therefore, the ability of the banking system to finance
projects efficiently is even more crucial. Bankruptcy laws that protect creditors allow
banks to lend to small and medium enterprises reducing the incidence of credit rationing
to larger firms.
In my model, stronger creditor protection increases the availability of credit to
entrepreneurs, which raises the demand for capital and increases the equilibrium interest
5
Most notably Rajan and Zingales (2003) and Singh (1995).
6
Additionally, the government is often the main borrower in the economy, providing formidable
competition to firms of all sizes for investment funds.
9
rate. Lenders benefit from financial development because of the increase in the lending
rate, while poor and middle-income entrepreneurs benefit because they are able to
produce when they otherwise could not in a financially repressed economy. Wealthy
agents, on the other hand, are harmed by financial development because they suffer an
increase in the cost of capital. These agents have the narrowly focused common
objective required (Olson 1965) to organize into a political interest group and block the
reform of the financial sector. The policymaker values aggregate welfare as well as the
campaign contributions that she receives from the interest group of wealthy agents but is
constrained in the amount of aggregate welfare that she can forgo in favor of campaign
contributions by the economy’s political institutions. Political institutions determine how
accountable the policymaker must be to the general electorate when evaluating her tastes
for campaign contributions aimed at maintaining financial repression. In equilibrium, the
level of financial development the policymaker chooses is dependent on both interest
group influence and the level of democratic accountability imposed on governments by
the country’s political institutions.
7
By using numerical simulation techniques, my analysis sheds light on the political
barriers to developing the financial system. The central conclusion of the theoretical
model is that political institutions that impose more accountability on policymakers
achieve higher levels of financial development. While special interest groups influence
financial policy, a country’s political structure determines the potency of elite influence
7
In a related work, Biasis and Mariotti (2003) model bankruptcy reform by assuming the government acts
like a benevolent social planner who always maximizes aggregate welfare irrespective of institutional
restraints. In contrast, I model the government more realistically as I model a self-interested actor who is
influenced by special interests groups and constrained by the political institutions that structure the
economy.
10
on reform. Determinants of accountability such as competitive elections, independent
judicial systems, and low corruption ensure that policymakers must value the well-being
of the citizenry when making policy decisions. Hence, politicians are unwilling to
capitalize on campaign contributions by withholding financial reforms that increase
social welfare.
Comparative static analysis of the model allows me to examine the critical
relationship between the economic environment, political accountability, and financial
reform. This relationship has not been significantly explored by any other theoretical or
empirical work on the determinants of financial development. The model validates the
intuition that financial reform is more easily attained in wealthier industrialized
economies. In contrast, it predicts that financial openness to capital inflows, e.g. foreign
direct investment, aggravates financial reform. I show that capital openness negatively
affects financial development by lowering the cost of capital and reducing the ability of
creditor rights protection to increase aggregate social welfare. Finally, I demonstrate that
special interest group members benefit more from financial repression when there is
relatively more wealth inequality in the economy. Thus, wealth inequality decreases the
likelihood that the policymaker will choose to develop the financial system.
The paper is organized as follows. Section 2 presents the set up of the simple
financial system and introduces three special cases of credit markets: perfect markets,
complete financial repression and imperfect credit markets. Section 3 analyzes the effect
of financial development on the welfare of the special interest group, politically
disorganized unconstrained and constrained firms and poor lenders. Section 4
summarizes the Grossman and Helpman model of interest group influence and provides
11
an application to political institutions. Sections 5 and 6 discuss the political equilibrium
and present five results that follow from the model. Section 7 concludes.
2. A Simple Model of Financial Development
2.1 Setup
The set up of the financial system is loosely based on Matsuyama (2000). I make
several crucial modifications to Matsuyama’s original framework in order to capture the
process of political reform in the credit system of an emerging economy. Furthermore,
my model assumes that agents face a decreasing returns to scale production technology in
contrast to the constant returns function in Matsuyama’s paper.
There exists a small closed economy populated with a continuum of agents.
Agents live for two periods. In the first period, individuals make their investment
decisions and in the second period, they consume final wealth. Additionally, there are
two sources of heterogeneity across agents: they are endowed with different amounts of
initial wealth, w, and some have access to an investment project. At the beginning of
period one, each individual receives a wealth endowment, 0 > w .
To focus this theoretical investigation on a financial system in a poor economy I
make two additional assumptions. The first poor country assumption introduces
significant wealth inequality into the economy by asserting that there are times as many
agents in the bottom half of the domain of the wealth distribution function (the poor
cohort) than at the top half the wealth distribution (rich cohort). Secondly, I assume that
every agent in the top half of the wealth distribution has access to an investment project
with a gross rate of return equal to, g . However, only a
?
1
of agents in the bottom
12
domain have access to investment projects. I make this assumption in order to avoid
results that depend on all the entrepreneurs being rich. Let N be the number of agents
with wealth, w. Hence, the number of agents with investment projects is equal to
}
=
M
N dw
M
N
0
1
. Agents without investment projects are called households. Agents with
investment projects are called entrepreneurs. These assumptions capture two realistic
characteristics of poor economies, wealth inequality and the social stratification of agents
into capitalists and laborers.
8
Wealth is uniformly distributed across households by
(
¸
(
¸
2
, 0 ~
M
U w . Let
) (w H denote the wealth distribution across households, where
M
w H
2
) ( ' = . Wealth is
uniformly distributed across entrepreneurs by [ ] M U w , 0 ~ . Let ) (w G denote the wealth
distribution across entrepreneurs, where
M
w G
1
) ( ' = .
The expression ) 3 (
8
) ( ' ) ( '
2
) 1 (
2 /
0 0
?
?
+ = +
?
} }
M
N dw w wG N dw w wH N
M M
gives the total
number of agents with wealth equal to w.
A critical characteristic of the project’s production function is that there exists a
minimum scale of investment, 1 ? k . The minimum investment scale implies that
entrepreneurs with initial wealth less than one must borrow from the economy’s credit
market in order to set up their projects. Entrepreneurs with wealth greater than one self-
8
See Biasis and Mariotti (2003).
13
finance their projects and lend their remaining endowment to other agents. The
production function is given by Equation (2.1)
9
)
`
¹
¹
´
¦
?
<
=
, 1
1 0
) (
2 / 1
k if gk
k if
k F . (2.1)
The production function places restrictions on the range of the parameter g.
10
As
long as g is greater than the gross interest rate, ) (? i , agents will want to invest at least
one unit of capital into their projects. The interest rate is a function of financial
development. It is endogenously determined by the credit market equilibrium condition
that aggregate demand for investment is equal to the aggregate supply of capital.
The credit market is characterized by a market imperfection. The legal structure
covering debt contracts (bankruptcy laws) is too weak to ensure that creditors will receive
the full amount owed to them if the borrower defaults. Let B be the loan amount. B is
equivalent to the capital the agent chooses to invest minus his wealth endowment,
w k B ? = . As described in Equation (2.2) the borrower will only repay his debt
obligation if the cost of repayment is less than the fraction of output lost by defaulting.
] 1 , 0 [ , ) ( ) (
2 / 1
? ? ? = ? ? ? gk w k i B i . (2.2)
9
The decreasing returns production function ensures that the policymaker faces a tradeoff between social
welfare and political rents when choosing financial policy. Under a CRS production technology an
increase in the welfare of poor agents will be offset by a corresponding decrease in the welfare of wealthy
agents, therefore aggregate social welfare will remain unchanged when financial policy is improved. With
DRS technology, aggregate social welfare is unambiguously increased when the policymaker chooses a
financial policy that increases access to credit to poorer entrepreneurs. This production function has two
important implications on the agent’s investment decisions. It ensures that the optimal capital level of
capital agents choose is finite rather than infinite as in the CRS case. DRS technology also implies that the
amount of wealth required to set up a project is increasing in the level of capital investment. The
increasing wealth requirement is an additional obstacle to poor entrepreneurs that want to invest optimally.
10
The gross return must be greater than zero for agents to invest in their projects. The model restricts
0
2 / 1
> ? k gk
for the function to be sensible. Since the minimum level of capital investment is assumed to
be 1 = k , the restriction implies
1 > g
.
14
The term is the fraction of output creditors can seize in case of default. Higher
values of correspond to stronger enforcement of creditor rights. The degree to which
the financial system protects creditor rights is the measure of financial development in
this model. Financial development and creditor rights enforcement will be used
interchangeably throughout this paper. Creditors will not lend more capital than they
recoup from the borrower in cases of default. Even though default does not occur in
equilibrium, the market imperfection constrains the amount of capital that agents can
borrow. Solving Equation (2.2) for the level of capital investment, k ,yields the
constrained level of capital, ) , ( w k
c
? for each agent.
2
2 2 2 2 2 2
) ( 2
) ( 4 ) ( 2
) , (
?
? ? ? ? ?
?
i
wi g g wi g
w k k
c
+ + +
= = when <1. (2.3)
Equation (2.3) is an increasing function of the agent’s wealth endowment, the
enforcement of creditor rights, and the productive capacity of the investment project. It is
a decreasing function of the interest rate. See Appendix A for derivatives.
Let ) (
*
? k be the optimal level of capital investment. Credit constrained agents
can only invest an amount ) ( ) , (
*
? ? k w k
c
< . Solving the inequality, ) ( ) , (
*
? ? k w k
c
< , for
initial wealth shows that the borrowing constraint will bind for agents whose wealth
endowment is less than ) ( ) 2 1 (
*
? ? k w ? < . Note that for 5 . ? ? , the constraint binds
only when initial wealth is less than zero. Since wealth is assumed nonnegative all agents
in the economy are able to borrow optimally whenever ] 1 , 5 [. ? ? . Therefore, agents are
15
indifferent to levels of financial development in this range. Without loss of generality, I
restrict the parameter space to be ] 5 ,. 0 [ ? ? .
11
Rearranging equation (2.2) shows that each borrower must use as collateral a
portion of final wealth equal to the proportion
|
|
.
|
\
|
?
k i
g
) (
1
?
?
of the investment k . The
wealth threshold required to borrow k units of capital is increasing in the scale of
investment. To invest k units of capital, an agent must have a wealth endowment at least
equal to
k i
g
k w
) (
1 ) (
?
?
? = . (2.4)
Given that the minimum amount of capital required to set up a project is 1 = k , the
threshold level of wealth required to become an entrepreneur is
) (
1 ) 1 (
?
?
i
g
w ? = . (2.5)
Less initial wealth is required to set up projects when there is a high level of financial
development and debt contracts are strongly enforced. As financial development
declines, agents are required to have a larger wealth endowment in order to borrow.
Let
F
w represent final wealth. Entrepreneurs with wealth equal to or greater than
the wealth threshold ) 1 ( w set up their projects. All other agents (households and poor
entrepreneurs) must lend their wealth endowment to the credit market. In this manner,
the level of financial development segments the society into wealthy entrepreneurs and
poor lenders.
11
Biasis and Mariotti (2003) restrict the policy space in a similar manner.
16
) 1 (
) 1 ( ) (
) ( ) , ( ) ( ) , (
) ( ) ( ) ( ) (
*
* * *
w
w w if w i
w k w if iw w k i w k g
k w w if iw k i k g
w
c c
F
¦
¦
¹
¦
¦
´
¦
<
? ? + ?
? + ?
=
?
? ? ?
? ? ?
. (2.6)
It is instructive to examine the equilibrium investment decisions, interest rates and
final aggregate wealth under the two extreme cases of financial development, perfect
credit markets and complete financial repression. I then investigate aggregate social
welfare under the intermediate case of imperfect credit markets.
2.2 Perfect Credit Markets: Case of 1 = ?
Consider, for a moment, the equilibrium that emerges under perfect capital
markets, 1 = ? . The creditors receive the full amount of repayment in cases of default and
do not limit the amount agents can borrow. Each entrepreneur chooses the level of
capital investment, k that maximizes final wealth.
1
2 / 1
? + ? = k to subject iw ik gk w
F
.
The Kuhn Tucker conditions for an optimum are
? + =
k
g
i
2
(ia)
0 ) 1 ( = ? k ? (ii)
1 ? k (iii)
where is the multiplier appended to the constraint that capital investment must be
greater than or equal to one. Equation (ia) implies that
2
2
) ( 4 ? ?
=
i
g
k . (ib)
17
The Kuhn Tucker conditions identify two solutions. The first solution occurs
when the capital constraint is nonbinding, 0 = ? . Then capital investment is optimal and
equal to
2
2
*
4i
g
k k = = .
When 0 ? ? the capital constraint is binding and 1 = k .
Lemma 1. Let 1 = ? . Then
2
1
g
i ? < in equilibrium.
Proof: Suppose 1 < i . Then lenders receive a negative return from lending, leading to
excess supply. Suppose
2
g
i > . Then final wealth is strictly decreasing in k. The return
to lending is greater than the return to borrowing for every agent, therefore there will be
an excess supply of capital in the economy.
When
2
g
i < Agents prefer to invest an amount
*
k into their projects. When
2
g
i = agents will invest exactly 1 unit of capital into their firms. The equilibrium
behavior of individual agents implies
¦
¦
)
¦
¦
`
¹
¦
¦
¹
¦
¦
´
¦
= ?
< =
+
?
}
}
} }
M
M
M M
g
i if dw
M
N
g
i if dw
M
Nk
dw
M
w
N dw
M
w
N
0
0
*
0
2 /
0
2
1
2
1
2
2
) 1 (?
. (2.8a)
The left-hand side represents aggregate credit supply. It is simply the summation
of all the initial wealth in the economy and is independent of the interest rate. The right
hand side represents credit demand. It is equal to the preferred level of investment
multiplied by the number of agents.
18
Equation (3.8) can be simplified to the expression
¦
¦
)
¦
¦
`
¹
¦
¦
¹
¦
¦
´
¦
= ?
< =
+
2
1
2
) 3 (
8
*
g
i if
g
i if k
M
? . (2.8b)
In the case where
2
g
i < we can use (2.8b) to obtain an explicit solution of the
equilibrium interest rate
) 3 (
2
*
? +
=
M
g
i
E
. (2.9)
Aggregate final wealth, when
*
E
i i = is equal to (2.10a)
N
M
g
dw
M
w
N dw
M
w
N
M
g
dw
M
N
M
M
g
dw
M
N
M
g W
M M
M M
F
2 / 1
0
2 /
0
0 0
2 / 1
) 3 (
8
2
2
) 1 (
) 3 (
2
1
4
3
) 3 (
2 1
) 3 (
8
|
.
|
\
|
+ =
|
|
.
|
\
|
+
?
+
+
|
.
|
\
|
+
?
|
.
|
\
|
+ =
} }
} }
+
?
?
?
?
?
.
When
2
g
i = , some entrepreneurs would not set up projects in equilibrium and in a
perfect credit markets it would be random who became an entrepreneur and who did not.
Additionally, in equilibrium entrepreneurs are indifferent between lending to the credit
market and setting up a firm because the return to lending is equal to the return to
entrepreneurship. Let ? be the proportion of individuals who set up firms in equilibrium.
The term is defined by,
¦
)
¦
`
¹
¦
¹
¦
´
¦
>
?
=
2 1
2
2
M if
M if
M
?
. When M>2, there is enough capital
in the economy for every agent to set up their projects.
19
Aggregate final wealth is equal to
N
M
M M g
dw
M
w
dw
M
w
N
g
M
N
g
dw
M
N g W
M M M M
F
|
|
.
|
\
|
|
.
|
\
| ?
+ + =
|
|
.
|
\
|
+
?
+ ? =
} } } }
+
1
) 3 (
8 2
2
2
) 1 (
2
1
2
1
0
2 /
0 1 1
? ?
?
? ?
. (2.10b)
2.3 Complete Financial Repression: Case of 0 = ?
When there is no protection against bankruptcy, households refuse to lend and no
credit is supplied to the market. Furthermore, the capital constraint in equation (2.3)
reduces to w w k
c
? ) , 0 ( . If their wealth endowment is above one, entrepreneurs sink all
of their initial wealth into their project. If the wealth endowment is below one, an
entrepreneur cannot set up his project and will not lend to the credit markets. All agents
that do not invest in projects simply eat their wealth endowment. Aggregate final wealth
in the completely repressed economy is
+
?
<
|
.
|
\
|
? =
=
}
F
M
F
W
N
M
M g
dw
M
w
g W
1
3
2
2 / 1
1
2 / 1
. (2.12)
As one would expect, aggregate social welfare is lower in the completely repressed
economy then in the perfect markets case
2.4 Imperfect Markets: Case of ) 5 . 0 , 0 ( ? ?
Under imperfect markets, only the wealthiest entrepreneurs can borrow optimally
at the equilibrium interest rate. Poor entrepreneurs are constrained by their wealth
20
endowment in the amount that they can borrow and subsequently invest into their
investment projects.
Lemma 2. Let 1 < ? . Then
2
) (
g
i ? ? in equilibrium.
Proof: Similar to Lemma 1, when ) (? i i = .
For algebraic simplicity and without loss of generality, I consider the interest rate
when ) (? i is strictly less than
2
g
. In this range of the interest rate, every entrepreneur
with access to a production technology would like to invest 1
) ( 4
) (
2
2
*
> =
?
?
i
g
k .
Imperfect markets exist in the model whenever ( ) 5 ,. 0 ? ? . Entrepreneurs with
initial wealth equal to ) (
) ( ) (
1 k w
k i
g
w = ? ?
? ?
?
will open a firm with investment scale
equal to ) (
*
? k . For entrepreneurs with initial wealth,
) (
1 ) (
*
*
?
?
k i
g
k w ? < the credit
constraint binds and they must limit investment to 1 ) , ( ? w k
c
? as described in Equation
(2.3). Agents that have access to a production technology but with initial wealth
) 1 ( w w < do not invest and merely lend their endowment to the credit market. Agents
without access to projects simply loan their wealth endowment to the economy’s credit
markets. The above discussion implies that in equilibrium
} } } }
+ =
|
|
.
|
\
|
+
?
M
k w
k w
w
c M M
dw
M
N k dw
M
w k
N dw
M
w
dw
M
w
N
) (
*
) (
) 1 ( 0
2 /
0
*
*
1
) (
) , ( 2
2
) 1 (
?
? ?
(2.13)
Equation (2.13) shows aggregate capital demand for constrained and unconstrained
borrowers. The equilibrium interest rate, ) (?
E
i solves
21
|
|
.
|
\
|
+ = +
} }
M
k w
k w
w
c
dw
M
k dw
M
w k
N
M
N
) (
*
) (
) 1 (
*
*
1
) (
) , (
) 3 (
8
?
?
? (2.14)
Sections 2.2 and 2.3 offer two extreme versions of financial development: perfect
credit markets and complete financial repression. Section 2.4 presents a more realistic,
and thus more compelling, example of financial development, the imperfect markets case.
However, what the model gains in realism when imposing imperfect markets it loses in
theoretical simplicity. Since theoretical results are hard to achieve in the intermediate
case it is useful to simulate a parametric version of the model. Below I use numerical
simulation techniques to solve Equation (2.14) for the equilibrium interest rate as a
function of creditor rights enforcement, . Once the equilibrium interest rate, ) (? i , is
computed I am able to analyze aggregate welfare over the range of creditor rights
enforcement, . Table 2.1 summarizes variables in the theoretical model.
Table 2.1: Summary of Variables
Variable Description
Creditor Rights Enforcement
) (? i Interest Rate
) 2 / , 0 ( ~ M U w
H
and ) 2 / , 0 ( ~ M U w
H
Initial Wealth Endowment for Households
(H) and Entrepreneurs (E)
Inequality Marker
) (
*
? k
Optimal Investment Scale
) , ( w k
c
?
Constrained Investment Scale
) (
*
k w
Optimal Wealth Threshold
) 1 ( w Minimum Wealth Threshold
) , (
sup
? w K
ply
Capital Supply
) , ( w K
Demand
?
Capital Demand
) , ( w W
F
?
Aggregate Final Welfare
) , ( w W
SIG
?
Special Interest Group (SIG) Final Wealth
22
Table 2.2 presents the parametric model. The first column of Table 2.2 identifies
the model’s theoretical restrictions. I choose exogenous initial parameter values that are
consistent with the theoretical restrictions. Creditor rights enforcement ranges from
[ ] 49 ,. 1 . ? ? . For each level of , I compute the endogenous variables listed in Column 3
by numerical simulation. The four remaining initial parameter values do not vary over
the range of imperfect markets, . The level of productivity, g , and the household and
entrepreneur wealth distributions are chosen such that 1 ) (
*
? ? k and
2
) ( 1
g
i < ? ? . A
wide range of parameters values, g , satisfy this condition; 4 = g , also meets the criteria
that the project has a reasonable rate of return.
12
Furthermore, M is deliberately chosen
so that 2 < M . Under this restriction, there is not enough capital in the economy for
every entrepreneur to invest in his project. This restriction is a realistic feature of an
emerging market.
13
The term is chosen to allow significant wealth inequality into the
parametric model. When 3 = ? there are 3 times as many agents in the bottom of the
domain of the wealth distribution function as in the top of the domain. Furthermore,
only
3
1 1
=
?
of all agents in the bottom domain are entrepreneurs and have the opportunity
to produce. Finally, the parameter N equals the number of agents with initial wealth w
and can take be any positive number. Without loss of generality, I let N=1.
12
Over a 5-year election cycle, a project with productivity g would have a gross annual rate of return of at
least .8 (depending on the optimal level of investment
) (
*
? k
).
13
Recall that in the perfect markets case, whenever 2 < M , then 2 / g i = . Entrepreneurs receive the same
rate of return whether they invest in their projects or lend to the credit markets. Therefore, they are
indifferent between investing and lending. Under imperfect markets, the rate of return to investing is
higher than the rate of return to lending, therefore, every entrepreneur wants to invest the amount
) (
*
? k
.
However, poor entrepreneurs are faced with a wealth constraint that forces them to invest sub optimally or
not at all.
23
Table 2.2: The Parametric Model
Theoretical Restrictions Initial Exogenous
Parameter Values
Endogenous Variables
Imperfect Markets:
) 5 ,. 0 ( ? ?
[ ] 49 ,. 1 . ? ?
) (? i
Minimum Capital Scale:
1 ) (
*
? ? k
4 = g
) (
*
? k
) , ( w k
c
?
Interest Rate Range:
2
) ( 1
g
i ? ? ?
4 . 1
) , 0 ( ~
) 2 / , 0 ( ~
= M
M U w
M U w
E
H
) (
) 1 (
*
k w
w
3 = ?
) , (
sup
? w K
ply
) , ( w K
Demand
?
1 = N
) , , (
) , , (
? ?
? ?
w W
w W
SIG
F
The first task of the numerical simulation is to calculate the interest rate as a
function of creditor rights enforcement, . The equilibrium interest rate equates the
aggregate supply of credit in the economy to the aggregate demand for credit as described
in Equation 2.14. Figure 2.1 graphs the equilibrium interest rate over the range of
consistent with the imperfect markets environment. The figure shows that the interest
rate in an increasing function of creditor rights enforcement.
24
Equilibrium Interest Rate
0
0.5
1
1.5
2
2.5
0 0.1 0.2 0.3 0.4 0.5 0.6
Creditor Rights Enforcement
I
n
t
e
r
e
s
t
R
a
t
e
Figure 2.1: Graph of endogenously determined interest rates as a function of Creditor Rights
Enforcement.
Figure 2.2 graphs the optimal capital investment for each entrepreneur. Recall
that the investment choice made by unconstrained entrepreneurs is equal to
2
2
*
) ( 4 ? i
g
k = .
Unconstrained entrepreneurs demand less capital as creditor protection improves and
lending becomes more profitable relative to investment due to the decreasing returns to
scale production function. Note that as financial development approaches the perfect
markets case optimal investment converges to one as the equilibrium interest rate
converges to
2
g
i = .
25
Optimal Capital Investment
0
0.5
1
1.5
2
2.5
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
C
a
p
i
t
a
l
Figure 2.2: Optimal Capital Investment for Decreasing Returns to Scale Production Function as a
Function of Creditor Rights Enforcement.
3. Welfare
The general equilibrium model generates four implications of imperfect credit
markets and a DRS production technology on aggregate social welfare. First, only
entrepreneurs above a certain wealth threshold can set up firms. Second, middle-income
entrepreneurs are constrained to invest sub optimally. Third, it is wealth maximizing for
rich entrepreneurs to invest more than is socially optimal. Fourth, the level of creditor
rights enforcement, , determines the equilibrium interest rate by controlling who
borrows and thusly the demand for credit.
In this section, I derive the welfare functions for the five constituent groups that
compose aggregate welfare: poor lenders, constrained entrepreneurs, unconstrained
entrepreneurs and SIG members. Poor lenders are composed of households without
projects and entrepreneurs that do not have enough wealth to borrow under imperfect
credit markets. The welfare of each group is a function of financial development. Figure
2.3 shows entrepreneurs segmented by initial wealth endowment.
26
w=0 ) 1 ( w w = ) (
*
k w w = w=M(1-x) M
|___________________|___________________|_______________________|______________|
Poor Lenders Constrained Entrepreneurs Unconstrained Entrepreneurs SIG Members
Figure 2.3: Entrepreneurs by wealth endowment
3.1 Politically Organized Unconstrained Firms (The SIG)
The special interest group, SIG, is composed of the agents in the top x% of the
wealth distribution. This elite group of citizens satisfies the conditions necessary for the
organization of a successful interest group. Olson (1965) argues that in order to
organize, interest groups must be composed of a small number of agents with narrowly
and uniformly defined interests. I show that this group of agents is harmed by financial
development and has the incentives and the resources (since they are the wealthiest
agents) to block reform.
SIG welfare is composed of the total profit of members of the SIG plus their
interest income. SIG profit is given by Equation (3.1).
}
?
? = ?
M
M x
SIG
M
dw
k i gk
) 1 (
* 2 / 1 *
) ( ) ( ) ( ) ( ? ? ? ? . (3.1)
Substituting
2
2
*
) ( 4
) (
?
?
i
g
k = into (3.1) reduces the profit function to
x
i
g
M
dw
i
g
M
M x
SIG
) ( 4 ) ( 4
) (
2
) 1 (
2
? ?
? = = ?
}
?
.
The change in SIG profit as financial development increases is equal to
0 ) ( '
) ( 16
) ( '
2
2
<
?
= ? x i
i
g
SIG
?
?
? (3.2)
27
SIG profit is decreasing in the level of financial development. To find the total change in
SIG welfare as financial development increases I add the interest income that SIG
members earn from lending to other agents.
}
?
|
|
.
|
\
| ?
? + = + =
M
M x
SIG
M
M x
M
M
i x
i
g
dw
M
w
i x
i
g
W
) 1 (
2 2 2 2 2
) 1 (
) (
) ( 4
) (
) ( 4
) ( ?
?
?
?
? . (3.3)
) ( ) 2 (
2 ) ( 4
2
2
?
?
i x x
M
x
i
g
? + =
The change in elite welfare is given by
) ( ' ) 2 (
2
) ( '
16
) ( '
2
2
? ? ? i x x
M
x i
i
g
W
SIG
? +
?
= . (3.4)
The lower the level of financial development, the lower is the cost of capital. As
long as the increase in capital costs is greater than the increase in interest income brought
about by the increased financial development, the wealthy are better off when credit
markets are repressed. Figure 2.4 graphs SIG welfare as the enforcement of creditor
rights increases using the parameter restrictions defined in Table 2. For the initial
conditions 4 . 1 = M and 25 . = x , the aggregate wealth endowment for SIG members is
equal to 30625 .
4 . 1
) 25 . 1 )( 4 . 1 (
=
}
?
dw
M
w
. Final aggregate wealth of the SIG members ranges
from a high of 1.528 at 1 . = ? to a low of 1.068 at 3 . = ? . Furthermore, the figure
demonstrates that aggregate final wealth for the SIG is lower at 49 . = ? than at 1 . = ? .
Hence, these entrepreneurs prefer weak creditor rights enforcement when the financial
system is imperfect.
28
SIG Welfare
1.1
1.11
1.12
1.13
1.14
1.15
1.16
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
S
I
G
F
i
n
a
l
W
e
a
l
t
h
Figure 2.4: Aggregate Welfare of the Special Interest Group as a Function of Creditor Rights
Enforcement
3.2 Politically Unorganized Unconstrained Entrepreneurs
Not all unconstrained firms are able to organize politically, only the top x% of the
wealth distribution. These unconstrained entrepreneurs that are politically disorganized
have wealth over the threshold necessary to invest the optimal level of capital, ) (
*
k w ,
and below that required to be in the SIG. The aggregate profit of politically disorganized
unconstrained firms is given by Equation (3.5).
}
?
? = ?
) 1 (
) (
* *
*
2 / 1
1
)) ( ) ( ) ( ( ) (
x M
k w
U
dw
M
k i gk ? ? ? ? (3.5)
)) ( ) ( ) ( (
) 2 1 (
) 1 (
* *
2 / 1
? ? ?
?
k i gk
M
x ?
(
¸
(
¸
?
? ? = .
Substituting ) (
*
? k into the wealth threshold yields
) 2 1 ( ) (
*
? ? = k w . (3.6)
Substituting for ) (
*
? k and ) (
*
k w into Equation (3.5) gives
29
) ( 4
) 2 1 (
) 1 (
) ( 4 ) ( 2
) 2 1 (
) 1 ( ) (
2 2 2
?
?
? ?
?
?
i
g
M
x
i
g
i
g
M
x
U
(
¸
(
¸
?
? ? =
(
¸
(
¸
?
(
¸
(
¸
?
? ? = ?
. (3.7)
The derivative of the aggregate unconstrained firm profit with respect to financial
development is equal to
) ( '
) ( 4
) 2 1 (
) 1 (
) ( 2
) ( '
2
2 2
?
?
?
?
? i
i
g
M
x
Mi
g
U (
¸
(
¸
?
(
¸
(
¸
?
? ? + = ? . (3.8)
The first term is the increase in profits due to the increase in the number of unconstrained
firms as financial development rises. The second term corresponds to the drop in profits
caused by increase in the cost of capital.
Aggregate welfare for the unconstrained agents is equal to profits plus interest
income.
}
?
+ ? =
) 1 (
) (
*
1
) ( ) ( ) (
x M
k w
U U
dw
M
w i W ? ? ? . (3.9)
Derivative of unconstrained welfare is
) (
) 2 1 ( 2
2
) 2 1 (
2
) 1 (
) ( ' ) ( ' ) ( '
2
?
? ?
? ? ? i
M M
x M
i W
U U
?
+
|
|
.
|
\
| ?
?
?
+ ? = . (3.10)
The second term is the increase in interest income due to the increase in the interest rate
and the third term is the increase in aggregate interest income due to the increase in the
number of unconstrained firms.
3.3 Constrained Entrepreneurs
Constrained entrepreneurs are forced to invest a sub optimal level of capital into
their investment project. The level of capital investment is a function, ) , ( w k
c
? , is a
30
increasing function of their wealth endowment and financial development. Aggregate
profit for constrained entrepreneurs is
}
? = ?
) (
) 1 (
*
1
)) , ( ) ( ) , ( ( ) , (
k w
w
c c
C
dw
M
w k i w gk w ? ? ? ? . (3.11)
where ) , ( w k
c
? is defined in Equation (3.3).
Let dw
M
w k
w K
k w
w
c
C
}
=
) (
) 1 (
*
) , (
) , (
?
? be aggregate credit demand for constrained firms. Then
the change in profit for constrained entrepreneurs is
0 )) , ( ) ( ) ( ' ) , ( (
) , ( 2
) , ( '
) , ( '
2 / 1
> + ? = ? w K i i w K
w K
w gK
w
C C
C
C
C
? ? ? ?
?
?
? . (3.12)
See Appendix B for detailed derivative. The first term represents the increase in
revenue from the increase in financial development and the second term represents lost
profit due to an increase in the cost of capital. Constrained entrepreneurs welfare is
interest income plus profits:
dw
M
w
i w w W
k
w
w
C C
}
+ ? =
*
1
) ( ) , ( ) , ( ? ? ? . (3.13)
It is increasing in the level of financial development:
0 ) ( ' )) ( ) ( ' (
) (
)) ( ( 2
8 4 ) ( '
3
> ? + ?
?
+ + ? = ? ? ? ?
?
? ?
? ?
c C
gi i g
i
i g
W . (3.14)
3.4 Poor Lenders
Because they lack access to capital, the poorest entrepreneurs in the economy are
unable to set up their investment projects. These agents, along with households, lend
their wealth to the credit markets and earn
31
|
|
.
|
\
|
?
+ =
} }
2 /
0
) 1 (
0
2
2
) 1 (
) ( ) (
M w
L
dw
M
w
dw
M
w
N i W
?
? ? (3.15)
}
?
? +
|
|
.
|
\
|
+ ? ? =
2 /
0
2
'
2
2
) 1 (
) (
) (
) ( '
) (
)) 1 ( 1 ( 2 ) (
M
L
dw
M
w
i
i
i g
i
g
w W
?
?
?
? ?
?
? . (3.16)
Poor lenders experience two effects when financial development increases. Firstly, they
earn a higher return on interest income from lending. Secondly, poor entrepreneurs at the
margin of the initial wealth threshold are able set up firms because the minimum wealth
endowment required to borrow capital has decreased.
3.5 Aggregate Welfare
Aggregate welfare is the sum of all income earned from lending at the
endogenous interest rate and the profit made by the entrepreneurs:
|
.
|
\
|
? + ? + ? + + = ) , ( ) ( ) ( ) 3 (
8
) ( ) ( w
M
i N W
C U SIG
? ? ? ? ? ? .
The total increase in aggregate income that results from an increase in financial
development is
) , ( ) ( ) ( ' ) , (
) , ( 2
) , ( '
) (
) ( 4
) 2 1 (
) ( 2
) 2 1 (
) 3 (
8
) ( ) ( '
2 / 1
2
2 2 2
w K i i w K
w K
w gK
i
i
g
M
M
i
g
M
M
i W
C C
C
C
? ? ? ?
?
?
?
?
?
?
?
? ? ?
+ ? +
?
(
¸
(
¸
(
¸
(
¸
?
? ?
?
+ + ? =
. (3.17)
Figure 2.5 graphs aggregate welfare as a function of financial development.
Social welfare improves when financial development increases. Recall the aggregate
wealth endowment is equal to dw
M
w
N dw
M
w
N
M M
} }
+
?
0
2 /
0
2
2
) 1 (?
. Under the initial
parameter specification the aggregate wealth endowment is equal to 1.04. Final
32
aggregate wealth at 1 . = ? is 2.86799. Final aggregate wealth at 49 . = ? is 4.088. When
credit markets are imperfect, strong creditor enforcement increase aggregate final wealth
by more than 40% over weak enforcement.
Aggregate Social Welfare
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
A
g
g
r
e
g
a
t
e
F
i
n
a
l
W
e
a
l
t
h
Figure 2.5: Social Welfare as a Function of Creditor Rights Enforcement
In this section, I have specified the welfare functions for each of the constituent
groups. The policymaker will choose the level of financial development that maximizes a
linear function of aggregate welfare and campaign contributions. I analyze the
politician’s utility function in Section 4.
4. Political Utility
In the Grossman and Helpman (GH) model, special interest groups try to
influence the policymaker’s policy choice by giving her political gifts. The SIG’s design
contribution schedules that associate gifts to the politician with every policy option
available to her in order to maximize its own objective function. Contributions are
assumed nonnegative. Mimicking the principal in the principal agent game the SIG
33
designs a payment scheme, C(), “to give the politician the appropriate incentives to act
on its behalf” (GH 2001).
The politician cares about the total level of political contributions and aggregate
well-being. She desires contributions because they can be used to finance campaign
spending (among other benefits). Social welfare is a concern to the politician because her
constituents are more likely to reelect her if she has delivered a high standard of living.
The policymaker maximizes a utility function ) , ( c G ? where is the policy choice made
by the policymaker and c is the political contributions. Grossman and Helpman state
that, “The utility function is meant to capture the policymaker’s personal preferences
over the various possible policy outcomes, as well as her concern for her future electoral
prospects. The policy will affect the politician’s chances of being reelected if voters
look retrospectively at her record when deciding whether to vote for her in subsequent
elections”.
14
Because the government uses contributions to pursue its own political gain,
G(.) is increasing in c and is a single peaked function for any level of c. Following GH, I
assume that the policymaker’s utility function is a linear function of her concern for the
welfare of the electorate and her desire for political gifts. The weights on aggregate
welfare and campaign contributions sum up to 1. Furthermore, I use the simplest version
of the GH model by limiting the interaction to that between the political entity and a
single interest group. Thus, the policymaker’s utility is given by
) ( ) 1 ( ) ( ) , ( ? ? ? ? ? C W c G ? + = (4.1)
14
A less cynical view of the government is that it cares about the overall welfare of its constituency and
since the policy variable, p, directly affects societal well-being it is a component in the government’s
utility.
34
where ) (? W gives the aggregate welfare as a function of the policy choice. The
term ) (? C represents contributions from special interest group.
One of the attributes of the GH model is that it allows for liberal interpretation of
the parameters. The GH interpretation of the weights on utility is the policymakers’
tastes for campaign finance. In addition to this interpretation the weights could be
thought of as the degree of corruption in the economy. In more corrupt societies
politicians find it easier to take bribes from elite rent seekers that benefit from regressive
financial policies than in less corrupt systems. In addition, could represent the capacity
of the political system to reform, or otherwise improve policies that increase aggregate
social welfare. In my application of the GH model to financial development, I interpret
the weights generically to be the degree of democratic accountability that the political
system requires of the government. In societies with high accountability, policymakers
must weight general welfare heavily, where the opposite is the case politicians are free to
pursue rent-seeking behavior by courting the favors of narrowly defined special interest
groups.
My model predicts that political systems with a high degree of democratic
accountability will be more likely to improve their financial policies. In the model a
politician is able to increase aggregate welfare by reforming financial policy. She
increases SIG welfare by maintaining the status quo level of financial repression. In a
political system with accountable (unaccountable) institutions the politician gives
aggregate (SIG) welfare a weight of one. In anti democratic political systems,
policymakers value campaign contributions with a weight of one. The model predicts
that proclivity of political institutions towards accountability to the general electorate will
35
determine the politician’s concern for aggregate welfare or SIG welfare when making her
policy decisions. Thus institutional details of the political system affect the governments’
incentives to reform the credit market.
Strictly speaking the policymaker described in my interpretation of the GH model
is always self-interested, without constraint she would prefer to maximize the rents of
political office. It is the institutional constraints inherent in the economy, which
determine the political advantage of trading off aggregate welfare for contributions and
vice versa. Institutions determine whether the capital that the politician earns is political
(people are happy with her policy performance) or monetary (she earns campaign funds
and other political gifts from interest groups), the politician acts in order to ensure her
ability to be reelected. As the benevolent social planner is a useful fiction to help
describe optimal government behavior, the completely self-interested policymaker is also
beneficial in this model when discussing the policy reform that occurs in equilibrium.
The political institutions determine the avenue by which the politician acquires the
“capital” necessary to be reelected.
Grossman and Helpman define the equilibrium policy ?
~
as the policy that
maximizes the interest groups utility function ) ,
~
( c U ? subject to the constraint that
) 0 ,
ˆ
( ) , ( ? ? G c G ? . The equilibrium interest group contribution c
~
makes the politician
just indifferent between the ?
~
and the policy she would pick if no contributions were
forthcoming from the SIG, ?
ˆ
. In this model favored policies that increase aggregate
welfare may be abandoned or modified by the policymaker in order to capture the rents
of political office. The equilibrium policy that results from this principle agent game is
36
pareto efficient in the sense that one actor cannot be made better of without the other
becoming worse off.
If I assume that the SIG utility function can be written as a function of its welfare
from the government’s policy choice minus what it pays in political contributions,
c W c U
SIG
? = ) ( ) , ( ? ? , then the politician’s objective function can be rewritten as
) ( ) 1 ( ) ( ) , ( ? ? ? ? ?
SIG
W W c G ? + = , (4.2)
The policymaker chooses the policy by maximizing a linear weighted function of
aggregate welfare and SIG welfare. This model provides the basis from which to
evaluate the ability of a small closed economy to enact reform given the strength of the
special interest within the economy and the policymaker’s predilection for political gifts
determined by the economy’s political institutions.
In Section 5, the GH model is applied to the case of reform in the credit market.
Improvements in financial development increase aggregate growth in the economy.
Elite agents, those individuals in the upper tail of the wealth distribution, experience a
decrease in welfare when the level of financial development is increased. The elites have
the ability and the incentive to form a SIG that lobbies the policymaker to block financial
development. The equilibrium policy that is chosen by the policymaker is dependent on
degree of accountability imposed by the society’s political institutions, the productive
capacity of the economy, the extent of wealth inequality, and the supply of capital.
Figure 2.6 gives the model’s timeline.
37
Date t=0 Date t=1 Date t=2
|_________________________________|_______________________________|
Politician receives Politician chooses Agents make their
contribution schedule investment decisions
from the interest group and consume final wealth
Figure 2.6: Sequence of events and decisions
5. Political Equilibrium
In this section, I investigate how the SIG uses political gifts to sway the
policymaker into enacting a low level of financial development. At the beginning of
period one, the politician decides whether to reform the credit market. Based on the
government’s financial policy, citizens decide how much capital to invest in their
projects. The differing effect of credit market efficiency on the returns of borrowers and
lenders divides the politician’s constituency into the poor who benefit from better
enforcement of legal sanctions and the rich who profit from inefficient credit markets.
Recall from Section 4 that the policymaker’s objective is to maximize a political
utility function
) ( ) 1 ( ) ( ) , ( ? ? ? ? ?
SIG
c W c G ? + = (4.1)
where ) (? W is aggregate final wealth and ) (?
SIG
c is the contribution the SIG gives the
policymaker for a given policy variable, . The contribution is decreasing in financial
development because the SIG is worse off by increasing creditor rights.
The political utility function has a first order condition equal to
0 ) ( ' ) 1 ( ) ( ' = ? + ? ? ? ?
SIG
c W . (5.1)
I assume that the contribution schedule is differential in . This
implies ) ( ' ) ( ' ? ?
SIG SIG
W c = . I can substitute ) ( ' ?
SIG
W for ) ( ' ?
SIG
c in the politician’s first
38
order condition, which yields Equation (6.2) below. Therefore, Equation (5.1) can be
rewritten as
0 ) ( ' ) 1 ( ) ( ' = ? + ? ? ? ?
SIG
W W . (5.2)
Substituting the first order conditions for SIG and aggregate welfare, Equations (3.4) and
(3.17) respectively, into the policymaker’s first order condition yields (5.3) below.
( )
0 ) ( ) 2 (
2
) (
) ( 16
) 1 (
) ( ) ( ) , (
) , ( 2
) , (
) (
) ( 4
) 2 1 (
) ( 2
) 2 1 (
) 3 (
8
) (
2
2
2 / 1 2
2 2 2
=
)
`
¹
¹
´
¦
? ? + ?
?
?
+
¦
)
¦
`
¹
¦
¹
¦
´
¦
? + ?
?
+ ?
|
|
.
|
\
|
|
|
.
|
\
| ?
? ?
|
|
.
|
\
|
|
.
|
\
| ?
+ + ?
? ?
?
?
? ? ?
?
?
?
?
?
?
?
? ? ?
i x x
M
x i
i
g
i i w K
w K
w gK
i
i
g
M
M
i
g
M
M
i
C
C
C
The policymaker chooses a that satisfies the first order condition for
maximizing a weighted sum of aggregate welfare and the welfare of the SIG. The policy
choice is dependent on the exogenous parameter values -- the productivity level g,
abundance of capital, M, the aggregate wealth of the special interest group, x, and the
level of political accountability imposed by institutions, . Financial policy is dependent
on the economy’s political institutions, , which determines the politician’s ability to
maximize rents to the detriment of social welfare.
Result 1: Political institutions that impose more democratic accountability on policy
makers will generate higher levels of financial development.
Political institutions that impose democratic accountability make it difficult for
the policymaker to succumb to the interests of elites that lobby for financial repression.
When accountability is low, political utility is decreasing in financial development. For
political systems that impose democratic institutions, the political utility function is
upward sloping in financial development. Figure 2.7 graphs political utility for different
39
levels of political accountability. Low values of are indicative of political institutions
that require governments have little accountability to the general electorate. As the
weight increases the politician’s concern for general welfare also increases. Anti
democratic political institutions result in a political utility functions that is a decreasing
function of financial development. Political systems that impose more accountability in
policy choices lead to a political utility function that is increasing in financial
development. Figure 2.7 also shows that there exists a political structure in which the
policymaker is exactly indifferent between low and high levels of financial development.
In the graphs this point occurs at a political weight of = ? .032824.
The parametric model establishes the importance of political accountability in
determining the extent financial development. However, other papers have been hesitant
to confirm a relationship between political institutions that impose accountability and
financial development. Beck et al. (2001) minimize the importance of political structure
in explaining financial structure. Using measures of political environment that include
competitiveness in elections, government openness, and inter-party competition, the
authors find “a weak, fragile link between political structure and finance development”.
Instead, Beck et al. (2002) attribute the change in financial development to legal origin
and initial endowment of colonies settled by Europeans. Taking a broader view Glaser et
al. (2004) criticize empirical techniques used in previous literature to relate political
institutions and economic growth. The authors’ main argument is that countries can
improve human capital without democratic accountability and that once these economies
become richer they can improve institutions.
40
Acemoglu et al. (2004) summarize a large literature connecting political
institutions and economics growth by stating “political institutions place all political
power in the hands of a single individual or a small group, economic institutions that
provide protection of property rights and equal opportunity for the rest of the population
are difficult to sustain.”
15
My model supports the authors’ conclusion that in accountable
political systems it is more difficult for politicians to ignore the welfare of the majority in
favor of the preferences of a small number of elite. Furthermore, by rigorously
formalizing the role of political structure in financial development, I am able to
demonstrate the significance of political institutions in generating changes in financial
policy and economic outcomes. Though the investigation of political structure poses an
empirical challenge to researches, its connection to economic growth should not be
disregarded.
Figure 2.7: Political Utility
15
The new institutional economics literature supporting a relationship between political structure and
economic growth include Buchanan and Tullock (1962), North (1981, 1990), Knack and Keefer (1995),
Hall and Jones (1999), Acemoglu, Johnson and Robinson (2001, 2002). Empirical research identifying
politics and firm behavior include See Hellman et al. (1998); Hellman and Shankermann (2000); Krozner
(1998) and (1999); Faccio (2002); Denizer et al. (1998); Morck et al. (1998); Gordan and Lei (1999);
Johnson and Titman (2002); and Laffont and Tirole (1991).
41
Political Accountability alpha = .01
1.125
1.13
1.135
1.14
1.145
1.15
1.155
1.16
1.165
1.17
1.175
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .02
1.15
1.155
1.16
1.165
1.17
1.175
1.18
1.185
1.19
0 0.2 0.4 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .03
1.175
1.18
1.185
1.19
1.195
1.2
1.205
1.21
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Acountability alpha = 0.0328204
1.18
1.185
1.19
1.195
1.2
1.205
1.21
1.215
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .04
1.195
1.2
1.205
1.21
1.215
1.22
1.225
1.23
1.235
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha =.05
1.215
1.22
1.225
1.23
1.235
1.24
1.245
1.25
1.255
1.26
1.265
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .06
1.23
1.24
1.25
1.26
1.27
1.28
1.29
1.3
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
t
i
y
Political Accountability alpha = .08
1.27
1.28
1.29
1.3
1.31
1.32
1.33
1.34
1.35
1.36
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .1
1.3
1.32
1.34
1.36
1.38
1.4
1.42
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
42
6. Comparative Statics Under Imperfect Credit Markets
Section 6 investigates the comparative static properties of the model by exploring
the effects of four shocks to the economic environment – productivity, wealth, openness,
and inequality. To evaluate the effects of a small shock to the economy on the level of
political accountability required for reform, I choose a range of parameter values around
the initial restrictions imposed in the parametric model given in Table 2.2. The ranges of
the parameter values are constrained by the theoretical restrictions also listed in Table
2.2. Table 2.33 shows the ranges of the exogenous parameters that I evaluate. For each
incremental change in the exogenous parameter values I calculate the level of political
accountability, , that makes the politician exactly indifferent between low financial
development (=.1) and high levels of financial development ( =.49).
Table 2.3: Parameter Ranges
Comparative Static Scenarios Exogenous Parameters
Change in Productivity ] 0 . 5 , 0 . 4 [ ? g
Percent Change in Wealth Per Capita w increased by 10% to 100%
Change in Openness to Financial Inflows ] 5 . , 0 [ ? f
Percent Change in Wealth Inequality increased by 5% to 25%
6.1 Change in Productivity
Result 2: Politicians are more likely to enforce creditor rights in more productive
economies.
In this scenario, the productivity parameter, g , is increased in increments of 1 . 0
from 4.0, the initial starting value to 5.0. All other exogenous parameters are held
constant at the initial values listed in Table 2. Figure 2.8 plots accountability points
that make the politician indifferent between the lowest level of creditor rights
enforcement under imperfect markets 1 . = ? and the highest, 49 . = ? . The figure shows
43
that as the economy’s productive capacity increases, policymakers weight aggregate
welfare more than SIG welfare and are therefore more likely to increase financial
development. Consider two economies. Economy A has a high rate of productivity and
Economy B has a low rate of productivity. Given that both economies have a similar
political structure, a policymaker is more likely to reform the financial system in
Economy A then in Economy B. According to this result, the constraints imposed by the
political system on policymakers are partially offset by exogenous improvements in
productive capacity. Additionally as g increases the wealth threshold required to borrow
enough capital to set up projects fall, increasing firm entry. Aggregate welfare rises when
productivity increases, causing the politician to value social welfare more even if it is
politically easy to acquire gifts from elites. Result 2 is consistent with Hall and Jones
(1999) who find evidence that social infrastructure, including political structure, is a
significant determinant capital accumulation and productivity.
Effect of Productivity (g) on Financial
Development
0.0280
0.0290
0.0300
0.0310
0.0320
0.0330
0.0340
0 0.2 0.4 0.6 0.8 1 1.2
Increase in Productivity (g)
P
o
l
i
t
i
c
a
l
A
c
c
o
u
n
t
a
b
i
l
i
t
y
Figure 2.8: Plot of points that make the policymaker just indifferent between high and low financial
development for different levels of productivity ranging from g = 4 to g=5.
44
6.2 Percent Change in Wealth Per Capita
Result 3: The likelihood of financial reform increases as wealth per capita increases.
In the second comparative static scenario, I increase the initial wealth
endowments per capita from 10% to 100% for both households and entrepreneurs. The
increase in initial wealth increases the level of available credit in the economy. As the
supply of capital increases the equilibrium interest rate decreases for each level of
creditor rights enforcement, . The loss in welfare that SIG members suffer due to
improvements in creditor rights enforcement is partially offset by the decrease in the cost
of capital. Additionally, constrained entrepreneurs experience welfare gains because the
wealth thresholds required for firm entry and optimal investment fall. Figure 2.9 plots
the levels of political accountability, , that makes the politician indifferent between low
and high financial development as the level of wealth per capita is increased. As Figure
2.9 demonstrates, it is easier for the politician to develop credit markets in wealthier
economies where the supply of credit is abundant.
Effect of Wealth on Financial Development
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0 0.2 0.4 0.6 0.8 1 1.2
% Increase in Wealth Per Capita (w)
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Figure 2.9: Plot of points that make the policymaker just indifferent between high and low financial
development for different levels of aggregate initial wealth as wealth per capita varies.
45
For very wealthy economies, the condition that 0 ? ? is violated. Politicians would have
to receive negative utility from improvements in aggregate social welfare in order to
choose a low level of financial development.
Result 3 is an illustration of path dependence in corporate ownership structures
discussed in Bebchuk and Roe (1999). The authors show that current corporate
ownership is dependent on patterns of past ownership. Furthermore, the authors argue
that special interest politics shapes corporate governance, stating “A country’s initial
pattern of corporate structures influences the power that various interest groups have in
the process producing corporate rules. If the initial pattern provides one group of players
with relatively more wealth and power, this group would have a better chance to have
corporate rules that it favors down the road.” In my theoretical model, as per capita
wealth increases, politically influential SIG members are harmed less by financial
development because the increase in available credit lowers production costs. The same
results hold in reverse. When the economy is poor and capital is scarce, the SIG suffers
more when credit markets rights are strongly enforced than in a capital abundant
economy. Result 3 suggests that special interests will be more of an obstacle to financial
reform in poor emerging markets than in industrialized nations.
6.3 Change in Financial Openness
Result 4: Capital openness has a negative effect on financial development.
So far, the simulations have been generated under the assumption of a closed
economy. I examine the effect of international capital supply on financial reform by now
assuming that capital is allowed to flow into the economy from abroad. In order to
simulate openness I increase the level of capital supply M, by adding f, where f
46
represents capital inflows and ranges from an increase of 10% to 60%. Capital supply
under an open economy is given by
} }
+
+ ? =
f M M
ply
dw
M
w
dw
M
w
K
0
2 /
0
sup
) 1 (? . Foreign
investors only supply capital to domestic agents in the model; they do not start projects.
Since the supply of capital increases while demand remains unchanged, the equilibrium
interest rate falls. I assume that the world interest rate,
w
i , is normalized to 0. As long as
the endogenous equilibrium interest rate is greater than
w
i foreigners will want to invest
capital in the domestic economy. Foreign interest income does not enter into the
policymaker’s political utility function. Her utility function is entirely composed of
domestic aggregate welfare and the welfare of the SIG.
Effect of Capital Inflow (Openness) on
Financial Development
0.032
0.033
0.034
0.035
0.036
0.037
0.038
0.039
0.04
0 0.1 0.2 0.3 0.4 0.5 0.6
Increases in Capital Inflows (f)
P
o
l
i
t
i
c
a
l
A
c
c
o
u
n
t
a
b
i
l
i
t
y
Figure 2.10: Plot of points that make the policymaker just indifferent between high and low
Financial Development for different degrees of capital openness
Other studies, namely Rajan and Zingales (2003), have suggested that capital
openness induces elite groups to change their demands for financial repression by
increasing the opportunities for international investment and growth. However, the
authors also explain that financial openness in the absence of trade openness is not
47
sufficient to change the preference of large domestic firms for financial repression. “…in
the absence of domestic or foreign competition in product markets, these [large] firms
will have little need to access external funds. Moreover, given the state of information
asymmetries across markets, it is unlikely that small domestic firms are likely to be
financed directly by foreign investors. If potential domestic entrants are unlikely to be
financed by foreigners, industrial incumbents will still retain an incentive to keep entrants
at bay by opposing financial development.” Rajan and Zingales conclude that both trade
and financial openness are necessary for elite groups to back financial development. In
my model, lower interest rates lead to higher profits for elite entrepreneurs. A politician
heavily influenced by the special interest group ( is low) may choose a lower level of
creditor right enforcement since the loss in aggregate welfare would be offset by an
increase in elite welfare. In this manner, the model suggests that increasing the capital
account has the unattended effect of decreasing access to finance for small and medium
entrepreneurs while increasing elite profits.
Trade openness provides competition in the product markets that lowers profits
and internal cash flow, forcing large domestic firms to depend on external finance.
Financial openness requires sound macroeconomic policies that reduce the government’s
ability to direct credit to favored firms. While the most profitable firms can tap into
foreign capital, other firms, now dependent on external finance, may support financial
development in order to increase access to credit.
In my model, the inadequacy of financial openness alone to provide an incentive
for SIG members to back enforcement of creditor rights is manifested by the reaction of
the interest rate to capital inflows. The additional capital decreases the equilibrium
48
interest rate because capital supply increases more than capital demand. The lower
interest rate increases the optimal level of capital firms would like to invest, while
lowering the return to lending. Wealthy agents have less incentive to lend in the credit
markets and more incentive to invest in their projects at the lower interest rate. The
decrease in the interest rate benefits SIG members, therefore aggregate social welfare
increases overall even though lower interest rates harm poor citizens who must lend their
wealth endowments. Figure 2.10 demonstrates how the “rich get richer and the poor get
poorer” impact of capital inflows into the domestic economy affects the nature of
financial reform. As foreign investment increases exogenously into the system, the
policymaker has less incentive to reform. Elites continue to pressure the politician to
maintain a low level of creditor enforcement in order to swallow up all the additional
credit and increase profits.
16
6.4 Percent Change in Wealth Inequality
Result 5: Wealth inequality has a negative effect on financial development.
Section 6.4 investigates the likelihood of financial reform under increasing wealth
inequality by increasing , the number of agents in the bottom domain of the wealth
distribution (poor). Household wealth is equal to
}
2 /
0
M
dw
M
w
? . The parameter is
increased by .05% to .25% from a starting value of 3. Figure 2.11 graphs levels of
political accountability that make the policymaker indifferent between high and low
16
It is important to note that along with capital, financial openness brings institutional improvements to the
domestic economy such as good corporate governance and foreign bank competition. The institutional
features of financial openness are not modeled in this paper. Result 4 suggests that the benefit of capital
openness is its effect on financial institutions rather than the corresponding increase in capital. According
to the model, capital inflows alone will not increase the level of financial development.
49
financial development as wealth inequality varies. The figure shows that the more
unequal wealth is distributed in a country the higher the level of political accountability,
, required for financial development.
Effect of Inequality on Financial Development
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0 0.05 0.1 0.15 0.2 0.25
% Increase in Inequality
P
o
l
i
t
i
c
a
l
U
t
i
i
l
i
t
y
Figure 2.11: Plot of points that make the policymaker just indifferent between high and low financial
development for different levels as wealth inequality varies.
Wealth inequality in the simulated economy reduces the equilibrium interest rate
and increases the return to setting up projects. SIG members benefit from the low cost of
capital that inequality brings about under financial repression. In Figure 10, I plot a
range of wealth inequality parameters, , for the two extreme levels of financial
development, 1 . = ? and 49 . = ? . Figure 2.12 shows elite entrepreneurs benefit more
from financial repression when wealth is more unequally distributed, i.e. as increases.
Elites also suffer more from financial development when increases.
50
Effect of Inequality on SIG Welfare
1.1
1.11
1.12
1.13
1.14
1.15
1.16
1.17
0 0.05 0.1 0.15 0.2 0.25
% Increase in Inequality
S
I
G
F
i
n
a
l
W
e
a
l
t
h
=.1
=.49
Figure 2.12: Plot of SIG welfare under different degrees if wealth inequality under financial
repression (=.1) and financial development (=.49)
Result 5 is consistent with other research exploring the connection between
inequality and financial development. Gradestein (2004) argues that in countries with
high aggregate income and more equally distributed wealth, influential elites are more
willing to support the enforcement of private property rights. Perotti and Volpin (2004)
show that inequality reduces the level of minority investor protection in equity markets.
Engerman and Sokoloff (2002) study economic institutions across the Americas and
argue “that with more extreme inequality or heterogeneity in the population were more
likely to develop institutional structures that greatly advantaged members of elite classes
(and disadvantaging the bulk of the population) by providing them with more political
influence and access to economic opportunities.” While these studies have provided an
empirical linkage between inequality and financial development, my theoretical model
provides some insight into why such a relationship exists. Initial economic inequality
leads to persistent political inequality. Thus, financial policy is formed to benefit the
economically elite and politically influential at the expense of economic growth. In this
manner, Result 5 is another example of the path dependence of financial policy.
51
7. Concluding Remarks
The central lesson of this chapter is that political structure - namely the degree of
democratic accountability imposed by a country’s institutions - significantly influences
the capacity of policymakers to enact financial reforms. The model supports previous
empirical and anecdotal evidence that special interest groups influence the degree of
financial intermediary development. However, the model’s conclusions shed doubt on
the supposition that all governments respond uniformly to the pressure of powerful
economic agents in society, demonstrating instead that a complete interest group
depiction of financial reform includes the role of idiosyncratic political institutions in
shaping the incentives and constraining the actions of governments.
I have presented a theoretical model of reform that unites these two political
determinants of financial development. First, an interest group uses political
contributions to influence government financial policy. Second, the politician’s
sensitivity to interest group pressure depends on whether or not political institutions
require her to be accountable to the general electorate. Furthermore, the level of
democratic accountability necessary for governments to improve financial development
is dependent on the country’s economic environment; including productivity, wealth per
capita, financial openness, and the degree of wealth inequality. Financial development
leads to popular political support from most citizens however; political repression
enlarges the campaign funds in the politician’s coffers. Both sources of political capital
help the policymaker in her quest to be reelected. Institutions determine the source of
political capital that is more advantageous to the policymaker. The model’s predictions
raise a variety of empirical questions. While the literature has considered some of them,
52
answers are far from conclusive. In conclusion, I simply discuss the questions I believe
deserve further exploration.
First, what is the likelihood that policymakers trade aggregate social welfare in
favor of interest groups? Democratic accountability, the main determinant of financial
development in my model, has several components. Tsebelis (2002) shows that political
systems with a large number of polarized veto players are less likely to enact reforms that
improve aggregate welfare. Lederman et al. (2001) suggest that political institutions that
enforce accountability tend to be less corrupt hence policymakers have less incentive to
engage in rent seeking behavior at the expense of aggregate welfare. Other aspects of
accountability such as competitive elections and independent judicial systems ensure that
policymakers must value the well-being of the citizenry when making policy decisions.
Empirical exploration into all the different components of political accountability and
their relationship to financial development is needed.
Second, do countries with high industrial concentration and heavy regulation of
new business entry tend to focus on aspects of financial development that are beneficial
to large industrial incumbents (equity markets) as opposed to reforms that increase new
business entry (credit markets)? Singh (1995) found that large corporations in
developing countries rely heavily on equity to finance the investment. Unlike the US and
the UK where equity markets developed as result of market forces, Singh finds that
emerging market governments have taken a proactive role in developing their stock
markets. Furthermore, Singh (1997) suggests that developing countries may enact
policies that increase equity issuance by large firms but neglect more substantive policy
measures to financial institutions.
53
Third, do capital inflows (FDI) reduce a country’s incentive to reform its credit
markets? Singh (1997) also discusses how less developed country governments may be
reacting to pressure by large firms to ensure cheap access to capital and by international
investors to open financial markets. By developing their equity markets, governments
can appease both interest groups and may chose to do so even if that means endangering
welfare-enhancing improvements to credit markets and property rights.
Fourth, are governments whose legal structure accommodates active intervention
in financial markets more responsive to interest group pressure? La Porta et al.
(1997,1998) argue that common law regimes are superior in producing efficient legal
systems to civil law. In contrast, Rajan and Zingales (1999, 2003) argue that civil
systems are more efficient than common law systems in adopting good policies because
all legislation emanates from the center. The centralized nature of civil law systems also
means that governments are more influenced by special interests when making their
policy decisions. Therefore, the government’s tendency to intervene in markets is
proxied by legal institutions: civil law countries are more likely to intervene in financial
policy than common law governments. A government whose legal structure
accommodates intervention to a greater extant may respond more willingly to interest
group pressure.
Fifth, in the theoretical model entrepreneur faces the same level of creditor rights
enforcement, . In reality, firms face different obstacles in attaining financing for
investments based on their size. A natural extension of the model is to explore the effects
of disparate credit market imperfections on the equilibrium interest rate and access to
capital.
54
Finally, what role does persistent inequality and low capital availability play in
sustaining financial repression? I have discussed the theoretical importance of path
dependence in explaining patterns of corporate ownership. More work is needed to asses
the empirical significance of poverty and inequality on financial development.
My paper is an attempt to provide a theoretical framework for analyzing these
questions. As emphasized by Acemoglu et al. (2004), “a theory of why different
countries have different economic institutions must be based on politics, on the structure
of political power, and the nature of political institutions… Constructing formal models
incorporating and extending these ideas is the most important task ahead.”
55
Chapter 3: An Empirical Analysis of Political Institutions and Credit
Market Development
1. Introduction
In the previous chapter, the central conclusion of my theoretical model reveals the
role of the political system in implementing financial reform: political systems that
impose more democratic accountability on policymakers achieve higher levels of
financial development. The theory shows that the level of accountability in the political
system determines the politician’s preference between special interest group
contributions and aggregate welfare. Accountable political systems limit government
protection of elite advantage based on insecure property rights by making it difficult for
politicians to accept monetary enticements from special interests; hence, politicians
choose financial policy that increases aggregate welfare.
Assumed in both the theory of special interest politics and its application to
financial development is that all governments respond uniformly to the pressure of the
powerful economic agents in society, as such the interest group depiction of financial
reform ignores the role of institutions in shaping the incentives and constraining the
actions of governments. Political scientists Garret and Lange (1995) recognize the
omission of institutions from a theory of interest group led policy reform, stating,
“Researchers have assumed that the effects of internationally generated changes in the
constellation of domestic economic preferences will be quickly and faithfully reflected in
changes in policies and institutional arrangements within countries. If one understands
which economic interests have gained economic strength, one knows which has gained
political power and in turn how the policy is going to change … [However], institutions
56
invariable outlive the constellation of interests that created them, and hence they provide
barriers to market driven change”.
Other researchers have explored the manner in which institutions affect the
overall quality of the financial system. Djankov, McLiesh, and Shleifer (2004) examine
whether legal rights that protect creditors against default and creditor registries that
collect information on the credit history of borrowers are associated with higher levels of
credit market development. The authors conclude that while both types of institutions are
important for intermediary development, creditor rights are more influential in rich
countries and credit registries play a more significant role in poor countries.
Additionally, Desai, Dyck, and Zingales (2004) find that the institutional details of the
tax system affect the quality of corporate governance. The higher the tax rate the more
incentive managers have to divert corporate profits. On the other hand, strong tax
enforcement reduces managerial diversion of profits and increases stock market value.
This chapter investigates whether the institutional details of the political system
play significant role in achieving financial development that is independent of the level of
democratic accountability between policymakers and voters. I find that political
institutions, as summarized by the number and cohesion of its veto players, are weakly
related to financial development. The chapter proceeds as follows. Section 2 discusses
the role of veto players in policy reform and democratic accountability. Section 3
presents the data. Section 4 discusses the four empirical strategies I use to examine the
relationship between veto players and financial development. Section 5 presents the
results of each empirical strategy. Section 6 concludes.
57
2. Veto Players under Different levels of Political Accountability
Political institutions play a critical role in sustaining democratic accountability. A
convenient method for summarizing a country’s political institutions is by the
characteristics of the veto players in the political system. Tsebelis (1995, 2002) defines
the concept of veto players as “individual or collective actors whose agreement (by
majority rule for collective actors) is required for change in the status quo”. He provides
a method of categorizing governments by the number and policy cohesion of veto
players. The veto player framework is antecedent to the idea of checks and balances
found in the American Constitution and federalist and French revolutionary writings and
has been utilized implicitly and explicitly in contemporary comparative politics. It
allows for comparisons across all political regime types including presidentialism and
parliamentarism, unicameralism and bicamerialism, and two-party and multi-party
systems.
An example of the veto player framework is found in the comparison of American
and British political systems. Often the US and British political systems are paired
together as having similar characteristics. However, the differences between the two
systems are more compelling, “presidential vs. parliamentary systems, bicameralism vs.
(de facto) unicameralism, undisciplined vs. disciplined parties, appointed vs. independent
bureaucracies and the presence vs. the absence of supreme courts”. Thus, the British
system is a single veto player political structure while the American system accords with
a multi veto player set up. (One needs only to concede the impossibility of Thatcher style
market reforms occurring in a system of frequent gridlock like the US to understand the
capability of the veto framework to explain differing levels of resistance to policy change
58
by governments). Simplifying the nexus of comparison to the ability of political systems
to reform policy reduces the confusion of multi-dimensional analysis. The categorization
of governments by the number and cohesion of veto players is the most important
contribution of the veto player framework advanced by Tsebelis.
Along with a method of categorizing political systems, Tsebelis also presents a
theory for predicting which systems have a higher capacity to implement reform.
Tsebelis summarizes his theory as follows, “A significant policy change has to be
approved by all veto players and it will be more difficult to achieve the larger the number
of players and the greater ideological distance between them. Empirical research,
measuring the effect of veto players on policy reforms in areas as diverse as labor
markets and trade openness, supports Tsebelis’ results for advanced democracies.
17
Other researchers have suggested that Tsebelis’s propositions may not be true in
the context of weaker democracies. Keefer (2001) states that the “absence of multiple
veto players in countries often means that some groups in society are less represented
than they otherwise would be.” Political scientists Andrews and Montinola (2004) agree,
adding,
In advanced, industrial democracies, institutions that enforce the rule of law are
well established, and there is a relatively clear link between policy change (that is,
passage) and implementation. … And reform turns mainly on whether legislative
veto players can agree on passage of policies. In emerging democracies,
agreement on policy is not the only potential obstacle to reform. The more
important task for reformers is preventing passage of corrupt legislation and
ensuring proper implementation of genuine reforms.
17
See Baun (1999) and Hellerberg and Basinger (1998).
59
Tsebelis’s predictions are based on the supposition that the most formidable
obstacle to policy change is policy coordination between veto players. Andrews and
Montinola show that since the institutional constraints on legislators in emerging
democracies are weak, the chief obstacle to reform for these countries is collusion among
veto players in taking bribes. A multi-veto player framework makes it more difficult for
a politician to exploit his position for personal gain and retain office. Other researches
have made similar distinctions about the differences in the impact of the number of veto
players in advanced versus emerging democracies.
18
Moreover, Tsebelis intentionally ignores whether political institutions achieve
good or bad policies in order to focus attention entirely on the capability of political
systems to change the status quo, regardless of the effects of the change on the economy.
According to Andrews and Montinola (2004), good reforms strengthen the rule of law,
while bad reforms lead to expropriation by the government. Applying this idea to
financial markets, good reforms protect private property and creditor rights. Bad reforms
violate creditor rights and retard financial development.
I examine whether political institutions have disparate effects on credit market in
advanced versus weak democracies. In advanced democracies, obstacles to policy
change occur in only one dimension, coordination between policymakers. Policy
changes that increase the rule of law are more successfully implemented when there are
fewer veto players that block reform. However, a multi-veto player framework is
desirable when the reforms are bad because there are more vetoes to policy change. In
contrast, for moderately or weakly democratic states, it is desirable to have a multi-veto
18
See Hellman (1998) and Moser (1999).
60
player system whether the reform is good or bad because politically preferences formed
by weakly accountable political systems cause the primary obstacle to policy reform to be
collusion on bribes rather than coordination on policy. Multi-veto players prevent
collusion and increase voter representation insuring that good reforms are implemented
by legislatures and bad reforms are blocked. Moreover, since financial systems begin
from a status quo of financial repression that benefit elite interests groups within the
country, reforms are more likely to be of good quality (otherwise policymakers could
maximize their bribes by maintaining the status quo).
Other researchers have shown that multiple veto player political system increases
the likelihood of good quality reforms in weak democracies. Keefer and Savage (2001)
show that checks and balances in the political system increase the ability of independent
central banks to decrease inflation. Keefer (2001) suggests that multiple veto players
increase government credibility. Henisz (2000) provides evidence that political
constraints on the executive increase economic growth. Finally, La Porta et al. (2002)
show that checks and balances in the political system are correlated with political
freedom.
Though the number of veto players has disparate predictions depending on the
level of democracy, policy cohesion of the veto players has the same effect on capacity
for policy change for all levels of democratic accountability. Cohesion deals with the
policy coordination problem of veto players and is unrelated to the bribery collusion
problem produced by weakly accountable political systems in emerging democracies.
Thus, as Tsebelis asserts, we should expect that policy cohesion among veto players
increases the likelihood of reform.
61
I offer the following two testable hypotheses on the relationship between
intermediary development, accountability, and political institutions.
H1: Financial development increases with the number of veto players for
emerging democracies.
H2: Veto players that are more politically polarized are less likely to generate
policy change that leads to financial development for all levels of democratic
accountability.
3. Data
The panel covers 157 countries over 21 years. A complete description of the data
is shown in Appendix C. The data years are restricted to 1980-2001 based on research by
Singh (1997). Singh observes that governments in less developed countries started from
a level of financial repression then undertook a great deal of financial sector reform in the
1980’s and 1990’s. I use this period of heavy reform in the developing world to test the
prediction that political institutions play a role in the ability of governments to resist
special interest group pressures and to improve the financial sector.
To proxy financial intermediary development, I use Private Credit, the amount of
credit extended to private firms by financial intermediaries divided by GDP, from the
Database on Financial Development and Structure by Beck et al. (2002). Private Credit
is the preferred measure of intermediary development in the finance and growth literature
because it isolates private bank lending from credit issued by central banks, government
lending and credit to government run enterprises (Levine et al. (1999)). According to the
World Business Environment Survey (WBES) only 28% of small firms receive financing
62
from banks, where small firms are defined as enterprises with 5 to 50 employees.
19
Medium firms are defined as enterprises with 51 to 500 employees and large firms
employ over 500 workers. Medium and large firms receive 40% and 47% of financing
for investments from banks. From this survey data, one can infer that most private credit
is acquired by medium and large firms, which also have this most political influence
within a country.
Following Tsebelis (1999) and Keefer (2003), I measure political institutions
along two dimensions – the number of veto players, Checks, and the distance in their
policy positions, Political Polarization. Both variables are from the Database of Political
Institutions (DPI). Checks is measured by assigning one check if the government is
controlled by different parties in the executive and legislative branch for presidential
systems and assigning a check for party in the governing coalition under parliamentary
systems. Checks are also added to detail the competitiveness of national elections,
including whether elections have closed lists or open lists. In my empirical analysis, I
use the Log of Checks variable, because I believe that the number of veto players in the
political system has a nonlinear relationship to private credit. For instance, the marginal
effect on the dependent variable of going from, say, two checks, to three checks is much
greater than the marginal effect of going from ten checks to eleven checks.
Political Polarization captures the distance in policy agreement between
policymakers in a particular year and is measured by assigning the four largest political
parties a policy position along a left to right scale of economic policy. The largest
difference between two political parties is given from 0 to 2 is the degree of polarization.
19
The WBES surveys approximately 10,000 firms in 80 countries on issues related to financing obstacles.
63
To test the disparate effects of veto players on different levels of democracy I use
as a measure of political accountability, the annual Freedom House World Country scores
Freedom Index. The Freedom Index scores are found by rating countries by their
government’s protection of civil liberties and political liberties. I assume that countries
that protect the rights of their citizenry along these two dimensions are more likely to
enforce political institutions that reinforce accountability. The index places countries into
three categories – Free Countries, Partly Free Countries, and Not Free Countries. I
examine whether the variables measuring political institutions will have a different effect
on financial development for each category of political accountability.
Additionally, I investigate the ability of political institutions to increase
intermediary development under different degrees of openness. Rajan and Zingales
(2003) hypothesize that openness to trade and financial flows lead members of elite
interest groups to support financial development instead of repression. I measure the
independent effects of both types of openness on financial development in the presence
of political variables. Foreign direct investment (FDI) and trade openness (Trade) are
taken from the World Bank indicators database (2003).
I also control for the level of economic development. As a proxy for economic
development I use the Log of GDP per capita in 1995, Lgdp95. I used the 1995 measure
as opposes to GDP per capita at the beginning of the sample because the panel is
unbalanced and I will have more country-year observations towards the end of the series.
Inflation is also included in the regression to control for its effect on Private Credit over
the data’s time interval.
64
The countries are separated into three groups, Free Countries, Partly Free
Countries, and Not Free Countries. I assume that in Free Countries politicians are the
most accountable to their constituencies and that they are the least accountable in Not
Free Countries. I then analyze the data in order to determine if a country’s political
institutions increase Private Credit within the three different levels of democratic
accountability. In Tables 3.1A – 3.1C, countries are grouped by their political system’s
promotion of civil and political liberties. Even within these groupings, there exists
considerable variation in the data. Note that among Free Countries, countries include
highly industrialized nations as well as developing countries. Table 3.2 summarizes the
data. Panel A displays the summary statistics for Free Countries, Panel B for Partly
Free Countries and Panel C for Not Free Countries. Overall, Free Countries have the
highest level of financial development as proxied by Private Credit. It is instructive to
compare countries with highly accountable political institutions to political systems with
weaker voter representation in policy reform. The amount of Private Credit is just over
54% of GDP in Free Countries, almost twice as much as the percentage of Private Credit
in Partly Free Countries and two and a half times the percentage in Not Free Countries.
Free Countries also have the greatest variance in financial development of the three
groups. Moreover, Free Countries tend to have more Checks in the political system (not
surprisingly, because these are also the most democratic countries) and have veto players
that are the most ideologically polarized.
For the smallest economies, namely the African countries of Ethiopia, Liberia,
Chad, Sudan and Niger, Private Credit can be less than 15% of GDP and in some cases
significantly less. For the most economically advanced countries, like the US, the UK,
65
Japan, and Germany credit extended to private corporations can total over 90% of GDP.
The other variables of interest demonstrate considerable range over the data set as well.
The number of veto players necessary to implement reform in the financial sector ranges
from a low of 1 to up to 18 Checks. Most countries in the sample have 5 or fewer Checks
within the political system; however, there are notable exceptions. India tops the list with
as many as 18 Checks required to pass reform legislation. Other countries with Checks
over 5 include France, Turkey, Denmark, and Thailand. Political Polarization between
veto players ranges from 0 to 2, 0 indicating complete policy cohesion. The US, UK,
Israel, Turkey, Central African Republic, and France are among the most polarized
political systems, while Algeria, The Bahamas, Canada and Chile are included in the
least.
The openness indicators, Trade and FDI also show tremendous variation in the
sample. Partly Free Countries on average fall in the middle of Free and Not Free
political regimes in terms of financial development, number of veto players, degree of
polarization, wealth per capita and financial and trade openness.
Also of interest is the correlation between the variables in the bottom half of each
panel. Both the Log of Checks and Polarization are positively correlated with Private
Credit for Free Countries. The Log of Checks has a positive correlation and Polarization
a negative correlation in Partly Free Countries. Both institutional variables have a
negative correlation with Private Credit in Not Free Countries. Overall, economic
development as proxied by Lgdp95 is highly correlated with the Private Credit, Log of
Checks and Polarization however; the correlation between economic development and
the number of veto players and political polarization diminishes significantly for the
66
Partly Free and Not Free Countries. Furthermore, Trade and FDI are negatively
associated with Private Credit in the Free Countries and positively related with Private
Credit in the Partly Free Countries. In the next two sections, I analyze whether these
differences in signs between the explanatory and dependent variables for the different
levels of democratic accountability are suggestive of the disparate effects of political
institutions on financial development.
4. Empirical Strategy
I test the hypothesis of the paper by using four different empirical strategies. In
Strategy 1, I cluster the standard errors at the country level as in Desai et al. (2004). By
clustering the standard errors I avoid assuming that observations within each country are
independent of each other. For Strategy 2, I cluster the standard errors by country and
then test whether liberalization plays a key factor in the relationship between the political
variables and the proportion of private credit to GDP. According to Dermirguc-Kunt
and Detragiache (1998) and Williamson and Mahar (1998), many countries had
liberalized interest rates by 1995. I test whether the effect of political institutions is more
strongly related to financial development before the completion of liberalization in most
countries or afterwards.
In Strategy 3, I average the data over 5-year intervals and I do not cluster the
standard errors at the country level. This strategy is similar to the approach used by
Levine, Beck, and Loayza (2000). The authors examine the relationship between
institutions on financial development and growth and find that differences on legal and
accounting system help explain differences in credit market development.
67
In Strategy 4, I use data on firm ownership as the dependent variable in order to
determine of political institutions affect the ownership arrangement (family owned versus
widely held) in listed corporations. Widely-held corporations may be less numerous in
countries with political systems where financial policy can be more easily influenced by
narrow interest groups represented by family owned enterprises.
For each empirical strategy, I present 4 tables. The first displays the OLS
regressions of two equations, the Basic Equation and the Openness Equation. The Basic
Equation includes the political variables, Log of Checks and Polarization plus the two
controls, Lgdp95 and Inflation. The Openness Equation includes the basic variable plus
Trade and FDI.
In the second table, I add Legal Origin to the regression equation in order to
determine the independent effects of the legal regime and political institutions on Private
Credit. Much attention in the literature has been given to the role of legal origin in
influencing financial development.
20
La Porta et al. (1998) explore the contribution of a
country’s legal origin in the formation of its financial structure and its corporate
governance institutions, finding that legal origin — be it English common law, or French,
German or Scandinavian civil law — partly determines the quality of investor protection
and the size of the stock market versus the banking sector. The paper concludes that
English common law systems generally have the strongest investor protection
enforcement, followed by Germany, Scandinavian, and lastly, French civil systems.
Beck et al. (2002) suggest that political institutions do not explain financial development
independently of legal origin. I test the authors’ conclusions when segmenting the data
into the different levels of democratic accountability.
20
See La Porta et al. (1997, 1998), Beck et al. (2002), and Levine et al. (2000).
68
Next, I present instrumental variables estimation of the Basic and Openness
equation. The use of Lgdp95 may be problematic because the variable is highly
correlated with the other variables in the regression equation and with the error term.
Instrumental variables estimation is commonly used when there is contemporaneous
correlation between an independent variable and the error term. Rodrik et al. (2002)
estimates the independent contributions of geography, institutions, and trade on economic
growth. In their search for instruments for their explanatory variables that are
uncorrelated with the error term and highly correlated with the regressor that is being
instrumented, the authors appeal to geography-based instruments identified in the finance
and growth literature.
Acemoglu et al. (2001) use an instrument for GDP that is based on the mortality
of settlers who first colonized countries from Europe. If the settlers found hospitable
living environment they settled the country and built good institutions that lead to
economic growth. If the colonizers were morbidly vulnerable to environmental
conditions they built extracting institutions that lead to economic stagnation. Therefore,
settler mortality is highly correlated with GDP; however, it is also exogenous to
institutions. The log of settler mortality, Lsettler, is a good but somewhat restrictive
instrument because it can only be used on a sample of former colonies. In my data set, 70
countries in the sample have data on settler mortality. I also employ a second instrument,
distance from the equator in degrees from Hall and Jones (1999). Both instruments proxy
for economic development by correlating with “the extent of Western European
influence.”
69
A second potential concern with the OLS analysis is omitted variable bias when
controlling for trade and financial openness in the regression equation. Policy reforms
that enlarge the amount of credit lent to private firms may also encourage the country to
increase its trade in goods and attract more foreign capital. Without including the
specific policy reforms into the regression equation it is impossible to discern how much
of the change in Private Credit is due to Trade and FDI alone. Frankel and Romer
(1999) provide an instrument for actual trade shares by estimating a gravity model of
bilateral trade. A similar geography based instrument is not as easily derived for
financial openness. The log of their measure, Log FR, is based on geographical
characteristics such as distance between trading partners. I use this measure as an
instrument for the Trade variable.
5. Results
5A. Strategy 1: Clustering Standard Errors by Country
Table 3.3 displays the OLS regressions with the standard errors clustered by
country for the three different levels of democratic accountability, Free Countries, Partly
Free Countries, and Not Free Countries. For each category of democratic accountability,
I estimate the Basic and Openness Equations. I find that for no group of countries is the
Log of Checks a significant determinant of Private Credit under this empirical strategy.
Political Polarization is a negative and significant at the 5 % level for the Partially Free
group of countries in Regression 4. The variable has a coefficient of -0.083. When
Trade and FDI are added in Regression 4 the coefficient on Polarization drops to -0.053
at a 10% level of significance. The proxy for economic development, Lgdp95, is positive
and significant for all levels of accountability. Inflation is negative and significant
70
determinant of Private Credit for all regressions except Regression 6. Trade is an
important factor for Partially Free countries whereas FDI is an important determinant in
Free Countries.
The inclusion of Legal Origin in Table 3.4 does not substantively change the
relationship between the political variables and Private Credit discussed in Table 3.3.
The Log of Checks in the political system is not a significant determinant of the
dependent variable. Polarization is negative and significant at the 5% and 10% levels in
Regressions 3 and 4, respectively with coefficients of approximately -0.05 in both
regressions. French Legal Origin is significant at the 5% level for Partially Free
Countries only. However, the significance of French Legal Origin is not robust to the
inclusion of indicators for financial openness. German Legal Origin is a strong
contributor to Private Credit for the Free Countries, while intermediary development is
hampered by a Socialist Legal Origin in the Free and Not Free groups.
The last two tables estimate the Basic and Openness equations using IV
techniques. Panel A shows the first stage regressions. Regressions 1, 3, and 5 use the
log of settler mortality (Lsettler) as an instrument for Lgdp95 and Regression 2, 4, and 6
instruments for Lgdp95 using the distance from the equator in degrees (Disteq). Panel B
of Table 3.5 gives the second stage regressions. Under both the instruments, Polarization
is negative and significant at the 5% level for Partly Free Countries. Its coefficients are -
0.077 and -0.075 in Regressions 3 and 4, respectively.
Political Polarization is robust to the inclusion of FDI and Trade in Table 3.6
when Disteq is used as an instrument for Lgdp95 and Log FR is used as an instrument for
trade openness. In Regression 4, Polarization has coefficient of -0.074 and is significant
71
at the 5 % level. FDI is a significant factor of financial development in Free Countries
while Trade is negative and significant. Unlike the OLS estimation, Trade is not a
significant component of Private Credit for the Partially Free Countries.
5B. Strategy 2: The Effect of Liberalization
The second empirical approach takes into account the potential role of financial
liberalization occurring in many of the countries during the period under analysis. I test
whether the relationship between the political variables and financial development
changes after the time by which most countries had liberalized their financial markets, by
including a dummy variable for country-year observations after 1995, Group B. I then
include interaction terms of the Group B dummy variable and the two political variables,
Log of Checks and Polarization. I define the interaction terms as LchecksB and
PolarizationB. Finally, I test whether coefficients of the interaction terms are equal to
zero to determine if the relationship between the political variables and financial
development is altered by liberalization. I am not able to test the equality of the
coefficients for the polarization interaction term in the Not Free Countries case. Post
1995 all the observations for Polarization are exactly equal to 0. Since there is no
variation in the explanatory variable, it is automatically dropped from the regression
estimation.
Table 3.7 presents the results of the OLS estimation of the Basic and Openness
Equations. Also shown in the table are the F-statistics for whether the interaction terms
are equal to zero. The Log of Checks is not a significant determinant of Private Credit in
any of the regressions. Polarization is significant at the 5% level in Regression 3. The
interaction dummies are not significant. The F-statistics suggest that financial
72
liberalization did not significantly alter the relationship between the political variables
and financial development of the credit market.
Table 3.8 adds Legal Origin to the Basic and Openness Equations. The results are
similar to the Legal Origin regressions in Table 3.4 that employ empirical Strategy 1.
Political Polarization and French Legal Origin are both negative and significant at the
5% level in the Basic Equation only. Their coefficients are -0.046 and -0.125,
respectively. German Legal Origin makes a positive contribution to Private Credit for
Free Countries while a Socialist legal regime is a deterrent to credit development in Free
and Not Free Countries. According to the F-statistics, the interaction terms are not
significantly different from zero.
The IV estimation shown in Table 3.9 shows that Polarization is a negative
determinant of Private Credit for both the Basic and Openness Equations for Partly Free
Countries. Its coefficients are -0.85 and -0.082 respectively. When I control for trade
and capital openness and instrument for trade openness in Table 3.10, Polarization is
only a significant determinant in Regression 4. As in the previous regression, for both
Tables 3.9 and 3.10, liberalization does not significantly change the relationship between
the political variables and private credit.
5C. Strategy 3: Averaging Data in 5-year intervals
Empirical Strategy 3 averages the data into 5 non-overlapping intervals from
1980-2000. The standard errors are robustly estimated. Table 3.11 shows that the Log of
Checks is positive and significant at the 5% level for the Partially Free Countries. The
coefficients are 0.09 and 0.098 in Regressions 3 and 4 respectively. This result contrast
to no relationship between Log of Checks and Private Credit found in the first two
73
empirical approaches. It also does not support the hypothesis the multiple veto players in
the political system is more valuable to intermediary development in weaker
democracies. Polarization is found to be a significant negative determinant of Private
Credit in Table 3.11 for the Partially Free Countries. Its coefficients are -0.166 and -
0.094 in the Basic and Openness regressions, respectively.
Table 3.12 adds Legal Origin to the regressions. The Log of Checks is negative
and significant at the 5% level for Free Countries. Its coefficients are -0.122 and -0.127
for the Basic and Openness Equations, respectively. The Log of Checks is not a
significant factor for Private Credit in the Partially Free and Not Free groups of
countries. Polarization is negatively related to Private Credit at the 5% level in
Regression 3 and at the 10% level in Regression 4 of the Partially Free states. Table
3.13 presents IV estimation. The Log of Checks remains positive and significant for Free
Countries. It is also positive for the Partially Free Countries when Lsettler is used as an
instrument for Lgdp95 in Regression 3. Table 3.14 includes trade instrumented by Log
FR and it adds FDI to each regression. The relationship between Log of Checks and
Private Credit is unchanged from the previous table, it is positive and significant for Free
Countries, however, neither of the political variables is significantly related to Private
Credit for the Partially Free and Not Free Countries.
5D. Strategy 4: Ownership as the Dependent Variable
Empirical Strategy 4 estimates the effect of political institutions on the percentage
of publicly listed firms that are privately held in each country. The ownership data is
from in Mork Wolfenzon and Yeung (2004). It covers 31 countries. There are four
variables collected in the Morck et al. data set – the percentage of widely held public
74
firms in the country when control is inferred at 10%, the percentage of widely held public
firms in the country when control is inferred at 20%, the percentage of family owned
firms in the country when control is inferred at 10%, and the percentage of family owned
firms in the country when control is inferred at 20%. After I merge the ownership data
with my averaged data set, I have 28 observations therefore; I do not divide the countries
into 3 groups as before. Instead, I include the index of democratic accountability directly
into the regressions. The Freedom Index ranges from 1 (Free Countries) to 3 (Not Free
Countries) as before. Since most of the ownership data is collected from the years 1999
and 2002, I average the explanatory variables over the time interval 1996-2000. The
basic equation includes Log of Checks, Polarization, the Freedom Index, and Lgdp95.
The Openness Equation includes the Basic Equation plus Trade and FDI. I repeat the
OLS and IV estimation as in the other 4 strategies except I do not include in the
regressions the proxies for openness in the IV estimation because that leaves me with
only 6 observations. All tables under this empirical approach display the same result. I
do not find a relationship between the political variables and the level of ownership.
6. Conclusion
This chapter uses a number of different empirical approaches to test two
hypotheses. First, the number of veto players is positively related to intermediary
development in weak democracies. Second, political polarization deters credit market
development in all countries. There is no support for the first hypothesis as I find no
consistent relationship between the Log of Checks and Private Credit. Moreover, there is
only weak support for the hypothesis that political polarization negatively effects
intermediary development. To quote noted philosopher, Cornel West, “democracy
75
matters” however, the manner in which a country organizes its political institutions does
not seem to effect intermediary development once the level of democratic accountability
between politicians and constituencies has been taken into account. Thus, the analysis
presented in this chapter does not find any empirical advantage to the arrangement of
political institutions in restricting special interest group influence on financial policy.
76
Table 3.1A : Free Country and Years
Argentina (1984-2000) Gambia (1980, 1989-1993) Nigeria (1980-1983)
Australia(1980-2000) Germany (1980-2001) Norway (1980-2001)
Austria (1980-2001) Ghana(1980-1981,2000-2001) Panama (1994-2001)
The Bahamas (1980-2001) Greece (1980-2001) Papa New Guinea (1980-1992,
1998-2001)
Bangladesh (1991-1992) Grenada (1985-2001)
Barbados (1980-2001) Guyana (1993-2001) Peru (1980-1988, 2001)
Belgium (1980-2001) Honduras (1982, 1984-1992,
1997-1998)
Philippines (1987-1989, 1996-
2001)
Belize (1981-2001) Hungary (1990-2001) Poland (1990-2001)
Benin (1991-2001) Iceland (1980-2001) Portugal (1980-2001)
Bolivia (1982-1994, 1996-
2001)
India (1980-1990, 1998-2001) Romania (1996-2001)
Botswana (1980-2001) Ireland (1980-2001) Samoa (1989-2001)
Brazil (1985-1992) Israel (1980-2001) Slovak Republic (1994-1994,
1998-2001)
Bulgaria (1991-2001) Italy (1980-2001) Slovenia (1991-2001)
Canada (1980-2001) Jamaica (1980-2001) Solomon Island (1980-1999)
Cape Verde (1991 -2001) Japan (1980-2001) South Africa (1994-2001
Chile (1990-2001) Republic of Korea (1988-
2001)
Spain (1980-2001)
Colombia (1980-1988) Latvia (1991, 1994-2001) Sri Lanka (1980-1982)
Costa Rica (1980-2001) Lithuania (1991-2001) St. Lucia (1980-2001)
Croatia (2000-2001) Luxembourg (1980-2001) St. Vincent and the
Grenadines (1980-2001)
Cyprus (1981-2001) Malawi (1994-1998) Suriname (1988, 2000-2001)
Czech Republic (1993-2001) Mali (1992-1993, 1995-2001) Sweden (1980-2001)
Denmark (1980-2001) Malta (1980-1092, 1987-2001) Switzerland (1980-2001)
Dominica (1980-2001) Mauritius (1981-2001) Thailand (1989-1990,1998-
2001)
Dominican Republic (1980-
1992, 1998-2002)
Mexico (2000-2001) Trinidad and Tobago (1980-
2000)
Ecuador (1980-1995, 1998-
1999)
Mongolia (1991-1999) United Kingdom (1980-2001)
El Salvador (1997-2001) Namibia (1990-2001) United States (1980-2001)
Estonia (1991, 1993-2001) Nepal (1991-1992) Uruguay (1985-2001)
Fiji (1980-1986, 1999) Netherlands (1980-2001) Vanuatu(1980-1982,1989-
2001)
France (1980-2001) New Zealand (1980-2001) Venezuela(1980-1991,1996-
1998)
77
Table 3.1B : Partially Free Country and Years
Algeria (1989-1991) Guinea-Bissau (1991-2001) Nigeria (1987-1992, 1998-
2001)
Angola (1991) Guyana (1980-1992) Oman (1992)
Argentina (1982-1983, 2001) Haiti (1986-1987,1990,1994-
1999)
Pakistan (1985-1998)
Armenia (1991-2001) Honduras (1980-1981, 1983,
1993-1996, 1999-2001)
Panama (1980-1987,1990-
1993)
Bahrain (1981-1992) Hungary (1984-1989) Papua New Guinea (1993-
1997)
Bangladesh (1980-1990, 1993-
2001)
India (1991-1997) Paraguay (1980-1987, 1989-
2001)
Belarus (1991-1995) Indonesia (1980-1992, 1998-
2001)
Peru (1989-2000)
Benin (1990) Iran, Islamic Rep. (1980,
1984-1987)
Philippines (1980-1986,1990-
1995)
Bhutan (1980-1991) Jordan (1984-1987, 1989-2001) Poland (1980-1981, 1983-
1989)
Bolivia (1995) Kazakhstan (1991-1993) Romania (1991-1995)
Brazil (1980-1984, 1993-
2001)
Kenya (1980-1986, 1992) Samoa (1980-1988)
Bulgaria (1990) Korea, Rep. (1980-1987) Senegal (1980-2001)
Burkina Faso (1980-1981,
1983, 1992-2001)
Kuwait (1980-1989, 1992-
2001)
Seychelles (1992-2001)
Burundi (1992) Kyrgyz Rep (1991-1999) Sierra Leone (1980-1991,
1996, 1998-2001)
Cape Verde (1987, 1990) Latvia (1992-1993) Singapore (1980-2001)
Central African Republic
(1991-2001)
Lebanon (1980-1987, 1991-
1994, 2001)
Slovak Republic (1993, 1996-
1997)
Chile (1980-1981, 1983-1989) Lesotho (1980-1987, 1991-
2001)
Solomon Islands (2000-2001)
Colombia (1989-2001) Liberia (1983-1988, 1997-
2001)
South Africa ( 1980, 1983-
1993)
Congo, Rep. (1991-1996,
2000-2001)
Madagascar (1982-2001) Sri Lanka (1983-2001)
Cote d’Ivoire (1980-1987,
1989-1992, 1999-2001)
Malawi (1999-2001) Sudan (1980-1983, 1986-
1988)
Croatia (1991-1999) Malaysia (1980-2001) Suriname (1987, 1989-1999)
Cyprus (1980) Maldives (1980-1987) Swaziland (1980-1992)
Djibouti (1980-1981,1984,
1999-2001)
Mali (1991, 1994) Thailand (1980-1988,1991-
1997)
Dominican Republic (1993-
1997)
Malta (1983-1986) Togo (1999-2001)
Ecuador (1996-1997,2000-
2001)
Mauritania (2000-2001) Tonga (1980-2001)
Egypt, Arab Rep. (1980-1992) Mauritius (1980) Trinidad and Tobago (2001)
El Salvador (1980-1996) Mexico (1980-1999) Tunisia (1980-1992)
Estonia (1992) Moldova (1991-2001) Turkey (1980-2001)
78
Table 3.1B : Partially Free Country and Years (Continue)
Ethiopia(1991-1992,1995-
2001)
Mongolia (1990) Uganda (1980-1990, 1994-
2001)
Fiji (1987-1998, 2000-2001) Morocco (1980-2001) Ukraine (1991-2001)
Gabon (1990-2001) Mozambique (1991-1992,
1994-2001)
Uruguay (1980-1984)
The Gambia (1981-1988,
2001)
Namibia (1989) Vanuatu (1983-1988)
Ghana (1992-1999) Nepal (1980-1990, 1993-2001) Venezuela (1992-1995,1999-
2001)
Grenada (1980, 1984) Nicaragua (1980-1997, 1999-
2001)
Zambia (1980-1990, 1993-
2001)
Guatemala (1980, 1984-2001) Niger (1991-1995,1999-2001) Zimbabwe (1980-2000)
79
Table 3.1C : Not Free Country and Years
Algeria (1980-1988, 1992-
2001)
The Gambia (1994-2000) Niger (1980-1990, 1996-1998)
Angola (1980-1990, 1992-
1998)
Ghana (1982 -1991) Nigeria (1984-1986, 1993-
1997)
Argentina (1980-1981) Grenada (1981-1983) Oman (1980-1991, 1993-
2001)
Bahrain (1993-2001) Guatemala (1981-1983) PRK (1980-2000)
Belarus (1996-2001) Guinea-Bissau (1980-1990) Pakistan (1980-1984, 1999-
2001)
Benin (1980-1989) Haiti (1980-1985, 1988-1993,
2000-2001)
Panama (1988-1989)
Bhutan (1992-2001) Hungary (1980-1983) Paraguay (1988)
Bolivia (1980-1981) Indonesia (1993-1997) Poland (1982)
Bulgaria (1980-1989) Iran (1981-1983, 1988-2001) Romania (1980-1990)
Burkina Faso (1982, 1984-
1991)
Jordan (1980-1983, 1988) Rwanda (1980-2001)
Burundi (1980-1991, 1993-
2001)
Kazakhstan (1994-2001) Saudi Arabia (1980-2001)
Cameroon (1980-2001) Kenya (1987-1991, 1993-
2001)
Seychelles (1980-1991)
Cape Verde (1980-1986,
1988-1989)
Kuwait (1990-1991) Sierra Leone (1992-
1995,1997)
Central African Republic
(1980-1990)
Kyrgyz Rep (2000-2001) South Africa (1981-1982)
Chad (1980-2001) Lao PDR (1980-2001) Sudan (1984-1985, 1989-
2001)
Chile (1982) Lesotho (1988-1990) Suriname (1980-1986)
China (1980-2001) Liberia (1980-1982, 1989-
1996)
Swaziland (1993-2001)
Congo, Dem Rep (1980-2001) Madagascar (1980-1981) Syria Arab Rep (1980-2000)
Congo, Rep (1980-1990,
1997-1999)
Malawi (1980-1993) Togo (1980-1998)
Djibouti (1982-1983, 1985-
1998)
Maldives (1988-2001) Tunisia (1993-2001)
Egypt, Arab Rep. (1993-2001) Mali (1980-1999) Uganda (1991-1993)
Equatorial Guinea (1980-
2001)
Mauritania (1980-1999) Vietnam (1980-2001)
Ethiopia (1980-1990, 1993-
1994)
Mongolia (1980-1989) Zimbabwe (2001)
Gabon (1980-1989) Mozambique (1980-
1990,1993)
80
Table 3.2: Summary Statistics
Panel A: Free Countries
Private
Credit
Log of
Checks
Polarization Trade FDI Lgdp95 Inflation
Observations 1198 1134 1093 1243 1238 1290 1280
Mean 0.542 1.160 .7282 83.534 2.921 8.68 18.033
Std. Dev. 0.376 .552 .914 43.575 4.799 1.368 65.09
Min 0.0178 0 0 10.079 -6.897 5.044 -32
Max 1.790 2.890 2 290.710 93.720 10.696 969
Correlations
Private
Credit
1.000
Log of
Checks
0.255 1.000
Polarization 0.281 0.513 1.000
Trade -0.030 0.042 -0.060 1.000
FDI -0.028 -0.035 -0.017 0.376 1.000
Lgdp95 0.6946 0.461 0.513 -0.085 -0.064 1.000
Inflation -.2019 -0.0786 -0.0480 -0.1014 -0.0771 -0.1068 1.00
Panel B: Partly Free Countries
Private
Credit
Log of
Checks
Polarization Trade FDI Lgdp95 Inflation
Observations 888 918 918 998 988 1066 1007
Mean 0.274 0.492 0.109 74.565 1.794 7.006 29.512
Std. Dev. 0.240 0.620 0.414 49.097 3.581 1.226 89.56
Min 0.001 0 0 8.953 -28.622 3.898 -26
Max 1.552 2.890 2 361.179 39.776 10.068 976
Correlations
Private
Credit
1.000
Log of
Checks
0.1701 1.000
Polarization -0.047 0.458 1.000
Trade 0.531 -0.039 -0.111 1.000
FDI 0.277 0.052 -0.028 0.436 1.000
Lgdp95 0.597 0.007 0.004 0.577 0.267 1.000
Inflation -0.194 0.077 0.097 -0.004 -0.076 -0.026 1.00
81
Panel C: Not Free Countries
Private
Credit
Log of
Checks
Polarization Trade FDI Lgdp95 Inflation
Observations 533 753 735 696 688 783 709
Mean 0.223 0.154 0.015 68.102 1.720 6.389 20.306
Std. Dev. 0.208 0.393 0.168 68.011 7.662 1.158 56.019
Min 0 0 0 6.320 -82.810 3.898 -29
Max 1.197 2.079 2 275.232 145.210 9.727 637
Correlations
Private
Credit
1.000
Log of
Checks
-0.049 1.000
Polarization -0.025 0.391 1.000
Trade 0.112 -0.028 -0.037 1.000
FDI 0.004 -0.005 -0.009 0.369 1.000
Lgdp95 0.419 -0.032 -0.026 0.350 -0.0350 1.000
Inflation -0.236 -0.042 -0.013 -0.085 -0.011 -0.166 1.000
82
Table 3.3: OLS Regression of the Basic and Openness Equations
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Standard errors are clustered
at the country level. Absolute value of robust t-statistics are in parenthesis.* significant at 10%;
** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of Checks -0.041 -0.033 0.048 0.051 0.009 0.023
(1.09) (0.89) (1.38) (1.49) (0.29) (0.7)
Polarization -0.006 -0.01 -0.083** -0.053* -0.021 -0.034
(0.19) (0.33) (2.50) (1.94) (1.19) (1.73)
Inflation -0.001*** -0.001*** -0.001** -0.001*** -0.001** -0.001
(3.94) (3.91) (2.59) (3.04) (2.03) (1.75)
Lgdp95 0.184*** 0.18*** 0.114*** 0.078*** 0.094*** 0.088***
(7.85) (7.53) (6.37) (6.40) (5.89) (2.89)
Trade 0 0.002*** -0.001
(0.61) (3.36) (0.74)
FDI 0.008** -0.001 0.009
(2.21) (0.29) (0.75)
Constant -1.007*** -0.966*** -0.515*** -0.414*** -0.346*** -0.277**
(5.88) (5.49) (4.51) (5.40) (3.50) (2.02)
Observations 1000 956 724 668 455 396
R-squared 0.48 0.47 0.44 0.52 0.32 0.19
83
Table 3.4: OLS Regression of the Basic and Openness equations and Legal Origin
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Standard errors are clustered
at the country level. Each equation includes dummy variables for French, German, Scandinavian,
and Socialist legal origins. Absolute value of robust t-statistics are in parenthesis. * significant at
10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The dependent
variable is Private
Credit.
1 2 3 4 5 6
Log of Checks -0.066 -0.059 0.001 0.041 0.009 0.036
(1.66) (1.53) (0.04) (1.31) (0.28) (0.94)
Polarization 0.008 0.003 -0.055** -0.05* -0.015 -0.014
(0.28) (0.12) (2.12) (1.91) (0.51) (0.42)
Lgdp95 0.174*** 0.165*** 0.135*** 0.082*** 0.099*** 0.067**
(7.79) (7.35) (7.16) (4.88) (4.78) (2.39)
Inflation -0.001*** -0.001*** -0.001** -0.001** -0.001 -0.001**
(3.30) (3.24 (2.05) (2.15) (1.67) (2.16)
Legor_fr -0.016 -0.027 -0.12** -0.017 -0.026 0.023
(0.27) (0.46) (2.38) (0.36) (0.58) (0.5)
Legor_ge 0.325*** 0.356*** -0.057 0.112 0*** 0***
(2.85) (3.17) (0.76) (1.57) (0.00) (0.00)
Legor_sc -0.103 -0.078 0*** 0*** 0*** 0***
(0.7) (0.52) (0.00) (0.00) (0.00) (0.00)
Legor_so -0.225*** -0.236*** -0.134 -0.08 -0.29*** -0.184**
(5.19) (5.70) (1.75) (1.48) (4.70) (2.06)
FDI 0.01*** -0.001 -0.005
(2.01) (0.21) (0.7)
Trade 0 0.002*** 0.001
(0.24) (3.24) (0.92)
Constant -0.897*** -0.833*** -0.569*** -0.422*** -0.377*** -0.262
(4.90) (4.02) (5.45) (4.50) (2.92) (1.57)
Observations 908 871 629 592 302 260
R-squared 0.53 0.54 0.48 0.52 0.49 0.31
84
Table 3.5: IV Regression of the Basic Equation
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. Regressions
A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B instruments
Lgdp95 with Distance from the Equator in Degrees (Disteq). Standard errors are clustered at the
country level. Absolute value of robust t-statistics are in parenthesis. * significant at 10%; **
significant at 5%; *** significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Country Partly Free Country Not Free Country
A B A B A B
1 2 3 4 5 6
Log of Checks 0.215*** 0.551*** 0.048 0.388*** 0.310 -0.039
(2.88) (7.04) (0.56) (3.88) (1.80) (0.23)
Polarization 0.778 0.212*** 0.60 0.025 -0.423 -0.505
(1.50) (3.68) (0.56) (0.17) (1.46) (1.33)
Inflation 0 -0.001* 0.002 -0.003 0.010*** -0.004**
(0.70) (1.84) (1.61) (1.59) (2.86) (2.29)
Instrument -0.791*** 5.49*** -0.664*** 2.89*** -2.86 5.015***
(23.59) (16.81) (16.59) (6.35) (21.58) (12.38)
Constant 11.219*** 6.22*** 10.08*** 6.29*** 7.448*** 5.50***
(63.22) (58.81) (48.03) (54.89) (21.58) (57.11)
Observations 491 526 478 600 246 407
R-squared 0.60 .0.48 0.42 0.08 0.11 0.28
Panel B: Second Stage Regressions of the Basic Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1
2 3 4 5 6
Lgdp95 0.179*** 0.238*** 0.161*** 0.148** 0.222* 0.139
(7.40) (5.03) (4.40) (2.18) (1.86) (2.39)*
Log of Checks -0.026 -0.085 0.029 0.034 -0.022 -0.004
(0.63) (1.44) (0.87) (0.78) (0.40) -0.13
Polarization 0.001 -0.023 -0.077** -0.075** 0.021 0.003
(0.02) (0.58) (2.10) (2.57) (0.46) -0.13
Inflation -001*** -0.001*** -0.001 -0.001 -0.004*** -0.001
(3.44) (3.54) (3.23) (3.27)** (2.76) -1.62
Constant -0.990*** -1.374*** -0.925*** -0.735 -1.11 -0.628
(5.16) (4.03) (3.94) -1.58 (1.58) -1.8
Observations 491 526 478 600 246 407
R-squared 0.49 0.5 0.39. 0.43 . 0.27
85
Table 3.6: IV Regression of the Openness Equation
The Openness Equation includes the Log of Checks, Polarization, Inflation, Lgdp95, Trade, and
FDI. Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Trade is instrumented
with the Log of the Frankel-Romer proxy of natural trade openness (Log FR). Standard errors
are clustered at the country level. Absolute value of robust t-statistics are in parenthesis. *
significant at 10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Panel A: First Stage Regressions of Lgdp95
Log of Checks 0.152* 0.512*** 0.015 0.401*** 0.223 0.195
(1.71) (6.48) (.17) (4.23) (1.22) (1.47)
Polarization 0.070 0.244*** 0.109 0.090 0.318 -0.374
(1.06) (4.18) (0.92) (0.66) (0.90) (1.31)
FDI -0.001 0.017 0.085*** 0.052*** 0.065* 0.062***
(0.007) (1.21) (4.33) (2.95) (1.96) (2.96)
Inflation 0.001* -0.006 0.002* -0.001 0.009** -0.001
(1.81) (1.23) (1.89) (.45) (2.57) (-0.39)
Instrument -0.820*** 5.841*** -0.562*** 2.621*** -0.234 3.858
(15.47) (16.75) (11.15) (6.07) (3.47)*** (13.35)
LogFrankRom 0.20*** .124** -0.013 0.143** 0.023* 0.277***
(3.41) (2.55) (1.35) (1.99) (1.86) (4.75)
Constant 10.816*** 5.755*** 9.50*** 5.742*** 7.026*** 4.574***
(41.29) (30.26) (34.31) (24.45) (18.62) (23.94)
Observations 364 515 397 561 233 354
R-square 0.48 0.48 0.33 0.10 0.10 0.33
Panel B: First Stage Regressions of Trade
Log of Checks 7.055*** 4.454** 0.165 7.623*** 11.286** 11.285***
(3.24) (2.28) (0.04) (2.81) (2.47) (2.79)
Polarization -4.202** -6.570*** -12.54* -8.399** -16.188* -13.067
(2.58) (4.55) (2.36) (2.15) (1.83) (1.50)
FDI 2.506 2.431*** 10.33*** 6.614*** 5.489888 6.824***
(5.91) (7.00) (11.84) (13.04) (6.59) (10.63)
Inflation -0.016 -0.001 -0.201*** -1.23*** -0.121 0.001
(1.20) (.45) (13.62) (2.88) (1.30) (0.03)
Instrument -5.359*** 2.621*** -1.91 -24.947** 1.353 40.343***
(4.12) (6.07) (.85) (2.02) (0.80) (4.25)
Log FR 30.393 .143** -0.126 33.933*** 0.372 24.130***
(21.29) (1.99) (0.30) (16.44) (1.20) (13.64)
Constant 2.212 5.742*** 65.08*** -
29.801***
41.401*** -
25.758***
(0.34) (24.45) (5.29 (4.43) (4.39) (4.44)
Observations 364 561 397 561 233 354
R-square 0.65 0.10 0.31 0.50 0.16 0.45
86
Panel C: Second Stage Regression of the Openness Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1 2 3 4 5 6
Lgdp95 0.155*** 0.222*** 0.113 0.183** 0.263 0.211*
(6.43) (4.88) (1.21) (2.14) (0.61) (1.88)
Trade -0.002** -0.003** -0.002 0.001 0.002 -0.008*
(2.21) (2.67) (0.09) (.51) (0.07) (1.90)
Lchecks 0.009 -0.044 0.012 0.020 -0.054 0.056
(0.30) (0.87) (0.41) (0.43) (0.11) (0.78)
Polarization -0.036 -0.048 -0.084 -0.074** 0.059 -0.122
(1.05) (1.36) (0.25) (2.03) (0.09) (1.43)
FDI 0.023*** 0.020*** 0.054 0.001 -0.026 0.051*
(2.96) (3.12) (0.17) (0.11) (0.11) (1.74)
Inflation -0.001*** -0.001*** -0.001 -0.001*** -0.004*** -0.001
(3.66) (4.09) (0.25) (2.74) (2.98) (1.08)
Constant -0.719 -1.110*** -0.409 -1.020* -1.45 -0.677
(3.84) (3.65) (0.39) (1.81) (0.35) (1.23)
Observations 364 515 397 561 233 354
R-sq 0.50 0.53 . 0.23 . .
87
Table 3.7: OLS Regression of the Basic and Openness Equations and Liberalization
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic Equation plus Trade and FDI. Standard errors are
clustered at the country level. Each equation includes a dummy variable for country-year
observation after 1995 (GroupB) and interactions terms between GroupB and Log of Checks
(LCHECKSB) and GroupB and Polarization (POLARIZB). F-Tests of equality of coefficient
between the political variables, Log of Checks and Polarization, and the interaction terms are
given below. Absolute value of robust t-statistics are in parenthesis.
* significant at 10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1 2 3 4 5 6
Log of checks -0.033 -0.031 0.026 0.031 -0.004 0.014
(0.81) (.76) (1.05) (1.13) (0.19) (0.56)
Polarization 0.003 0.002 -0.068** -0.042 -0.008 -0.022
(0.11) (0.08) (2.28) (1.58) (0.53) (1.09)
LCHECKSB 0.018 0.029 0.069 0.059 0.048 0.022
(0.43) (0.67) (1.35) (1.43) (0.82) (0.31)
POLARIZB 0.021 -0.035 -0.051 -0.038
(0.76) (1.18) (1.02) (0.98)
GROUPB 0.129** 0.117** 0.000 -0.006 0.012 0.032
(2.55) (2.31) (0.01) (0.18) (0.29) (0.65)
Lgdp95 0.184*** 0.180*** 0.114*** 0.078*** 0.093 0.087***
(7.77) (7.59) (6.18) (6.18) (5.59) (2.96)
Inflation -0.001*** -0.001*** -0.001*** -0.001*** -0.061** -0.001*
(3.96) (3.94) (2.55) (2.96) (2.03) (1.80)
Trade -0.001 0.002*** -0.001
(0.64) (3.44) (0.74)
FDI 0.005 -0.002 0.009
(1.58) (0.36) (0.74)
Constant 1.070*** 1.010*** -0.516 -0.410*** -0.344 -0.276
(6.12) (5.73) (4.23) (4.91) (3.41) (2.06)
Observations 1000 956 724 668 455 396
R-squared 0.51 0.50 0.45 0.53 0.32 0.19
F-statistic
a
F(2,80) =
0.31
F(2, 77) =
0.73
F(2,79) =
0.91
F(2,74) =
1.02
F(1,58) =
0.67
F(1,53) =
0.09
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
88
Table 3.8: OLS Regressions of the Basic and Openness Equations, Legal Origin, and
Liberalization
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Standard errors are clustered
at the country level. Each regression includes a dummy variable for country-year observation
after 1995 (GroupB) and interactions terms between GroupB and Log of Checks (LCHECKSB)
and GroupB and Polarization (POLARIZB). Each regression includes dummy variables for
French, German, Scandinavian, and Socialist legal origins. F-Tests of equality of coefficient
between the political variables and the interaction terms are given below. Absolute value of
robust t-statistics are in parenthesis. * significant at 10%; ** significant at 5%; *** significant at
1%
Free Countries Partly Free Countries Not Free Countries
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of Checks -0.051 -0.052 -0.026 0.017 -0.01 0.007
(1.13) (1.2) (0.98) (0.63) (0.4) (0.25)
Polarization 0.015 0.015 -0.046** -0.043 0.007 0.02
(0.53) (0.54) (2.01) (1.69) (0.31) (0.92)
LCHECKSB -0.02 -0.005 0.081 0.069 0.059 0.069
(0.49) (0.12) (1.48) (1.52) (0.92) (1)
POLARIZB -0.01 -0.024 -0.042 -0.032 0*** 0***
(0.36) (0.85) (0.82) (0.78) (0.00) (0.00)
GROUPB 0.188*** 0.168*** 0.001 -0.006 0.025 0.041
(3.37) (3.06) (0.03) (0.17) (0.66) (1.05)
Lgdp95 0.173*** 0.165*** 0.135*** 0.082*** 0.098*** 0.065**
(7.67) (7.40) (7.06) (4.71) (4.77) (2.49)
Inflation -0.001*** -0.001*** -0.001* -0.001** -0.001 -0.001**
(3.10) (3.11) (1.95) (2.01) (1.73) (2.37)
Legor_fr -0.022 -0.033 -0.125** -0.022 -0.027 0.024
(0.39) (0.57) (2.38) (0.45) (0.6) (0.54)
Legor_ge 0.32*** 0.339*** -0.038 0.124 0*** 0***
(2.83) (3.07) (0.54) (1.77) (0.00) (0.00)
Legor_sc -0.108 -0.086 0*** 0*** 0*** 0***
(0.72) (0.57) (0.00) (0.00) (0.00) (0.00)**
Legor_so -0.29*** -0.291*** -0.131* -0.077 -0.281*** -0.17**
(5.60) (6.01) (1.95) (1.45) (4.44) (2.02)
FDI 0.005 -0.002 -0.008
(1.48) (0.34) (1.11)
Trade 0 0.002*** 0.001
(0.32) (3.33) (0.93)
Constant -0.963*** -0.88*** -0.566*** -0.418*** -0.376*** -0.262
(5.18) (4.30) (5.12) (4.18) (3.01) (1.67)
Observations 908 871 629 592 302 260
R-squared 0.57 0.57 0.49 0.53 0.5 0.34
F-statistic
a
0.25 0.47 1.32 1.26 0.84 1.00
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
89
Table 3.9: IV Regression of the Basic Equation and Liberalization
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. Regressions
A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B instruments
Lgdp95 with Distance from the Equator in Degrees (Disteq). Standard errors are clustered at the
country level. The Basic Equation includes the Log of Checks, Polarization, Inflation, and
Lgdp95. Each regression includes a dummy variable for country-year observation after 1995
(GroupB) and interactions terms between GROUPB and Log of Checks (LCHECKSB) and
GROUPB and Polarization (POLARIZB). F-Tests of equality of coefficient between the political
variables and the interaction terms are given below. Absolute value of robust t-statistics are in
parenthesis.*significant at 10%; **significant at 5%; ***significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Log of Checks 0.192** 0.466*** 0.024 0.273** 0.248 0.045
(2.04) (4.45) (0.24) (2.28) (1.17) (0.23)
Polarization 0.049 0.214*** 0.071 0.064 -0.363 -0.528
(0.77) (3.05) (0.48) (0.33) (1.19) (1.37)
LCHECKSB 0.052 0.155 0.089 0.376 0.168 -0.298
(0.35) (0.99) (0.55) (1.76) (0.46) (0.75)
POLARIZB 0.096 0.007 -0.006 -0.118
(0.87) (0.05) (0.03) (0.39)
GROUPB -0.116 -0.296 -0.221* -0.369*** 0.05 0.269***
(0.70) (1.73) (1.88) (2.64) (0.34) (2.30)
Inflation 0.000 -0.001*** 0.001 -0.003* 0.010*** -0.004**
(0.65) (2.06) (1.22) (1.82) (2.88) (2.41)
Instrument -0.790*** 5.527*** -0.661*** 2.813*** -0.284*** 4.875***
(23.51) (16.68) (16.52) (6.13) (4.48) (11.95)
Constant 11.264 6.362*** 10.143*** 6.425 7.421 5.441***
(59.45) (47.89) (47.85) (50.94) (21.09) (54.97)
Observations 491 526 478 600 246 407
R-squared 0.60 0.48 0.43 0.08 0.12 0.29
90
Panel B: Second Stage Regression of the Basic Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1 2 3 4 5 6
Lgdp95 0.178*** 0.232*** 0.161*** 0.152*** 0.221* 0.140**
(7.73) (4.86) (4.36) (2.12) (1.83) (2.41)
Log of Checks 0.011 -0.045 0.011 0.029 -0.036 -0.007
(0.28) (0.75) (0.33) (0.88) (0.80) (0.27)
Polarization -0.10 -0.20 -0.085** -0.082** 0.032 0.005
(0.37) (0.52) (2.15) (2.18) (0.80) (0.19)
LCHECKSB -0.06 -0.045 0.056 0.10 0.038 0.015
(1.19) (1.06) (0.92) (0.18) (0.52) (0.37)
POLARIZB 0.024 -0.015 0.004 0.010
(0.57) (0.33) (0.06) (0.19)
GROUPB 0.207*** 0.188*** -0.014 0.015 0.007 -0.010
(3.11) (3.66) (0.37) (0.33) (0.12) (0.29)
Inflation -0.001*** -0.001 -0.001*** -0.001 -0.004** -0.001
(3.34) (3.26) (3.30) (2.85) (2.64) (1.59)
Constant -1.08*** -1.419*** -0.831*** -0.768 1.105 0.632*
(5.75) (4.11) (3.31) (1.53) (1.56) (1.80)
Observations 491 526 478 600 246 407
R-squared 0.55 0.54 0.39 0.42 . .0.27
F-statistic
a
F(2,39)=
0.86
F(2,46)=
0.69
F(2,49)=
1.15
F(2,63)=
0.11
F(1,33)=
0.27
F(1,50)=
0.14
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
91
Table 3.10: IV Regression of the Openness Equation and Liberalization
The Openness Equation includes the Log of Checks, Polarization, Inflation, Lgdp95, Trade, and
FDI. Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Trade is instrumented
with the Log of the Frankel-Romer proxy of natural trade openness (Log FR). Standard errors
are clustered at the country level. The Basic Equation includes the Log of Checks, Polarization,
Inflation, and Lgdp95. Each regression includes a dummy variable for country-year observation
after 1995 (GroupB) and interactions terms between GROUPB and Log of Checks (LCHECKSB)
and GROUPB and Polarization (POLARIZB). F-Tests of equality of coefficient between the
political variables and the interaction terms are given below. Absolute value of robust t-statistics
are in parenthesis. * significant at 10%; ** significant at 5%; *** significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Log of
Checks
-0.033 0.401*** -0.002 0.269** 0.242 0.222
(0.81) (3.82) (0.02) (2.39) (1.09) (1.45)
Polarization 0.003 0.265*** 0.156 -0.156 -0.332 -0.397
(0.11) (3.74) (0.96) (0.90) (0.90) (1.35)
LCHECKSB 0.018 0.207 0.081 0.415** -0.065 -0.109
(0.43) (1.31) (0.43) (2.13) (0.16) (0.36)
POLARIZB 0.021 -0.034 -0.086 -0.207 Dropped Dropped
(0.76) (0.028) (0.36) (0.75)
GROUPB 0.129 -0.398** -0.252** -0.311** 0.018 -0.004
(2.55) (2.23) (2.00) (2.39) (0.12) (0.04)
FDI 0.005 0.030** 0.092*** 0.057*** 0.065* 0.063***
(0.23) (2.02) (4.62) (3.12) (1.90) (2.88)
Inflation 0.001* -0.001 0.002 -0.001 0.009** 0.000
(1.80) (1.53) (1.39) (0.62) (2.54) (0.37)
Instrument -0.820 5.935*** -0.564**** 2.573*** -0.233*** 3.855***
(15.44) (16.92) (11.17) (5.94) (3.42) (12.21)
Log FR 0.195*** 0.114** -0.016 0.147** 0.024* 0.276***
(3.35) (2.34) (1.68) (2.02) (1.73) (4.73)
Constant 10.884*** 5.918*** 9.60 5.838*** 7.014*** 4.574***
(4.02) (28.71) (34.19) (24.02) (18.15) (23.83)
Observations 364 515 397 561 233 354
R-squared 0.48 0.48 0.34 0.11 0.09 0.45
92
Panel B: First Stage Regressions of Trade
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Log of
Checks
7.380*** 7.998*** 5.118 3.943 10.074* 11.538**
(2.68)** (3.08) (1.03) (1.22) (1.82) (2.48)
Polarization -5.049 -7.348*** -10.158 -6.060 14.360 -12.843
(2.55) (4.20) (1.41) (1.19) (1.55) (1.43)
LCHECKSB -0.906 -7.702* -14.319 11.596** 3.148 -1.216
(0.22) (1.97) (1.70) (2.07) (0.31) (0.13)
POLARIZB 2.630 1.769 -1.802 -6.911
(0.77) (0.58) (0.17) (0.87)
GROUPB -1.524 9.094** 1.178 -7.601** 4.54 2.866
(0.31) (2.06) (0.21) (2.04) (1.23) (0.96)
FDI 2.648*** 2.237*** 10.693*** 6.706*** 5.241*** 6.76***
(5.22) (6.00) (12.04) (12.85) (6.16) (10.10)
Inflation -0.017 -0.006 -0.223*** -0.129*** -0.116 -0.001
(1.23) (0.045) (3.93) (2.95) (1.24) (0.03)
Instrument 5.353 15.281 -2.24 26.028** 1.510 39.240***
(1.433) (1.76) (1.00) (2.09) (0.89) (4.10)
Log FR 30.363*** 27.678*** -0.233 34.07*** -0.489 24.108***
(21.19) (22.98) (0.55) (16.35) (1.44) (13.60)
Constant -1.959*** -20.341*** 66.797*** 27.6*** 39.80*** -26.142***
(6.691) (3.99) (5.36) (3.96) (4.13) (4.49)
R-squared 0.64 0.61 0.32 .050 0.16 .
Observations 364 515 397 561 233 354
93
Panel C: Second Stage Regressions of the Openness Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The
dependent
variable is
Private
Credit.
1 2 3 4 5 6
Lgdp95 0.154*** 0.212*** 0.098*** 0.188** 0.270 0.208*
(6.81) (4.94) (2.90) (2.09) (0.63) (1.90)
Trade -0.002** -0.003** 0.002 0.001 0.003 -0.008*
(2.19) (2.54) (0.31) (0.53) (0.12) (1.91)
Log of
Checks
0.033 -0.003 0.006 0.018 -0.073 0.064
(1.48) (0.07) (0.14) (0.52) (0.19) (0.87)
Polarization -0.052** -0.044 -0.053 -0.085* 0.081 -0.126
(2.05) (1.42) (0.72) (1.88) (0.17) (1.44)
LCHECKSB -0.062 -0.053 0.019 -0.002 0.022 -0.032
(1.49) (1.26) (0.18) (0.02) (0.21) (0.26)
POLARIZB 0.049 -0.014 0.057 0.022
(1.12) (0.35) (1.34) (0.38)
GROUPB 0.107* 0.181*** -0.038 0.024 -0.011 0.031
(1.79) (3.16) (0.81) (0.45) (0.07) (0.51)
FDI 0.016** 0.012* 0.000 0.000 -0.031 0.049
(2.25) (1.92) (0) (0.02) (0.18) (1.77)
Inflation -0.001*** -0.001 0.000 -0.001 -0.004** -0.001
(3.73) (3.85) (0.30) (2.31) (2.34) (1.11)
Constant 0.746*** -1.119 -0.551* -1.063 -1.54 -0.665
(4.13) (3.88) (1.74) (1.76) (0.41) (1.23)
Observations 364 515 397 561 233 354
R-squared 0.52 0.57 0.43 0.20 . .
F-statistic
a
F(2,32)=
1.59
F(2,45)=
0.92
F(2,42)=
0.93
F(2,60)=
0.13
F(1,30)=
0.05
F(1,46)=
0.07
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
94
Table 3.11: OLS Regression of the Basic and Openness Equations
The Basic equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Data is averaged over 5-year
intervals. Absolute value of robust t-statistics are in parenthesis. * significant at 10%; **
significant at 5%; *** significant at 1%.
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The Dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of checks -0.0868 -0.089 0.09** 0.098** 0.068 0.083
(1.77) (1.7) (2.12) (2.34) (0.94) (0.9)
Polarization -0.012 -0.012 -0.166*** -0.094** -0.132 -0.113
(0.45) (0.42) (2.64) (2.06) (1.52) (1.25)
Inflation -0.001*** -0.001*** -0.001*** -0.001*** 0 0
(3.87) (3.63) (4.23) (5.28) (1.49) (1.65)
Lgdp95 0.205*** 0.202*** 0.126*** 0.07*** 0.11*** 0.103**
(10.55) (9.47) (8.01) (4.62) (7.08) (2.65)
Trade 0 0.002*** -0.001
(0.8) (4.65) (1.26)
FDI 0.01 -0.004 0.006
(1.7) (0.71) (0.89)
Constant -1.132*** -1.094*** -0.613*** -0.39*** -0.459*** -0.36**
(8.47) (7.01) (6.00) (4.58) (5.03) (2.09)
Observations 187 180 109 102 70 61
R-squared 0.52 0.51 0.51 0.64 0.35 0.2
95
Table 3.12: OLS Regression of the Basic and Openness Equations and Legal Origin
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Data is averaged over 5-year
intervals. Each regression includes dummy variables for French, German, Scandinavian, and
Socialist legal origins. Absolute value of robust t-statistics are in parenthesis. * significant at
10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The Dependent
variable is Private
Credit.
1 2 3 4 5 6
Log of Checks -0.122** -0.127** 0.006 0.087 0.081 0.144
(2.07) (2.22) (0.13) (1.84) (1.08) (1.4)
Polarization -0.003 0.004 -0.099** -0.09 0.283 -0.058
(0.08) (0.13) (2.01) (1.82) (1.8) (0.22)
Lgdp95 0.203*** 0.189*** 0.152*** 0.074*** 0.113*** 0.052*
(8.64) (7.64) (9.65) (3.61) (5.04) (1.84)
Inflation -0.001*** -0.001** -0.001 0 0 0**
(2.80) (2.40) (1.43) (1.23) (1.18) (2.32)
Legor_fr 0.014 -0.015 -0.159*** -0.014 -0.038 0.037
(0.3) (0.32) (3.93) (0.33) (0.69) (0.64)
Legor_ge 0.309*** 0.337*** -0.102 0.128* 0*** 0***
(3.65) (3.94) (1.75) (1.87) (0.00) (0.00)
Legor_sc -0.103 -0.092 0*** 0*** 0*** 0***
(1.08) (0.98) (0.00) (0.00) (0.00) (0.00)
Legor_so -0.154*** -0.197*** -0.159 -0.172 0*** 0***
(2.94) (3.96) (1.21) (1.23) (0.00) (0.00)
FDI 0.013 -0.005 -0.008
(1.66) (1) (0.18)
Trade 0 0.002*** 0.001
(0.36) (3.92) (0.99)
Constant -1.101*** -0.988*** -0.665*** -0.41*** -0.486*** -0.245
(6.04) (4.59) (6.69) (3.91) (3.26) (1.48)
Observations 168 163 92 88 41 35
R-squared 0.58 0.58 0.57 0.64 0.61 0.38
96
Table 3.13: IV Regression of the Basic Equation
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. Regressions
A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B instruments
Lgdp95 with Distance from the Equator in Degrees (Disteq). Data is averaged over 5-year
intervals. Absolute value of robust t-statistics are in parenthesis. * significant at 10%; **
significant at 5%; *** significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of
checks
0.499** 1.047*** -0.062 0.400 0.650 -0.291
(2.21) (4.81) (0.22) (1.21) (1.08) (0.45)
Polarization 0.111 0.279** -0.526 -0.278 -0.918 -3.535
(0.72) (1.99) (1.40) (0.51) (0.44) (0.70)
Inflation 0.001 -0.002 -0.002 -0.005 0.008 -0.002
(0.49) (1.08) (0.77) (1.56) (0.61) (0.99)
Instrument -0.733*** 4.123*** -
0.781***
0.126 -0.100 5.25***
(8.13) (6.40) (6.25) (0.10) (0.54) (5.04)
Constant 10.705*** 6.117*** 10.745 6.87*** 6.236*** 5.370***
(21.708) (23.03) (16.32) (19.70) (6.12) (21.49)
Observations 85 93 69 87 33 62
R-squared .56 .54 .42 -0.002 -0.069 .32
Panel B: Second Stage Regressions of the Basic Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Lgdp95 0.202*** 0.272*** 0.204*** 1.175 0.554 0.146***
(6.02) (6.62) (5.72) (.10) (.50) (3.01)
Lchecks -0.067 -0.191** 0.059 -0.346 -0.186 0.035
(1.06) (2.22) (1.15) (0.07) (0.23) (.46)
Polarization -0.009 0.023 -0.099 0.162 0.398 0.299
(0.25) (0.63) (1.09) (0.05) (0.35) (1.48)
Inflation -0.002** -0.002** -0.001 0.004 -0.007 -0.000
(2.49) (2.83) (.13) (0.07) (0.72) (1.03)
Constant -1.141*** -1.556*** -1.160*** -7.847 -2.990 -0.672**
(4.69) (5.79) (4.63) (0.09) (0.48) (2.35)
Observations 85 93 69 87 33 62
R-squared .51 .57 .40 . . .2
97
Table 3.14: IV Regression of the Openness Equation
The Openness Equation includes the Log of Checks, Polarization, Inflation, Lgdp95, Trade, and
FDI. Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Trade is instrumented
with the Log of the Frankel-Romer proxy of natural trade openness (Log FR). Data is averaged
over 5 year intervals. Absolute value of robust t-statistics are in parenthesis. * significant at 10%;
** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Panel A: First Stage Regressions of Lgdp95
Log of
Checks
0.586* 1.092*** -0.124 0.595* 0.445 0.348
(1.93) (4.66) (0.41) (1.84) (0.63) (0.86)
Polarization 0.034 0.289** -0.168 -0.049 -2.341 -2.381
(0.17) (2.00) (0.40) (0.10) (0.48) (0.76)
FDI 0.011 0.009 0.212*** 0.068* 0.208 0.104
(0.22) (0.27) (3.31) (1.73) (1.26) (1.92)
Inflation 0.002 -0.002 -0.001 -0.003 -0.011 -0.000
(0.67) (0.95) (0.33) (1.14) 90.55) (0.31)
Instrument -0.704*** 4.11*** -0.409** 0.076 -0.058 4.239***
(4.07) (5.84) (2.26) (0.06) (0.30) (6.35)
Log FR 0.084 -0.017 -0.015 0.332 0.026 0.287**
(0.56) (0.17) (0.60) (1.66) (0.66) (2.38)
Constant 10.235*** 6.073*** 8.702*** 5.600*** 5.940 4.268***
(12.67) (14.19) (8.87) (8.18) (5.46) (10.54)
Observations 63 91 57 83 31 54
R-squared 0.39 0.53 0.40 0.05 -0.06 0.48
Panel B: First Stage Regressions of Trade
Log of
Checks
3.207 3.92 -9.896 19.684* 17.239 11.386
(0.41) (0.55) (0.73) (1.90) (0.99) (.78)
Polarization -6.642 10.93** -7.365 -12.552 21.460 59.679
(1.27) (2.49) (0.39) (0.78) (0.18) (.53)
FDI 3.167** 2.772** 23.419*** 8.360*** 13.604*** 10.703***
(2.55) (2.71) (8.12) (6.65) (3.36) (5.47)
Inflation -0.014 -0.001 -0.005 -0.128 -0.160 0.023
(0.23) (0.01) (0.05) (1.44) (0.32) (0.48)
Instrument -15.704*** 25.232 9.248 57.455 2.019 25.835
(3.51) (1.18) (1.14) (1.45) (0.43) (1.08)
Log FR 35.30*** 27.922*** -0.705 45.484*** -0.831 20.535***
(9.07) (9.13) (0.64) (7.14) (0.85) (4.75)
Constant 32.91 -16.505 0.018 -61.634*** 31.318 -20.366
(1.57) (-1.27) (0) (2.83) (1.17) (1.40)
Observations 63 91 57 83 31 54
R-squared 0.68 0.060 0.61 0.59 0.36 0.441
98
Panel C: Second Stage Regressions of the Openness Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Lgdp95 0.215*** 0.267*** 0.067 0.694 -6.050 0.180**
(4.57) (5.56) (1.36) (0.45) (0.02) (2.48)
Trade -0.002** -0.002** 0.000 -0.003 -0.197 -0.010***
(2.02) (2.47) (0.32) (0.27) (0.02) (3.23)
Log of Checks -0.039 -0.149 -0.001 -0.187 6.199 0.087
(0.43) (1.40) (0.02) (0.28) (0.02) (0.38)
Polarization -0.067) -0.051 -0.40 -0.130 -9.907 0.720
(1.47) (1.27) (0.73) (0.40) (0.02) (1.32)
FDI 0.028** 0.023*** 0.040 -0.006 3.961 0.107***
(2.48) (2.92) (1.42) (0.15) (0.02) (3.37)
Inflation -0.002** -0.002*** -0.00 0.001 -0.103 -0.000
(2.55) (2.74) (1.58) (0.22) (0.02) (.98)
Constant 1.153*** -1.409*** -0.281 -4.186 42.525 -0.443***
(3.27) (4.59) (1.86) (0.44) (0.02) (11.19)
Observations 63 91 57 83 3131 54
R-squared 0.48 0.60 0.69 . . .
99
Table 3.15: OLS Estimation of Ownership
The dependent variable is the percentage of the publicly listed firms in each country that are
widely held. The Basic equation includes the Log of Checks, Polarization, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Explanatory variables are
averaged over the years 1996-2000. Absolute value of robust t-statistics are in parenthesis. *
significant at 10%; ** significant at 5%; *** significant at 1%
Widely-held10 Widely-held20
Basic Openness Basic Openness
1 2 3 4
Log of Checks 16.076 25.756 10.437 21.396
(0.96) (1.45) (0.56) (1.14)
Polarization -9.407 -11.815 -12.228 -14.42
(1.13) (1.41) (1.37) (1.58)
Democracy -9.053 5.045 14.778 0.539
(0.89) (0.33) (.11) (0.03)
Lgdp95 7.228 8.921* 12.931 14.733*
(1.45) (1.78) (1.67) (1.92)
Trade -0.098 -0.104
(1.20) (0.98)
FDI -0.710* -0.842
(1.78) (1.50)
Constant -48.556 78.850 -74.574 -107.537
(1.31) (1.51) (1.17) (1.26)
Observations 28 28 28
R-squared 0.21 0.31 0.29 0.38
100
Table 3.16: OLS Estimation of Ownership and Legal Origin
The dependent variable is the percentage of the publicly listed firms in each country that are
widely held. The Basic equation includes the Log of Checks, Polarization, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Each regression includes
dummy variables for French, German, Scandinavian, and Socialist legal origins. Explanatory
variables are averaged over the years 1996-2000. Absolute value of robust t-statistics are in
parenthesis. * significant at 10%; ** significant at 5%; *** significant at 1%
Widely-held10 Widely-held20
1 2 3 4
Lchecks -2.816 0.162 -4.273 -0.080
(0.12) (0.01) (0.19) (0.00)
Polarization 1.024 -1.042 -4.419 -6.656
(0.09) (0.11) (0.37) (0.57)
Democracy 76.613 2.250 -21.993 -2.911
(1.14) (0.13) (1.41) (0.13)
Trade Openness -0.182* -0.178
(1.85) (1.56)
FDI 0.70 -0.073
(0.12) (0.10)
Lgdp95 6.885 10.452 10.486 14.115
(0.89) (1.41) (1.15) (1.53)
Legor_fr -23.057 -26.854 -22.144 -25.307
(1.56) (1.65) (1.43) (1.52)
Legor_ge -9.666 -15.588 -2.499 -8.606
(0.53) (0.84) (0.12) (0.43)
Legor_sc -38.691** 42.617** -21.637 -25.260
(2.33) (2.56) (1.27) (1.52)
Constant -8.444 -48.152 -20.735 -62.556
(0.13) (0.66) (0.25) (0.60)
Observations 28 28 28 28
R-squared 0.48 0.53 0.40 0.49
101
Table 3.17: IV Estimation of Ownership
The dependent variable is the percentage of the publicly listed firms in each country that are
widely held. The Basic equation includes the Log of Checks, Polarization, and Lgdp95.
Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Explanatory variables
are averaged over the years 1996-2000. Absolute value of robust t-statistics are in
parenthesis.*significant at 10%; **significant at 5%; ***significant at 1%
Panel A: First Stage Regressions of Lgdp95
Widely-held10 Widely-held20
1 2 3 4
Lchecks -0.088 2.445** -0.088 2.445**
(0.09) (2.44) (0.09) (2.44)
Polarization 0.308 -0.203 0.308 -0.203
(0.85) (0.44) (0.85) (0.44)
Democracy -0.582 0.784 -0.582 0.784
(1.03) (0.84) (1.03) (0.84)
Lsettler/Disteq -0.697* 1.294 -0.697* 1.294
(2.02) (0.43) (2.02) (0.43)
Constant 12.22*** 5.440** 12.22*** 5.440**
(5.45) (2.69) (5.45) (2.69)
Observations 11 11 11 11
R-squared 0.70 0.52 0.70 0.52
Panel B: Second Stage Regressions of the Basic Equation
Widely-held10 Widely-held20
1 2 3 4
Lgdp95 -4.979 24.764 5.13 21.715
(0.23) (0.32) (0.28) (0.30)
Lchecks 44.393 -33.891 37.589 -26.314
(0.89) (0.16) (1.03) (0.13)
Polarization 7.724 -1.598 8.802 -9.671
(0.44) (0.07) (0.78) (0.49)
Democracy -6.009 -20.508 -0.633 -29.740
(0.19) (0.5) (0.02) (0.56)
Constant 31.800 -144.603 -53.14 -98.926
(0.16) (0.33) (0.27) (0.24)
Observations 11 12 11 12
R-squared 0.53 1.30 0.67 0.43
102
Chapter 4: Securities Laws, Firm Size, and Capital Issuance
1. Introduction
The dictum that institutions matter no longer surprises the vast majority of
economists. Institutional disparities in country organization go a long way in explaining
differences in firm behavior. This chapter seeks to add to the literature investigating the
role of institutions in firm level outcomes by examining the effects of securities laws on
capital issuance, the means by which firm finance their growth.
Several studies have found a role for legal efficiency in the ability of firms to
finance their growth from external sources. Rajan and Zingales (1998) pursue the
relationship between financial development and firm size. The authors find that
“industrial sectors that are relatively more in need of external finance develop
disproportionately faster in countries with more developed financial markets”. Kumar,
Rajan and Zingales (2001) investigate the linkage between firm size and judicial
efficiency. The empirical study mingles institutional and technological explanations for
the determinants of firm size in fifteen European countries. The researchers show that (1)
countries with more efficient judicial systems have larger firms (2) larger firms tend to
operate in capital-intensive industries; however, (3) capital intensive firms in efficient
judicial systems tend to be smaller.
Laeven and Woodruff (2004) focus on Mexican firms, which share the same
national commercial laws, however, are exposed to legal environments with varying
degrees of efficiency and enforcement at the state level. The authors’ data set allows
them to examine firms that are both large (corporations) and small (single person
proprietorships). States with better legal enforcement are found to have larger firms.
103
Laeven and Woodruff hypothesize that “better legal systems increase the investment of
firm owners by reducing the idiosyncratic risk they face.” Reduction of risk is associated
with the benefits of financial development – minority shareholder protection, better
informational exchange between management and investors and improved contracting
among firms and their suppliers. These benefits lead to greater capital accumulation,
higher levels of external finance, and growth in firm size.
Finally, Beck, Demirguc-Kunt, and Levine (2002) explore judicial independence
from the central government and the ability of the legal system to adapt to changing
circumstances. Their paper improves on previous research by incorporating the World
Business Survey (WBES). This data source covers 4000 small, medium and large firms
in 38 countries. The firms are asked about the financing obstacles in general, in meeting
collateral requirements, bureaucratic paperwork and acquiring long-term loans. The
authors show that firms in common law countries are much more able to garner external
finance than those in civil law countries. A limitation of their study is that their results
rely on “perceptions of financing obstacles” from the firm survey as opposed to actual
restrictions. These perceptions may be caused by cultural or environmental factors that
are independent of legal attributes.
Our paper contributes to the literature in legal institutions and firm size by
investigating the role of security laws on the ability of firms to raise external finance by
issuing capital. The paper takes advantage of two unique data sets in order to accomplish
this task. The first data set is compiled by La Porta, Lopez de Silanes, and Shleifer
(2004) and quantifies the different laws regulating the issuance of new equity in 49
countries. We refer to the authors as LLS through out the chapter. Securities regulations
104
are divided into laws governing public and private enforcement. Private security laws aid
in the ability of private agents to contract among themselves, while public security laws
regulate the organization of a public enforcers analogous to the Securities and Exchange
Commission in the US. LLS conclude that while securities laws that enhance private
enforcement of contracts improve stock market development, laws governing public
enforcement have no effect on development in general, or the ability of small firms to
raise external finance in particular.
The second data set features firm accounting information and primary issues and
is compiled by Knill (2004). It includes small, medium, and large firms of 54 countries
from the years 1996-2003. This data set allows us to analyze access to finance at the firm
level and to examine the role of securities legislation on capital issuances of publicly
listed firms of all sizes.
Our first objective is to investigate if a country’s securities laws are a significant
determinant of the probability that the firm issues a security. Second, we asses if
securities laws have asymmetric influence on the issuance of small and large firms and
we determine if the asymmetry is found in both industrialized countries and emerging
markets. Finally, we attempt to confirm at the firm level the conclusion of LLS that only
private enforcement matters in increasing access to capital. We find that the composition
of laws governing security issuance in each country significantly affects the probability
that a firm will issue equity. Furthermore, in contrast to LLS, we show that public
enforcement of security laws is a much more important determinant of firm access to
capital markets than private enforcement. It is our conclusion that aggregate
macroeconomic indicators of stock market size used in the LLS paper belie the
105
importance of a strong regulator in increasing access to capital, particularly for emerging
market firms. Section 2 discusses our economic approach to estimating access to finance.
Section 3 describes the data. Section 4 presents the empirical testing strategy and Section
5 provides the results. Section 6 provides a robustness check of our results and Section 7
concludes.
2. Methodology
A challenge to research examining access to external finance is that firm level
analysis is limited to a few large publicly listed firms per country. Our data set allows us
to investigate the impact of security laws on a wide representation of listed firms and
countries. We investigate whether the probability of capital issuance is dependent on a
country’s security laws, firm characteristics, and macroeconomic factors.
We begin by calculating the need for external funds for firms in our sample in
period t. We believe that firms that need external funds to make investments and do not
issue securities are more financially constrained than the other firms in our data set.
Higgens (1977) presents a financial planning model that calculates a firm’s need for
external funds in period t. This equation relates the firm growth rate to its need for
external funds and derives the external funds necessary from the “percentage of sales”.
Demirguc-Kunt and Maksimovic (2002) also use this financial planning model to identify
firms that require external finance to meet their investment needs. The ‘external funds
necessary’ (or EFN) for firm i in period t is calculated as follows:
) )( ( ) )( / ( ) )( / (
, , , 1 , , , , 1 , , 1 , 1 , , t i t i t i t i t i t i t i t i t i t i t i t i
RR S M S S S L S S S A EFN ? ? ? ? =
? ? ? ?
(1a)
where A
t-1
is the total assets of the firm in time t-1, S
t-1
and S
t
are the sales of the firm in
times t-1 and t respectively, L
t
is the liabilities of the firm in time t, M
t
is the profit
106
margin of the firm as defined by net income divided by sales for time t, RR is the
retention ratio for the firm. As noted by Demirguc-Kunt and Maksimovic (2002) two
simplifying assumptions are made in order for this methodology to be implemented.
First, both the asset utilization (A/S) and the profit margin of the firm must remain
constant per unit of sale. Second, the use of the formula to discern additional funds
necessary depends on true values of assets being reported (relative to their depreciable
basis).
Using firm-specific information to measure S
t
is problematic because Sales in
period t could be determined by weak securities laws that reduce access to external
finance, which, is the very situation that we are testing. Therefore, we approximate
industry wide rates of desired growth in Sales from t-1 to t using US firm level data. We
assume that the US industry growth rates are a good proxy for desired sales growth in
other countries using the logic of Rajan and Zingales (1998). The authors justify the
external dependence of US firms as a proxy for optimal demand for external funds
because the US has the most advanced capital markets in the world. Thus, US firms are
operating in as close to a frictionless financial environment as there is in the world.
Furthermore, Rajan and Zingales reason that industry wide measures of external
dependence are reasonable by stating, “Therefore, much of the demand for external funds
is likely to arise as a result of technological shocks that raise an industry’s investment
opportunities beyond what external funds can support. To the extent these shocks are
worldwide; the need for funds of the US firms is a good proxy.” We assume that since
US firms can access external finance optimally, they are able to increase their sales at an
optimal rate as well. Moreover, optimal rates of growth persist across industries in other
107
countries. For each US firm in our data set, we measure growth in Sales from year t-1 to
year t. Next, we average individual firm growth rates in the same industry and size
classification. The term
US
m j
g
,
is the average growth rate in Sales for firms in industry j of
size m. We measure separate growth rates for each size classification m, where m is
either small or large, because small and large firms in the same industry can grow at
drastically different rates.
21
We use
US
m j
g
,
to approximate desired growth in Sales for a firm outside the US in
industry j and size m. The predicted value of Desired Sales for firm i is equal to
) (
1 , ,
,
*
?
=
t i
US
m j
t i S g S (2)
We substitute t i S ,
*
for actual Sales in period t into equation (1a). Equation (1b) measures
the external funds necessary to increase Sales in year t-1 to Desired Sales in year t.
) )( ( ) )( / ( ) )( / (
,
*
, , 1 ,
*
,
*
, , 1 ,
*
, 1 , 1 ,
*
, t i t i t i t i t i t i t i t i t i t i t i t i
RR S M S S S L S S S A EFN ? ? ? ? =
? ? ? ?
(1b)
We drop all US firms from the sample to avoid bias in our empirical model.
Our econometric model is based on the findings of Korajczyk and Levy who
examine capital structure choice in the presence of financing constraints for US firms
(2003). They find that a firm’s choice of security issuance (either debt or equity) is based
on macroeconomic conditions and firm specific information. We use the independent
variables found in the paper to test the effect of securities laws on access to external
funds. However, our econometric approach differs slightly from Korajczyk and Levy
(2003) because our dependent variable lumps the decision to issue debt or equity into one
action. Thus, we are not modeling capital structure choice and can focus our estimation
21
Size is not a concern for Rajan and Zingales (1998) because they only have data for large
publicly traded firms.
108
entirely on financing constraints that prevent a firm in the panel from acquiring external
finance when it wants to.
We use a probit model to analyze the event that a firm i issues a security in time t.
Y=1 if the firm issues a security, and equals 0 if the firm does not issue.
The model we want to estimate is then
) (
) 1 ( Pr
6 5 4 , 3 1 , 2 1 , 1
,
t j k t i t k t i
t i
T I SECURITIES EFN Z F
Y ob
? ? ? ? ? ? ? + + + + + + ?
= =
? ?
(3)
? represents the standard normal distribution. The constant term is defined by ? .
1 , ? t i
F
is a vector of lagged firm-specific variables such as cash flow, debt/asset level, size,
profitability, risk, uniqueness of assets, trade credit and asset tangibility.
1 , ? t k
Z is a
vector of lagged macroeconomic variables such as GDP, corruption, shareholder rights,
the efficiency of the judiciary and the availability of domestic credit and foreign capital.
*
,t i
EFN , as described above, measures the external funds needed to increase Sales in year
t-1 to Desired Sales in year t. SECURITIES is the primary variable of interest and
represents the composition of securities laws in country k.
j
I represents the industry
indicators and
t
T is time dummies. Equation (3) allows us to investigate if the
composition of securities laws in Country k increases the probability that a firm will issue
a security when it requires external funds to finance its investments.
In addition, to estimating Equation (3) for the entire sample of non-US firms in
the data, we also estimate the equation separately for the economically advanced G 10
countries and emerging markets. We divide the data set into these two sub samples in
order to investigate whether differences in institutional quality of the legal system
109
between G10 and emerging market countries affect the impact of securities laws on
capital issuance.
3. Data
Knill (2004) collects observations for common stock, non-convertible debt,
convertible debt, non-convertible preferred stock and convertible preferred stock issued
domestically. The data covers all domestic issues of securities around the world from the
time period 1/1/1996 through 3/31/2003 and is collected from the SDC Global New
issues database. Table 4.1 shows the amount of each type of security issued by country.
Financials for the companies issuing domestically in a given year are from REUTERS.
This data set enables us to have a much richer sample of global new issues around the
world of both smaller and larger firms than afforded by SDC Platinum alone. REUTERS
provides financial information on all publicly traded firms for the majority of countries in
the world and does not suffer from the bias toward large firms to the extent that other
international databases such as Worldscope/Datastream/Research Insight do. In fact,
REUTERS even covers pink sheets and OTC/Bulletin Board firms whereas the others do
not. As such, the coverage is much more comprehensive and the average firm size is
much smaller.
In order to address sample selection bias from a data set composed entirely of
firms that issue domestic capital in a given year, Knill also collects data on firms not
issuing capital during this period of time to represent those public companies that either
cannot issue capital or are internally ‘wealthy’ to the point where there are financially
unconstrained. For emerging market country firm-year observations of the non-issuing
firms, financials are collected for 1996-2003 for the most exhaustive list of firms for each
110
country as possible from REUTERS, collecting the exact same data utilized for the issuer
dataset. Developed country firm-year observations are collected from Worldscope, due
to the inability of REUTERS to provide such large amounts of data given the
practitioner-oriented setup of this information-rich database. This is not believed to
cause bias due to the careful matching of accounting information and the quality of
information that is provided for the countries. Therefore, our data set includes security
issuance (or non-issuance) and financial accounting information for all listed firms
available in the REUTERS and Worldscope databases.
Following Korajczyk and Levy (2003), financial services are excluded due to the
special circumstances of their asset base and utility firms (Macro Industry: Financial
Services, Real Estate and Energy and Power) due to the abnormal stability and
predictability of cash flow. Knill also excludes those firms that have gone bankrupt due
to the special set of issues that are included in capital structure determination when a
company is failing
22
. This follows the methodology of Asquith et al. (1994) who find
that such situations generally cause a major restructuring of capital structure outside of
the scope of financial constraint relaxation. Lastly, Initial Public Offerings (IPO)s are
excluded. Welch (2002) finds that the firms who undertake IPOs find themselves in
unique environment, similar to those of the other deletions, which hosts a different set of
issues than the post-IPO period. Including these firms would bias the results.
The only firms not covered in REUTERS are those that have gone bankrupt or
have merged with another firm. The first group has deliberately been excluded from the
sample as mentioned above. The second group would only be a problem if the issuing
22
Firms going bankrupt would have additional difficulty obtaining capital, which would
convolute results.
111
company had acquired a firm in the sense that the capital structure may have changed
thus changing the financial environment examined. A cross-examination of the Global
New Issues database with the Mergers & Acquisitions database provides information on
whether any of the firms in the sample have been involved in a merger or acquisition.
Our dependent variable is Capital issuance. Capital Issuance is a dummy
variable, which equals 1 if firm i issued some type of capital (common stock, non-
convertible debt, convertible debt, non-convertible preferred stock and convertible
preferred stock) in time t and 0 otherwise. We explain the probability of capital issuance
by including as independent variables indexes of securities laws, firm specific variables
and macroeconomic indicators. We discuss each type of explanatory variable below. A
detailed list of variables and definitions is given in Appendix D.
Securities laws fall into two broad categories, public enforcement and private
enforcement. Private Enforcement is calculated in the La Porta et al. (2004) data set by
combining indexes of disclosure and liability. The disclosure index covers five distinct
areas (1) insiders compensation; (2) ownership by large shareholders; (3) inside
ownership; (4) contracts outside the normal course of business; and (5) transactions with
related parties. The liability index is the average of four levels of accountability in the
issuance of stock. These levels of accountability summarizes the burden proof for the
issuer of the stock (bdn_iss), the company directory (bdn_dir), the distributor (bdn_dis),
and the accountant (bdn_acc).
Public Enforcement regulates the supervisory behavior of the main regulatory
government authority in charge of the stock market. Public Enforcement is calculated by
the combining the following four sub indexes: (1) supervisory index; (2) investigative
112
powers index; (3) non criminal sanctions index (orders), and (4) criminal sanctions
index. The supervisor index is assessed on its independence from the central government
and that all hirings and firings are given due process (appointment), whether the
supervisor only regulates stock markets and not banks too (focus), and if the supervisor
has the authority to regulate primary offerings and listings on the stock market
(regulatory powers). The investigative index assesses whether the supervisor can
subpoena documents and witnesses. The orders index involves non criminal sanctions
for violations of disclosure standards, such as compensating investors for losses, or
instituting recommendations of the supervisor. The orders can be given to the issuer
(ord_iss), distributor (ord_dis) or accountant (ord_acc). Finally, the criminal index is a
measure the supervisor’s ability to impose criminal sanctions against the director
(crim_dir), distributor (crim_dis), issuer (crim_iss) and accountant (crim_acc).
Along with securities laws, access to capital is also explained by firm specific
variables. REUTERS obtains firm financial statistics for listed firms from country stock
exchanges. Variable definitions vary from country to country and are measured in
different currencies. To make variables comparable across countries, firm variables are
scaled by total assets unless otherwise noted.
As many empiricists have attributed size as a determinant of capital structure, we
assign size categories based on Total Assets. Korajczyk and Levy (2003) and Baker and
Wurgler (2002) find a positive relationship between leverage and size. Titman and
Wessels (1988) find that size influences not only the extent of leverage but also the type.
113
Our proxy for this follows both Titman and Wessels (1988) and Rajan and Zingales
(1995) and is calculated as total assets/GDP
23
.
The ratio of Cash to Total Assets describes the feasibility of using internal funds
to finance growth. According to the pecking order of capital structure choice (Myers,
1984) firms prefer to finance investments from internal funds, then bank loans, and
finally from capital issuance. We expect that the availability of internal funds will
decrease the likelihood that a firm would issue a security in a given year.
Uniqueness of assets is included based on the theory that a high uniqueness
increases the expected costs of bankruptcy. As assets of a company become more
distinctive, the ability to sell those assets when necessary (i.e. the liquidity of those
assets) decreases thus increasing liquidity risk. Within-country industry averages are
used in those cases where there is missing data. This should not be problematic due to
the uniformity of the nature of assets in firms within the same industry.
Profitability of firms would be an obvious influence on firms inasmuch as this
impacts how well a firm can either pay interest and/or dividends. Titman and Wessels
(1988) provide two measurements for this variable that are applicable universally. They
are operating income divided by sales and operating income divided by total assets. We
only provide results for profitability based on sales for brevity. We also include standard
deviation of the firm’s profitability ratio over the three years prior to issue (Risk) as an
independent variable. Firms with riskier profitability are less likely to issue capital.
Also relevant to capital structure determination is Asset Tangibility. This refers to
how palpable the assets of a firm are and relates to capital structure concerns through its
23
This is done annually so that firms may switch size groupings over years. The analyses are also done
using average size of the eight year periods. As results are unchanged, they are omitted for brevity.
114
limitations on debt levels due to the ability to provide collateral. A firm has less
collateral the less tangible its assets are. This, arguably, could be said to increase the
probability of bankruptcy due to the inability to obtain funds when there are especially
needed. This follows logically from the fact that a company without material assets
would not be able to liquidate to obtain the necessary funds to pay off debtors if it were
necessary. This variable is created by calculating fixed assets divide by book value of
assets (following Rajan and Zingales 1995). Once again, within-country industry
averages are used in those cases where there is missing data. For the same reasons given
above justifying the rationale for industry average substitution as proxies for uniqueness
of assets, industry averages are suitable proxies here.
To correct for any additional access a firm might have in other nations which
might affect financial constraints (Lins et al., 1999) it is vital to include an indication of
whether a firm has listings in other countries (i.e. ADR on a U.S. stock exchange). We
include a dummy variable for Cross listing that takes on a value of 1 if a firm is listed on
an exchange outside of its nation and 0 otherwise. Finally, differences in industry
classification are avoided by using as industry indicator the SDC Platinum Macro
industry code as our categorization. The industry dummy is included to account for any
industry fixed effects.
Given the fact that there are over 20,000 firms in our sample, it is not surprising
that the range of firm-level statistics such as cash, uniqueness, profitability and risk span
a range that is considerable in size. Smaller firms seem to have much more leverage than
their larger peers (Rajan and Zingales, 1995). Profitability and risk for the smaller firms
are considerably larger, reflecting the higher growth rate of the smaller firms (and based
115
on the fact that the figure is scaled by assets, controlling for size). The majority of the
sample is not cross listed. Table 4.2 summarizes the firm specific variables and
demonstrates the incredible range of variation in the financial accounting data.
Based on results from such papers as Korajcyk and Levy (2003) and Booth,
Demirguc-Kunt, and Maksimovic (1999), we include macroeconomic factors to capture
their impact on capital structures in different countries. All of our country-level control
variables are averaged over the years t-1 through t-3. Unless otherwise stated,
macroeconomic indicators are obtained via the International Finance Corporation (IFC).
GDP growth is the percentage growth in gross domestic product per capita and is
included to control for business cycle effects on issuance. Securities issuance tends to
occur counter cyclically because in good times firms have enough cash on hand to fund
investments internally, while in bad times they are more likely to raise money for
external capital markets.
We also control for the availability of domestic and foreign capital in each
country in order to isolate the role of securities laws on access to capital from both the
size of a country’s domestic financial markets and the openness of the country to foreign
funds. Domestic capital is the sum of market capital and private domestic credit, scaled
by GDP. Market Cap is equal to the Listed shares * their value and domestic credit is
credit extended to banks and other financial intermediaries scaled. Foreign capital is all
foreign investment – Foreign direct investment, Foreign portfolio investment, and foreign
bank lending and official flows – scaled by GDP.
Additionally, LLS note that better law enforcement could be associated with
increased access to capital regardless of the content of the securities laws. To focus the
116
regression analysis on the relationship between the composition of securities legislation
and capital issuance we control for the quality of the country’s legal institutions. Judicial
Efficiency controls for the efficiency of the legal system in executing the laws.
Corruption controls for the degree to which securities laws are fairly executed. Higher
values of this variable is associated with less corruption. Finally, Anti Director Rights
controls for investor protection given by corporate laws at the firm level as opposed to
national securities laws.
Table 4.3 summarizes the macroeconomic variables. There is considerable
variation in all the macroeconomic controls, especially GDP growth and the availability
of domestic and foreign capital. Table 4.4 presents the correlations between capital
issuance and the firm characteristics. Profitability, Uniqueness, Asset Tangibility, and
Cash on hand are all negatively correlated with our capital issuance indicator. Table 4.5
displays the pair wise correlations between the dependent variable, capital issuance
indicator, and the macro characteristics. The variables governing the legal environment,
Anti Director Rights, Judicial Efficiency, and the Private Enforcement index, are all
negatively correlated with issuance. Only Public Enforcement manifests a positive
correlation with securities issuance, foreshadowing our regression results.
4. Microeconomic Testing Strategy of LLS (2004)
Our objective is to test the accuracy of LLS conclusions at the micro level.
The authors present three different theories of optimal public policy in the regulation of
the private transactions between the issuers of securities (firms) and investors. The first
theory states that law simply codifies existing market arrangements and therefore serves
no purpose in improving securities transactions. The second theory asserts that the
117
codification of arrangements covering disclosure and liability aid in the efficiency of
private transactions by clarifying liability rules and standardizing security contracts.
Both of these theories assert that public enforcement of securities laws by the government
is at best unnecessary and at worse harmful to investors and firms. The last theory
presented by LLS states that government regulation of securities markets corrects market
failure due to information asymmetry by imposing sanctions and securing information by
subpoena. The authors find that law matters in supporting stock market development,
however only security laws enforcing private transactions are important. They argue that
public enforcement of securities laws are irrelevant and have no effect on macro
indicators of stock market development.
In contrast to LLS, we find that the firm level results are more in line with the
theory that government regulation is needed to support trade and to reduce the transaction
cost of issuance. We show that public enforcement of securities is especially important
in improving access to capital for emerging market firms.
We examine the LLS conclusion by estimating the probit model defined in
Equation 3. We regress the probability of capital issuance for each firm-year observation
on Public Enforcement, Private Enforcement, firm characteristics and macro controls.
We use the STATA cluster command in each regression by firm. This command helps us
avoid spurious results caused by omitted variable bias at the firm level. For each
regression, we calculate the marginal coefficients so that each coefficient represents the
increase in probability of capital issuance explained by the independent variable.
Our large sample size allows us to analyze the data in the two different
dimensions of firm size and institutional quality. We show that securities laws have
118
disparate effects on small and large firms and for firms in developed as opposed to
emerging economies. We test the null hypothesis that the coefficients for small firms and
large firms are equal in the whole sample of countries and the G10 and emerging markets
sub samples by computing the likelihood ratio statistic
( )
2 1
ln ln ln 2
U U R
L L L LR ? ? ? = (4)
where ln L
R
is the restricted log-likelihood function of the entire sample of firm year
observations, and ln L
U1
and ln L
U2
are the unrestricted log-likelihood functions for small
firms and large firms, respectively. The likelihood ratio is distributed chi-squared with
the degrees of freedom equal to the number of restrictions. We calculate the test statistic
for each of the three sub samples – all firm year observations, G10 firms, and emerging
market firms.
5. Results
Before looking at the individual regressions, we can make some general
comments about how the firm specific variables affect capital issuance. Cash on hand,
Uniqueness of Assets, and Profit Risk are consistently negative determinants of issuance
over the different sub samples. This is an expected result since the availability of internal
funds reduces the desire for external finance while uniqueness and risk have also been
shown to decrease issuance. Additionally, we find that External Finance Need (EFN) is a
more important determinant of capital issuance for the small firms in our sample than for
large firms. Large firms have more access to capital and can issue either to garner
investment capital or to optimize their existing capital structure. Small firms have less
access and tend only to go to the capital markets when they need funds.
119
There are also differences in the effects of the firm-specific variables between
firm located in a G10 or emerging market country. For instance, Profitability is a
positive variable for large emerging market firms. Profitability determines the ability of
the firm to pay dividends an interest and in countries with weak investor protection, profit
is an important signal of the ability (if not willingness) to pay. Cross listing is only
positive for large G10 firms. There may be benefits to small and EM firms from issuing
capital in foreign as opposed to domestic markets as a signal of financial well-being,
which reduces their domestic issues.
We also note some general trends in the data in regards to the macroeconomic
variables. In almost all of the regressions, the coefficients enter significantly at the 1%
level. Growth in GDP is a negative determinant of capital issuance in G10 countries and
positive in the emerging markets. For the G10 countries, this result is consistent with the
tendency of capital issuance to be counter cyclical. Since GDP growth controls for the
business cycle in good times (GDP growth is high) firms have enough internal funds to
finance investment and may not need to access capital markets. Additionally, the
availability of Domestic Capital often enters the regression equation as a negative
component of issuance. Since this variable is largely composed bank loans, Domestic
Capital maybe proxying for alternative sources of capital other than the securities
markets. Foreign Capital flows are positively related to capital issuance in emerging
market firms and negatively related to issuance for G10 firms. This result is independent
of firm size.
Corruption, Anti Director Rights, and an Efficient Judiciary control for the
fairness and effectiveness of investor protection. The probability of issuance is higher in
120
less corrupt countries. Except for large firms in the G10, Anti Director Rights have a
negative effect on issuance and an Efficient Judiciary has a positive effect. Interestingly,
for the large G10 firms Anti Director Rights are positively related to issuance and an
Efficient Judiciary has a negative relationship.
We use the two aggregate securities law indexes, Public and Private Enforcement,
to summarize effects of securities laws on capital issuance. Recall that the Public
Enforcement index aggregates supervisor independence, investigative powers, orders and
criminal sanction. The Private Enforcement index is composed of disclosure and burden
of proof. Table 4.6A displays the effects of Public and Private Enforcement for all firms
in our data set. The first regression includes Public Enforcement as the main explanatory
variable of security laws in each country. The second regression contains Private
Enforcement only. Finally, the third regression displays a horserace between Public and
Private Enforcement.
Regression 1 generates our primary result; Public Enforcement has a large
positive and significant role in capital issuance. The variable’s coefficient is 0.289 and is
significant at the 1% level. Regression 1 shows that capital issuance is pro cyclical over
the whole sample of firm year observations as GDP_growth is positive and significant.
The availability of domestic capital is a negative determinant of capital issuance and is
significant at the 1% level. Corruption plays a negligible role in issuance however the
other two measures of the fairness and efficiency of the legal system, Anti Director
Rights and Efficient Judiciary are significant components of capital issuance. Protection
of shareholders by corporate laws (anti director rights) is negative, while an efficient
judiciary makes a positive contribution.
121
Regression 2 examines the role of Private Enforcement on our dependent
variable. We find that enforcement of private transactions plays a negligible role in
capital issuance. In the final regression in Table 4.8A, we include both Public and
Private Enforcement in the regression in order to discern the independent contribution of
each variable on capital issuance. The coefficient on Public Enforcement remains
unchanged from Regression 1. It is positive and significant at the 1% level and has a
coefficient of 0.289. Private Enforcement is statistically insignificant.
In the next two tables, Tables 4.7B and 4.7C, we investigate whether securities
laws have disparate effect on small and large firms. A firm is categorized as small (large)
if its Total Assets are below (above) the median for all firms in its country of origin. We
find that for small firms presented in Table 4.6B, Public Enforcement has a larger impact
on capital issuance than that of the whole sample. The coefficient on Public Enforcement
is 0.445 (compared to 0.289 in Table 4.6A). Regression 2 reveals that Private
Enforcement of securities laws is a significant and negative determinant of capital
issuance for small firms. This result is repeated in Regression 3 when both indexes of
enforcement are included. The coefficient on Private Enforcement is -0.301 and is
significant at the 1% level. Public Enforcement on the other hand has a positive
coefficient of 0.453 and is also significant at the 1% level.
The dominance of Public over Private Enforcement does not hold for the large
firms in the sample. In Regression 1 of Table 4.6C, Public Enforcement is shown to have
an insignificant effect on issuance. Regression 2 shows that Private Enforcement is
positive for large firms and significant at the 1% level with a coefficient of 0.198. The
122
horse race between both securities indexes presented in Regression 3 shows that both
variables makes a significant contribution to capital issuance for the large firms overall.
We wrap up the analysis of Tables 4.7A-C by testing whether the differences in
small and large firms are statistically significant. The likelihood ratio test statistic for the
equality of coefficient of small and large firms in Regression 1 of Tables 4.7A-C is given
by Equation 4.
( )
10 . 4150
) 69 . 15060 02 . 6059 76 . 23194 ( 2
ln ln ln 2
2 1
=
+ + ? ? =
? ? ? =
U U R
L L L LR
The test statistic is distributed chi-square with 56 degrees of freedom. At a 1% level of
significance the critical value from of the chi-square distribution is 83.52. The null
hypothesis of equality of coefficients between small and large firms is resoundingly
rejected. We calculate the likelihood ratio statistic for Regressions 2 and 3 at 3333.82
and 2208.68 respectively. The critical values of chi-square at a 1% level of significance
for Regressions 2 and 3 are 83.52 and 85.95, respectively. These test statistics support
our finding that securities laws have disparate effects on small and large firms. Public
Enforcement increases issuance in small firms and Private Enforcement decreases the
probability of issuance. In contrast, both enforcement indexes make a positive
contribution to the probability of large firm issuance overall.
Next, we isolate G10 firms in our sample. The sub sample is composed of all the
countries in the G10 except for the US, which was dropped from the sample to prevent
biasness in the results as described in the methodology section. Overall G10 countries
have better institutional quality than the other countries in the data set, allowing us to
investigate the role of security laws in a good institutional environment. We divide firms
123
into small and large groups as before. In Table 4.7A, we find that Public Enforcement
has a smaller effect on issuance for small G10 firms than for the whole sample. Its
coefficient is -0.190 compared to 0.289 for the entire sample. Regression 2 manifests a
significant negative relationship between Private Enforcement and capital issuance. In
Regression 3, the results indicate that when both indexes are included in the regression
Public Enforcement is a positive factor in the issuance of small G10 firms while Private
Enforcement is a significant deterrent to issuance.
Table 4.6B demonstrates the difference between small and large G10 firms’
response to security laws. In Regression 1, Public Enforcement has a coefficient of
-0.401 and is significant at the 1% level. Private Enforcement is also shown to have a
negative effect on issuance in Regression 2. We also observe in Regression 3 that laws
regulating both private and public enforcement have negative impacts on large firm
issuance. The coefficients on Public and Private Enforcement are -0.422 and -0.340,
respectively. Calculations of the likelihood ratio statistics for all three regressions in
Tables 4.8A & B reject equality of coefficients between small and large G10 firms at the
1% level. The likelihood ratio statistics are 2740.96, 2610.24, and 2736.66 for
Regressions 1, 2 and 3, respectively. The critical values of the test from the chi-square
distribution are 83.52 for Regression 1 and 2 and 85.95 for Regression 3.
Finally, we investigate the role of securities indexes in firms whose country origin
is not in the G10. We define those countries as emerging markets. Table 4.8A displays
the results for all firm year observations in emerging market countries. For small
emerging market firms Public Enforcement is positive and significant at 0.459 as shown
in Regression 1 of Table 4.8A. Private Enforcement enters Regression 2 negatively and
124
significantly at the 5% level. Regression 3 displays the result that the largest impact of
the securities indexes on capital issuance is for small emerging market firms. The
coefficients are 1.001 and -1.054 for Public and Private Enforcement, respectively. Both
variables are significant at the 1% level.
Table 4.8B shows that large emerging market firms behave similarly to small
emerging market firms in their response to the public regulation of securities laws.
However, in contrast to the other sub samples of our data set, Private Enforcement is
positive and significant in Regression 2. When both security laws indexes are included in
Regression 3, only Public Enforcement retains a positive relationship with issuance.
Securities laws have a similar effect on small and large emerging market firms.
However, likelihood ratio tests strongly reject the null hypotheses of equality of
coefficients between all the explanatory variables. The statistics for Regressions 1-3 in
Tables 4.8A and 4.8B are 742.86, 750.64, and 811.08, respectively. The critical values of
the test from the chi-square distribution are 83.52 for Regression 1 and 2 and 85.95 for
Regression 3.
In the last series of regressions, we estimate the impact of each separate sub
index of Public and Private Enforcement in six individual regression equations for each
of the three sub samples. We report the results of these regressions for the securities laws
indexes only in Appendix E. Over the whole sample of firms supervisor characteristics,
and the supervisor’s investigative powers are positive determinants of capital issuance.
However, the significance and sign of the security indexes depend on the sub sample
examined. For instance, the supervisor characteristics, investigative power, criminal and
non-criminal sanctions significantly decrease the probability of issuance for large G10
125
firms. In contrast, all firms in emerging markets increase their securities issuance when
there is a strong supervisor with investigative powers and the authority to impose
sanctions.
6. Robustness Check
A reason for skepticism of the results we present may be found in the large z-
statistics given by the enforcement variables. For example, in the all-firm regressions
presented in Table 4.6A, in Regression 3, Public Enforcement has a z-statistic equal to
21.01. A reader may question the plausibility of these results given the cross-country
nature of the panel. A useful robustness check is to cluster the standard errors by country
in order to account for the cross-country nature of the regression analysis. Table 4.9A
repeats the analysis for all firms in Table 4.6A with the standard errors clustered at the
country level, and Table 4.9B presents these results for the regressions in which both
Public and Private Enforcement are included for each firm category. As Table 4.9A
demonstrates Public Enforcement remains significant at the 1% level in Regressions 1
and 3, while Private Enforcement remains insignificant in Regressions 2 and 3. Though
clustering the standard errors by country leads to a sizable reduction in the z-statistics for
most of the variables, the significant relationship between public enforcement of
securities laws and capital issuance remains unchanged.
Table 4.9 suggests that overall, small firms increase issuance when securities laws
are publicly enforced. There is not significant relationship between securities law
enforcement and capital issuance for large firms overall when standard errors are
clustered by country. Additionally, the significance of Public Enforcement for small G10
firms disappears, however this analysis supports our original conclusion that these firms
126
are harmed by private enforcement of securities laws. Also supported by the robustness
check, is that Public Enforcement is a negative and significant at the 1% level for large
G10 firms and is positive and significant at the 1% level for small and large Emerging
Market firms.
7. Conclusions
The main conclusion of LLS is that private enforcement of securities regulation
increases the level of stock market development while public enforcement has a
negligible effect on stock markets.
Public enforcement plays, at best, a modest role in the development of stock
markets. Specifically, there is no evidence that such factors as Supervisor’s
independence or focus work. Both the Supervisor’s investigative powers and the
strength of both criminal and non-criminal sanctions only matter for a narrow set
of outcomes. In contrast, the development of stock markets is strongly associated
with measures of private enforcement such as extensive disclosure requirements
and a relatively low burden of proof on investors seeking to recover damages
resulting from omissions of material information from the prospectus.
Our paper questions this conclusion by examining the impact of securities
regulation at the firm level. Our data set allows us to investigate effect of securities
regulation on listed small and large firms and in developing and industrialized countries.
We find that securities laws have disparate effects on capital issuance between small and
large firms in G10 and emerging market countries. Private Enforcement of securities
laws is found to be a deterrent to capital issuance for small firms and firms in emerging
127
markets. Public Enforcement significantly increases the probability of issuance for all
emerging market firms.
Our results suggest that private enforcement of securities laws is not the panacea
to stock market development described in LLS. Instead, our results encourage the
establishment of a strong regulatory authority in countries with weak institutions, while
LLS support strengthening laws regulating private transactions. Both courses of actions
have dramatically different effects on capital issuance, therefore securities laws should be
structured in a way that increases the availability of capital for all firms.
128
Table 4.1: Security Issuance by Country
Country Debt Conv. Debt Equity Preferred Conv. Preferred Total
Argentina 29 10 61 2 102
Australia 21 58 8245 48 8372
Austria 2 91 93
Bangladesh 5 5
Belgium 173 173
Bermuda 10 1 11
Bolivia 6 1 7
Brazil 94 25 51 35 205
Canada 26 14 40
Chile 37 160 197
China 7 1291 1298
Colombia 23 32 55
Costa Rica 3 3
Czech Republic 4 4
Denmark 1 192 193
Finland 6 1 224 231
France 48 11 1207 1266
Germany 6 1 585 7 599
Greece 2 133 135
Hong Kong 4 5 900 909
Hungary 16 16
India 125 179 304
Indonesia 40 128 168
Ireland 41 41
Israel 8 8
Italy 3 203 1 207
Japan 2149 239 1951 4339
Luxembourg 7 1 8
Malaysia 64 2 418 1 485
Mexico 91 1 33 125
Netherlands 10 1 136 6 153
New Zealand 2 5 42 3 52
Norway 1 1 102 104
Pakistan 22 22
Papua New Guinea 6 6
Peru 143 3 146
Philippines 18 42 60
Poland 2 32 34
129
Table 4.1: Security Issuance by Country (cont.)
Country Debt Conv. Debt Equity Preferred Conv. Preferred Total
Portugal 46 1 47
Singapore 59 314 373
South Africa 4 4
South Korea 397 9 406
Spain 5 98 103
Sri Lanka 11 11
Sweden 22 236 258
Switzerland 51 7 104 1 163
Taiwan 739 2 316 1057
Thailand 71 2 77 150
Turkey 11 11
US 42 121 3438 3620 17 7238
United Kingdom 7 1855 12 1874
Venezuela 19 38 1 58
Total 3947 497 23692 3764 17 31929
130
Table 4.2: Summary Statistics of Firm Specific Variables
Variable Obs Mean Std. Dev. Min Max
Capital Issuance 108280 0.22 0.41 0.00 1.00
Cash/TA 71742 0.14 0.16 0.00 1.14
Cross listing 108280 0.10 0.30 0.00 1.00
EFN 68977 -1.83E+10 1.96E+12 -4.70E+14 2.37E+12
Risk 103990 -3.36 1.52 -13.82 9.38
Profitability 108280 -0.05 1.31 -165.94 15.48
Uniqueness 71742 0.01 0.41 -76.53 48.80
Asset Tangibility 108280 0.39 0.48 0.00 73.42
Table 4.3: Summary Statistics of Macroeconomic Variables
Variable Obs Mean Std. Dev Min Max
GDP Growth 330 0.030 0.032 -0.131 0.111
Domestic Capital 284 154.943 104.721 28.755 536.873
Foreign Capital 313 0.030 5.719 -27.002 23.839
Corruption 336 3.848 1.296 1 6
Criminal 340 0.502 0.255 0 1
Anti-director 340 3.085 1.346 0 5
Judicial Efficiency 336 7.813 2.143 2.5 10
Private Enforcement Index 340 0.558 0.213 0.11 1
Public Enforcement Index 340 0.501 0.232 0 0.896
131
Table 4.4: Firm Specific Correlations (with Capital Issuance Indicator)
Capital
Issuance Cash/TA Cross listing EFN Risk Profitability Uniqueness
Cash/TA -0.022*** 1.000
Cross listing 0.038*** -0.002 1.000
EFN 0.002 0.003 -0.018*** 1.000
Risk 0.015*** 0.179*** 0.070*** -0.013*** 1.000
Profitability -0.036*** -0.014*** 0.007* 0.000 -0.036*** 1.000
Uniqueness -0.009** -0.006 -0.002 0.000 -0.002 -0.013*** 1.000
Asset Tangibility -0.008** -0.226*** 0.030*** -0.004 -0.086*** 0.008** 0.000
Table 4.5 Macroeconomic Variable Correlations (with Capital Issuance Indicator)
Capital
Issuance
GDP
Growth
Domestic
Capital
Foreign
Capital
Anti-
director
Judicial
Efficiency
Private
Enforcement
GDP Growth 0.072*** 1.000
Domestic Capital -0.135*** 0.041*** 1.000
Foreign Capital 0.029*** 0.157*** 0.106*** 1.000
Anti-director -0.156*** 0.183*** 0.531*** 0.280*** 1.000
Judicial Efficiency -0.044*** -0.010*** 0.534*** 0.060*** 0.580*** 1.000
Private Enforcement -0.171*** 0.213*** 0.488*** 0.325*** 0.825*** 0.462*** 1.000
Public Enforcement 0.002 0.407*** 0.072*** 0.355*** 0.492*** 0.065*** 0.554***
132
Table 4.6A: All Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and standard errors are clustered by firm. All country-level control variables
are averaged over the years t-1 through t-3. Industry and time fixed effects have been suppressed.
Absolute value of z statistics in brackets * significant at 10%; ** significant at 5%; ***
significant at 1%.
1 2 3
Cash_TA -0.039* -0.065** -0.039*
[1.66] [2.57] [1.64]
Uniqueness 0.000** 0.000** 0.000**
[2.08] [2.13] [2.09]
Asset_Tang 0.009 0.003 0.009
[1.49] [0.39] [1.57]
EFN 0.018** 0.020** 0.018**
[2.30] [2.24] [2.30]
Profitability 0.000 0.000 0.000
[1.49] [1.39] [1.49]
LRisk -0.035*** -0.020*** -0.035***
[12.29] [6.39] [12.35]
Cross listing -0.030*** -0.014* -0.03***
[3.81] [1.80] [3.85]
GDP_growth 0.011*** 0.022*** 0.012***
[8.43] [13.73] [8.39]
Domestic Capital -0.001*** -0.001*** -0.001***
[18.70] [13.83] [18.45]
Foreign Capital 0.009*** 0.018*** 0.009***
[12.63] [14.89] [12.58]
Corruption 0.003 0.034*** 0.003
[1.04] [8.24] [0.88]
Anti Director Rights -0.006*** -0.035*** -0.055***
[16.80] [8.24] [12.17]
Judicial Efficiency 0.060*** 0.025*** 0.061***
[17.40] [8.71] [17.48]
Public Enforcement 0.289*** 0.289***
[21.02] [21.01]
Private Enforcement -0.016 -0.027
[0.54] [0.89]
Observations 45179 45179 45179
Log Likelihood -23194.76 -23756.63 -23194.06
Pseudo R-squared 0.08 0.06 0.08
133
Table 4.6B Small Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and standard errors are clustered by firm. Size is determined by Total Assets
divided by GDP. Firms with Total Assets below the median for each country are defined as
small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.043** -0.063** -0.036*
[2.14] [2.34] [1.89]
Uniqueness 0.000* 0.000 0.000**
[1.94] [0.43] [1.98]
Asset_Tang -0.125*** -0.197*** -0.112***
[6.64] [7.58] [6.32]
EFN 0.006* .009* 0.006*
[1.93] [1.81] [1.93]
Profitability -0.001*** -0.001** -0.001***
[3.13] [2.53] [3.25]
LRisk -0.009*** 0.016*** -0.011***
[3.67] [5.34] [4.56]
Cross Listing -0.021** -0.024** -0.019**
[2.21] [2.17] [2.16]
GDP_growth 0.012*** 0.031*** 0.012***
[8.64] [12.89] [8.84]
Domestic Capital 0.000*** 0.000** -0.001***
[9.87] [2.52] [9.11]
Foreign Capital 0.006*** 0.011*** 0.005***
[8.05] [9.86] [7.23]
Corruption 0.016*** 0.055*** 0.009**
[4.24] [11.81] [2.44]
Anti Director Rights -0.088*** -0.040 -0.056***
[23.14] [7.13] [13.35]
Judicial Efficiency 0.040*** 0.005 .043***
[11.14] [1.47] [11.80]
Public Enforcement 0.445*** 0.453***
[26.67] [27.18]
Private Enforcement -0.204*** -0.309***
[5.23] [10.05]
Observations 18708 18708 18708
Log Likelihood -6059.02 -7021.24 -5960.72
Pseudo R-squared 0.33 0.22 0.34
134
Table 4.6C: Large Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Size is
determined by Total Assets divided by GDP. Firms with Total Assets above the median for each
country are defined as large. Probit estimation is used and standard errors are clustered by
firm. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.082** -0.103*** -0.086**
[2.40] [3.07] [2.54]
Uniqueness 0.000** 0.000** 0.000**
[1.99] [1.99] [1.97]
Asset_Tang 0.027 0.018** 0.024
[1.24] [2.38] [1.30]
EFN 0.033* 0.033* 0.033*
[1.83] [1.91] [1.81]
Profitability 0.000 0.000 0.000
[1.41] [1.40] [1.37]
LRisk -0.037*** -0.032*** -0.036***
[1.54] [10.01] [11.37]
Cross Listing -0.014 -0.005 -0.011
[1.45] [0.56] [1.16]
GDP_growth 0.008*** 0.010*** 0.007***
[3.93] [4.99] [3.49]
Domestic Capital -0.001*** -0.001*** -0.001***
[13.60] [13.62] [13.81]
Foreign Capital 0.008*** 0.009*** 0.008***
[8.02] [8.72] [7.95]
Corruption 0.017*** 0.030*** 0.020***
[3.86] [6.67] [4.45]
Anti Director Rights -0.015*** -0.027*** -0.035***
[3.15] [4.63] [5.65]
Judicial Efficiency 0.033*** 0.019*** 0.030***
[7.84] [5.20] [7.51]
Public Enforcement 0.087 0.089***
[1.43] [5.23]
Private Enforcement 0.198*** 0.208***
[4.73] [4.92]
Observations 26295 26295 26295
Log Likelihood -15060.69 -15068.48 -15041.76
Pseudo R-squared 0.05 0.05 0.05
135
Table 4.7A: G10 Small Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Size is
determined by Total Assets divided by GDP. Firms with Total Assets below the median for each
country are defined as small. . Probit estimation is used and standard errors are clustered by
firm. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.009* -0.012*** -0.009*
[1.92] [2.75] [1.87]
Uniqueness -0.009** -0.010** -0.009**
[2.13] [2.27] [2.10]
Asset_Tang -0.009 -0.007* -0.006
[1.62] [1.90] [.1.25]
EFN 0.003*** 0.004*** 0.003***
[3.16] [3.01] [3.20]
Profitability -0.001** -0.001** -0.001**
[2.54] [1.92] [2.47]
LRisk -0.005*** -0.005*** -0.005***
[6.58] [6.65] [6.72]
Cross Listing 0.000 -0.001 -0.001
[0.34] [0.26] [0.42]
GDP_growth -0.657*** -0.340 -0.446***
[4.59] [0.63] [3.81]
Domestic Capital 0.000*** 0.000*** 0.000***
[8.71] [7.63] [8.24]
Foreign Capital -0.002*** -0.002*** -0.002***
[9.01] [9.09] [9.06]
Corruption 0.009*** 0.009*** .006***
[6.99] [6.00] [6.28]
Anti Director Rights -0.005*** 0.002 -0.001
[3.10] [1.34] [1.66]
Judicial Efficiency 0.003 0.001*** 0.005
[0.91] [3.68] [1.20]
Public Enforcement 0.019 0.017***
[7.67] [7.29]
Private Enforcement -0.080** -0.071***
[4.88] [2.77]
Observations 12207 12207 12207
Log Likelihood -1652 -1692.05 -1647.69
Pseudo R-squared 0.39 0.37 0.41
136
Table 4.7B G10 Large Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Size is
determined by Total Assets divided by GDP. Firms with Total Assets above the median for each
country are defined as large. Probit estimation is used and standard errors are clustered by
firm. All country-level control variables are averaged over the years t-1 through t-3. Industry and
time fixed effects have been suppressed. Absolute value of z statistics in brackets * significant at
10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.069 -0.041 -0.067
[1.58] [1.30] [1.58]
Uniqueness -0.116** -0.113** -0.118**
[2.18] [2.02] [2.18]
Asset_Tang 0.109*** 0.118*** 0.113***
[4.49] [4.63] [4.49]
EFN 0.025 0.025 0.025
[0.94] [1.01] [0.94]
Profitability -0.134*** -0.127*** -0.134***
[3.22] [3.23] [3.22]
Lrisk -0.053*** -0.060*** -0.053***
[12.15] [12.65] [12.15]
Cross Listing 0.033** 0.011** 0.031***
[2.49] [2.03] [2.49]
GDP_growth -6.39*** -4.238*** -6.016***
[9.27] [9.74] [9.19]
Domestic Capital 0.001*** 0.000*** -0.001***
[7.93] [10.42] [7.5]
Foreign Capital -0.014*** 0.000*** -0.013***
[8.81] [8.71] [8.80]
Corruption 0.097*** 0.007*** 0.090***
[10.64] [9.75] [10.46]
Anti Director Rights 0.080*** -0.012*** 0.110***
[7.51] [4.13] [6.04]
Judicial Efficiency -0.139*** -0.009*** -0.140***
[7.73] [6.05] [7.67]
Public Enforcement -0.401*** -0.422***
[5.36] [4.75]
Private Enforcement -0.139*** -0.340
[2.61] [0.03]
Observations 18471 18471 18471
Log Likelihood -9894.86 -9913.74 -9894.86
Pseudo R-squared .09 .09 .09
137
Table 4.8A: Small Emerging Market Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and standard errors are clustered by firm. Size is determined by Total Assets
divided by GDP. Firms with Total Assets below the median for each country are defined as
small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.173*** -0.152*** -0.182***
[3.73] [3.33] [3.92]
Uniqueness 0.000* 0.000** 0.000*
[1.92] [2.04] [1.83]
Asset_Tang -0.410*** -0.449*** -0.381***
[8.90] [9.53] [8.41]
EFN 0.007* 0.008* 0.006*
[1.93] [1.92] [1.87]
Profitability -0.009*** -0.009** -0.008**
[2.19] [2.19] [2.14]
Lrisk -0.011* -0.006 -0.017***
[1.98] [1.18] [3.17]
Cross Listing [d) -0.084*** -0.106*** -0.051**
[3.40] [4.45] [2.00]
GDP_growth 0.003 0.004 0.007**
[1.09] [1.76] [2.49]
Domestic Capital 0.000 0.000 0.000
[2.38] [0.31] [0.13]
Foreign Capital 0.005 0.005 0.000
[0.76] [0.76] [0.02]
Corruption 0.127*** 0.118*** .127***
[3.23] [3.54] [0.08]
Anti Director Rights -0.059*** 0.006 -0.049***
[4.58] [0.46] [3.74]
Judicial Efficiency 0.022*** 0.028*** 0.026***
[2.92] [3.62] [3.32]
Public Enforcement 0.459*** 1.001***
[8.49] [12.12]
Private Enforcement -0.166** -1.054***
[2.16] [9.53]
Observations 6501 6501 6501
Log Likelihood 3323.59 3374.17 3234.88
Pseudo R-squared 0.25 0.24 0.27
138
Table 4.8B: Large Emerging Market Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. . Probit
estimation is used and standard errors are clustered by firm. Size is determined by Total Assets
divided by GDP. Firms with Total Assets below the median for each country are defined as
small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.130** -0.127** -0.133**
[2.17] [2.13] [2.24]
Uniqueness 0.000** 0.000** 0.000**
[2.51] [2.35] [2.55]
Asset_Tang -0.087** -0.100*** -0.086**
[2.36] [2.69] [2.34]
EFN 0.005 0.009 0.003
[0.26] [0.45] [0.19]
Profitability 0.000* 0.000* 0.000*
[1.81] [1.85] [1.81]
Lrisk -0.008* -0.004 -0.009**
[1.82] [1.03] [2.01]
Cross Listing (d) -0.049*** -0.054*** -0.047***
[3.21] [3.56] [3.09]
GDP_growth 0.009*** 0.009*** 0.010***
[4.03] [4.02] [2.98]
Domestic Capital -0.001*** -0.001*** -0.001***
[5.79] [5.32] [4.38]
Foreign Capital 0.006** 0.008** 0.005***
[4.79] [5.68] [4.91]
Corruption 0.068*** 0.070*** 0.066***
[5.79] [6.08] [3.74]
Anti Director Rights -0.019 0.002 -0.015
[1.36] [0.13] [1.06]
Judicial Efficiency -0.005 -0.005 -0.006
[0.76] [0.70] [0.78]
Public Enforcement 0.268*** 0.360***
[6.15] [5.78]
Private Enforcement 0.136** -0.194**
[2.22] [2.22]
Observations 7824 7824 7824
Log Likelihood 4546.79 4572.79 4542.18
Pseudo R-squared 0.08 0.07 0.08
139
Table 4.9A: All Firms – Standard Errors Clustered by Country
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and regressions are clustered by country. All country-level control variables
are averaged over the years t-1 through t-3. Industry and time fixed effects have been suppressed.
Absolute value of z statistics in brackets * significant at 10%; ** significant at 5%; ***
significant at 1%.
1 2 3
Cash_TA -0.039 -0.065** -0.039
[0.54] [0.83] [0.54]
Uniqueness 0.000 0.000** 0.000
[1.27] [1.07] [1.27]
Asset_Tang 0.009 0.003 0.009
[0.29] [0.15] [0.29]
EFN 0.018** 0.020** 0.018***
[2.65] [2.66] [2.65]
Profitability 0.000 0.000 0.000
[1.17] [1.13] [1.16]
LRisk -0.035 -0.020 -0.035*
[1.77] [0.96] [1.76]
Cross listing -0.030 -0.014* -0.03
[1.31] [0.62] [1.31]
GDP_growth 0.011** 0.022*** 0.012**
[2.32] [3.36] [2.44]
Domestic Capital -0.001*** -0.001* -0.001**
[2.62] [1.68] [2.49]
Foreign Capital 0.009*** 0.012 0.009***
[2.27] [1.70] [2.20]
Corruption 0.003 0.034 0.003
[1.17] [1.17] [0.16]
Anti Director Rights -0.006** -0.035 -0.055***
[2.57] [1.07] [1.64]
Judicial Efficiency 0.060*** 0.025* 0.061***
[3.02] [1.81] [3.22]
Public Enforcement 0.289*** 0.289***
[3.35] [3.36]
Private Enforcement -0.016 -0.027
[0.09] [0.16]
N 45179 45179 45179
Log Likelihood -23194.76 -23756.63 -23194.06
Pseudo R-sq 0.08 0.06 0.08
140
Table 4.9B: All Firm Groupings – Standard Errors Clustered by Country
The dependent variable is dummy variable for whether the firm issues capital in period t. . Probit
estimation is used and standard errors are clustered by country. Size is determined by Total
Assets divided by GDP. Firms with Total Assets below the median for each country are defined
as small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
All Firms G10 Firms Emerging Firms
Small Large Small Large Small Large
Cash_TA -0.036 -0.086* -0.009* -0.067 -0.182** -0.133**
[0.75] [1.68] [1.38] [1.04] [1.99] [2.38]
Uniqueness 0.000 0.000* -0.009** -0.118* 0.000 0.000
[0.80] [1.66] [1.62] [1.89] [0.57] [0.73]
Asset_Tang -0.112* 0.024 -0.006 0.113* -0.381*** -0.086*
[1.94] [0.40] [0.68] [1.69] [4.95] [1.86]
EFN 0.006*** 0.033 0.003** 0.025 0.006*** 0.003
[2.66] [1.38] [2.08] [0.76] [2.73] [0.23]
Profitability -0.001*** 0.000 -0.001** -0.134 -0.008** 0.000
[2.89] [0.91] [2.22] [1.63] [3.82] [1.05]
Lrisk -0.011* -0.036** -0.005* -0.053*** -0.017*** -0.009
[1.85] [2.17] [1.86] [3.75] [2.99] [0.90]
Cross Listing -0.019 -0.011 -0.001 0.031*** -0.051* -0.047***
[1.46] [0.86] [0.25] [3.00] [1.80] [3.49]
GDP_growth 0.012*** 0.007 -0.446** -6.016*** 0.007** 0.010**
[3.84] [1.38] [2.40] [1.76] [1.26] [2.34]
Domestic Capital -0.001 -0.001*** 0.000*** -0.001 0.000 -0.001***
[1.60] [3.08] [2.70] [1.21] [0.05] [2.91]
Foreign Capital 0.005* 0.008** -0.002** -0.013* 0.000 0.005***
[1.78] [2.11] [2.37] [1.65] [0.02] [1.67]
Corruption 0.009** 0.020 0.006** 0.090* 0.127*** 0.066***
[0.43] [1.37] [2.18] [1.82] [3.97] [2.98]
Anti Director Rights -0.056*** -0.035 -0.001 0.110** -0.049 -0.015
[2.90] [1.12] [0.23] [2.41] [1.44] [0.90]
Judicial Efficiency 0.043*** 0.030*** 0.005* -0.140* 0.026 -0.006
[2.66] [2.63] [1.76] [1.70] [0.30] [0.43]
Public Enforcement 0.453*** 0.089 0.017 -0.422*** 1.001*** 0.360***
[4.02] [1.40] [1.43] [3.10] [4.21] [3.90]
Private Enforcement -0.309** 0.208 -0.071** -0.340 -1.054*** -0.194
[2.17] [1.29] [2.42] [1.25] [4.21] [1.16]
Observations 18708 26295 12207 18471 6501 7824
Log Likelihood -5960.72 -15041.76 -1647.69 -9894.86 -3234.88 -4542.18
Pseudo R-squared 0.34 0.05 0.41 0.09 0.27 0.08
141
Appendices
Appendix A: Derivatives of the Constrained Capital Equation
2
2 2 2 2 2 2
) ( 2
) ( 4 ) ( 2
?
? ? ? ? ?
i
wi g g wi g
k k
c
+ + +
= = when <1. (2.3)
A1. Derivative of constrained capital with respect to financial development,
2
2 2 2
2
2 2 2 2
3
2 2 2 2 2 2
) ( 2
) ( 4 2
) ( ' ) ( 8 2 (
) ( ' ) ( 4 ) ( 4 2
) (
) ( 4 ) ( ' ) ( 2
?
? ?
? ? ? ?
? ? ? ? ?
?
? ? ? ? ? ?
?
i
wi g
i wi g g
i wi wi g g g
i
wi g i g wi g k
c
+
+
+ + + +
+
+ + + ?
=
?
?
A2. Derivative of constrained capital with respect to productivity, g
2
2 2 2
2 2 2
2 2
2
) ( 2
) ( 4
) ( 4
2
?
? ? ?
? ?
?
?
i
wi g
wi g
g
g
g
k
c
+ +
+
+
=
?
?
A.3 Derivative of constrained capital with respect to initial wealth, w
2
2 2 2
2
2
) ( 2
) ( 4
) ( 2
) ( 2
?
? ?
? ?
?
i
wi g
i g
i
w
k
c
+
+
=
?
?
142
Appendix B: Derivatives of the Firm Profit Equations with respect to
B.1 Large firm profit
B.2 Derivative of large firm profit with respect to
B.3 Constrained firm capital demand
B.4 Constrained firm profit
143
B.5 Derivative of constrained firm profit with respect to
144
Appendix C: Description of Variables, Chapter 3
Variable Definition Source
Private Credit Private credit by deposit money banks and
other financial institutions to GDP,
calculated using the following deflation
method: {(0.5)*[F
t
/P_e
t
+ F
t-1
/P_e
t-
1
]}/[GDP
t
/P_a
t
] where F is credit to the
private sector, P_e is end-of period CPI, and
P_a is average annual CPI
Beck, Demirguc-Kunt, and
Levine (2003)
Checks The numbers of veto players in the political
system, adjusting for whether these veto
players are independent of each other as
determined by the level of electoral
competitiveness, their respective party
affiliations and the electoral rules.
Database of Political
Institutions
Beck, Clark, Groff, Keefer,
and Walsh. (2001)
Political
Polarization
Maximum polarization between the
executive party and the four principle parties
of the legislature
Database of Political
Institutions
Keefer (2002)
Lgdp95 The natural log of GDP per capita is gross
domestic product divided by midyear
population. GDP is the sum of gross value
added by all resident producers in the
economy plus any product taxes and minus
any subsidies not included in the value of the
products. It is calculated without making
deductions for depreciation of fabricated
assets or for depletion and degradation of
natural resources. Data are in constant U.S.
dollars.
World Bank Indicators (2003)
Trade Trade is the sum of exports and imports of
goods and services measured as a share of
gross domestic product.
World Bank Indicators (2003)
FDI Foreign direct investment is net inflows of
investment to acquire a lasting management
interest (10 percent or more of voting stock)
in an enterprise operating in an economy
other than that of the investor. It is the sum
of equity capital, reinvestment of earnings,
other long-term capital, and short-term
capital as shown in the balance of payments.
World Bank Indicators (2003)
145
Inflation Inflation as measured by the annual growth rate
of the GDP implicit deflator shows the rate of
price change in the economy as a whole.
World Bank Indicators (2003)
Natural
Openness
The Natural logarithm of aggregated fitted
values if bilateral trade equation with geographic
variables
Frankel and Romer (1999)
Lsettler Natural logarithm of estimated European
settlers’ mortality rate.
Acemoglu, Johnson, and
Robinson (2001)
Disteq Distance from Equator of capital city measured
as abs(Latitude)/90.
World Bank (2002)
Engfrac Fraction of Population speaking English as a 1
st
language
Hall and Jones (1999)
Legor_uk British Legal Origin La Porta et al. (1998)
Legor_fr French Legal Origin La Porta et al. (1998)
Legor_ge German Legal Origin La Porta et al. (1998)
Legor_sc Scandanvian Legal Origin La Porta et al. (1998)
Legor_so Socialist Legal Origin La Porta et al. (1998)
GROUPB A dummy variable that equals 1 if the country-
year observation is greater than 1995 and 0
otherwise.
LCHECKSB An interaction term equal to Log of Checks *
GROUPB
POLARIZB An interaction term equal to Polarization *
GROUPB
Widely-Held10 The proportion of firms in a country that are
widely held, where control is inferred at 10%
Morck, Wolfenzon, and Yeung
(2004)
Widely-Held20 The proportion of firms in a country that are
widely held, where control is inferred at 20%
Morck, Wolfenzon, and Yeung
(2004)
146
Appendix D: Description of Variable, Chapter 4
Variable Description
Securities Law, Source and definitions: La Porta et al. (2004)
Disclose The index of disclosure equals the arithmetic mean of: (1) Prospect; (2)
Compensation; (3) Shareholders; (4) Inside ownership; (5) Contracts Irregular; (6)
and Transactions.
Prospectus Equals one if the law prohibits selling securities that are going to be listed on the
largest stock exchange of the country without delivering a prospectus to potential
investors; equals zero otherwise.
Compensa An index of prospectus disclosure requirements regarding the compensation of
directors and key officers. Equals one if the law or the listing rules require that the
compensation of each director and key officer be reported in the prospectus of a
newly-listed firm; equals one-half if only the aggregate compensation of directors
and key officers must be reported in the prospectus of a newly-listed firm; equals
zero when there is no requirement to disclose the compensation of directors and
key officers in the prospectus for a newly-listed firm.
Sharehol An index of disclosure requirements regarding the Issuer’s equity ownership
structure. Equals one if the law or the listing rules require disclosing the name and
ownership stake of each shareholder who, directly or indirectly, controls ten
percent or more of the Issuer’s voting securities; equals one-half if reporting
requirements for the Issuer’s 10% shareholders do not include indirect ownership
or if only their aggregate ownership needs to be disclosed; equals zero when the
law does not require disclosing the name and ownership stake of the Issuer’s 10%
shareholders. No distinction is drawn between large-shareholder reporting
requirements imposed on firms and those imposed on large shareholders
themselves.
Insideow An index of prospectus disclosure requirements regarding the equity ownership of
the Issuer’s shares by its directors and key officers. Equals one if the law or the
listing rules require that the ownership of the Issuer’s shares by each of its director
and key officers be disclosed in the prospectus; equals one-half if only the
aggregate number of the Issuer’s shares owned by its directors and key officers
must be disclosed in the prospectus; equals zero when the ownership of Issuer’s
shares by its directors and key officers need not be disclosed in the prospectus.
Contract An index of prospectus disclosure requirements regarding the Issuer’s contracts
outside the ordinary course of business. Equals one if the law or the listing rules
require that the terms of material contracts made by the Issuer outside the ordinary
course of its business be disclosed in the prospectus; equals one-half if the terms of
only some material contracts made outside the ordinary course of business must be
disclosed; equals zero otherwise.
Transact An index of the prospectus disclosure requirements regarding transaction between
the Issuer and its directors, officers, and/or large shareholders (i.e., “related
parties”). Equals one if the law or the listing rules require that all transactions in
which related parties have, or will have, an interest be disclosed in the prospectus;
equals one-half if only some transactions between the Issuer and related parties
must be disclosed in the prospectus; equals zero if transactions between the Issuer
and related parties need not be disclosed in the prospectus.
bdn_proof The index of burden of proof equals the arithmetic mean of: (1) Burden director;
(2) Burden distributor; and (3) Burden accountant.
147
bdn_dire Index of the procedural difficulty in recovering losses from the Issuer’s directors in
a civil liability case for losses due to misleading statements in the prospectus.
Equals one when investors are only required to prove that the prospectus contains a
misleading statement. Equals two-thirds when investors must also prove that they
relied on the prospectus and/or that their loss was caused by the misleading
statement. Equals one-third when investors prove that the director acted with
negligence and that they either relied on the prospectus or that their loss was
caused by the misleading statement or both. Equals zero if restitution from
directors is unavailable or the liability standard is intent or gross negligence.
bdn_dist Index of the procedural difficulty in recovering losses from the Distributor in a
civil liability case for losses due to misleading statements in the prospectus.
Equals one when investors are only required to prove that the prospectus contains a
misleading statement. Equals two-thirds when investors must also prove that they
relied on the prospectus and/or that their loss was caused by the misleading
statement. Equals one-third when investors prove that the Distributor acted with
negligence and that they either relied on the prospectus or that their loss was
caused by the misleading statement or both. Equals zero if restitution from the
Distributor is unavailable or the liability standard is intent or gross negligence.
bdn_acco Index of the procedural difficulty in recovering losses from the Accountant in a
civil liability case for losses due to misleading statements in the audited financial
information accompanying the prospectus. Equals one when investors are only
required to prove that the audited financial information accompanying the
prospectus contains a misleading statement. Equals two-thirds when investors
must also prove that they relied on the prospectus and/or that their loss was caused
by the misleading accounting information. Equals one-third when investors prove
that the Accountant acted with negligence and that they either relied on the
prospectus or that their loss was caused by the misleading statement or both.
Equals zero if restitution from the Accountant is unavailable or the liability
standard is intent or gross negligence.
Supervisor The index of characteristics of the Supervisor equals the arithmetic mean of: (1)
Appointment; (2) Tenure; (3) Focus; and (4) Rules.
Appoint Equals one if a majority of the members of the Supervisor are unilaterally
appointed by the Executive branch of government; equals zero otherwise.
Tenure Equals one if members of the Supervisor cannot be dismissed at the will of the
appointing authority; equals zero otherwise.
Focus Equals one if separate government agencies or official authorities are in charge of
supervising commercial banks and stock exchanges; equals zero otherwise.
Rules Equals one if the Supervisor can generally issue regulations regarding primary
offerings and/or listing rules on stock exchanges without prior approval of other
governmental authorities. Equals one-half if the Supervisor can generally issue
regulations regarding primary offerings and/or listing rules on stock exchanges
only with the prior approval of other governmental authorities. Equals zero
otherwise.
Investing The index of investigative powers equals the arithmetic mean of: (1) Document;
and (2) Witness.
Document An index of the power of the Supervisor to command documents when
investigating a violation of securities laws. Equals one if the Supervisor can
generally issue an administrative order commanding all persons to turn over
documents; equals one-half if the Supervisor can generally issue an administrative
148
order commanding publicly-traded corporations and/or their directors to turn over
documents; equals zero otherwise.
Witness An index of the power of the Supervisor to subpoena the testimony of witnesses
when investigating a violation of securities laws. Equals one if the Supervisor can
generally subpoena all persons to give testimony; equals one-half if the Supervisor
can generally subpoena the directors of publicly-traded corporations to give
testimony; equals zero otherwise.
Orders The index of orders equals the arithmetic mean of: (1) Orders issuer; (2) Orders
distributor; and (3) Orders accountant.
ord_iss An index aggregating stop and do orders that may be directed at the Issuer in case
of a defective prospectus. The index is formed by averaging the sub-indexes of
orders to stop and to do. The sub-index of orders to stop equals one if the Issuer
may be ordered to refrain from a broad range of actions; equals one-half if the
Issuer may only be ordered to desist from limited actions; equals zero otherwise.
The sub-index of orders to do equals one if the Issuer may be ordered to perform a
broad range of actions to rectify the violation; equals one-half if the Issuer may
only be ordered to perform limited actions; equals zero otherwise.
ord_dis An index aggregating stop and do orders that may be directed at the Distributor in
case of a defective prospectus. The index is formed by averaging the sub-indexes
of orders to stop and to do. The sub-index of orders to stop equals one if the
Distributor may be ordered to refrain from a broad range of actions; equals one-
half if the Distributor may only be ordered to desist from limited actions; equals
zero otherwise. The sub-index of orders to do equals one if the Distributor may be
ordered to perform a broad range of actions to rectify the violation; equals one-half
if the Distributor may only be ordered to perform limited actions; equals zero
otherwise.
ord_acc An index aggregating stop and do orders that may be directed at the Accountant in
case of a defective prospectus. The index is formed by averaging the sub-indexes
of orders to stop and to do. The sub-index of orders to stop equals one if the
Accountant may be ordered to refrain from a broad range of actions; equals one-
half if the Accountant may only be ordered to desist from limited actions; equals
zero otherwise. The sub-index of orders to do equals one if the Accountant may be
ordered to perform a broad range of actions to rectify the violation; equals one-half
if the Accountant may only be ordered to perform limited actions; equals zero
otherwise.
Criminal The index of criminal sanctions equals the arithmetic mean of: (1) Criminal
director; (2) Criminal distributor; and (3) Criminal accountant.
Crim_dir An index of criminal sanctions applicable to the Issuer’s directors and key officers
when the prospectus omits material information. The sub-index for directors
equals zero when directors cannot be held criminally liable when the prospectus is
misleading. Equals one-half if directors can be held criminally liable when aware
that the prospectus is misleading. Equals one if directors can also be held
criminally liable when negligently unaware that the prospectus is misleading. The
sub-index for key officers is constructed analogously.
Crim_dis An index of criminal sanctions applicable to the Distributor (or its officers) when
the prospectus omits material information. Equals zero if the Distributor cannot be
held criminally liable when the prospectus is misleading. Equals one-half if the
Distributor can be held criminally liable when aware that the prospectus is
misleading. Equals one if the Distributor can also be held criminally liable when
149
negligently unaware that the prospectus is misleading.
Crim_acc An index of criminal sanctions applicable to the Accountant (or its officers) when
the financial statements accompanying the prospectus omit material information.
Equals zero if the Accountant cannot be held criminally liable when the financial
statements accompanying the prospectus are misleading. Equals one-half if the
Accountant can be held criminally liable when aware that the financial statement
accompanying the prospectus are misleading. Equals one if the Accountant can
also be held criminally liable when negligently unaware that the financial
statements accompanying the prospectus are misleading.
Private
Enforcement
The index of private enforcement equals the arithmetic mean of: (1) Disclosure
Index; and (2) Burden of proof index.
Public
Enforcement
The index of public enforcement equals the arithmetic mean of: (1) Supervisor
characteristics index; (2) Investigative powers index; (3) Orders index; and (4)
Criminal index.
Firm Characteristics, Source: Knill (2004)
Asset
tangibility
Fixed assets divided by the book value of total assets; industry average is used in
cases of missing data FA/TA
Capital
Issuance
A dummy variable which equals 1 if firm i issued some type of capital (equity,
debt, convertible, etc.) in time t and 0 otherwise
Cross listing A dummy variable which takes on a value of 1 if the firm has stock listed on
additional exchanges and a 0 otherwise
Desired
Growth in
Sales
For each US firm in our data set, we measure growth in Sales from year t-1 to year
t. Next, we average individual firm growth rates in the same industry and size
classification. The term
US
m j
g
,
is the average growth rate in Sales for firms in
industry j of size m. We use
US
m j
g
,
to approximate desired growth in Sales for a
firm outside the US in industry j and size m.
EFN The ‘external funds necessary’ (or EFN) for firm i in period t is calculated as
follows:
) )( ( ) )( / ( ) )( / (
, , , 1 , , , , 1 , , 1 , 1 , , t i t i t i t i t i t i t i t i t i t i t i t i
RR S M S S S L S S S A EFN ? ? ? ? =
? ? ? ?
where A
t-1
is the total assets of the firm in time t-1, S
t-1
and S
t
are the sales of the
firm in times t-1 and t respectively, L
t
is the liabilities of the firm in time t, M
t
is
the profit margin of the firm as defined by net income divided by sales for time t,
RR is the retention ratio for the firm.
Growth in
assets
Growth in total assets (TA
t
– TA
t-1
)/TA
t-1
/(Year
t
-Year
t-1
)
Industry Macro Industry Code from SDC Platinum
Profitability Operating income divided by sales OpInc/Sales
Risk Standard deviation of the firm’s profitability ratio over the three years prior to
issue; industry average is used in cases of missing data SD(ROA
t
, ROA
t-1
, ROA
t-2
)
Uniqueness
of assets
Selling expense divided by sales; industry average is used in cases of missing data
SellExp/Sales
150
Macroeconomic Characteristics
Anti
Director
This index of Anti-director rights is formed by adding one when: (1) the country
allows shareholders to mail their proxy vote; (2) shareholders are not required to
deposit their shares prior to the General Shareholders’ Meeting; (3)cumulative
voting or proportional representation of minorities on the board of directors is
allowed; (4) an oppressed
minorities mechanism is in place; (5) the minimum percentage of share capital that
entitles a shareholder to call for an Extraordinary Shareholders’ Meeting is less
than or equal to ten percent (the sample median); or (6) when shareholders have
preemptive rights that can only be waved by a shareholders meeting. The range for
the index is
from zero to six. Source: La Porta et al. (1998)
Corruption An assigned value from 0 to 6 of perceived Corruption in a country, 0 being the
most Corrupt and 6 the least. The index is based on the likelihood of solicited
bribes from a country in relation to such factors of business as exchange controls,
tax assessment, and loan protection. Source: International Country Risk Guide
Dom
Credit_GDP
Credit provided by monetary authorities and deposit money banks, as well as other
banking institutions (where data is available). It includes all credit to various
sectors on a gross basis, with the exception of credit to the central government,
which is net. Source: WDI
Efficient
Judiciary
Assessment of the “efficiency and integrity of the legal environment as it affects
business, particularly foreign firms” produced by the country risk rating agency
International Country Risk (ICR). Average between 1980 and 1983. Scale from 0
to10, with lower scores representing lower efficiency levels. Source: International
Country Risk
GDP
Growth
GDP per capital growth (%). Source: WDI
Inflation Inflation levels expressed in percent averaged annually over the period 1996-2002.
Source: WDI
MktCap
Percent
Listed shares * their value, scaled by GDP
Domestic
Capital
Sum of all sources of capital in the domestic economy,
equal to MktCap Percent + Domestic Credit
Foreign
Capital
All foreign investment (Foreign direct investment + Foreign portfolio investment +
other (foreign bank lending and official flows)), scaled by GDP
151
Appendix E: Regression of Capital Issuance on Individual Sub Indexes of
Public and Private Enforcement
The dependent variable is dummy variable for whether the firm issues capital in period t.
Variable definitions are given in Appendix 3. Size is determined by Total Assets divided by
GDP. Firms with Total Assets below(above) the median for each country are defined as small
(large). Probit estimation is used and regressions are clustered by country. We estimate
) ( Pr ) 1 ( Pr
, 6 5 4 , 3 1 , 2 1 , 1 , t i t j k t i t k t i t i
T I SECURITIES EFN Z F ob Y ob ? ? ? ? ? ? ? ? + + + + + + + = =
? ?
A,
where SECURITIES is either supervisor, investig, orders, criminal, disclosure, or burden of
proof. Country-level control variables are averaged over the years t-1 through t-3. Only security
law indexes are reported. Absolute value of z statistics are in parenthesis ** significant at 5%;
*** significant at 1%.
Supervisor Investig Orders Criminal Disclosure Burden of
Proof
Whole
Sample
All Firms 0.325*** 0.091*** 0.090 0.147*** 0.80** 0.011
(23.24) (9.55) (10.33) (12.90) (2.26) (0.57)
Small 0.470*** 0.259*** 0.234*** 0.332*** -0.040 -0.116***
(24.76) (21.27) (19.87) (25.61) (0.89) (5.55)
Large 0.120*** 0.050*** -0.037*** -0.045*** 0.198 0.096***
(6.75) (4.02) (3.27) (2.82) (4.47) (3.72)
G10 Firms
Small 0.007 0.014*** 0.012*** 0.018** 0.225*** -0.041***
(1.19) (5.17) (4.00) (2.53) (7.80) (5.98)
Large -0.315*** -0.247*** -0.251*** -0.685*** -0.087 -0.08*
(6.05) (11.26) (9.76) (13.95) (0.42) (1.96)
Emerging
Markets
Small 0.853*** 0.242*** 0.252*** 0.066 -0.214*** -0.023
(15.20) (6.29) (7.30) (0.81) (2.80) (0.13)
Large 0.370*** 0.184*** 0.120*** 0.088*** 0.115*** 0.076
(8.01) (6.61) (5.00) (2.88) (3.08) (1.62)
152
Bibliography
Abiad, Abdul and Ashoka, Mody. “Financial Reform: What Shakes It? What Shapes It?,
IMF Working Paper , No. 03/70, (Apr. 2003).
Acemoglu, Daren, Johnson, Simon, and Robinson, James. “Instituions as the
Fundemental Cause of Long Run Growth”, NBER Working Paper 10481, (May, 2004).
Acemoglu, Daren, Johnson, Simon, Robinson, James, and Yared, Pierre. “Income and
Democracy”, Working Paper, (Feb., 2005)
Aghion, Phillippe, Alesina, Alberto, and Trebbi, Francesco. “Endogenous Political
Institutions”, NBER working Paper Series, No. 9006, (Jun., 2002).
Aivazlan, Varouj, Booth, Lawerence, Demirguc-Kent, Asli, and Maksimovic Vojislav.
“Capital Structure in Developing Countries”. Journal of Finance, Vol. LXI, No. 1, (Feb.,
2001).
Andrews, Donald and Fair, Ray. “Inference in Nonlinear Econometric Models with
Structural Change”, Review of Economic Studies, Vol. 55, No.4, (Oct. 1998), pp. 615-
639.
Andrews, Josephine and Montinola, Grabiella. “Veto Players and the Rule of Law in
Emerging Democracies”, Comparative Political Studies, Vol. 37, No.1, (Feb., 2004),
pp.55-87.
Ayyagari, Meghana, Demirguc-Kunt, Asli, and Maksimovic, Vojislav. “How Well Do
Institutional Theories Explain Firms’ Perceptions of Property Rights” Working Paper
(Feb., 2005).
Bebchuk, Lucian and Roe, Mark. “A Theory of Path Dependence in Corporate
Governance and Ownership”. Stanford Law Review, (Nov 1999).
Beck, Thorsten, Demirguc-Kunt, Asli, and Levine, Ross. “Law and Firms’ Access to
Finance. World Bank Working Paper, (2002).
Beck, T., Demirguc-Kunt, A. and Maksimovic, V. “Financial and Legal constraints to
Firm Growth: Does size matter?” The Journal of Finance, forthcoming.
Beck, Thorsten & Demirguc-Kunt, Asli and Levine, Ross. "Law, Endowments, and
Finance," Journal of Financial Economics, Vol. 70, No. 2, (2003) pp. 137-181.
Beck, Thornsten, Clarke, George, Groff, Alberto, Keefer, Philip, and Walsh, Patrick,
“New Tools in Comparative Political Economy: The Database of Political Institutions”.
World Bank Economic Review, Vol. 15 No.1. (September 2001), pp. 165-176.
153
Becker, Gary S. “A Theory of Competition Among Pressure Groups for Political
Influence”. Quarterly Journal of Economics, Vol. 98 (Aug., 1983) pp. 371 -400.
Beim, David. O and Calomiris, Charles W. Emerging Financial Markets. McGraw-Hill:
New York, (2001).
Bencivenga, Valerie R. and Smith, Bruce D. “Deficits, Inflation and the Banking System
in Developing Countries: The Optimal Degree of Financial Repression”. Oxford
Economics Paper. Vol. 44, Iss.4 (Oct., 1992), pp. 767-790.
Bernheim, B. Douglas and Whinston, Michael D. “Common Agency”, Econometrica
Vol. 54, No. 4 (July 1986) pp. 923-942.
Biasis, Bruno and Mariotti, Thomas “Credit, Wages and Bankruptcy Laws”, Working
Paper (Sep. 2003).
Blackburn, Keith and Hung, Victor. “A Theory of Growth, Financial Development, and
Trade” Economica, New Series, Vol. 65, No. 257, (Feb. 1998), pp. 107-124.
Booth, L. Aivazian, V. Demirguc-Kunt, A. and Maksimovic, V. “Capital Structure in
Developing Countries”, Forthcoming, Journal of Finance.
Buffie, Edward F. “Capital Rationing and Capital Accumulation”, Economica, New
Series, Vol. 58, Iss. 231, (Aug. 1991) pp. 299-316.
Claessens, Stijin, Djankov, Simon, and Lang Larry H.P. “The Separation of Ownership
and control in East Asian Corporations”, Journal of Financial Economic,(2002) Vol. 58
pp. 81-112.
Demirguc-Kunt, Asli, and Maksimovic, Vojislav. “Law, Finance, and Firm Growth”.
Journal of Finance, vol. 63, No. 6, (Dec., 1998).
Demirguc-Kunt, Asli, and Maksimovic, Vojislav. “Stock Market Development and Firm
Finance Choices”, World Bank Policy Research Working Paper 1461, (May 1995).
Desai, Mihir, Dyck, A., and Zingales, Luigi. “Theft and Taxes”, NBER Working Paper
W10978 (Dec. 2004).
Djankov, Simon, La Porta, Raphael, and Shleifer, Andrei. “Regulation of Entry”,
Quarterly Journal of Economics, (2002) Vol. 117 pp. 1 -37.
Djankov, Simeon, McLiesh, Caralee, and Shleifer Andrei. “Private Credit in 129
Countries”, NBER Working Paper No. W11078, (Jan 2005).
154
Djankov, Simon, and Murrell, Peter. “Enterprise Restructuring in Transition: A
Quantitative Survey”. Journal of Economic Literature, (2002).
Drazen, Alan. Political Economy in Macroeconomics, Princeton UP, (2000).
Dyck A. and Zingales L. “Private Benefits of Control: An International Comparison”.
The Journal of Finance, April 2004, vol. 59, no. 2, pp. 537-600.
Easterly, William and Levine, Ross. “Africa’s Growth Tragedy: Policies and Ethnic
Divisions”, The Quarterly Journal of Economics, Vol. 1212, No.4, (Nov., 1997), pp.
1203-1250.
Easterly, William Levine, Ross. “Tropics, Germs, and Crops: How Endowments
Influence Economic Development. Journal of Monetary Economics, Vol. 50, Iss. 1, (Jan.,
2003), pp. 3-39.
Engerman, Stanley, and Kenneth Sokoloff, 2002. “Factor Endowments, Inequality, and
Paths of Development among New World Economies”. NBER Working Paper,No. 9259.
Esfahani, Hadi Saleki. “Institutions and Government Controls”. Journal of Development
Economics. Vol. 63 (2000) pp. 197 -229.
Faccio, Mara. “Politically-Connected Firms: Can They Squeeze the State” Working
Paper 2002.
Fisman, Raymond. “Estimating the Value of Political Connections”, American Economic
Review, Vol. 91, pp. 1095 -1102.
Frankel, Jeffery, and Romer, David. “Does Trade Cause Growth?”, The American
Economic Review, Vol. 89, No. 3, (Jun., 1999), pp. 379-399.
Glen, Jack, and Pinto, Brian. “Debt or Equity? How Firms in Developing Countries
Choose”. IFC Discussion Paper No. 22 (1994).
Gordan, Robert H., and Lei, Wei. “Government as a Discriminating Monopolist in the
Financial Market: The Case of China”. Working Paper (Jun., 2001).
Gradstein, Mark, 2004. “Inequality, Democracy and the Emergence of Institutions”.
CEPR Discussion Paper No. 4187.
Greene, William. Econometric Analysis, Prentice-Hall (1997), Saddle River, NJ.
Greif, Avner, Milgrom, Paul, & Weingast, Barry R. (1994). Coordination, Commitment,
and Enforcement: The case of the merchant guild. Journal of Political Economy, 102(4),
745-777.
155
Grossman, Gene M, and Helpman, Elhanan. Special Interest Politics MIT (2001),
Cambridge, MA, pp. 225-279.
Harris, Milton, and Raviv, Artur. “The Theory of Capital Structure”. Journal of Finance,
Vol. 46, Iss. 1 (Mar., 1991), pp. 297-355.
Hellman, Joel. “Winners Take All: The Politics of Partial Reform in Postcommunist
Transitions”, World Politics, 50, (1998) pp. 203-234.
Henisz, Witold. “The Institutional Environment for Economic Growth”, Economics and
Politics, Vol. 12, No. 1, (Jan., 2002), pp. 1-31.
Henry, Peter Blair. “Do Stock Market Liberalizations Cause Investment Booms”. Journal
of Financial Economics, 58, (2000) pp. 301-334.
Himmelburg, Charles, Hubbard, R. Glen, and Love, Inessa. “Investor Protection,
Ownership and the Cost of Capital”, World Bank Working Paper Series No. 2834.
Islam, Romeen and Montenegro, Claudio. “What Determines the Quality of Institutions”,
Background Paper for the World Develepment Report, 2002: Building Institutions for
Markets, (Jan., 2002).
Jensen, Michael C., and Meckling, William H. “A Theory of the Firm: Managerial
Behavior, Agency Costs, and Ownership Structure”. Journal of Financial Economics.
Vol. 3, No. 4 (Oct., 1976) pp. 305-360.
Johnson, Simon and Todd Mitton. “Cronyism and Capital Controls: Evidence from
Malaysia”. Journal of Finance, (Jan., 2002).
Kaufman, Daniel, Kraay, Aart, and Zoido-Labaton, Pablo, “Governance Matters”, World
Bank Policy Research Working Paper, (1999) No. 2196.
Keefer, Philip, “When Do Special Interests Run Rampant? Disentangling the Role in
Banking Crises Elections, Incomplete Information, and Checks and Balances,” World
Bank Research Working Paper 2543.
Keefer, Philip and Stasavage, David. "The Limits of Delegation: Veto Players, Central
Bank Independence and the Credibility of Monetary Policy." American Political Science
Review (August 2003)
Keefer, Philip and Stasavage, David. “Bureaucratic Delegation and Political Institions:
When Are Independent Central Banks Irrelevant?” (June 2001) World Bank Policy
Working Paper 2356.
Korajczyk, R. and Levy, A. “Capital Structure Choice: Macroeconomic Conditions and
financial constraints”, Journal of Financial Economics Vol. 68, Iss. 1, (2003) pp. 75-109.
156
Kroszner, Randall, “On the Political Economy of Banking and Financial Regulatory
Reform in Emerging Markets”. Working Paper, (Apr., 1998).
Krishnan, Sharma. “The Underlying Constraints of Corporate Bond Market Development
in South East Asia”. DESA Discussion Paper, No.14.
Laeven, L. and Woodruff, C. “The Quality of the Legal System, Firm Ownership, and
Firm Size”, World Bank Working Paper.
La Porta, Raphael, Lopez-de-Silanes, Florencio, Pop_ELeches, Cristian, and Shleifer,
Andrei. “The Guarantees if Freedom”, NBER Working Paper 8157, (Feb. 2002).
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, and Vishny, Robert W.
“Law and Finance”. The Journal of Political Economy, Vol. 106, No. 6. (Dec., 1998), pp.
1113-1155.
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, “Corporate Governance
Around the World”. Journal of Finance, Vol. 54, No.2 April (1999) pp.471-518.
La Porta, Raphael, Lopez-de-Silanes, Florencio, and Shleifer, Andrei “Government
Ownership of Banks”. Journal of Finance. Vol 57, Issue 1 (Jan. 2002).
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, and Vishny, Robert W.
“Investor Protection and Corporate Valuation, Journal of Finance, Vol. 57, No. 3, June
(2002), pp. 1147-1170.
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, and Vishny, Robert W.
“Investor Protection and Corporate Valuation”. Journal of Finance. Vol. LVII, No. 3
(Jun., 2002).
La Porta, Rafael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, & Vishny, Robert.
(1999). The Quality of Government. Journal of Law, Economics, & Organization, 15,
222-279.
Lau, Laurence J., Qian, Yingyi, Gerard Roland. “Reform Without Losers: An
Interpretation of China’s Dual Track Approach to Transition”, Journal of Political
Economy.
Levine, Ross. “Financial Development and Economic Growth: Views and Agenda”.
Journal of Economic Literature, Vol. 35, No. 2. (Jun., 1997), pp. 688-726.
Lederman, Daniel, Loayza, Norman, and Soares, Rodrigo. “Accountability and
Corruption: Political Institutions Matter”. Economics and Politics, Vol. 7, Iss.1, (March,
2005), pp. 1-35.
157
Levine, Ross. “The Legal Environment, Banks, and Long Run Growth”, Journal of
Money, Credit, and Banking, Vol. 30, No.3, Part 2: Comparative Financial Systems
(Aug. 1998), pp. 596-613.
Levine Ross, and Zervos, Sara. “Stock Markets, Banks, and Economic Growth”,
American Economic Review, Vol. 88, Iss. 3, (Jun., 1998), pp. 537-558.
Levine, Ross and Demiriguc-Kunt, Asli. Financial Structure and Economic Growth:
Across Country Comparison of Banks, Markets and Development. MIT, Cambridge
(2001).
Lukauskas, Arvid J. “The Political economy of Financial Restriction: The Case of
Spain”. Comparative Politics. Vol. 27, Iss. 1 (Oct 194), pp. 67-89.
Matsuyama, Kiminori. “Endogeneous Inequality”. Review of Economic Studies, Vol. 67,
pp. 743-759 (2000).
Meyer, Stuart C. “The Capital Structure Puzzle”. The Journal of Finance, Vol. 39, Iss. 3,
(Jul. 1984), pp. 575-592.
Morck, Randall, Yeung, Bernard, and Yu, Wayne. “The Information Content of Stock
Markets: Why do Emerging Markets Have Synchronous Stock Price Movement”.
Journal of Financial Economics, 58, (2000), pp. 215 -260.
Moser, Peter. “Checks and Balances, and the Supply of Central Bank Independence”,
European Economic Review, 43,(1999), 1569-1593.
Nejadmalayeri, A. “Interest Rates, Taxes, Bankruptcy and Financing Decisions”,
Working Paper, University of Nevada.
Olson, Mancur, 1965. The Logic of Collective Action: Public Goods and the Theory of
Groups. Harvard University Press: Cambridge, MA.
Pagano, Marco and Volpin, Paolo F. “The Political Economy of Finance”. CEPR
Discussion Paper, No. 3231 (Feb., 2002).
Perotti, Enrico and Volpin, Paolo. “Lobbying on Entry”. Working Paper, (May 2004).
Persson, Torsten. “Do Political Insitutions Shape Economic Policy?”, Econometrica,
Vol. 70, No. 3 (May 2002), pp. 883-905.
Rajan, Raghuran G. , and Zingales, Luigi. ,Financial Dependence and Growth”. American
Economic Review, Vol. 88, Iss. 3, (Jun., 1998), pp. 559-586.
158
Rajan, Raghuran G. , and Zingales, Luigi. “The Great Reversals: The Political Economy
of Financial Development in the 20
th
Century”. (Jul. 2002) Journal of Financial
Economics, forthcoming.
Rojas_Suarez, Liliana and Wiesbrod, Steven. “Building Stability in Latin American
Financial Markets”, IADB Workin Paper Series 320, (1996).
Roderick, Dani. “Understanding Economic Policy Reform”. Journal of Economic
Literature.Vol. 34, Iss. 1, (Mar., 1996), pp. 9-41.
Rodrik, Dani, Subramanian, Arvind, and Trebbi, Francesco. “Institutions Rule, The
Primacy of Institutions Over Geography and Integration in Economic Development”,
Journal of Economic Growth, Vol. 9, No. 2 (Jun. 2004).
Shleifer, Andrei, and Vishny, Robert W. “A Survey of Corporate Governance”. The
Journal of Finance, Vol. 52, No. 2. (Jun., 1997), pp. 737-783.
Shleifer, Andrei, and Vishny, Robert W. “Large Shareholders and Corporate Control”.
Journal of Political Economy, Vol. 93, Iss. 3. Part 1, (Jun., 1986) pp. 462-488.
Shleifer, A.; Wolfenzon, D. “Investor Protection and Equity Markets”. Journal of
Financial Economics (2002), vol. 66, no. 1, pp. 3-27.
Singh, Ajit. “Corporate Financial Patterns in Industrializing Countries: A Comparative
International Study”, IFC Technical Paper No. 2 (Apr., 1995).
Stultz, Rene and Williamson, Rohan. “Culture, Openness, and Finance” NBER Working
Paper No. 8222, (April 2001)
Tsebelis, George. “Decision Making in Political Systems: Veto Players in
Presidentialism, Parliametarism, Multicameralism and Multipartyism, British Journal of
Political Science, Vol. 25, No. 3. (Jul., 1995), pp. 289-325.
Tsebelis, George. (1999). Veto players and law production in parliamentary democracies:
An Empirical analysis. American Political Science Review, 93(3), 591-608.
Tsebelis, George. (2002).Veto Players: How Political Institutions Work. Princeton, NJ:
Princeton University Press.
Tirole, Jean. “Corporate Governance”, Econometrica, Vol. 60, Iss. 1 (2001), pp. 1-35.
Titman, Sheridan and Roberto Wessels. “The Determinants of Capital Structure Choice”.
Journal of Finance, Vol. 43, Iss. 1 (Mar., 1998) pp. 1-19.
159
Tybout, James R. “Manufacturing Firms in Developing Countries: How Well Do They
Do and Why?”. Journal of Economic Literature. Vol. XXXVIII, (Mar., 2000), pp. 11–
44.
Wurgler, Jeffrey. “Financial Markets and the Allocation of Capital”. Journal of Financial
Economics, Vol. 58, (2000), pp.1
doc_390815765.pdf
A development finance institution (DFI) is an alternative financial institution which includes microfinance institutions, community development financial institution and revolving loan funds.
ABSTRACT
Title: AN INTEREST GROUP THEORY OF
FINANCIAL DEVELOPMENT
Nela N. Thomas Richardson, Ph.D., 2005
Directed By: Professor Peter Murrell, Department of
Economics
My work contributes to current explanations of the variance in financial
development across countries by considering the role of political and legal structure in
determining the effect of private interests on financial policy and illustrating the political
obstacles that policymakers face when reforming the financial system.
In Chapter 2, I present a political economy model to study the role of politics in
the process of financial development across emerging markets. The model concludes that
both special interest groups and political structure affect the level of credit market
development chosen in equilibrium. When policymakers are constrained by political
institutions that require democratic accountability, they are more likely to improve the
level of creditor rights enforcement in the financial system. Financial reform is also more
likely to occur in wealthy and highly productive economies. In contrast, the model shows
that openness to international capital inflows impedes financial development.
Furthermore elite special interest group members benefit more from financial repression
when wealth is unequally distributed; hence, income inequality provides a further
obstacle to financial reform in emerging markets.
Chapter 3 empirically investigates the role of political institutions in
implementing financial reform under three different levels of democratic accountability,
Free Countries, Partly Free Countries and Not Free Countries. I find that the institutional
details of the political system, as summarized by the number and cohesion of its veto
players – individuals whose consent is required for policy change – are weakly associated
with credit market development in Partly Free Countries.
Chapter 4 (co-authored with A. Knill) investigates the role of security laws on the
ability of firms to raise external finance by issuing capital. We find that securities laws
have disparate effects on capital issuance between small and large firms in G10 and
emerging market countries. Private enforcement of securities laws that codify existing
market arrangements is found to be a deterrent to capital issuance for small firms and
firms in emerging markets. Public enforcement of securities laws by government
regulation significantly increases the probability of issuance for emerging market firms.
AN INTEREST GROUP THEORY OF FINANCIAL DEVELOPMENT
By
Nela N. Thomas Richardson
Dissertation submitted to the Faculty of the Graduate School of the
University of Maryland, College Park, in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
2005
Advisory Committee:
Professor Peter Murrell, Chair
Professor Roger Betancourt
Professor Vojislov Maksimovic
Professor Carmen Reinhart
Associate Professor Peter Coughlin
© Copyright by
Nela N. Thomas Richardson
2005
ii
Dedication
I dedicate my dissertation to my son, Andrew Dubois Thomas Richardson, born during
the writing of this thesis. Drew, my wish for you is that your future is filled with
opportunities, abundance, music, happiness, and laughter and that you experience in your
life the joy that you have brought to mine.
iii
Acknowledgements
I am grateful to my advisor Peter Murrell for spending hours poring over my
theoretical model, for his expeditious and detailed feedback, and his guidance and
support. I also benefited greatly from the comments provided by Roger Betancourt,
Vojislov Maksimovic, David Smith, and John Rust. I thank seminar participants at the
University of Maryland, Syracuse University, the Financial Management Association
2003 Doctoral Workshop, and the Federal Reserve Board International Finance Division
for helpful comments. I also recognize the National Science Foundation for financially
supporting the first three years of my graduate education.
I wish to express my eternal gratitude and appreciation for my mother and role
model in all things, Mary Thomas, for instilling in me the tools necessary to complete
such an arduous task as a dissertation and encouraging me throughout this process.
It has been a long journey. My partner through all of it has been my husband and
soul mate, Christopher Richardson. I am profoundly thankful for his constant and
unwavering support, advice, love, encouragement, and most of all, patience.
iv
Table of Contents
ABSTRACT......................................................................................................................... i
Dedication........................................................................................................................... ii
Acknowledgements............................................................................................................ iii
Table of Contents............................................................................................................... iv
List of Tables ..................................................................................................................... vi
List of Figures ................................................................................................................... vii
Chapter 1: Introduction...................................................................................................... 1
Chapter 2: An Interest Group Theory of Credit Market Development.............................. 7
1. Introduction............................................................................................................. 7
2. A Simple Model of Financial Development ......................................................... 11
2.1 Setup ............................................................................................................. 11
2.2 Perfect Credit Markets: Case of 1 = ? .......................................................... 16
2.3 Complete Financial Repression: Case of 0 = ? .......................................... 19
2.4 Imperfect Markets: Case of ) 5 . 0 , 0 ( ? ? ....................................................... 19
3. Welfare.................................................................................................................. 25
3.1 Politically Organized Unconstrained Firms (The SIG) ................................ 26
3.2 Politically Unorganized Unconstrained Entrepreneurs................................. 28
3.3 Constrained Entrepreneurs............................................................................ 29
3.4 Poor Lenders ................................................................................................. 30
3.5 Aggregate Welfare........................................................................................ 31
4. Political Utility...................................................................................................... 32
5. Political Equilibrium............................................................................................. 37
6. Comparative Statics Under Imperfect Credit Markets.......................................... 42
6.1 Change in Productivity ................................................................................. 42
6.2 Percent Change in Wealth Per Capita........................................................... 44
6.3 Change in Financial Openness...................................................................... 45
6.4 Percent Change in Wealth Inequality ........................................................... 48
7. Concluding Remarks............................................................................................. 51
Chapter 3: An Empirical Analysis of Political Institutions and Credit Market
Development ..................................................................................................................... 55
1. Introduction........................................................................................................... 55
2. Veto Players under Different levels of Political Accountability .......................... 57
3. Data....................................................................................................................... 61
4. Empirical Strategy ................................................................................................ 66
5. Results................................................................................................................... 69
5A. Strategy 1: Clustering Standard Errors by Country..................................... 69
5B. Strategy 2: The Effect of Liberalization ...................................................... 71
5C. Strategy 3: Averaging Data in 5-year intervals............................................ 72
5D. Strategy 4: Ownership as the Dependent Variable ....................................... 73
6. Conclusion ............................................................................................................ 74
Chapter 4: Securities Laws, Firm Size, and Capital Issuance........................................ 102
1. Introduction......................................................................................................... 102
v
2. Methodology....................................................................................................... 105
3. Data..................................................................................................................... 109
4. Microeconomic Testing Strategy of LLS (2004)................................................ 116
5. Results................................................................................................................. 118
6. Robustness Check............................................................................................... 125
7. Conclusions......................................................................................................... 126
Appendices...................................................................................................................... 141
Appendix A: Derivatives of the Constrained Capital Equation................................. 141
Appendix B: Derivatives of the Firm Profit Equations with respect to .................. 142
Appendix C: Description of Variables, Chapter 3..................................................... 144
Appendix D: Description of Variable, Chapter 4 ...................................................... 146
Appendix E: Regression of Capital Issuance on Individual Sub Indexes of Public and
Private Enforcement.................................................................................................... 151
Bibliography ................................................................................................................... 152
vi
List of Tables
Table 2.1: Summary of Variables .................................................................................... 21
Table 2.2: The Parametric Model .................................................................................... 23
Table 2.3: Parameter Ranges ........................................................................................... 42
Table 3.1A : Free Country and Years .............................................................................. 76
Table 3.1B : Partially Free Country and Years................................................................. 77
Table 3.1C : Not Free Country and Years ........................................................................ 79
Table 3.2: Summary Statistics ......................................................................................... 80
Table 3.3: OLS Regression of the Basic and Openness Equations................................... 82
Table 3.4: OLS Regression of the Basic and Openness equations and Legal Origin....... 83
Table 3.5: IV Regression of the Basic Equation............................................................... 84
Table 3.6: IV Regression of the Openness Equation ........................................................ 85
Table 3.7: OLS Regression of the Basic and Openness Equations and Liberalization .... 87
Table 3.8: OLS Regressions of the Basic and Openness Equations, Legal Origin, and
Liberalization............................................................................................................ 88
Table 3.9: IV Regression of the Basic Equation and Liberalization................................. 89
Table 3.10: IV Regression of the Openness Equation and Liberalization........................ 91
Table 3.11: OLS Regression of the Basic and Openness Equations................................. 94
Table 3.12: OLS Regression of the Basic and Openness Equations and Legal Origin .... 95
Table 3.13: IV Regression of the Basic Equation............................................................. 96
Table 3.14: IV Regression of the Openness Equation ...................................................... 97
Table 3.15: OLS Estimation of Ownership...................................................................... 99
Table 3.16: OLS Estimation of Ownership and Legal Origin ....................................... 100
Table 3.17: IV Estimation of Ownership....................................................................... 101
Table 4.1: Security Issuance by Country....................................................................... 128
Table 4.2: Summary Statistics of Firm Specific Variables............................................ 130
Table 4.3: Summary Statistics of Macroeconomic Variables......................................... 130
Table 4.4: Firm Specific Correlations (with Capital Issuance Indicator) ...................... 131
Table 4.5 Macroeconomic Variable Correlations (with Capital Issuance Indicator) ..... 131
Table 4.6A: All Firms .................................................................................................... 132
Table 4.6B Small Firms .................................................................................................. 133
Table 4.6C: Large Firms ................................................................................................. 134
Table 4.7A: G10 Small Firms......................................................................................... 135
Table 4.7B G10 Large Firms .......................................................................................... 136
Table 4.8A: Small Emerging Market Firms .................................................................. 137
Table 4.8B: Large Emerging Market Firms.................................................................... 138
Table 4.9A: All Firms – Standard Errors Clustered by Country ................................... 139
Table 4.9B: All Firm Groupings – Standard Errors Clustered by Country ................... 140
vii
List of Figures
Figure 2.1: Graph of endogenously determined interest rates as a function of Creditor
Rights Enforcement. ................................................................................................. 24
Figure 2.2: Optimal Capital Investment for Decreasing Returns to Scale Production
Function as a Function of Creditor Rights Enforcement. ......................................... 25
Figure 2.3: Entrepreneurs by wealth endowment ............................................................ 26
Figure 2.4: Aggregate Welfare of the Special Interest Group as a Function of Creditor
Rights Enforcement .................................................................................................. 28
Figure 2.5: Social Welfare as a Function of Creditor Rights Enforcement ..................... 32
Figure 2.6: Sequence of events and decisions ................................................................. 37
Figure 2.7: Political Utility .............................................................................................. 40
Figure 2.8: Plot of points that make the policymaker just indifferent between high and
low financial development for different levels of productivity ranging from g = 4 to
g=5. ........................................................................................................................... 43
Figure 2.9: Plot of points that make the policymaker just indifferent between high and
low financial development for different levels of aggregate initial wealth as wealth
per capita varies. ....................................................................................................... 44
Figure 2.10: Plot of points that make the policymaker just indifferent between high and
low Financial Development for different degrees of capital openness..................... 46
Figure 2.11: Plot of points that make the policymaker just indifferent between high and
low financial development for different levels as wealth inequality varies............ 49
Figure 2.12: Plot of SIG welfare under different degrees if wealth inequality under
financial repression (=.1) and financial development (=.49) ................................ 50
1
Chapter 1: Introduction
The purpose of this dissertation is to examine the institutions that underpin
financial systems. Political and legal institutions function in tandem to shape financial
policy, in part, by determining the ability of those who benefit from financial repression
to block development. I find that legal institutions that fail to protect private property
rights generate a dichotomy between entrenched corporate interests and aggregate
welfare. For this reason, politicians are influenced by powerful, elite special interest
groups when forming financial policy.
My thesis contributes to current explanations of
the variance in financial development across countries by analyzing the effect of private
interests on financial policy and examining the institutional obstacles that policymakers
face when reforming the financial system.
The context of this dissertation is embedded in a large literature examining the
role of finance on growth. In a recent revitalization of this research, Levine (1997)
demonstrates the critical causal relationship between financial structure and economic
development by explaining the purpose of financial systems within society, how they
operate, and the mechanisms by which they affect and are affected by economic growth.
Shleifer and Vishny (1997) produce a second watershed paper under the finance and
development umbrella. Their article examines the different ways economies deal with
the problem of corporate governance: the set of laws and institutions created to ensure
that firms share their profits with the suppliers of capital. The authors focus on two
differentiating features of corporate governance systems across countries: ownership
concentration and the degree of legal protection of investors. Shareholder rights
determine the competency of the financial system to allocate society’s resources
2
efficiently. High ownership concentration is a market response to the agency costs that
plague financial systems with weak corporate governance institutions. A juxtaposition of
Shleifer and Vishny (1997) and Levine (1997) reveals that the effectiveness of a financial
system to generate economic growth is dependent on the institutional structure in which
the system is embedded.
La Porta et al. (1998) offers a rationale for the success of financial systems in
promoting growth known as the “law and finance” view of financial development. The
authors explore the contribution of a country’s legal origin in the formation of its
financial structure and its corporate governance institutions, finding that legal origin —
be it English common law, or French, German or Scandinavian civil law — partly
determines the quality of investor protection and the size of the stock market versus the
banking sector. The paper concludes that English common law systems generally have
the strongest investor protection enforcement, followed by Germany, Scandinavian, and
lastly, French civil systems. Beck et al. (2001) support these results, finding that legal
origin has a considerable influence on access to bank credit.
Though the law and finance view is the leading explanation for the variance in the
proficiency of financial systems across countries, the literature also recognizes a
relationship between political institutions and financial system development. Rajan and
Zingales (2002) analyze the importance of interest groups as opposed to legal origin or
culture in influencing financial development. They propose and test a theory that firms
are more willing to support financial liberalization in times of trade openness and
increased international competition. Firms are more resistant to financial development
when the economy is relatively more closed. The authors explain La Porta et al.’s (1998)
3
results by suggesting that in Civil law countries, like France, it is easier for governments
to implement new policies when swayed by interest groups to do so.
Biasis and Mariotti (2003) provide a theoretical model that implements the story
told by Rajan and Zingales. The authors show that soft bankruptcy laws, which are
indicative of low levels of financial development, may actually increase social welfare by
reducing the potential for inefficient liquidation caused by imperfect credit markets.
However, imperfect sanctions against default impose a collateral requirement that
prevents poor agents from accessing the credit market, thus, the gains to rich
entrepreneurs are bought at the expense of the poor. Within their model, the authors
point out that the “[a]gents with different initial resources typically have different
preferences towards the bankruptcy law. Hence different laws can be chosen in different
countries, reflecting the political influence of the different social classes, and possibly at
odds with social welfare.”
Other research in the political economy aspects of financial development include
Pagano and Volpin (2002), who survey the literature on corporate governance structures
by examining the ability of political economy methodology to analyze the economic
regulations and financial institutions that result from the balance of power between the
constituents of society. The main insights of the political economy approach is that it
explains international differences in financial policy by describing “which constituencies
are assuming a certain regulatory outcome, why they are currently dictating the rules, and
how and why the balance of power can shift against them.” From the political science
literature Haggard et al. (1993) give a detailed multi-country case study of the influence
of political institutions on financial repression and liberalization. The authors identify
4
several key factors that affect financial structure development, including macro instability
as a political liability, the existence of conservative central banks, new government
installation, and the presence of countervailing interest groups.
Other researchers, for example Denizer et al. (1998), have shown that through a
combination of powerful elites and low inter-party competition, interest groups have been
able to instigate and perpetuate implicit subsidies through government allocation of credit
to favored firms and industries. Though research on the politics and economics of
financial development is scarce, several papers suggest that the effects of political
economy on financial development deserve further investigation.
1
In addition, empirical studies, like those undertaken by Singh (1995) and Glen
and Pinto (1994), underscore the importance of policy reform on firms’ financial
behavior. Singh credits the intense intervention in financial markets by developing-
country governments for the tendency of large firms to rely primarily on new equity
issues to finance investments instead of the internal revenues or bank loans predicted by
theory. The experience of firms in developing countries – notoriously lacking in legal
rules, corporate governance institutions and legal enforcement – reveals that there must
be influences other than legal structure on the size of the stock market and the degree of
financial development. The paper by Glen and Pinto offers a qualitative framework to
analyze how the issues of financing cost, riskiness, disclosure and fear of loss of control
affect the firm’s capital structure decision. The authors connect government policies —
including capital controls, tax incentives and interest rate ceilings — to financing
decisions made by firms.
1
See Hellman et al. (1998); Hellman and Shankermann (2000); Krozner (1998, 1999); Faccio (2002);
Morck et al. (1998); Gordan and Lei (1999); Johnson and Titman (2002); and Laffont and Tirole (1991).
5
Booth et al. (1999) highlight differences in financing behavior between emerging
market and industrialized firms. The researchers find that firms in developing countries
are affected by the same sort of financial variables when making their capital structure
decisions as firms in industrialized countries. The crucial difference is that knowing the
country of an emerging market firm’s origin is as least as important in explaining its
capital decisions as the firm’s characteristics. This observation provides further
evidence, albeit indirect, that a country’s institutional factors, such as political
institutions, go a long way in accounting for firms’ financial decision-making.
Chapters 2 and 3 of my thesis explore the role of institutions in developing credit
markets. Credit market development is particularly critical in emerging markets because
firms receive the lion’s share of their investments funds from banks.
2
Chapter 2 presents
a general equilibrium model that analyzes the effect of political institutions on financial
intermediary development in an emerging market economy. The model finds that the
amount of political accountability imposed by the country’s political system affects the
level of credit market development that a policymaker chooses in equilibrium.
Chapter 3 explores whether there exists a connection between the institutional
details of the political system and intermediary development. I develop four empirical
strategies to analyze this relationship and determine the sensitivity of the influence of
political institutions on credit access to different legal regimes and levels of financial
openness. Though Chapter 2 identifies the importance of democratic accountability in
increasing accesses to credit, Chapter 3 finds that the detailed characteristics of the
2
For example, Rojas-Suarez and Wiesbrod (1996) study capital market development in seven Latin
American countries. The authors find that only in Chile is equity a substantial source of finance. For other
countries, firms are almost entirely dependent on bank loans to make investments.
6
political system play a weak role in credit development once political constraints on the
executive have been taken into account.
While most corporations receive much of their financing through bank loans,
beginning in the 1990’s large corporations in developing countries have relied heavily on
new issues to fund investments. Chapter 4 (coauthored with A. Knill) investigates the
role of securities laws on the ability of firms to raise external funds by issuing capital.
We find that securities laws have disparate effects on capital issuance between small and
large firms in G10 and emerging market countries.
7
Chapter 2: An Interest Group Theory of Credit Market Development
1. Introduction
Financial development, the capacity of the financial system to efficiently allocate
capital from saver to investor, enhances a country’s ability to generate economic growth.
3
Complicating this relationship between financial development and growth is the
acknowledgement that some economic agents benefit from dysfunctional financial
systems. The disparate effects of financial development on the economic prosperity of
different segments of society are part of the explanation for why financial systems vary
dramatically across countries.
The finance and growth literature has uncovered the role of economic institutions
in determining a country’s level of financial development and made the important
connection between legal institutions, property rights, and the enforcement of financial
contracts.
4
However, the role of institutions in financial development is not limited to
economic functions; political institutions affect the ability of those who benefit from
repression of financial markets to persuade their governments to restrict financial
development. In order to understand completely the role of institutions in explaining
differences in financial structures across countries it is necessary to consider both
channels of institutional influence.
3
See also Levine and King (1993), Levine (1997) Levine and Zervos (1998) for landmark studies on
financial development and growth.
4
La Porta et al. (1998) is the first to offer a rationale for the success of financial systems in promoting
growth known as the “law and finance” view of financial development. The authors explore the
contribution of a country’s legal origin – be it English Common Law or French and German Civil Law – in
the formation of its financial structure and its corporate governance institutions, finding that legal origin
partly determines the quality of investor protection.
8
The recent history of emerging markets illustrates the political obstacles faced by
governments that attempt to develop their financial system. In these countries, the
decades of the 1980s and 1990s were characterized, largely, by reform of the banking
system. With varying degrees of success, reforms focused on liberalizing financial prices
(interest rates) and reducing government-directed credit programs to favored corporations
and industries. Several studies present anecdotal evidence of how high capital costs,
increased firm competition, and reductions in government subsidies strengthen large
corporations’ political resistance to financial development.
5
To examine the role political institutions in achieving financial development in
emerging economies, I present a model in which a single policymaker sets financial
policy by determining the level of creditor rights enforcement. The enforcement of
creditor rights is a natural focus for examining government financial policy due to the
importance of banking regulation in emerging markets. Firms in emerging markets are
much more limited in their capacity to raise external finance than firms in developed
countries
6
. They do not have access to the range of securities and financial options
available in industrialized nations. Therefore, the ability of the banking system to finance
projects efficiently is even more crucial. Bankruptcy laws that protect creditors allow
banks to lend to small and medium enterprises reducing the incidence of credit rationing
to larger firms.
In my model, stronger creditor protection increases the availability of credit to
entrepreneurs, which raises the demand for capital and increases the equilibrium interest
5
Most notably Rajan and Zingales (2003) and Singh (1995).
6
Additionally, the government is often the main borrower in the economy, providing formidable
competition to firms of all sizes for investment funds.
9
rate. Lenders benefit from financial development because of the increase in the lending
rate, while poor and middle-income entrepreneurs benefit because they are able to
produce when they otherwise could not in a financially repressed economy. Wealthy
agents, on the other hand, are harmed by financial development because they suffer an
increase in the cost of capital. These agents have the narrowly focused common
objective required (Olson 1965) to organize into a political interest group and block the
reform of the financial sector. The policymaker values aggregate welfare as well as the
campaign contributions that she receives from the interest group of wealthy agents but is
constrained in the amount of aggregate welfare that she can forgo in favor of campaign
contributions by the economy’s political institutions. Political institutions determine how
accountable the policymaker must be to the general electorate when evaluating her tastes
for campaign contributions aimed at maintaining financial repression. In equilibrium, the
level of financial development the policymaker chooses is dependent on both interest
group influence and the level of democratic accountability imposed on governments by
the country’s political institutions.
7
By using numerical simulation techniques, my analysis sheds light on the political
barriers to developing the financial system. The central conclusion of the theoretical
model is that political institutions that impose more accountability on policymakers
achieve higher levels of financial development. While special interest groups influence
financial policy, a country’s political structure determines the potency of elite influence
7
In a related work, Biasis and Mariotti (2003) model bankruptcy reform by assuming the government acts
like a benevolent social planner who always maximizes aggregate welfare irrespective of institutional
restraints. In contrast, I model the government more realistically as I model a self-interested actor who is
influenced by special interests groups and constrained by the political institutions that structure the
economy.
10
on reform. Determinants of accountability such as competitive elections, independent
judicial systems, and low corruption ensure that policymakers must value the well-being
of the citizenry when making policy decisions. Hence, politicians are unwilling to
capitalize on campaign contributions by withholding financial reforms that increase
social welfare.
Comparative static analysis of the model allows me to examine the critical
relationship between the economic environment, political accountability, and financial
reform. This relationship has not been significantly explored by any other theoretical or
empirical work on the determinants of financial development. The model validates the
intuition that financial reform is more easily attained in wealthier industrialized
economies. In contrast, it predicts that financial openness to capital inflows, e.g. foreign
direct investment, aggravates financial reform. I show that capital openness negatively
affects financial development by lowering the cost of capital and reducing the ability of
creditor rights protection to increase aggregate social welfare. Finally, I demonstrate that
special interest group members benefit more from financial repression when there is
relatively more wealth inequality in the economy. Thus, wealth inequality decreases the
likelihood that the policymaker will choose to develop the financial system.
The paper is organized as follows. Section 2 presents the set up of the simple
financial system and introduces three special cases of credit markets: perfect markets,
complete financial repression and imperfect credit markets. Section 3 analyzes the effect
of financial development on the welfare of the special interest group, politically
disorganized unconstrained and constrained firms and poor lenders. Section 4
summarizes the Grossman and Helpman model of interest group influence and provides
11
an application to political institutions. Sections 5 and 6 discuss the political equilibrium
and present five results that follow from the model. Section 7 concludes.
2. A Simple Model of Financial Development
2.1 Setup
The set up of the financial system is loosely based on Matsuyama (2000). I make
several crucial modifications to Matsuyama’s original framework in order to capture the
process of political reform in the credit system of an emerging economy. Furthermore,
my model assumes that agents face a decreasing returns to scale production technology in
contrast to the constant returns function in Matsuyama’s paper.
There exists a small closed economy populated with a continuum of agents.
Agents live for two periods. In the first period, individuals make their investment
decisions and in the second period, they consume final wealth. Additionally, there are
two sources of heterogeneity across agents: they are endowed with different amounts of
initial wealth, w, and some have access to an investment project. At the beginning of
period one, each individual receives a wealth endowment, 0 > w .
To focus this theoretical investigation on a financial system in a poor economy I
make two additional assumptions. The first poor country assumption introduces
significant wealth inequality into the economy by asserting that there are times as many
agents in the bottom half of the domain of the wealth distribution function (the poor
cohort) than at the top half the wealth distribution (rich cohort). Secondly, I assume that
every agent in the top half of the wealth distribution has access to an investment project
with a gross rate of return equal to, g . However, only a
?
1
of agents in the bottom
12
domain have access to investment projects. I make this assumption in order to avoid
results that depend on all the entrepreneurs being rich. Let N be the number of agents
with wealth, w. Hence, the number of agents with investment projects is equal to
}
=
M
N dw
M
N
0
1
. Agents without investment projects are called households. Agents with
investment projects are called entrepreneurs. These assumptions capture two realistic
characteristics of poor economies, wealth inequality and the social stratification of agents
into capitalists and laborers.
8
Wealth is uniformly distributed across households by
(
¸
(
¸
2
, 0 ~
M
U w . Let
) (w H denote the wealth distribution across households, where
M
w H
2
) ( ' = . Wealth is
uniformly distributed across entrepreneurs by [ ] M U w , 0 ~ . Let ) (w G denote the wealth
distribution across entrepreneurs, where
M
w G
1
) ( ' = .
The expression ) 3 (
8
) ( ' ) ( '
2
) 1 (
2 /
0 0
?
?
+ = +
?
} }
M
N dw w wG N dw w wH N
M M
gives the total
number of agents with wealth equal to w.
A critical characteristic of the project’s production function is that there exists a
minimum scale of investment, 1 ? k . The minimum investment scale implies that
entrepreneurs with initial wealth less than one must borrow from the economy’s credit
market in order to set up their projects. Entrepreneurs with wealth greater than one self-
8
See Biasis and Mariotti (2003).
13
finance their projects and lend their remaining endowment to other agents. The
production function is given by Equation (2.1)
9
)
`
¹
¹
´
¦
?
<
=
, 1
1 0
) (
2 / 1
k if gk
k if
k F . (2.1)
The production function places restrictions on the range of the parameter g.
10
As
long as g is greater than the gross interest rate, ) (? i , agents will want to invest at least
one unit of capital into their projects. The interest rate is a function of financial
development. It is endogenously determined by the credit market equilibrium condition
that aggregate demand for investment is equal to the aggregate supply of capital.
The credit market is characterized by a market imperfection. The legal structure
covering debt contracts (bankruptcy laws) is too weak to ensure that creditors will receive
the full amount owed to them if the borrower defaults. Let B be the loan amount. B is
equivalent to the capital the agent chooses to invest minus his wealth endowment,
w k B ? = . As described in Equation (2.2) the borrower will only repay his debt
obligation if the cost of repayment is less than the fraction of output lost by defaulting.
] 1 , 0 [ , ) ( ) (
2 / 1
? ? ? = ? ? ? gk w k i B i . (2.2)
9
The decreasing returns production function ensures that the policymaker faces a tradeoff between social
welfare and political rents when choosing financial policy. Under a CRS production technology an
increase in the welfare of poor agents will be offset by a corresponding decrease in the welfare of wealthy
agents, therefore aggregate social welfare will remain unchanged when financial policy is improved. With
DRS technology, aggregate social welfare is unambiguously increased when the policymaker chooses a
financial policy that increases access to credit to poorer entrepreneurs. This production function has two
important implications on the agent’s investment decisions. It ensures that the optimal capital level of
capital agents choose is finite rather than infinite as in the CRS case. DRS technology also implies that the
amount of wealth required to set up a project is increasing in the level of capital investment. The
increasing wealth requirement is an additional obstacle to poor entrepreneurs that want to invest optimally.
10
The gross return must be greater than zero for agents to invest in their projects. The model restricts
0
2 / 1
> ? k gk
for the function to be sensible. Since the minimum level of capital investment is assumed to
be 1 = k , the restriction implies
1 > g
.
14
The term is the fraction of output creditors can seize in case of default. Higher
values of correspond to stronger enforcement of creditor rights. The degree to which
the financial system protects creditor rights is the measure of financial development in
this model. Financial development and creditor rights enforcement will be used
interchangeably throughout this paper. Creditors will not lend more capital than they
recoup from the borrower in cases of default. Even though default does not occur in
equilibrium, the market imperfection constrains the amount of capital that agents can
borrow. Solving Equation (2.2) for the level of capital investment, k ,yields the
constrained level of capital, ) , ( w k
c
? for each agent.
2
2 2 2 2 2 2
) ( 2
) ( 4 ) ( 2
) , (
?
? ? ? ? ?
?
i
wi g g wi g
w k k
c
+ + +
= = when <1. (2.3)
Equation (2.3) is an increasing function of the agent’s wealth endowment, the
enforcement of creditor rights, and the productive capacity of the investment project. It is
a decreasing function of the interest rate. See Appendix A for derivatives.
Let ) (
*
? k be the optimal level of capital investment. Credit constrained agents
can only invest an amount ) ( ) , (
*
? ? k w k
c
< . Solving the inequality, ) ( ) , (
*
? ? k w k
c
< , for
initial wealth shows that the borrowing constraint will bind for agents whose wealth
endowment is less than ) ( ) 2 1 (
*
? ? k w ? < . Note that for 5 . ? ? , the constraint binds
only when initial wealth is less than zero. Since wealth is assumed nonnegative all agents
in the economy are able to borrow optimally whenever ] 1 , 5 [. ? ? . Therefore, agents are
15
indifferent to levels of financial development in this range. Without loss of generality, I
restrict the parameter space to be ] 5 ,. 0 [ ? ? .
11
Rearranging equation (2.2) shows that each borrower must use as collateral a
portion of final wealth equal to the proportion
|
|
.
|
\
|
?
k i
g
) (
1
?
?
of the investment k . The
wealth threshold required to borrow k units of capital is increasing in the scale of
investment. To invest k units of capital, an agent must have a wealth endowment at least
equal to
k i
g
k w
) (
1 ) (
?
?
? = . (2.4)
Given that the minimum amount of capital required to set up a project is 1 = k , the
threshold level of wealth required to become an entrepreneur is
) (
1 ) 1 (
?
?
i
g
w ? = . (2.5)
Less initial wealth is required to set up projects when there is a high level of financial
development and debt contracts are strongly enforced. As financial development
declines, agents are required to have a larger wealth endowment in order to borrow.
Let
F
w represent final wealth. Entrepreneurs with wealth equal to or greater than
the wealth threshold ) 1 ( w set up their projects. All other agents (households and poor
entrepreneurs) must lend their wealth endowment to the credit market. In this manner,
the level of financial development segments the society into wealthy entrepreneurs and
poor lenders.
11
Biasis and Mariotti (2003) restrict the policy space in a similar manner.
16
) 1 (
) 1 ( ) (
) ( ) , ( ) ( ) , (
) ( ) ( ) ( ) (
*
* * *
w
w w if w i
w k w if iw w k i w k g
k w w if iw k i k g
w
c c
F
¦
¦
¹
¦
¦
´
¦
<
? ? + ?
? + ?
=
?
? ? ?
? ? ?
. (2.6)
It is instructive to examine the equilibrium investment decisions, interest rates and
final aggregate wealth under the two extreme cases of financial development, perfect
credit markets and complete financial repression. I then investigate aggregate social
welfare under the intermediate case of imperfect credit markets.
2.2 Perfect Credit Markets: Case of 1 = ?
Consider, for a moment, the equilibrium that emerges under perfect capital
markets, 1 = ? . The creditors receive the full amount of repayment in cases of default and
do not limit the amount agents can borrow. Each entrepreneur chooses the level of
capital investment, k that maximizes final wealth.
1
2 / 1
? + ? = k to subject iw ik gk w
F
.
The Kuhn Tucker conditions for an optimum are
? + =
k
g
i
2
(ia)
0 ) 1 ( = ? k ? (ii)
1 ? k (iii)
where is the multiplier appended to the constraint that capital investment must be
greater than or equal to one. Equation (ia) implies that
2
2
) ( 4 ? ?
=
i
g
k . (ib)
17
The Kuhn Tucker conditions identify two solutions. The first solution occurs
when the capital constraint is nonbinding, 0 = ? . Then capital investment is optimal and
equal to
2
2
*
4i
g
k k = = .
When 0 ? ? the capital constraint is binding and 1 = k .
Lemma 1. Let 1 = ? . Then
2
1
g
i ? < in equilibrium.
Proof: Suppose 1 < i . Then lenders receive a negative return from lending, leading to
excess supply. Suppose
2
g
i > . Then final wealth is strictly decreasing in k. The return
to lending is greater than the return to borrowing for every agent, therefore there will be
an excess supply of capital in the economy.
When
2
g
i < Agents prefer to invest an amount
*
k into their projects. When
2
g
i = agents will invest exactly 1 unit of capital into their firms. The equilibrium
behavior of individual agents implies
¦
¦
)
¦
¦
`
¹
¦
¦
¹
¦
¦
´
¦
= ?
< =
+
?
}
}
} }
M
M
M M
g
i if dw
M
N
g
i if dw
M
Nk
dw
M
w
N dw
M
w
N
0
0
*
0
2 /
0
2
1
2
1
2
2
) 1 (?
. (2.8a)
The left-hand side represents aggregate credit supply. It is simply the summation
of all the initial wealth in the economy and is independent of the interest rate. The right
hand side represents credit demand. It is equal to the preferred level of investment
multiplied by the number of agents.
18
Equation (3.8) can be simplified to the expression
¦
¦
)
¦
¦
`
¹
¦
¦
¹
¦
¦
´
¦
= ?
< =
+
2
1
2
) 3 (
8
*
g
i if
g
i if k
M
? . (2.8b)
In the case where
2
g
i < we can use (2.8b) to obtain an explicit solution of the
equilibrium interest rate
) 3 (
2
*
? +
=
M
g
i
E
. (2.9)
Aggregate final wealth, when
*
E
i i = is equal to (2.10a)
N
M
g
dw
M
w
N dw
M
w
N
M
g
dw
M
N
M
M
g
dw
M
N
M
g W
M M
M M
F
2 / 1
0
2 /
0
0 0
2 / 1
) 3 (
8
2
2
) 1 (
) 3 (
2
1
4
3
) 3 (
2 1
) 3 (
8
|
.
|
\
|
+ =
|
|
.
|
\
|
+
?
+
+
|
.
|
\
|
+
?
|
.
|
\
|
+ =
} }
} }
+
?
?
?
?
?
.
When
2
g
i = , some entrepreneurs would not set up projects in equilibrium and in a
perfect credit markets it would be random who became an entrepreneur and who did not.
Additionally, in equilibrium entrepreneurs are indifferent between lending to the credit
market and setting up a firm because the return to lending is equal to the return to
entrepreneurship. Let ? be the proportion of individuals who set up firms in equilibrium.
The term is defined by,
¦
)
¦
`
¹
¦
¹
¦
´
¦
>
?
=
2 1
2
2
M if
M if
M
?
. When M>2, there is enough capital
in the economy for every agent to set up their projects.
19
Aggregate final wealth is equal to
N
M
M M g
dw
M
w
dw
M
w
N
g
M
N
g
dw
M
N g W
M M M M
F
|
|
.
|
\
|
|
.
|
\
| ?
+ + =
|
|
.
|
\
|
+
?
+ ? =
} } } }
+
1
) 3 (
8 2
2
2
) 1 (
2
1
2
1
0
2 /
0 1 1
? ?
?
? ?
. (2.10b)
2.3 Complete Financial Repression: Case of 0 = ?
When there is no protection against bankruptcy, households refuse to lend and no
credit is supplied to the market. Furthermore, the capital constraint in equation (2.3)
reduces to w w k
c
? ) , 0 ( . If their wealth endowment is above one, entrepreneurs sink all
of their initial wealth into their project. If the wealth endowment is below one, an
entrepreneur cannot set up his project and will not lend to the credit markets. All agents
that do not invest in projects simply eat their wealth endowment. Aggregate final wealth
in the completely repressed economy is
+
?
<
|
.
|
\
|
? =
=
}
F
M
F
W
N
M
M g
dw
M
w
g W
1
3
2
2 / 1
1
2 / 1
. (2.12)
As one would expect, aggregate social welfare is lower in the completely repressed
economy then in the perfect markets case
2.4 Imperfect Markets: Case of ) 5 . 0 , 0 ( ? ?
Under imperfect markets, only the wealthiest entrepreneurs can borrow optimally
at the equilibrium interest rate. Poor entrepreneurs are constrained by their wealth
20
endowment in the amount that they can borrow and subsequently invest into their
investment projects.
Lemma 2. Let 1 < ? . Then
2
) (
g
i ? ? in equilibrium.
Proof: Similar to Lemma 1, when ) (? i i = .
For algebraic simplicity and without loss of generality, I consider the interest rate
when ) (? i is strictly less than
2
g
. In this range of the interest rate, every entrepreneur
with access to a production technology would like to invest 1
) ( 4
) (
2
2
*
> =
?
?
i
g
k .
Imperfect markets exist in the model whenever ( ) 5 ,. 0 ? ? . Entrepreneurs with
initial wealth equal to ) (
) ( ) (
1 k w
k i
g
w = ? ?
? ?
?
will open a firm with investment scale
equal to ) (
*
? k . For entrepreneurs with initial wealth,
) (
1 ) (
*
*
?
?
k i
g
k w ? < the credit
constraint binds and they must limit investment to 1 ) , ( ? w k
c
? as described in Equation
(2.3). Agents that have access to a production technology but with initial wealth
) 1 ( w w < do not invest and merely lend their endowment to the credit market. Agents
without access to projects simply loan their wealth endowment to the economy’s credit
markets. The above discussion implies that in equilibrium
} } } }
+ =
|
|
.
|
\
|
+
?
M
k w
k w
w
c M M
dw
M
N k dw
M
w k
N dw
M
w
dw
M
w
N
) (
*
) (
) 1 ( 0
2 /
0
*
*
1
) (
) , ( 2
2
) 1 (
?
? ?
(2.13)
Equation (2.13) shows aggregate capital demand for constrained and unconstrained
borrowers. The equilibrium interest rate, ) (?
E
i solves
21
|
|
.
|
\
|
+ = +
} }
M
k w
k w
w
c
dw
M
k dw
M
w k
N
M
N
) (
*
) (
) 1 (
*
*
1
) (
) , (
) 3 (
8
?
?
? (2.14)
Sections 2.2 and 2.3 offer two extreme versions of financial development: perfect
credit markets and complete financial repression. Section 2.4 presents a more realistic,
and thus more compelling, example of financial development, the imperfect markets case.
However, what the model gains in realism when imposing imperfect markets it loses in
theoretical simplicity. Since theoretical results are hard to achieve in the intermediate
case it is useful to simulate a parametric version of the model. Below I use numerical
simulation techniques to solve Equation (2.14) for the equilibrium interest rate as a
function of creditor rights enforcement, . Once the equilibrium interest rate, ) (? i , is
computed I am able to analyze aggregate welfare over the range of creditor rights
enforcement, . Table 2.1 summarizes variables in the theoretical model.
Table 2.1: Summary of Variables
Variable Description
Creditor Rights Enforcement
) (? i Interest Rate
) 2 / , 0 ( ~ M U w
H
and ) 2 / , 0 ( ~ M U w
H
Initial Wealth Endowment for Households
(H) and Entrepreneurs (E)
Inequality Marker
) (
*
? k
Optimal Investment Scale
) , ( w k
c
?
Constrained Investment Scale
) (
*
k w
Optimal Wealth Threshold
) 1 ( w Minimum Wealth Threshold
) , (
sup
? w K
ply
Capital Supply
) , ( w K
Demand
?
Capital Demand
) , ( w W
F
?
Aggregate Final Welfare
) , ( w W
SIG
?
Special Interest Group (SIG) Final Wealth
22
Table 2.2 presents the parametric model. The first column of Table 2.2 identifies
the model’s theoretical restrictions. I choose exogenous initial parameter values that are
consistent with the theoretical restrictions. Creditor rights enforcement ranges from
[ ] 49 ,. 1 . ? ? . For each level of , I compute the endogenous variables listed in Column 3
by numerical simulation. The four remaining initial parameter values do not vary over
the range of imperfect markets, . The level of productivity, g , and the household and
entrepreneur wealth distributions are chosen such that 1 ) (
*
? ? k and
2
) ( 1
g
i < ? ? . A
wide range of parameters values, g , satisfy this condition; 4 = g , also meets the criteria
that the project has a reasonable rate of return.
12
Furthermore, M is deliberately chosen
so that 2 < M . Under this restriction, there is not enough capital in the economy for
every entrepreneur to invest in his project. This restriction is a realistic feature of an
emerging market.
13
The term is chosen to allow significant wealth inequality into the
parametric model. When 3 = ? there are 3 times as many agents in the bottom of the
domain of the wealth distribution function as in the top of the domain. Furthermore,
only
3
1 1
=
?
of all agents in the bottom domain are entrepreneurs and have the opportunity
to produce. Finally, the parameter N equals the number of agents with initial wealth w
and can take be any positive number. Without loss of generality, I let N=1.
12
Over a 5-year election cycle, a project with productivity g would have a gross annual rate of return of at
least .8 (depending on the optimal level of investment
) (
*
? k
).
13
Recall that in the perfect markets case, whenever 2 < M , then 2 / g i = . Entrepreneurs receive the same
rate of return whether they invest in their projects or lend to the credit markets. Therefore, they are
indifferent between investing and lending. Under imperfect markets, the rate of return to investing is
higher than the rate of return to lending, therefore, every entrepreneur wants to invest the amount
) (
*
? k
.
However, poor entrepreneurs are faced with a wealth constraint that forces them to invest sub optimally or
not at all.
23
Table 2.2: The Parametric Model
Theoretical Restrictions Initial Exogenous
Parameter Values
Endogenous Variables
Imperfect Markets:
) 5 ,. 0 ( ? ?
[ ] 49 ,. 1 . ? ?
) (? i
Minimum Capital Scale:
1 ) (
*
? ? k
4 = g
) (
*
? k
) , ( w k
c
?
Interest Rate Range:
2
) ( 1
g
i ? ? ?
4 . 1
) , 0 ( ~
) 2 / , 0 ( ~
= M
M U w
M U w
E
H
) (
) 1 (
*
k w
w
3 = ?
) , (
sup
? w K
ply
) , ( w K
Demand
?
1 = N
) , , (
) , , (
? ?
? ?
w W
w W
SIG
F
The first task of the numerical simulation is to calculate the interest rate as a
function of creditor rights enforcement, . The equilibrium interest rate equates the
aggregate supply of credit in the economy to the aggregate demand for credit as described
in Equation 2.14. Figure 2.1 graphs the equilibrium interest rate over the range of
consistent with the imperfect markets environment. The figure shows that the interest
rate in an increasing function of creditor rights enforcement.
24
Equilibrium Interest Rate
0
0.5
1
1.5
2
2.5
0 0.1 0.2 0.3 0.4 0.5 0.6
Creditor Rights Enforcement
I
n
t
e
r
e
s
t
R
a
t
e
Figure 2.1: Graph of endogenously determined interest rates as a function of Creditor Rights
Enforcement.
Figure 2.2 graphs the optimal capital investment for each entrepreneur. Recall
that the investment choice made by unconstrained entrepreneurs is equal to
2
2
*
) ( 4 ? i
g
k = .
Unconstrained entrepreneurs demand less capital as creditor protection improves and
lending becomes more profitable relative to investment due to the decreasing returns to
scale production function. Note that as financial development approaches the perfect
markets case optimal investment converges to one as the equilibrium interest rate
converges to
2
g
i = .
25
Optimal Capital Investment
0
0.5
1
1.5
2
2.5
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
C
a
p
i
t
a
l
Figure 2.2: Optimal Capital Investment for Decreasing Returns to Scale Production Function as a
Function of Creditor Rights Enforcement.
3. Welfare
The general equilibrium model generates four implications of imperfect credit
markets and a DRS production technology on aggregate social welfare. First, only
entrepreneurs above a certain wealth threshold can set up firms. Second, middle-income
entrepreneurs are constrained to invest sub optimally. Third, it is wealth maximizing for
rich entrepreneurs to invest more than is socially optimal. Fourth, the level of creditor
rights enforcement, , determines the equilibrium interest rate by controlling who
borrows and thusly the demand for credit.
In this section, I derive the welfare functions for the five constituent groups that
compose aggregate welfare: poor lenders, constrained entrepreneurs, unconstrained
entrepreneurs and SIG members. Poor lenders are composed of households without
projects and entrepreneurs that do not have enough wealth to borrow under imperfect
credit markets. The welfare of each group is a function of financial development. Figure
2.3 shows entrepreneurs segmented by initial wealth endowment.
26
w=0 ) 1 ( w w = ) (
*
k w w = w=M(1-x) M
|___________________|___________________|_______________________|______________|
Poor Lenders Constrained Entrepreneurs Unconstrained Entrepreneurs SIG Members
Figure 2.3: Entrepreneurs by wealth endowment
3.1 Politically Organized Unconstrained Firms (The SIG)
The special interest group, SIG, is composed of the agents in the top x% of the
wealth distribution. This elite group of citizens satisfies the conditions necessary for the
organization of a successful interest group. Olson (1965) argues that in order to
organize, interest groups must be composed of a small number of agents with narrowly
and uniformly defined interests. I show that this group of agents is harmed by financial
development and has the incentives and the resources (since they are the wealthiest
agents) to block reform.
SIG welfare is composed of the total profit of members of the SIG plus their
interest income. SIG profit is given by Equation (3.1).
}
?
? = ?
M
M x
SIG
M
dw
k i gk
) 1 (
* 2 / 1 *
) ( ) ( ) ( ) ( ? ? ? ? . (3.1)
Substituting
2
2
*
) ( 4
) (
?
?
i
g
k = into (3.1) reduces the profit function to
x
i
g
M
dw
i
g
M
M x
SIG
) ( 4 ) ( 4
) (
2
) 1 (
2
? ?
? = = ?
}
?
.
The change in SIG profit as financial development increases is equal to
0 ) ( '
) ( 16
) ( '
2
2
<
?
= ? x i
i
g
SIG
?
?
? (3.2)
27
SIG profit is decreasing in the level of financial development. To find the total change in
SIG welfare as financial development increases I add the interest income that SIG
members earn from lending to other agents.
}
?
|
|
.
|
\
| ?
? + = + =
M
M x
SIG
M
M x
M
M
i x
i
g
dw
M
w
i x
i
g
W
) 1 (
2 2 2 2 2
) 1 (
) (
) ( 4
) (
) ( 4
) ( ?
?
?
?
? . (3.3)
) ( ) 2 (
2 ) ( 4
2
2
?
?
i x x
M
x
i
g
? + =
The change in elite welfare is given by
) ( ' ) 2 (
2
) ( '
16
) ( '
2
2
? ? ? i x x
M
x i
i
g
W
SIG
? +
?
= . (3.4)
The lower the level of financial development, the lower is the cost of capital. As
long as the increase in capital costs is greater than the increase in interest income brought
about by the increased financial development, the wealthy are better off when credit
markets are repressed. Figure 2.4 graphs SIG welfare as the enforcement of creditor
rights increases using the parameter restrictions defined in Table 2. For the initial
conditions 4 . 1 = M and 25 . = x , the aggregate wealth endowment for SIG members is
equal to 30625 .
4 . 1
) 25 . 1 )( 4 . 1 (
=
}
?
dw
M
w
. Final aggregate wealth of the SIG members ranges
from a high of 1.528 at 1 . = ? to a low of 1.068 at 3 . = ? . Furthermore, the figure
demonstrates that aggregate final wealth for the SIG is lower at 49 . = ? than at 1 . = ? .
Hence, these entrepreneurs prefer weak creditor rights enforcement when the financial
system is imperfect.
28
SIG Welfare
1.1
1.11
1.12
1.13
1.14
1.15
1.16
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
S
I
G
F
i
n
a
l
W
e
a
l
t
h
Figure 2.4: Aggregate Welfare of the Special Interest Group as a Function of Creditor Rights
Enforcement
3.2 Politically Unorganized Unconstrained Entrepreneurs
Not all unconstrained firms are able to organize politically, only the top x% of the
wealth distribution. These unconstrained entrepreneurs that are politically disorganized
have wealth over the threshold necessary to invest the optimal level of capital, ) (
*
k w ,
and below that required to be in the SIG. The aggregate profit of politically disorganized
unconstrained firms is given by Equation (3.5).
}
?
? = ?
) 1 (
) (
* *
*
2 / 1
1
)) ( ) ( ) ( ( ) (
x M
k w
U
dw
M
k i gk ? ? ? ? (3.5)
)) ( ) ( ) ( (
) 2 1 (
) 1 (
* *
2 / 1
? ? ?
?
k i gk
M
x ?
(
¸
(
¸
?
? ? = .
Substituting ) (
*
? k into the wealth threshold yields
) 2 1 ( ) (
*
? ? = k w . (3.6)
Substituting for ) (
*
? k and ) (
*
k w into Equation (3.5) gives
29
) ( 4
) 2 1 (
) 1 (
) ( 4 ) ( 2
) 2 1 (
) 1 ( ) (
2 2 2
?
?
? ?
?
?
i
g
M
x
i
g
i
g
M
x
U
(
¸
(
¸
?
? ? =
(
¸
(
¸
?
(
¸
(
¸
?
? ? = ?
. (3.7)
The derivative of the aggregate unconstrained firm profit with respect to financial
development is equal to
) ( '
) ( 4
) 2 1 (
) 1 (
) ( 2
) ( '
2
2 2
?
?
?
?
? i
i
g
M
x
Mi
g
U (
¸
(
¸
?
(
¸
(
¸
?
? ? + = ? . (3.8)
The first term is the increase in profits due to the increase in the number of unconstrained
firms as financial development rises. The second term corresponds to the drop in profits
caused by increase in the cost of capital.
Aggregate welfare for the unconstrained agents is equal to profits plus interest
income.
}
?
+ ? =
) 1 (
) (
*
1
) ( ) ( ) (
x M
k w
U U
dw
M
w i W ? ? ? . (3.9)
Derivative of unconstrained welfare is
) (
) 2 1 ( 2
2
) 2 1 (
2
) 1 (
) ( ' ) ( ' ) ( '
2
?
? ?
? ? ? i
M M
x M
i W
U U
?
+
|
|
.
|
\
| ?
?
?
+ ? = . (3.10)
The second term is the increase in interest income due to the increase in the interest rate
and the third term is the increase in aggregate interest income due to the increase in the
number of unconstrained firms.
3.3 Constrained Entrepreneurs
Constrained entrepreneurs are forced to invest a sub optimal level of capital into
their investment project. The level of capital investment is a function, ) , ( w k
c
? , is a
30
increasing function of their wealth endowment and financial development. Aggregate
profit for constrained entrepreneurs is
}
? = ?
) (
) 1 (
*
1
)) , ( ) ( ) , ( ( ) , (
k w
w
c c
C
dw
M
w k i w gk w ? ? ? ? . (3.11)
where ) , ( w k
c
? is defined in Equation (3.3).
Let dw
M
w k
w K
k w
w
c
C
}
=
) (
) 1 (
*
) , (
) , (
?
? be aggregate credit demand for constrained firms. Then
the change in profit for constrained entrepreneurs is
0 )) , ( ) ( ) ( ' ) , ( (
) , ( 2
) , ( '
) , ( '
2 / 1
> + ? = ? w K i i w K
w K
w gK
w
C C
C
C
C
? ? ? ?
?
?
? . (3.12)
See Appendix B for detailed derivative. The first term represents the increase in
revenue from the increase in financial development and the second term represents lost
profit due to an increase in the cost of capital. Constrained entrepreneurs welfare is
interest income plus profits:
dw
M
w
i w w W
k
w
w
C C
}
+ ? =
*
1
) ( ) , ( ) , ( ? ? ? . (3.13)
It is increasing in the level of financial development:
0 ) ( ' )) ( ) ( ' (
) (
)) ( ( 2
8 4 ) ( '
3
> ? + ?
?
+ + ? = ? ? ? ?
?
? ?
? ?
c C
gi i g
i
i g
W . (3.14)
3.4 Poor Lenders
Because they lack access to capital, the poorest entrepreneurs in the economy are
unable to set up their investment projects. These agents, along with households, lend
their wealth to the credit markets and earn
31
|
|
.
|
\
|
?
+ =
} }
2 /
0
) 1 (
0
2
2
) 1 (
) ( ) (
M w
L
dw
M
w
dw
M
w
N i W
?
? ? (3.15)
}
?
? +
|
|
.
|
\
|
+ ? ? =
2 /
0
2
'
2
2
) 1 (
) (
) (
) ( '
) (
)) 1 ( 1 ( 2 ) (
M
L
dw
M
w
i
i
i g
i
g
w W
?
?
?
? ?
?
? . (3.16)
Poor lenders experience two effects when financial development increases. Firstly, they
earn a higher return on interest income from lending. Secondly, poor entrepreneurs at the
margin of the initial wealth threshold are able set up firms because the minimum wealth
endowment required to borrow capital has decreased.
3.5 Aggregate Welfare
Aggregate welfare is the sum of all income earned from lending at the
endogenous interest rate and the profit made by the entrepreneurs:
|
.
|
\
|
? + ? + ? + + = ) , ( ) ( ) ( ) 3 (
8
) ( ) ( w
M
i N W
C U SIG
? ? ? ? ? ? .
The total increase in aggregate income that results from an increase in financial
development is
) , ( ) ( ) ( ' ) , (
) , ( 2
) , ( '
) (
) ( 4
) 2 1 (
) ( 2
) 2 1 (
) 3 (
8
) ( ) ( '
2 / 1
2
2 2 2
w K i i w K
w K
w gK
i
i
g
M
M
i
g
M
M
i W
C C
C
C
? ? ? ?
?
?
?
?
?
?
?
? ? ?
+ ? +
?
(
¸
(
¸
(
¸
(
¸
?
? ?
?
+ + ? =
. (3.17)
Figure 2.5 graphs aggregate welfare as a function of financial development.
Social welfare improves when financial development increases. Recall the aggregate
wealth endowment is equal to dw
M
w
N dw
M
w
N
M M
} }
+
?
0
2 /
0
2
2
) 1 (?
. Under the initial
parameter specification the aggregate wealth endowment is equal to 1.04. Final
32
aggregate wealth at 1 . = ? is 2.86799. Final aggregate wealth at 49 . = ? is 4.088. When
credit markets are imperfect, strong creditor enforcement increase aggregate final wealth
by more than 40% over weak enforcement.
Aggregate Social Welfare
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
A
g
g
r
e
g
a
t
e
F
i
n
a
l
W
e
a
l
t
h
Figure 2.5: Social Welfare as a Function of Creditor Rights Enforcement
In this section, I have specified the welfare functions for each of the constituent
groups. The policymaker will choose the level of financial development that maximizes a
linear function of aggregate welfare and campaign contributions. I analyze the
politician’s utility function in Section 4.
4. Political Utility
In the Grossman and Helpman (GH) model, special interest groups try to
influence the policymaker’s policy choice by giving her political gifts. The SIG’s design
contribution schedules that associate gifts to the politician with every policy option
available to her in order to maximize its own objective function. Contributions are
assumed nonnegative. Mimicking the principal in the principal agent game the SIG
33
designs a payment scheme, C(), “to give the politician the appropriate incentives to act
on its behalf” (GH 2001).
The politician cares about the total level of political contributions and aggregate
well-being. She desires contributions because they can be used to finance campaign
spending (among other benefits). Social welfare is a concern to the politician because her
constituents are more likely to reelect her if she has delivered a high standard of living.
The policymaker maximizes a utility function ) , ( c G ? where is the policy choice made
by the policymaker and c is the political contributions. Grossman and Helpman state
that, “The utility function is meant to capture the policymaker’s personal preferences
over the various possible policy outcomes, as well as her concern for her future electoral
prospects. The policy will affect the politician’s chances of being reelected if voters
look retrospectively at her record when deciding whether to vote for her in subsequent
elections”.
14
Because the government uses contributions to pursue its own political gain,
G(.) is increasing in c and is a single peaked function for any level of c. Following GH, I
assume that the policymaker’s utility function is a linear function of her concern for the
welfare of the electorate and her desire for political gifts. The weights on aggregate
welfare and campaign contributions sum up to 1. Furthermore, I use the simplest version
of the GH model by limiting the interaction to that between the political entity and a
single interest group. Thus, the policymaker’s utility is given by
) ( ) 1 ( ) ( ) , ( ? ? ? ? ? C W c G ? + = (4.1)
14
A less cynical view of the government is that it cares about the overall welfare of its constituency and
since the policy variable, p, directly affects societal well-being it is a component in the government’s
utility.
34
where ) (? W gives the aggregate welfare as a function of the policy choice. The
term ) (? C represents contributions from special interest group.
One of the attributes of the GH model is that it allows for liberal interpretation of
the parameters. The GH interpretation of the weights on utility is the policymakers’
tastes for campaign finance. In addition to this interpretation the weights could be
thought of as the degree of corruption in the economy. In more corrupt societies
politicians find it easier to take bribes from elite rent seekers that benefit from regressive
financial policies than in less corrupt systems. In addition, could represent the capacity
of the political system to reform, or otherwise improve policies that increase aggregate
social welfare. In my application of the GH model to financial development, I interpret
the weights generically to be the degree of democratic accountability that the political
system requires of the government. In societies with high accountability, policymakers
must weight general welfare heavily, where the opposite is the case politicians are free to
pursue rent-seeking behavior by courting the favors of narrowly defined special interest
groups.
My model predicts that political systems with a high degree of democratic
accountability will be more likely to improve their financial policies. In the model a
politician is able to increase aggregate welfare by reforming financial policy. She
increases SIG welfare by maintaining the status quo level of financial repression. In a
political system with accountable (unaccountable) institutions the politician gives
aggregate (SIG) welfare a weight of one. In anti democratic political systems,
policymakers value campaign contributions with a weight of one. The model predicts
that proclivity of political institutions towards accountability to the general electorate will
35
determine the politician’s concern for aggregate welfare or SIG welfare when making her
policy decisions. Thus institutional details of the political system affect the governments’
incentives to reform the credit market.
Strictly speaking the policymaker described in my interpretation of the GH model
is always self-interested, without constraint she would prefer to maximize the rents of
political office. It is the institutional constraints inherent in the economy, which
determine the political advantage of trading off aggregate welfare for contributions and
vice versa. Institutions determine whether the capital that the politician earns is political
(people are happy with her policy performance) or monetary (she earns campaign funds
and other political gifts from interest groups), the politician acts in order to ensure her
ability to be reelected. As the benevolent social planner is a useful fiction to help
describe optimal government behavior, the completely self-interested policymaker is also
beneficial in this model when discussing the policy reform that occurs in equilibrium.
The political institutions determine the avenue by which the politician acquires the
“capital” necessary to be reelected.
Grossman and Helpman define the equilibrium policy ?
~
as the policy that
maximizes the interest groups utility function ) ,
~
( c U ? subject to the constraint that
) 0 ,
ˆ
( ) , ( ? ? G c G ? . The equilibrium interest group contribution c
~
makes the politician
just indifferent between the ?
~
and the policy she would pick if no contributions were
forthcoming from the SIG, ?
ˆ
. In this model favored policies that increase aggregate
welfare may be abandoned or modified by the policymaker in order to capture the rents
of political office. The equilibrium policy that results from this principle agent game is
36
pareto efficient in the sense that one actor cannot be made better of without the other
becoming worse off.
If I assume that the SIG utility function can be written as a function of its welfare
from the government’s policy choice minus what it pays in political contributions,
c W c U
SIG
? = ) ( ) , ( ? ? , then the politician’s objective function can be rewritten as
) ( ) 1 ( ) ( ) , ( ? ? ? ? ?
SIG
W W c G ? + = , (4.2)
The policymaker chooses the policy by maximizing a linear weighted function of
aggregate welfare and SIG welfare. This model provides the basis from which to
evaluate the ability of a small closed economy to enact reform given the strength of the
special interest within the economy and the policymaker’s predilection for political gifts
determined by the economy’s political institutions.
In Section 5, the GH model is applied to the case of reform in the credit market.
Improvements in financial development increase aggregate growth in the economy.
Elite agents, those individuals in the upper tail of the wealth distribution, experience a
decrease in welfare when the level of financial development is increased. The elites have
the ability and the incentive to form a SIG that lobbies the policymaker to block financial
development. The equilibrium policy that is chosen by the policymaker is dependent on
degree of accountability imposed by the society’s political institutions, the productive
capacity of the economy, the extent of wealth inequality, and the supply of capital.
Figure 2.6 gives the model’s timeline.
37
Date t=0 Date t=1 Date t=2
|_________________________________|_______________________________|
Politician receives Politician chooses Agents make their
contribution schedule investment decisions
from the interest group and consume final wealth
Figure 2.6: Sequence of events and decisions
5. Political Equilibrium
In this section, I investigate how the SIG uses political gifts to sway the
policymaker into enacting a low level of financial development. At the beginning of
period one, the politician decides whether to reform the credit market. Based on the
government’s financial policy, citizens decide how much capital to invest in their
projects. The differing effect of credit market efficiency on the returns of borrowers and
lenders divides the politician’s constituency into the poor who benefit from better
enforcement of legal sanctions and the rich who profit from inefficient credit markets.
Recall from Section 4 that the policymaker’s objective is to maximize a political
utility function
) ( ) 1 ( ) ( ) , ( ? ? ? ? ?
SIG
c W c G ? + = (4.1)
where ) (? W is aggregate final wealth and ) (?
SIG
c is the contribution the SIG gives the
policymaker for a given policy variable, . The contribution is decreasing in financial
development because the SIG is worse off by increasing creditor rights.
The political utility function has a first order condition equal to
0 ) ( ' ) 1 ( ) ( ' = ? + ? ? ? ?
SIG
c W . (5.1)
I assume that the contribution schedule is differential in . This
implies ) ( ' ) ( ' ? ?
SIG SIG
W c = . I can substitute ) ( ' ?
SIG
W for ) ( ' ?
SIG
c in the politician’s first
38
order condition, which yields Equation (6.2) below. Therefore, Equation (5.1) can be
rewritten as
0 ) ( ' ) 1 ( ) ( ' = ? + ? ? ? ?
SIG
W W . (5.2)
Substituting the first order conditions for SIG and aggregate welfare, Equations (3.4) and
(3.17) respectively, into the policymaker’s first order condition yields (5.3) below.
( )
0 ) ( ) 2 (
2
) (
) ( 16
) 1 (
) ( ) ( ) , (
) , ( 2
) , (
) (
) ( 4
) 2 1 (
) ( 2
) 2 1 (
) 3 (
8
) (
2
2
2 / 1 2
2 2 2
=
)
`
¹
¹
´
¦
? ? + ?
?
?
+
¦
)
¦
`
¹
¦
¹
¦
´
¦
? + ?
?
+ ?
|
|
.
|
\
|
|
|
.
|
\
| ?
? ?
|
|
.
|
\
|
|
.
|
\
| ?
+ + ?
? ?
?
?
? ? ?
?
?
?
?
?
?
?
? ? ?
i x x
M
x i
i
g
i i w K
w K
w gK
i
i
g
M
M
i
g
M
M
i
C
C
C
The policymaker chooses a that satisfies the first order condition for
maximizing a weighted sum of aggregate welfare and the welfare of the SIG. The policy
choice is dependent on the exogenous parameter values -- the productivity level g,
abundance of capital, M, the aggregate wealth of the special interest group, x, and the
level of political accountability imposed by institutions, . Financial policy is dependent
on the economy’s political institutions, , which determines the politician’s ability to
maximize rents to the detriment of social welfare.
Result 1: Political institutions that impose more democratic accountability on policy
makers will generate higher levels of financial development.
Political institutions that impose democratic accountability make it difficult for
the policymaker to succumb to the interests of elites that lobby for financial repression.
When accountability is low, political utility is decreasing in financial development. For
political systems that impose democratic institutions, the political utility function is
upward sloping in financial development. Figure 2.7 graphs political utility for different
39
levels of political accountability. Low values of are indicative of political institutions
that require governments have little accountability to the general electorate. As the
weight increases the politician’s concern for general welfare also increases. Anti
democratic political institutions result in a political utility functions that is a decreasing
function of financial development. Political systems that impose more accountability in
policy choices lead to a political utility function that is increasing in financial
development. Figure 2.7 also shows that there exists a political structure in which the
policymaker is exactly indifferent between low and high levels of financial development.
In the graphs this point occurs at a political weight of = ? .032824.
The parametric model establishes the importance of political accountability in
determining the extent financial development. However, other papers have been hesitant
to confirm a relationship between political institutions that impose accountability and
financial development. Beck et al. (2001) minimize the importance of political structure
in explaining financial structure. Using measures of political environment that include
competitiveness in elections, government openness, and inter-party competition, the
authors find “a weak, fragile link between political structure and finance development”.
Instead, Beck et al. (2002) attribute the change in financial development to legal origin
and initial endowment of colonies settled by Europeans. Taking a broader view Glaser et
al. (2004) criticize empirical techniques used in previous literature to relate political
institutions and economic growth. The authors’ main argument is that countries can
improve human capital without democratic accountability and that once these economies
become richer they can improve institutions.
40
Acemoglu et al. (2004) summarize a large literature connecting political
institutions and economics growth by stating “political institutions place all political
power in the hands of a single individual or a small group, economic institutions that
provide protection of property rights and equal opportunity for the rest of the population
are difficult to sustain.”
15
My model supports the authors’ conclusion that in accountable
political systems it is more difficult for politicians to ignore the welfare of the majority in
favor of the preferences of a small number of elite. Furthermore, by rigorously
formalizing the role of political structure in financial development, I am able to
demonstrate the significance of political institutions in generating changes in financial
policy and economic outcomes. Though the investigation of political structure poses an
empirical challenge to researches, its connection to economic growth should not be
disregarded.
Figure 2.7: Political Utility
15
The new institutional economics literature supporting a relationship between political structure and
economic growth include Buchanan and Tullock (1962), North (1981, 1990), Knack and Keefer (1995),
Hall and Jones (1999), Acemoglu, Johnson and Robinson (2001, 2002). Empirical research identifying
politics and firm behavior include See Hellman et al. (1998); Hellman and Shankermann (2000); Krozner
(1998) and (1999); Faccio (2002); Denizer et al. (1998); Morck et al. (1998); Gordan and Lei (1999);
Johnson and Titman (2002); and Laffont and Tirole (1991).
41
Political Accountability alpha = .01
1.125
1.13
1.135
1.14
1.145
1.15
1.155
1.16
1.165
1.17
1.175
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .02
1.15
1.155
1.16
1.165
1.17
1.175
1.18
1.185
1.19
0 0.2 0.4 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .03
1.175
1.18
1.185
1.19
1.195
1.2
1.205
1.21
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Acountability alpha = 0.0328204
1.18
1.185
1.19
1.195
1.2
1.205
1.21
1.215
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .04
1.195
1.2
1.205
1.21
1.215
1.22
1.225
1.23
1.235
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha =.05
1.215
1.22
1.225
1.23
1.235
1.24
1.245
1.25
1.255
1.26
1.265
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .06
1.23
1.24
1.25
1.26
1.27
1.28
1.29
1.3
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
t
i
y
Political Accountability alpha = .08
1.27
1.28
1.29
1.3
1.31
1.32
1.33
1.34
1.35
1.36
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Political Accountability alpha = .1
1.3
1.32
1.34
1.36
1.38
1.4
1.42
0 0.1 0.2 0.3 0.4 0.5 0.6
Enforcement of Creditor Rights
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
42
6. Comparative Statics Under Imperfect Credit Markets
Section 6 investigates the comparative static properties of the model by exploring
the effects of four shocks to the economic environment – productivity, wealth, openness,
and inequality. To evaluate the effects of a small shock to the economy on the level of
political accountability required for reform, I choose a range of parameter values around
the initial restrictions imposed in the parametric model given in Table 2.2. The ranges of
the parameter values are constrained by the theoretical restrictions also listed in Table
2.2. Table 2.33 shows the ranges of the exogenous parameters that I evaluate. For each
incremental change in the exogenous parameter values I calculate the level of political
accountability, , that makes the politician exactly indifferent between low financial
development (=.1) and high levels of financial development ( =.49).
Table 2.3: Parameter Ranges
Comparative Static Scenarios Exogenous Parameters
Change in Productivity ] 0 . 5 , 0 . 4 [ ? g
Percent Change in Wealth Per Capita w increased by 10% to 100%
Change in Openness to Financial Inflows ] 5 . , 0 [ ? f
Percent Change in Wealth Inequality increased by 5% to 25%
6.1 Change in Productivity
Result 2: Politicians are more likely to enforce creditor rights in more productive
economies.
In this scenario, the productivity parameter, g , is increased in increments of 1 . 0
from 4.0, the initial starting value to 5.0. All other exogenous parameters are held
constant at the initial values listed in Table 2. Figure 2.8 plots accountability points
that make the politician indifferent between the lowest level of creditor rights
enforcement under imperfect markets 1 . = ? and the highest, 49 . = ? . The figure shows
43
that as the economy’s productive capacity increases, policymakers weight aggregate
welfare more than SIG welfare and are therefore more likely to increase financial
development. Consider two economies. Economy A has a high rate of productivity and
Economy B has a low rate of productivity. Given that both economies have a similar
political structure, a policymaker is more likely to reform the financial system in
Economy A then in Economy B. According to this result, the constraints imposed by the
political system on policymakers are partially offset by exogenous improvements in
productive capacity. Additionally as g increases the wealth threshold required to borrow
enough capital to set up projects fall, increasing firm entry. Aggregate welfare rises when
productivity increases, causing the politician to value social welfare more even if it is
politically easy to acquire gifts from elites. Result 2 is consistent with Hall and Jones
(1999) who find evidence that social infrastructure, including political structure, is a
significant determinant capital accumulation and productivity.
Effect of Productivity (g) on Financial
Development
0.0280
0.0290
0.0300
0.0310
0.0320
0.0330
0.0340
0 0.2 0.4 0.6 0.8 1 1.2
Increase in Productivity (g)
P
o
l
i
t
i
c
a
l
A
c
c
o
u
n
t
a
b
i
l
i
t
y
Figure 2.8: Plot of points that make the policymaker just indifferent between high and low financial
development for different levels of productivity ranging from g = 4 to g=5.
44
6.2 Percent Change in Wealth Per Capita
Result 3: The likelihood of financial reform increases as wealth per capita increases.
In the second comparative static scenario, I increase the initial wealth
endowments per capita from 10% to 100% for both households and entrepreneurs. The
increase in initial wealth increases the level of available credit in the economy. As the
supply of capital increases the equilibrium interest rate decreases for each level of
creditor rights enforcement, . The loss in welfare that SIG members suffer due to
improvements in creditor rights enforcement is partially offset by the decrease in the cost
of capital. Additionally, constrained entrepreneurs experience welfare gains because the
wealth thresholds required for firm entry and optimal investment fall. Figure 2.9 plots
the levels of political accountability, , that makes the politician indifferent between low
and high financial development as the level of wealth per capita is increased. As Figure
2.9 demonstrates, it is easier for the politician to develop credit markets in wealthier
economies where the supply of credit is abundant.
Effect of Wealth on Financial Development
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0 0.2 0.4 0.6 0.8 1 1.2
% Increase in Wealth Per Capita (w)
P
o
l
i
t
i
c
a
l
U
t
i
l
i
t
y
Figure 2.9: Plot of points that make the policymaker just indifferent between high and low financial
development for different levels of aggregate initial wealth as wealth per capita varies.
45
For very wealthy economies, the condition that 0 ? ? is violated. Politicians would have
to receive negative utility from improvements in aggregate social welfare in order to
choose a low level of financial development.
Result 3 is an illustration of path dependence in corporate ownership structures
discussed in Bebchuk and Roe (1999). The authors show that current corporate
ownership is dependent on patterns of past ownership. Furthermore, the authors argue
that special interest politics shapes corporate governance, stating “A country’s initial
pattern of corporate structures influences the power that various interest groups have in
the process producing corporate rules. If the initial pattern provides one group of players
with relatively more wealth and power, this group would have a better chance to have
corporate rules that it favors down the road.” In my theoretical model, as per capita
wealth increases, politically influential SIG members are harmed less by financial
development because the increase in available credit lowers production costs. The same
results hold in reverse. When the economy is poor and capital is scarce, the SIG suffers
more when credit markets rights are strongly enforced than in a capital abundant
economy. Result 3 suggests that special interests will be more of an obstacle to financial
reform in poor emerging markets than in industrialized nations.
6.3 Change in Financial Openness
Result 4: Capital openness has a negative effect on financial development.
So far, the simulations have been generated under the assumption of a closed
economy. I examine the effect of international capital supply on financial reform by now
assuming that capital is allowed to flow into the economy from abroad. In order to
simulate openness I increase the level of capital supply M, by adding f, where f
46
represents capital inflows and ranges from an increase of 10% to 60%. Capital supply
under an open economy is given by
} }
+
+ ? =
f M M
ply
dw
M
w
dw
M
w
K
0
2 /
0
sup
) 1 (? . Foreign
investors only supply capital to domestic agents in the model; they do not start projects.
Since the supply of capital increases while demand remains unchanged, the equilibrium
interest rate falls. I assume that the world interest rate,
w
i , is normalized to 0. As long as
the endogenous equilibrium interest rate is greater than
w
i foreigners will want to invest
capital in the domestic economy. Foreign interest income does not enter into the
policymaker’s political utility function. Her utility function is entirely composed of
domestic aggregate welfare and the welfare of the SIG.
Effect of Capital Inflow (Openness) on
Financial Development
0.032
0.033
0.034
0.035
0.036
0.037
0.038
0.039
0.04
0 0.1 0.2 0.3 0.4 0.5 0.6
Increases in Capital Inflows (f)
P
o
l
i
t
i
c
a
l
A
c
c
o
u
n
t
a
b
i
l
i
t
y
Figure 2.10: Plot of points that make the policymaker just indifferent between high and low
Financial Development for different degrees of capital openness
Other studies, namely Rajan and Zingales (2003), have suggested that capital
openness induces elite groups to change their demands for financial repression by
increasing the opportunities for international investment and growth. However, the
authors also explain that financial openness in the absence of trade openness is not
47
sufficient to change the preference of large domestic firms for financial repression. “…in
the absence of domestic or foreign competition in product markets, these [large] firms
will have little need to access external funds. Moreover, given the state of information
asymmetries across markets, it is unlikely that small domestic firms are likely to be
financed directly by foreign investors. If potential domestic entrants are unlikely to be
financed by foreigners, industrial incumbents will still retain an incentive to keep entrants
at bay by opposing financial development.” Rajan and Zingales conclude that both trade
and financial openness are necessary for elite groups to back financial development. In
my model, lower interest rates lead to higher profits for elite entrepreneurs. A politician
heavily influenced by the special interest group ( is low) may choose a lower level of
creditor right enforcement since the loss in aggregate welfare would be offset by an
increase in elite welfare. In this manner, the model suggests that increasing the capital
account has the unattended effect of decreasing access to finance for small and medium
entrepreneurs while increasing elite profits.
Trade openness provides competition in the product markets that lowers profits
and internal cash flow, forcing large domestic firms to depend on external finance.
Financial openness requires sound macroeconomic policies that reduce the government’s
ability to direct credit to favored firms. While the most profitable firms can tap into
foreign capital, other firms, now dependent on external finance, may support financial
development in order to increase access to credit.
In my model, the inadequacy of financial openness alone to provide an incentive
for SIG members to back enforcement of creditor rights is manifested by the reaction of
the interest rate to capital inflows. The additional capital decreases the equilibrium
48
interest rate because capital supply increases more than capital demand. The lower
interest rate increases the optimal level of capital firms would like to invest, while
lowering the return to lending. Wealthy agents have less incentive to lend in the credit
markets and more incentive to invest in their projects at the lower interest rate. The
decrease in the interest rate benefits SIG members, therefore aggregate social welfare
increases overall even though lower interest rates harm poor citizens who must lend their
wealth endowments. Figure 2.10 demonstrates how the “rich get richer and the poor get
poorer” impact of capital inflows into the domestic economy affects the nature of
financial reform. As foreign investment increases exogenously into the system, the
policymaker has less incentive to reform. Elites continue to pressure the politician to
maintain a low level of creditor enforcement in order to swallow up all the additional
credit and increase profits.
16
6.4 Percent Change in Wealth Inequality
Result 5: Wealth inequality has a negative effect on financial development.
Section 6.4 investigates the likelihood of financial reform under increasing wealth
inequality by increasing , the number of agents in the bottom domain of the wealth
distribution (poor). Household wealth is equal to
}
2 /
0
M
dw
M
w
? . The parameter is
increased by .05% to .25% from a starting value of 3. Figure 2.11 graphs levels of
political accountability that make the policymaker indifferent between high and low
16
It is important to note that along with capital, financial openness brings institutional improvements to the
domestic economy such as good corporate governance and foreign bank competition. The institutional
features of financial openness are not modeled in this paper. Result 4 suggests that the benefit of capital
openness is its effect on financial institutions rather than the corresponding increase in capital. According
to the model, capital inflows alone will not increase the level of financial development.
49
financial development as wealth inequality varies. The figure shows that the more
unequal wealth is distributed in a country the higher the level of political accountability,
, required for financial development.
Effect of Inequality on Financial Development
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0 0.05 0.1 0.15 0.2 0.25
% Increase in Inequality
P
o
l
i
t
i
c
a
l
U
t
i
i
l
i
t
y
Figure 2.11: Plot of points that make the policymaker just indifferent between high and low financial
development for different levels as wealth inequality varies.
Wealth inequality in the simulated economy reduces the equilibrium interest rate
and increases the return to setting up projects. SIG members benefit from the low cost of
capital that inequality brings about under financial repression. In Figure 10, I plot a
range of wealth inequality parameters, , for the two extreme levels of financial
development, 1 . = ? and 49 . = ? . Figure 2.12 shows elite entrepreneurs benefit more
from financial repression when wealth is more unequally distributed, i.e. as increases.
Elites also suffer more from financial development when increases.
50
Effect of Inequality on SIG Welfare
1.1
1.11
1.12
1.13
1.14
1.15
1.16
1.17
0 0.05 0.1 0.15 0.2 0.25
% Increase in Inequality
S
I
G
F
i
n
a
l
W
e
a
l
t
h
=.1
=.49
Figure 2.12: Plot of SIG welfare under different degrees if wealth inequality under financial
repression (=.1) and financial development (=.49)
Result 5 is consistent with other research exploring the connection between
inequality and financial development. Gradestein (2004) argues that in countries with
high aggregate income and more equally distributed wealth, influential elites are more
willing to support the enforcement of private property rights. Perotti and Volpin (2004)
show that inequality reduces the level of minority investor protection in equity markets.
Engerman and Sokoloff (2002) study economic institutions across the Americas and
argue “that with more extreme inequality or heterogeneity in the population were more
likely to develop institutional structures that greatly advantaged members of elite classes
(and disadvantaging the bulk of the population) by providing them with more political
influence and access to economic opportunities.” While these studies have provided an
empirical linkage between inequality and financial development, my theoretical model
provides some insight into why such a relationship exists. Initial economic inequality
leads to persistent political inequality. Thus, financial policy is formed to benefit the
economically elite and politically influential at the expense of economic growth. In this
manner, Result 5 is another example of the path dependence of financial policy.
51
7. Concluding Remarks
The central lesson of this chapter is that political structure - namely the degree of
democratic accountability imposed by a country’s institutions - significantly influences
the capacity of policymakers to enact financial reforms. The model supports previous
empirical and anecdotal evidence that special interest groups influence the degree of
financial intermediary development. However, the model’s conclusions shed doubt on
the supposition that all governments respond uniformly to the pressure of powerful
economic agents in society, demonstrating instead that a complete interest group
depiction of financial reform includes the role of idiosyncratic political institutions in
shaping the incentives and constraining the actions of governments.
I have presented a theoretical model of reform that unites these two political
determinants of financial development. First, an interest group uses political
contributions to influence government financial policy. Second, the politician’s
sensitivity to interest group pressure depends on whether or not political institutions
require her to be accountable to the general electorate. Furthermore, the level of
democratic accountability necessary for governments to improve financial development
is dependent on the country’s economic environment; including productivity, wealth per
capita, financial openness, and the degree of wealth inequality. Financial development
leads to popular political support from most citizens however; political repression
enlarges the campaign funds in the politician’s coffers. Both sources of political capital
help the policymaker in her quest to be reelected. Institutions determine the source of
political capital that is more advantageous to the policymaker. The model’s predictions
raise a variety of empirical questions. While the literature has considered some of them,
52
answers are far from conclusive. In conclusion, I simply discuss the questions I believe
deserve further exploration.
First, what is the likelihood that policymakers trade aggregate social welfare in
favor of interest groups? Democratic accountability, the main determinant of financial
development in my model, has several components. Tsebelis (2002) shows that political
systems with a large number of polarized veto players are less likely to enact reforms that
improve aggregate welfare. Lederman et al. (2001) suggest that political institutions that
enforce accountability tend to be less corrupt hence policymakers have less incentive to
engage in rent seeking behavior at the expense of aggregate welfare. Other aspects of
accountability such as competitive elections and independent judicial systems ensure that
policymakers must value the well-being of the citizenry when making policy decisions.
Empirical exploration into all the different components of political accountability and
their relationship to financial development is needed.
Second, do countries with high industrial concentration and heavy regulation of
new business entry tend to focus on aspects of financial development that are beneficial
to large industrial incumbents (equity markets) as opposed to reforms that increase new
business entry (credit markets)? Singh (1995) found that large corporations in
developing countries rely heavily on equity to finance the investment. Unlike the US and
the UK where equity markets developed as result of market forces, Singh finds that
emerging market governments have taken a proactive role in developing their stock
markets. Furthermore, Singh (1997) suggests that developing countries may enact
policies that increase equity issuance by large firms but neglect more substantive policy
measures to financial institutions.
53
Third, do capital inflows (FDI) reduce a country’s incentive to reform its credit
markets? Singh (1997) also discusses how less developed country governments may be
reacting to pressure by large firms to ensure cheap access to capital and by international
investors to open financial markets. By developing their equity markets, governments
can appease both interest groups and may chose to do so even if that means endangering
welfare-enhancing improvements to credit markets and property rights.
Fourth, are governments whose legal structure accommodates active intervention
in financial markets more responsive to interest group pressure? La Porta et al.
(1997,1998) argue that common law regimes are superior in producing efficient legal
systems to civil law. In contrast, Rajan and Zingales (1999, 2003) argue that civil
systems are more efficient than common law systems in adopting good policies because
all legislation emanates from the center. The centralized nature of civil law systems also
means that governments are more influenced by special interests when making their
policy decisions. Therefore, the government’s tendency to intervene in markets is
proxied by legal institutions: civil law countries are more likely to intervene in financial
policy than common law governments. A government whose legal structure
accommodates intervention to a greater extant may respond more willingly to interest
group pressure.
Fifth, in the theoretical model entrepreneur faces the same level of creditor rights
enforcement, . In reality, firms face different obstacles in attaining financing for
investments based on their size. A natural extension of the model is to explore the effects
of disparate credit market imperfections on the equilibrium interest rate and access to
capital.
54
Finally, what role does persistent inequality and low capital availability play in
sustaining financial repression? I have discussed the theoretical importance of path
dependence in explaining patterns of corporate ownership. More work is needed to asses
the empirical significance of poverty and inequality on financial development.
My paper is an attempt to provide a theoretical framework for analyzing these
questions. As emphasized by Acemoglu et al. (2004), “a theory of why different
countries have different economic institutions must be based on politics, on the structure
of political power, and the nature of political institutions… Constructing formal models
incorporating and extending these ideas is the most important task ahead.”
55
Chapter 3: An Empirical Analysis of Political Institutions and Credit
Market Development
1. Introduction
In the previous chapter, the central conclusion of my theoretical model reveals the
role of the political system in implementing financial reform: political systems that
impose more democratic accountability on policymakers achieve higher levels of
financial development. The theory shows that the level of accountability in the political
system determines the politician’s preference between special interest group
contributions and aggregate welfare. Accountable political systems limit government
protection of elite advantage based on insecure property rights by making it difficult for
politicians to accept monetary enticements from special interests; hence, politicians
choose financial policy that increases aggregate welfare.
Assumed in both the theory of special interest politics and its application to
financial development is that all governments respond uniformly to the pressure of the
powerful economic agents in society, as such the interest group depiction of financial
reform ignores the role of institutions in shaping the incentives and constraining the
actions of governments. Political scientists Garret and Lange (1995) recognize the
omission of institutions from a theory of interest group led policy reform, stating,
“Researchers have assumed that the effects of internationally generated changes in the
constellation of domestic economic preferences will be quickly and faithfully reflected in
changes in policies and institutional arrangements within countries. If one understands
which economic interests have gained economic strength, one knows which has gained
political power and in turn how the policy is going to change … [However], institutions
56
invariable outlive the constellation of interests that created them, and hence they provide
barriers to market driven change”.
Other researchers have explored the manner in which institutions affect the
overall quality of the financial system. Djankov, McLiesh, and Shleifer (2004) examine
whether legal rights that protect creditors against default and creditor registries that
collect information on the credit history of borrowers are associated with higher levels of
credit market development. The authors conclude that while both types of institutions are
important for intermediary development, creditor rights are more influential in rich
countries and credit registries play a more significant role in poor countries.
Additionally, Desai, Dyck, and Zingales (2004) find that the institutional details of the
tax system affect the quality of corporate governance. The higher the tax rate the more
incentive managers have to divert corporate profits. On the other hand, strong tax
enforcement reduces managerial diversion of profits and increases stock market value.
This chapter investigates whether the institutional details of the political system
play significant role in achieving financial development that is independent of the level of
democratic accountability between policymakers and voters. I find that political
institutions, as summarized by the number and cohesion of its veto players, are weakly
related to financial development. The chapter proceeds as follows. Section 2 discusses
the role of veto players in policy reform and democratic accountability. Section 3
presents the data. Section 4 discusses the four empirical strategies I use to examine the
relationship between veto players and financial development. Section 5 presents the
results of each empirical strategy. Section 6 concludes.
57
2. Veto Players under Different levels of Political Accountability
Political institutions play a critical role in sustaining democratic accountability. A
convenient method for summarizing a country’s political institutions is by the
characteristics of the veto players in the political system. Tsebelis (1995, 2002) defines
the concept of veto players as “individual or collective actors whose agreement (by
majority rule for collective actors) is required for change in the status quo”. He provides
a method of categorizing governments by the number and policy cohesion of veto
players. The veto player framework is antecedent to the idea of checks and balances
found in the American Constitution and federalist and French revolutionary writings and
has been utilized implicitly and explicitly in contemporary comparative politics. It
allows for comparisons across all political regime types including presidentialism and
parliamentarism, unicameralism and bicamerialism, and two-party and multi-party
systems.
An example of the veto player framework is found in the comparison of American
and British political systems. Often the US and British political systems are paired
together as having similar characteristics. However, the differences between the two
systems are more compelling, “presidential vs. parliamentary systems, bicameralism vs.
(de facto) unicameralism, undisciplined vs. disciplined parties, appointed vs. independent
bureaucracies and the presence vs. the absence of supreme courts”. Thus, the British
system is a single veto player political structure while the American system accords with
a multi veto player set up. (One needs only to concede the impossibility of Thatcher style
market reforms occurring in a system of frequent gridlock like the US to understand the
capability of the veto framework to explain differing levels of resistance to policy change
58
by governments). Simplifying the nexus of comparison to the ability of political systems
to reform policy reduces the confusion of multi-dimensional analysis. The categorization
of governments by the number and cohesion of veto players is the most important
contribution of the veto player framework advanced by Tsebelis.
Along with a method of categorizing political systems, Tsebelis also presents a
theory for predicting which systems have a higher capacity to implement reform.
Tsebelis summarizes his theory as follows, “A significant policy change has to be
approved by all veto players and it will be more difficult to achieve the larger the number
of players and the greater ideological distance between them. Empirical research,
measuring the effect of veto players on policy reforms in areas as diverse as labor
markets and trade openness, supports Tsebelis’ results for advanced democracies.
17
Other researchers have suggested that Tsebelis’s propositions may not be true in
the context of weaker democracies. Keefer (2001) states that the “absence of multiple
veto players in countries often means that some groups in society are less represented
than they otherwise would be.” Political scientists Andrews and Montinola (2004) agree,
adding,
In advanced, industrial democracies, institutions that enforce the rule of law are
well established, and there is a relatively clear link between policy change (that is,
passage) and implementation. … And reform turns mainly on whether legislative
veto players can agree on passage of policies. In emerging democracies,
agreement on policy is not the only potential obstacle to reform. The more
important task for reformers is preventing passage of corrupt legislation and
ensuring proper implementation of genuine reforms.
17
See Baun (1999) and Hellerberg and Basinger (1998).
59
Tsebelis’s predictions are based on the supposition that the most formidable
obstacle to policy change is policy coordination between veto players. Andrews and
Montinola show that since the institutional constraints on legislators in emerging
democracies are weak, the chief obstacle to reform for these countries is collusion among
veto players in taking bribes. A multi-veto player framework makes it more difficult for
a politician to exploit his position for personal gain and retain office. Other researches
have made similar distinctions about the differences in the impact of the number of veto
players in advanced versus emerging democracies.
18
Moreover, Tsebelis intentionally ignores whether political institutions achieve
good or bad policies in order to focus attention entirely on the capability of political
systems to change the status quo, regardless of the effects of the change on the economy.
According to Andrews and Montinola (2004), good reforms strengthen the rule of law,
while bad reforms lead to expropriation by the government. Applying this idea to
financial markets, good reforms protect private property and creditor rights. Bad reforms
violate creditor rights and retard financial development.
I examine whether political institutions have disparate effects on credit market in
advanced versus weak democracies. In advanced democracies, obstacles to policy
change occur in only one dimension, coordination between policymakers. Policy
changes that increase the rule of law are more successfully implemented when there are
fewer veto players that block reform. However, a multi-veto player framework is
desirable when the reforms are bad because there are more vetoes to policy change. In
contrast, for moderately or weakly democratic states, it is desirable to have a multi-veto
18
See Hellman (1998) and Moser (1999).
60
player system whether the reform is good or bad because politically preferences formed
by weakly accountable political systems cause the primary obstacle to policy reform to be
collusion on bribes rather than coordination on policy. Multi-veto players prevent
collusion and increase voter representation insuring that good reforms are implemented
by legislatures and bad reforms are blocked. Moreover, since financial systems begin
from a status quo of financial repression that benefit elite interests groups within the
country, reforms are more likely to be of good quality (otherwise policymakers could
maximize their bribes by maintaining the status quo).
Other researchers have shown that multiple veto player political system increases
the likelihood of good quality reforms in weak democracies. Keefer and Savage (2001)
show that checks and balances in the political system increase the ability of independent
central banks to decrease inflation. Keefer (2001) suggests that multiple veto players
increase government credibility. Henisz (2000) provides evidence that political
constraints on the executive increase economic growth. Finally, La Porta et al. (2002)
show that checks and balances in the political system are correlated with political
freedom.
Though the number of veto players has disparate predictions depending on the
level of democracy, policy cohesion of the veto players has the same effect on capacity
for policy change for all levels of democratic accountability. Cohesion deals with the
policy coordination problem of veto players and is unrelated to the bribery collusion
problem produced by weakly accountable political systems in emerging democracies.
Thus, as Tsebelis asserts, we should expect that policy cohesion among veto players
increases the likelihood of reform.
61
I offer the following two testable hypotheses on the relationship between
intermediary development, accountability, and political institutions.
H1: Financial development increases with the number of veto players for
emerging democracies.
H2: Veto players that are more politically polarized are less likely to generate
policy change that leads to financial development for all levels of democratic
accountability.
3. Data
The panel covers 157 countries over 21 years. A complete description of the data
is shown in Appendix C. The data years are restricted to 1980-2001 based on research by
Singh (1997). Singh observes that governments in less developed countries started from
a level of financial repression then undertook a great deal of financial sector reform in the
1980’s and 1990’s. I use this period of heavy reform in the developing world to test the
prediction that political institutions play a role in the ability of governments to resist
special interest group pressures and to improve the financial sector.
To proxy financial intermediary development, I use Private Credit, the amount of
credit extended to private firms by financial intermediaries divided by GDP, from the
Database on Financial Development and Structure by Beck et al. (2002). Private Credit
is the preferred measure of intermediary development in the finance and growth literature
because it isolates private bank lending from credit issued by central banks, government
lending and credit to government run enterprises (Levine et al. (1999)). According to the
World Business Environment Survey (WBES) only 28% of small firms receive financing
62
from banks, where small firms are defined as enterprises with 5 to 50 employees.
19
Medium firms are defined as enterprises with 51 to 500 employees and large firms
employ over 500 workers. Medium and large firms receive 40% and 47% of financing
for investments from banks. From this survey data, one can infer that most private credit
is acquired by medium and large firms, which also have this most political influence
within a country.
Following Tsebelis (1999) and Keefer (2003), I measure political institutions
along two dimensions – the number of veto players, Checks, and the distance in their
policy positions, Political Polarization. Both variables are from the Database of Political
Institutions (DPI). Checks is measured by assigning one check if the government is
controlled by different parties in the executive and legislative branch for presidential
systems and assigning a check for party in the governing coalition under parliamentary
systems. Checks are also added to detail the competitiveness of national elections,
including whether elections have closed lists or open lists. In my empirical analysis, I
use the Log of Checks variable, because I believe that the number of veto players in the
political system has a nonlinear relationship to private credit. For instance, the marginal
effect on the dependent variable of going from, say, two checks, to three checks is much
greater than the marginal effect of going from ten checks to eleven checks.
Political Polarization captures the distance in policy agreement between
policymakers in a particular year and is measured by assigning the four largest political
parties a policy position along a left to right scale of economic policy. The largest
difference between two political parties is given from 0 to 2 is the degree of polarization.
19
The WBES surveys approximately 10,000 firms in 80 countries on issues related to financing obstacles.
63
To test the disparate effects of veto players on different levels of democracy I use
as a measure of political accountability, the annual Freedom House World Country scores
Freedom Index. The Freedom Index scores are found by rating countries by their
government’s protection of civil liberties and political liberties. I assume that countries
that protect the rights of their citizenry along these two dimensions are more likely to
enforce political institutions that reinforce accountability. The index places countries into
three categories – Free Countries, Partly Free Countries, and Not Free Countries. I
examine whether the variables measuring political institutions will have a different effect
on financial development for each category of political accountability.
Additionally, I investigate the ability of political institutions to increase
intermediary development under different degrees of openness. Rajan and Zingales
(2003) hypothesize that openness to trade and financial flows lead members of elite
interest groups to support financial development instead of repression. I measure the
independent effects of both types of openness on financial development in the presence
of political variables. Foreign direct investment (FDI) and trade openness (Trade) are
taken from the World Bank indicators database (2003).
I also control for the level of economic development. As a proxy for economic
development I use the Log of GDP per capita in 1995, Lgdp95. I used the 1995 measure
as opposes to GDP per capita at the beginning of the sample because the panel is
unbalanced and I will have more country-year observations towards the end of the series.
Inflation is also included in the regression to control for its effect on Private Credit over
the data’s time interval.
64
The countries are separated into three groups, Free Countries, Partly Free
Countries, and Not Free Countries. I assume that in Free Countries politicians are the
most accountable to their constituencies and that they are the least accountable in Not
Free Countries. I then analyze the data in order to determine if a country’s political
institutions increase Private Credit within the three different levels of democratic
accountability. In Tables 3.1A – 3.1C, countries are grouped by their political system’s
promotion of civil and political liberties. Even within these groupings, there exists
considerable variation in the data. Note that among Free Countries, countries include
highly industrialized nations as well as developing countries. Table 3.2 summarizes the
data. Panel A displays the summary statistics for Free Countries, Panel B for Partly
Free Countries and Panel C for Not Free Countries. Overall, Free Countries have the
highest level of financial development as proxied by Private Credit. It is instructive to
compare countries with highly accountable political institutions to political systems with
weaker voter representation in policy reform. The amount of Private Credit is just over
54% of GDP in Free Countries, almost twice as much as the percentage of Private Credit
in Partly Free Countries and two and a half times the percentage in Not Free Countries.
Free Countries also have the greatest variance in financial development of the three
groups. Moreover, Free Countries tend to have more Checks in the political system (not
surprisingly, because these are also the most democratic countries) and have veto players
that are the most ideologically polarized.
For the smallest economies, namely the African countries of Ethiopia, Liberia,
Chad, Sudan and Niger, Private Credit can be less than 15% of GDP and in some cases
significantly less. For the most economically advanced countries, like the US, the UK,
65
Japan, and Germany credit extended to private corporations can total over 90% of GDP.
The other variables of interest demonstrate considerable range over the data set as well.
The number of veto players necessary to implement reform in the financial sector ranges
from a low of 1 to up to 18 Checks. Most countries in the sample have 5 or fewer Checks
within the political system; however, there are notable exceptions. India tops the list with
as many as 18 Checks required to pass reform legislation. Other countries with Checks
over 5 include France, Turkey, Denmark, and Thailand. Political Polarization between
veto players ranges from 0 to 2, 0 indicating complete policy cohesion. The US, UK,
Israel, Turkey, Central African Republic, and France are among the most polarized
political systems, while Algeria, The Bahamas, Canada and Chile are included in the
least.
The openness indicators, Trade and FDI also show tremendous variation in the
sample. Partly Free Countries on average fall in the middle of Free and Not Free
political regimes in terms of financial development, number of veto players, degree of
polarization, wealth per capita and financial and trade openness.
Also of interest is the correlation between the variables in the bottom half of each
panel. Both the Log of Checks and Polarization are positively correlated with Private
Credit for Free Countries. The Log of Checks has a positive correlation and Polarization
a negative correlation in Partly Free Countries. Both institutional variables have a
negative correlation with Private Credit in Not Free Countries. Overall, economic
development as proxied by Lgdp95 is highly correlated with the Private Credit, Log of
Checks and Polarization however; the correlation between economic development and
the number of veto players and political polarization diminishes significantly for the
66
Partly Free and Not Free Countries. Furthermore, Trade and FDI are negatively
associated with Private Credit in the Free Countries and positively related with Private
Credit in the Partly Free Countries. In the next two sections, I analyze whether these
differences in signs between the explanatory and dependent variables for the different
levels of democratic accountability are suggestive of the disparate effects of political
institutions on financial development.
4. Empirical Strategy
I test the hypothesis of the paper by using four different empirical strategies. In
Strategy 1, I cluster the standard errors at the country level as in Desai et al. (2004). By
clustering the standard errors I avoid assuming that observations within each country are
independent of each other. For Strategy 2, I cluster the standard errors by country and
then test whether liberalization plays a key factor in the relationship between the political
variables and the proportion of private credit to GDP. According to Dermirguc-Kunt
and Detragiache (1998) and Williamson and Mahar (1998), many countries had
liberalized interest rates by 1995. I test whether the effect of political institutions is more
strongly related to financial development before the completion of liberalization in most
countries or afterwards.
In Strategy 3, I average the data over 5-year intervals and I do not cluster the
standard errors at the country level. This strategy is similar to the approach used by
Levine, Beck, and Loayza (2000). The authors examine the relationship between
institutions on financial development and growth and find that differences on legal and
accounting system help explain differences in credit market development.
67
In Strategy 4, I use data on firm ownership as the dependent variable in order to
determine of political institutions affect the ownership arrangement (family owned versus
widely held) in listed corporations. Widely-held corporations may be less numerous in
countries with political systems where financial policy can be more easily influenced by
narrow interest groups represented by family owned enterprises.
For each empirical strategy, I present 4 tables. The first displays the OLS
regressions of two equations, the Basic Equation and the Openness Equation. The Basic
Equation includes the political variables, Log of Checks and Polarization plus the two
controls, Lgdp95 and Inflation. The Openness Equation includes the basic variable plus
Trade and FDI.
In the second table, I add Legal Origin to the regression equation in order to
determine the independent effects of the legal regime and political institutions on Private
Credit. Much attention in the literature has been given to the role of legal origin in
influencing financial development.
20
La Porta et al. (1998) explore the contribution of a
country’s legal origin in the formation of its financial structure and its corporate
governance institutions, finding that legal origin — be it English common law, or French,
German or Scandinavian civil law — partly determines the quality of investor protection
and the size of the stock market versus the banking sector. The paper concludes that
English common law systems generally have the strongest investor protection
enforcement, followed by Germany, Scandinavian, and lastly, French civil systems.
Beck et al. (2002) suggest that political institutions do not explain financial development
independently of legal origin. I test the authors’ conclusions when segmenting the data
into the different levels of democratic accountability.
20
See La Porta et al. (1997, 1998), Beck et al. (2002), and Levine et al. (2000).
68
Next, I present instrumental variables estimation of the Basic and Openness
equation. The use of Lgdp95 may be problematic because the variable is highly
correlated with the other variables in the regression equation and with the error term.
Instrumental variables estimation is commonly used when there is contemporaneous
correlation between an independent variable and the error term. Rodrik et al. (2002)
estimates the independent contributions of geography, institutions, and trade on economic
growth. In their search for instruments for their explanatory variables that are
uncorrelated with the error term and highly correlated with the regressor that is being
instrumented, the authors appeal to geography-based instruments identified in the finance
and growth literature.
Acemoglu et al. (2001) use an instrument for GDP that is based on the mortality
of settlers who first colonized countries from Europe. If the settlers found hospitable
living environment they settled the country and built good institutions that lead to
economic growth. If the colonizers were morbidly vulnerable to environmental
conditions they built extracting institutions that lead to economic stagnation. Therefore,
settler mortality is highly correlated with GDP; however, it is also exogenous to
institutions. The log of settler mortality, Lsettler, is a good but somewhat restrictive
instrument because it can only be used on a sample of former colonies. In my data set, 70
countries in the sample have data on settler mortality. I also employ a second instrument,
distance from the equator in degrees from Hall and Jones (1999). Both instruments proxy
for economic development by correlating with “the extent of Western European
influence.”
69
A second potential concern with the OLS analysis is omitted variable bias when
controlling for trade and financial openness in the regression equation. Policy reforms
that enlarge the amount of credit lent to private firms may also encourage the country to
increase its trade in goods and attract more foreign capital. Without including the
specific policy reforms into the regression equation it is impossible to discern how much
of the change in Private Credit is due to Trade and FDI alone. Frankel and Romer
(1999) provide an instrument for actual trade shares by estimating a gravity model of
bilateral trade. A similar geography based instrument is not as easily derived for
financial openness. The log of their measure, Log FR, is based on geographical
characteristics such as distance between trading partners. I use this measure as an
instrument for the Trade variable.
5. Results
5A. Strategy 1: Clustering Standard Errors by Country
Table 3.3 displays the OLS regressions with the standard errors clustered by
country for the three different levels of democratic accountability, Free Countries, Partly
Free Countries, and Not Free Countries. For each category of democratic accountability,
I estimate the Basic and Openness Equations. I find that for no group of countries is the
Log of Checks a significant determinant of Private Credit under this empirical strategy.
Political Polarization is a negative and significant at the 5 % level for the Partially Free
group of countries in Regression 4. The variable has a coefficient of -0.083. When
Trade and FDI are added in Regression 4 the coefficient on Polarization drops to -0.053
at a 10% level of significance. The proxy for economic development, Lgdp95, is positive
and significant for all levels of accountability. Inflation is negative and significant
70
determinant of Private Credit for all regressions except Regression 6. Trade is an
important factor for Partially Free countries whereas FDI is an important determinant in
Free Countries.
The inclusion of Legal Origin in Table 3.4 does not substantively change the
relationship between the political variables and Private Credit discussed in Table 3.3.
The Log of Checks in the political system is not a significant determinant of the
dependent variable. Polarization is negative and significant at the 5% and 10% levels in
Regressions 3 and 4, respectively with coefficients of approximately -0.05 in both
regressions. French Legal Origin is significant at the 5% level for Partially Free
Countries only. However, the significance of French Legal Origin is not robust to the
inclusion of indicators for financial openness. German Legal Origin is a strong
contributor to Private Credit for the Free Countries, while intermediary development is
hampered by a Socialist Legal Origin in the Free and Not Free groups.
The last two tables estimate the Basic and Openness equations using IV
techniques. Panel A shows the first stage regressions. Regressions 1, 3, and 5 use the
log of settler mortality (Lsettler) as an instrument for Lgdp95 and Regression 2, 4, and 6
instruments for Lgdp95 using the distance from the equator in degrees (Disteq). Panel B
of Table 3.5 gives the second stage regressions. Under both the instruments, Polarization
is negative and significant at the 5% level for Partly Free Countries. Its coefficients are -
0.077 and -0.075 in Regressions 3 and 4, respectively.
Political Polarization is robust to the inclusion of FDI and Trade in Table 3.6
when Disteq is used as an instrument for Lgdp95 and Log FR is used as an instrument for
trade openness. In Regression 4, Polarization has coefficient of -0.074 and is significant
71
at the 5 % level. FDI is a significant factor of financial development in Free Countries
while Trade is negative and significant. Unlike the OLS estimation, Trade is not a
significant component of Private Credit for the Partially Free Countries.
5B. Strategy 2: The Effect of Liberalization
The second empirical approach takes into account the potential role of financial
liberalization occurring in many of the countries during the period under analysis. I test
whether the relationship between the political variables and financial development
changes after the time by which most countries had liberalized their financial markets, by
including a dummy variable for country-year observations after 1995, Group B. I then
include interaction terms of the Group B dummy variable and the two political variables,
Log of Checks and Polarization. I define the interaction terms as LchecksB and
PolarizationB. Finally, I test whether coefficients of the interaction terms are equal to
zero to determine if the relationship between the political variables and financial
development is altered by liberalization. I am not able to test the equality of the
coefficients for the polarization interaction term in the Not Free Countries case. Post
1995 all the observations for Polarization are exactly equal to 0. Since there is no
variation in the explanatory variable, it is automatically dropped from the regression
estimation.
Table 3.7 presents the results of the OLS estimation of the Basic and Openness
Equations. Also shown in the table are the F-statistics for whether the interaction terms
are equal to zero. The Log of Checks is not a significant determinant of Private Credit in
any of the regressions. Polarization is significant at the 5% level in Regression 3. The
interaction dummies are not significant. The F-statistics suggest that financial
72
liberalization did not significantly alter the relationship between the political variables
and financial development of the credit market.
Table 3.8 adds Legal Origin to the Basic and Openness Equations. The results are
similar to the Legal Origin regressions in Table 3.4 that employ empirical Strategy 1.
Political Polarization and French Legal Origin are both negative and significant at the
5% level in the Basic Equation only. Their coefficients are -0.046 and -0.125,
respectively. German Legal Origin makes a positive contribution to Private Credit for
Free Countries while a Socialist legal regime is a deterrent to credit development in Free
and Not Free Countries. According to the F-statistics, the interaction terms are not
significantly different from zero.
The IV estimation shown in Table 3.9 shows that Polarization is a negative
determinant of Private Credit for both the Basic and Openness Equations for Partly Free
Countries. Its coefficients are -0.85 and -0.082 respectively. When I control for trade
and capital openness and instrument for trade openness in Table 3.10, Polarization is
only a significant determinant in Regression 4. As in the previous regression, for both
Tables 3.9 and 3.10, liberalization does not significantly change the relationship between
the political variables and private credit.
5C. Strategy 3: Averaging Data in 5-year intervals
Empirical Strategy 3 averages the data into 5 non-overlapping intervals from
1980-2000. The standard errors are robustly estimated. Table 3.11 shows that the Log of
Checks is positive and significant at the 5% level for the Partially Free Countries. The
coefficients are 0.09 and 0.098 in Regressions 3 and 4 respectively. This result contrast
to no relationship between Log of Checks and Private Credit found in the first two
73
empirical approaches. It also does not support the hypothesis the multiple veto players in
the political system is more valuable to intermediary development in weaker
democracies. Polarization is found to be a significant negative determinant of Private
Credit in Table 3.11 for the Partially Free Countries. Its coefficients are -0.166 and -
0.094 in the Basic and Openness regressions, respectively.
Table 3.12 adds Legal Origin to the regressions. The Log of Checks is negative
and significant at the 5% level for Free Countries. Its coefficients are -0.122 and -0.127
for the Basic and Openness Equations, respectively. The Log of Checks is not a
significant factor for Private Credit in the Partially Free and Not Free groups of
countries. Polarization is negatively related to Private Credit at the 5% level in
Regression 3 and at the 10% level in Regression 4 of the Partially Free states. Table
3.13 presents IV estimation. The Log of Checks remains positive and significant for Free
Countries. It is also positive for the Partially Free Countries when Lsettler is used as an
instrument for Lgdp95 in Regression 3. Table 3.14 includes trade instrumented by Log
FR and it adds FDI to each regression. The relationship between Log of Checks and
Private Credit is unchanged from the previous table, it is positive and significant for Free
Countries, however, neither of the political variables is significantly related to Private
Credit for the Partially Free and Not Free Countries.
5D. Strategy 4: Ownership as the Dependent Variable
Empirical Strategy 4 estimates the effect of political institutions on the percentage
of publicly listed firms that are privately held in each country. The ownership data is
from in Mork Wolfenzon and Yeung (2004). It covers 31 countries. There are four
variables collected in the Morck et al. data set – the percentage of widely held public
74
firms in the country when control is inferred at 10%, the percentage of widely held public
firms in the country when control is inferred at 20%, the percentage of family owned
firms in the country when control is inferred at 10%, and the percentage of family owned
firms in the country when control is inferred at 20%. After I merge the ownership data
with my averaged data set, I have 28 observations therefore; I do not divide the countries
into 3 groups as before. Instead, I include the index of democratic accountability directly
into the regressions. The Freedom Index ranges from 1 (Free Countries) to 3 (Not Free
Countries) as before. Since most of the ownership data is collected from the years 1999
and 2002, I average the explanatory variables over the time interval 1996-2000. The
basic equation includes Log of Checks, Polarization, the Freedom Index, and Lgdp95.
The Openness Equation includes the Basic Equation plus Trade and FDI. I repeat the
OLS and IV estimation as in the other 4 strategies except I do not include in the
regressions the proxies for openness in the IV estimation because that leaves me with
only 6 observations. All tables under this empirical approach display the same result. I
do not find a relationship between the political variables and the level of ownership.
6. Conclusion
This chapter uses a number of different empirical approaches to test two
hypotheses. First, the number of veto players is positively related to intermediary
development in weak democracies. Second, political polarization deters credit market
development in all countries. There is no support for the first hypothesis as I find no
consistent relationship between the Log of Checks and Private Credit. Moreover, there is
only weak support for the hypothesis that political polarization negatively effects
intermediary development. To quote noted philosopher, Cornel West, “democracy
75
matters” however, the manner in which a country organizes its political institutions does
not seem to effect intermediary development once the level of democratic accountability
between politicians and constituencies has been taken into account. Thus, the analysis
presented in this chapter does not find any empirical advantage to the arrangement of
political institutions in restricting special interest group influence on financial policy.
76
Table 3.1A : Free Country and Years
Argentina (1984-2000) Gambia (1980, 1989-1993) Nigeria (1980-1983)
Australia(1980-2000) Germany (1980-2001) Norway (1980-2001)
Austria (1980-2001) Ghana(1980-1981,2000-2001) Panama (1994-2001)
The Bahamas (1980-2001) Greece (1980-2001) Papa New Guinea (1980-1992,
1998-2001)
Bangladesh (1991-1992) Grenada (1985-2001)
Barbados (1980-2001) Guyana (1993-2001) Peru (1980-1988, 2001)
Belgium (1980-2001) Honduras (1982, 1984-1992,
1997-1998)
Philippines (1987-1989, 1996-
2001)
Belize (1981-2001) Hungary (1990-2001) Poland (1990-2001)
Benin (1991-2001) Iceland (1980-2001) Portugal (1980-2001)
Bolivia (1982-1994, 1996-
2001)
India (1980-1990, 1998-2001) Romania (1996-2001)
Botswana (1980-2001) Ireland (1980-2001) Samoa (1989-2001)
Brazil (1985-1992) Israel (1980-2001) Slovak Republic (1994-1994,
1998-2001)
Bulgaria (1991-2001) Italy (1980-2001) Slovenia (1991-2001)
Canada (1980-2001) Jamaica (1980-2001) Solomon Island (1980-1999)
Cape Verde (1991 -2001) Japan (1980-2001) South Africa (1994-2001
Chile (1990-2001) Republic of Korea (1988-
2001)
Spain (1980-2001)
Colombia (1980-1988) Latvia (1991, 1994-2001) Sri Lanka (1980-1982)
Costa Rica (1980-2001) Lithuania (1991-2001) St. Lucia (1980-2001)
Croatia (2000-2001) Luxembourg (1980-2001) St. Vincent and the
Grenadines (1980-2001)
Cyprus (1981-2001) Malawi (1994-1998) Suriname (1988, 2000-2001)
Czech Republic (1993-2001) Mali (1992-1993, 1995-2001) Sweden (1980-2001)
Denmark (1980-2001) Malta (1980-1092, 1987-2001) Switzerland (1980-2001)
Dominica (1980-2001) Mauritius (1981-2001) Thailand (1989-1990,1998-
2001)
Dominican Republic (1980-
1992, 1998-2002)
Mexico (2000-2001) Trinidad and Tobago (1980-
2000)
Ecuador (1980-1995, 1998-
1999)
Mongolia (1991-1999) United Kingdom (1980-2001)
El Salvador (1997-2001) Namibia (1990-2001) United States (1980-2001)
Estonia (1991, 1993-2001) Nepal (1991-1992) Uruguay (1985-2001)
Fiji (1980-1986, 1999) Netherlands (1980-2001) Vanuatu(1980-1982,1989-
2001)
France (1980-2001) New Zealand (1980-2001) Venezuela(1980-1991,1996-
1998)
77
Table 3.1B : Partially Free Country and Years
Algeria (1989-1991) Guinea-Bissau (1991-2001) Nigeria (1987-1992, 1998-
2001)
Angola (1991) Guyana (1980-1992) Oman (1992)
Argentina (1982-1983, 2001) Haiti (1986-1987,1990,1994-
1999)
Pakistan (1985-1998)
Armenia (1991-2001) Honduras (1980-1981, 1983,
1993-1996, 1999-2001)
Panama (1980-1987,1990-
1993)
Bahrain (1981-1992) Hungary (1984-1989) Papua New Guinea (1993-
1997)
Bangladesh (1980-1990, 1993-
2001)
India (1991-1997) Paraguay (1980-1987, 1989-
2001)
Belarus (1991-1995) Indonesia (1980-1992, 1998-
2001)
Peru (1989-2000)
Benin (1990) Iran, Islamic Rep. (1980,
1984-1987)
Philippines (1980-1986,1990-
1995)
Bhutan (1980-1991) Jordan (1984-1987, 1989-2001) Poland (1980-1981, 1983-
1989)
Bolivia (1995) Kazakhstan (1991-1993) Romania (1991-1995)
Brazil (1980-1984, 1993-
2001)
Kenya (1980-1986, 1992) Samoa (1980-1988)
Bulgaria (1990) Korea, Rep. (1980-1987) Senegal (1980-2001)
Burkina Faso (1980-1981,
1983, 1992-2001)
Kuwait (1980-1989, 1992-
2001)
Seychelles (1992-2001)
Burundi (1992) Kyrgyz Rep (1991-1999) Sierra Leone (1980-1991,
1996, 1998-2001)
Cape Verde (1987, 1990) Latvia (1992-1993) Singapore (1980-2001)
Central African Republic
(1991-2001)
Lebanon (1980-1987, 1991-
1994, 2001)
Slovak Republic (1993, 1996-
1997)
Chile (1980-1981, 1983-1989) Lesotho (1980-1987, 1991-
2001)
Solomon Islands (2000-2001)
Colombia (1989-2001) Liberia (1983-1988, 1997-
2001)
South Africa ( 1980, 1983-
1993)
Congo, Rep. (1991-1996,
2000-2001)
Madagascar (1982-2001) Sri Lanka (1983-2001)
Cote d’Ivoire (1980-1987,
1989-1992, 1999-2001)
Malawi (1999-2001) Sudan (1980-1983, 1986-
1988)
Croatia (1991-1999) Malaysia (1980-2001) Suriname (1987, 1989-1999)
Cyprus (1980) Maldives (1980-1987) Swaziland (1980-1992)
Djibouti (1980-1981,1984,
1999-2001)
Mali (1991, 1994) Thailand (1980-1988,1991-
1997)
Dominican Republic (1993-
1997)
Malta (1983-1986) Togo (1999-2001)
Ecuador (1996-1997,2000-
2001)
Mauritania (2000-2001) Tonga (1980-2001)
Egypt, Arab Rep. (1980-1992) Mauritius (1980) Trinidad and Tobago (2001)
El Salvador (1980-1996) Mexico (1980-1999) Tunisia (1980-1992)
Estonia (1992) Moldova (1991-2001) Turkey (1980-2001)
78
Table 3.1B : Partially Free Country and Years (Continue)
Ethiopia(1991-1992,1995-
2001)
Mongolia (1990) Uganda (1980-1990, 1994-
2001)
Fiji (1987-1998, 2000-2001) Morocco (1980-2001) Ukraine (1991-2001)
Gabon (1990-2001) Mozambique (1991-1992,
1994-2001)
Uruguay (1980-1984)
The Gambia (1981-1988,
2001)
Namibia (1989) Vanuatu (1983-1988)
Ghana (1992-1999) Nepal (1980-1990, 1993-2001) Venezuela (1992-1995,1999-
2001)
Grenada (1980, 1984) Nicaragua (1980-1997, 1999-
2001)
Zambia (1980-1990, 1993-
2001)
Guatemala (1980, 1984-2001) Niger (1991-1995,1999-2001) Zimbabwe (1980-2000)
79
Table 3.1C : Not Free Country and Years
Algeria (1980-1988, 1992-
2001)
The Gambia (1994-2000) Niger (1980-1990, 1996-1998)
Angola (1980-1990, 1992-
1998)
Ghana (1982 -1991) Nigeria (1984-1986, 1993-
1997)
Argentina (1980-1981) Grenada (1981-1983) Oman (1980-1991, 1993-
2001)
Bahrain (1993-2001) Guatemala (1981-1983) PRK (1980-2000)
Belarus (1996-2001) Guinea-Bissau (1980-1990) Pakistan (1980-1984, 1999-
2001)
Benin (1980-1989) Haiti (1980-1985, 1988-1993,
2000-2001)
Panama (1988-1989)
Bhutan (1992-2001) Hungary (1980-1983) Paraguay (1988)
Bolivia (1980-1981) Indonesia (1993-1997) Poland (1982)
Bulgaria (1980-1989) Iran (1981-1983, 1988-2001) Romania (1980-1990)
Burkina Faso (1982, 1984-
1991)
Jordan (1980-1983, 1988) Rwanda (1980-2001)
Burundi (1980-1991, 1993-
2001)
Kazakhstan (1994-2001) Saudi Arabia (1980-2001)
Cameroon (1980-2001) Kenya (1987-1991, 1993-
2001)
Seychelles (1980-1991)
Cape Verde (1980-1986,
1988-1989)
Kuwait (1990-1991) Sierra Leone (1992-
1995,1997)
Central African Republic
(1980-1990)
Kyrgyz Rep (2000-2001) South Africa (1981-1982)
Chad (1980-2001) Lao PDR (1980-2001) Sudan (1984-1985, 1989-
2001)
Chile (1982) Lesotho (1988-1990) Suriname (1980-1986)
China (1980-2001) Liberia (1980-1982, 1989-
1996)
Swaziland (1993-2001)
Congo, Dem Rep (1980-2001) Madagascar (1980-1981) Syria Arab Rep (1980-2000)
Congo, Rep (1980-1990,
1997-1999)
Malawi (1980-1993) Togo (1980-1998)
Djibouti (1982-1983, 1985-
1998)
Maldives (1988-2001) Tunisia (1993-2001)
Egypt, Arab Rep. (1993-2001) Mali (1980-1999) Uganda (1991-1993)
Equatorial Guinea (1980-
2001)
Mauritania (1980-1999) Vietnam (1980-2001)
Ethiopia (1980-1990, 1993-
1994)
Mongolia (1980-1989) Zimbabwe (2001)
Gabon (1980-1989) Mozambique (1980-
1990,1993)
80
Table 3.2: Summary Statistics
Panel A: Free Countries
Private
Credit
Log of
Checks
Polarization Trade FDI Lgdp95 Inflation
Observations 1198 1134 1093 1243 1238 1290 1280
Mean 0.542 1.160 .7282 83.534 2.921 8.68 18.033
Std. Dev. 0.376 .552 .914 43.575 4.799 1.368 65.09
Min 0.0178 0 0 10.079 -6.897 5.044 -32
Max 1.790 2.890 2 290.710 93.720 10.696 969
Correlations
Private
Credit
1.000
Log of
Checks
0.255 1.000
Polarization 0.281 0.513 1.000
Trade -0.030 0.042 -0.060 1.000
FDI -0.028 -0.035 -0.017 0.376 1.000
Lgdp95 0.6946 0.461 0.513 -0.085 -0.064 1.000
Inflation -.2019 -0.0786 -0.0480 -0.1014 -0.0771 -0.1068 1.00
Panel B: Partly Free Countries
Private
Credit
Log of
Checks
Polarization Trade FDI Lgdp95 Inflation
Observations 888 918 918 998 988 1066 1007
Mean 0.274 0.492 0.109 74.565 1.794 7.006 29.512
Std. Dev. 0.240 0.620 0.414 49.097 3.581 1.226 89.56
Min 0.001 0 0 8.953 -28.622 3.898 -26
Max 1.552 2.890 2 361.179 39.776 10.068 976
Correlations
Private
Credit
1.000
Log of
Checks
0.1701 1.000
Polarization -0.047 0.458 1.000
Trade 0.531 -0.039 -0.111 1.000
FDI 0.277 0.052 -0.028 0.436 1.000
Lgdp95 0.597 0.007 0.004 0.577 0.267 1.000
Inflation -0.194 0.077 0.097 -0.004 -0.076 -0.026 1.00
81
Panel C: Not Free Countries
Private
Credit
Log of
Checks
Polarization Trade FDI Lgdp95 Inflation
Observations 533 753 735 696 688 783 709
Mean 0.223 0.154 0.015 68.102 1.720 6.389 20.306
Std. Dev. 0.208 0.393 0.168 68.011 7.662 1.158 56.019
Min 0 0 0 6.320 -82.810 3.898 -29
Max 1.197 2.079 2 275.232 145.210 9.727 637
Correlations
Private
Credit
1.000
Log of
Checks
-0.049 1.000
Polarization -0.025 0.391 1.000
Trade 0.112 -0.028 -0.037 1.000
FDI 0.004 -0.005 -0.009 0.369 1.000
Lgdp95 0.419 -0.032 -0.026 0.350 -0.0350 1.000
Inflation -0.236 -0.042 -0.013 -0.085 -0.011 -0.166 1.000
82
Table 3.3: OLS Regression of the Basic and Openness Equations
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Standard errors are clustered
at the country level. Absolute value of robust t-statistics are in parenthesis.* significant at 10%;
** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of Checks -0.041 -0.033 0.048 0.051 0.009 0.023
(1.09) (0.89) (1.38) (1.49) (0.29) (0.7)
Polarization -0.006 -0.01 -0.083** -0.053* -0.021 -0.034
(0.19) (0.33) (2.50) (1.94) (1.19) (1.73)
Inflation -0.001*** -0.001*** -0.001** -0.001*** -0.001** -0.001
(3.94) (3.91) (2.59) (3.04) (2.03) (1.75)
Lgdp95 0.184*** 0.18*** 0.114*** 0.078*** 0.094*** 0.088***
(7.85) (7.53) (6.37) (6.40) (5.89) (2.89)
Trade 0 0.002*** -0.001
(0.61) (3.36) (0.74)
FDI 0.008** -0.001 0.009
(2.21) (0.29) (0.75)
Constant -1.007*** -0.966*** -0.515*** -0.414*** -0.346*** -0.277**
(5.88) (5.49) (4.51) (5.40) (3.50) (2.02)
Observations 1000 956 724 668 455 396
R-squared 0.48 0.47 0.44 0.52 0.32 0.19
83
Table 3.4: OLS Regression of the Basic and Openness equations and Legal Origin
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Standard errors are clustered
at the country level. Each equation includes dummy variables for French, German, Scandinavian,
and Socialist legal origins. Absolute value of robust t-statistics are in parenthesis. * significant at
10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The dependent
variable is Private
Credit.
1 2 3 4 5 6
Log of Checks -0.066 -0.059 0.001 0.041 0.009 0.036
(1.66) (1.53) (0.04) (1.31) (0.28) (0.94)
Polarization 0.008 0.003 -0.055** -0.05* -0.015 -0.014
(0.28) (0.12) (2.12) (1.91) (0.51) (0.42)
Lgdp95 0.174*** 0.165*** 0.135*** 0.082*** 0.099*** 0.067**
(7.79) (7.35) (7.16) (4.88) (4.78) (2.39)
Inflation -0.001*** -0.001*** -0.001** -0.001** -0.001 -0.001**
(3.30) (3.24 (2.05) (2.15) (1.67) (2.16)
Legor_fr -0.016 -0.027 -0.12** -0.017 -0.026 0.023
(0.27) (0.46) (2.38) (0.36) (0.58) (0.5)
Legor_ge 0.325*** 0.356*** -0.057 0.112 0*** 0***
(2.85) (3.17) (0.76) (1.57) (0.00) (0.00)
Legor_sc -0.103 -0.078 0*** 0*** 0*** 0***
(0.7) (0.52) (0.00) (0.00) (0.00) (0.00)
Legor_so -0.225*** -0.236*** -0.134 -0.08 -0.29*** -0.184**
(5.19) (5.70) (1.75) (1.48) (4.70) (2.06)
FDI 0.01*** -0.001 -0.005
(2.01) (0.21) (0.7)
Trade 0 0.002*** 0.001
(0.24) (3.24) (0.92)
Constant -0.897*** -0.833*** -0.569*** -0.422*** -0.377*** -0.262
(4.90) (4.02) (5.45) (4.50) (2.92) (1.57)
Observations 908 871 629 592 302 260
R-squared 0.53 0.54 0.48 0.52 0.49 0.31
84
Table 3.5: IV Regression of the Basic Equation
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. Regressions
A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B instruments
Lgdp95 with Distance from the Equator in Degrees (Disteq). Standard errors are clustered at the
country level. Absolute value of robust t-statistics are in parenthesis. * significant at 10%; **
significant at 5%; *** significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Country Partly Free Country Not Free Country
A B A B A B
1 2 3 4 5 6
Log of Checks 0.215*** 0.551*** 0.048 0.388*** 0.310 -0.039
(2.88) (7.04) (0.56) (3.88) (1.80) (0.23)
Polarization 0.778 0.212*** 0.60 0.025 -0.423 -0.505
(1.50) (3.68) (0.56) (0.17) (1.46) (1.33)
Inflation 0 -0.001* 0.002 -0.003 0.010*** -0.004**
(0.70) (1.84) (1.61) (1.59) (2.86) (2.29)
Instrument -0.791*** 5.49*** -0.664*** 2.89*** -2.86 5.015***
(23.59) (16.81) (16.59) (6.35) (21.58) (12.38)
Constant 11.219*** 6.22*** 10.08*** 6.29*** 7.448*** 5.50***
(63.22) (58.81) (48.03) (54.89) (21.58) (57.11)
Observations 491 526 478 600 246 407
R-squared 0.60 .0.48 0.42 0.08 0.11 0.28
Panel B: Second Stage Regressions of the Basic Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1
2 3 4 5 6
Lgdp95 0.179*** 0.238*** 0.161*** 0.148** 0.222* 0.139
(7.40) (5.03) (4.40) (2.18) (1.86) (2.39)*
Log of Checks -0.026 -0.085 0.029 0.034 -0.022 -0.004
(0.63) (1.44) (0.87) (0.78) (0.40) -0.13
Polarization 0.001 -0.023 -0.077** -0.075** 0.021 0.003
(0.02) (0.58) (2.10) (2.57) (0.46) -0.13
Inflation -001*** -0.001*** -0.001 -0.001 -0.004*** -0.001
(3.44) (3.54) (3.23) (3.27)** (2.76) -1.62
Constant -0.990*** -1.374*** -0.925*** -0.735 -1.11 -0.628
(5.16) (4.03) (3.94) -1.58 (1.58) -1.8
Observations 491 526 478 600 246 407
R-squared 0.49 0.5 0.39. 0.43 . 0.27
85
Table 3.6: IV Regression of the Openness Equation
The Openness Equation includes the Log of Checks, Polarization, Inflation, Lgdp95, Trade, and
FDI. Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Trade is instrumented
with the Log of the Frankel-Romer proxy of natural trade openness (Log FR). Standard errors
are clustered at the country level. Absolute value of robust t-statistics are in parenthesis. *
significant at 10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Panel A: First Stage Regressions of Lgdp95
Log of Checks 0.152* 0.512*** 0.015 0.401*** 0.223 0.195
(1.71) (6.48) (.17) (4.23) (1.22) (1.47)
Polarization 0.070 0.244*** 0.109 0.090 0.318 -0.374
(1.06) (4.18) (0.92) (0.66) (0.90) (1.31)
FDI -0.001 0.017 0.085*** 0.052*** 0.065* 0.062***
(0.007) (1.21) (4.33) (2.95) (1.96) (2.96)
Inflation 0.001* -0.006 0.002* -0.001 0.009** -0.001
(1.81) (1.23) (1.89) (.45) (2.57) (-0.39)
Instrument -0.820*** 5.841*** -0.562*** 2.621*** -0.234 3.858
(15.47) (16.75) (11.15) (6.07) (3.47)*** (13.35)
LogFrankRom 0.20*** .124** -0.013 0.143** 0.023* 0.277***
(3.41) (2.55) (1.35) (1.99) (1.86) (4.75)
Constant 10.816*** 5.755*** 9.50*** 5.742*** 7.026*** 4.574***
(41.29) (30.26) (34.31) (24.45) (18.62) (23.94)
Observations 364 515 397 561 233 354
R-square 0.48 0.48 0.33 0.10 0.10 0.33
Panel B: First Stage Regressions of Trade
Log of Checks 7.055*** 4.454** 0.165 7.623*** 11.286** 11.285***
(3.24) (2.28) (0.04) (2.81) (2.47) (2.79)
Polarization -4.202** -6.570*** -12.54* -8.399** -16.188* -13.067
(2.58) (4.55) (2.36) (2.15) (1.83) (1.50)
FDI 2.506 2.431*** 10.33*** 6.614*** 5.489888 6.824***
(5.91) (7.00) (11.84) (13.04) (6.59) (10.63)
Inflation -0.016 -0.001 -0.201*** -1.23*** -0.121 0.001
(1.20) (.45) (13.62) (2.88) (1.30) (0.03)
Instrument -5.359*** 2.621*** -1.91 -24.947** 1.353 40.343***
(4.12) (6.07) (.85) (2.02) (0.80) (4.25)
Log FR 30.393 .143** -0.126 33.933*** 0.372 24.130***
(21.29) (1.99) (0.30) (16.44) (1.20) (13.64)
Constant 2.212 5.742*** 65.08*** -
29.801***
41.401*** -
25.758***
(0.34) (24.45) (5.29 (4.43) (4.39) (4.44)
Observations 364 561 397 561 233 354
R-square 0.65 0.10 0.31 0.50 0.16 0.45
86
Panel C: Second Stage Regression of the Openness Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1 2 3 4 5 6
Lgdp95 0.155*** 0.222*** 0.113 0.183** 0.263 0.211*
(6.43) (4.88) (1.21) (2.14) (0.61) (1.88)
Trade -0.002** -0.003** -0.002 0.001 0.002 -0.008*
(2.21) (2.67) (0.09) (.51) (0.07) (1.90)
Lchecks 0.009 -0.044 0.012 0.020 -0.054 0.056
(0.30) (0.87) (0.41) (0.43) (0.11) (0.78)
Polarization -0.036 -0.048 -0.084 -0.074** 0.059 -0.122
(1.05) (1.36) (0.25) (2.03) (0.09) (1.43)
FDI 0.023*** 0.020*** 0.054 0.001 -0.026 0.051*
(2.96) (3.12) (0.17) (0.11) (0.11) (1.74)
Inflation -0.001*** -0.001*** -0.001 -0.001*** -0.004*** -0.001
(3.66) (4.09) (0.25) (2.74) (2.98) (1.08)
Constant -0.719 -1.110*** -0.409 -1.020* -1.45 -0.677
(3.84) (3.65) (0.39) (1.81) (0.35) (1.23)
Observations 364 515 397 561 233 354
R-sq 0.50 0.53 . 0.23 . .
87
Table 3.7: OLS Regression of the Basic and Openness Equations and Liberalization
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic Equation plus Trade and FDI. Standard errors are
clustered at the country level. Each equation includes a dummy variable for country-year
observation after 1995 (GroupB) and interactions terms between GroupB and Log of Checks
(LCHECKSB) and GroupB and Polarization (POLARIZB). F-Tests of equality of coefficient
between the political variables, Log of Checks and Polarization, and the interaction terms are
given below. Absolute value of robust t-statistics are in parenthesis.
* significant at 10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1 2 3 4 5 6
Log of checks -0.033 -0.031 0.026 0.031 -0.004 0.014
(0.81) (.76) (1.05) (1.13) (0.19) (0.56)
Polarization 0.003 0.002 -0.068** -0.042 -0.008 -0.022
(0.11) (0.08) (2.28) (1.58) (0.53) (1.09)
LCHECKSB 0.018 0.029 0.069 0.059 0.048 0.022
(0.43) (0.67) (1.35) (1.43) (0.82) (0.31)
POLARIZB 0.021 -0.035 -0.051 -0.038
(0.76) (1.18) (1.02) (0.98)
GROUPB 0.129** 0.117** 0.000 -0.006 0.012 0.032
(2.55) (2.31) (0.01) (0.18) (0.29) (0.65)
Lgdp95 0.184*** 0.180*** 0.114*** 0.078*** 0.093 0.087***
(7.77) (7.59) (6.18) (6.18) (5.59) (2.96)
Inflation -0.001*** -0.001*** -0.001*** -0.001*** -0.061** -0.001*
(3.96) (3.94) (2.55) (2.96) (2.03) (1.80)
Trade -0.001 0.002*** -0.001
(0.64) (3.44) (0.74)
FDI 0.005 -0.002 0.009
(1.58) (0.36) (0.74)
Constant 1.070*** 1.010*** -0.516 -0.410*** -0.344 -0.276
(6.12) (5.73) (4.23) (4.91) (3.41) (2.06)
Observations 1000 956 724 668 455 396
R-squared 0.51 0.50 0.45 0.53 0.32 0.19
F-statistic
a
F(2,80) =
0.31
F(2, 77) =
0.73
F(2,79) =
0.91
F(2,74) =
1.02
F(1,58) =
0.67
F(1,53) =
0.09
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
88
Table 3.8: OLS Regressions of the Basic and Openness Equations, Legal Origin, and
Liberalization
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Standard errors are clustered
at the country level. Each regression includes a dummy variable for country-year observation
after 1995 (GroupB) and interactions terms between GroupB and Log of Checks (LCHECKSB)
and GroupB and Polarization (POLARIZB). Each regression includes dummy variables for
French, German, Scandinavian, and Socialist legal origins. F-Tests of equality of coefficient
between the political variables and the interaction terms are given below. Absolute value of
robust t-statistics are in parenthesis. * significant at 10%; ** significant at 5%; *** significant at
1%
Free Countries Partly Free Countries Not Free Countries
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of Checks -0.051 -0.052 -0.026 0.017 -0.01 0.007
(1.13) (1.2) (0.98) (0.63) (0.4) (0.25)
Polarization 0.015 0.015 -0.046** -0.043 0.007 0.02
(0.53) (0.54) (2.01) (1.69) (0.31) (0.92)
LCHECKSB -0.02 -0.005 0.081 0.069 0.059 0.069
(0.49) (0.12) (1.48) (1.52) (0.92) (1)
POLARIZB -0.01 -0.024 -0.042 -0.032 0*** 0***
(0.36) (0.85) (0.82) (0.78) (0.00) (0.00)
GROUPB 0.188*** 0.168*** 0.001 -0.006 0.025 0.041
(3.37) (3.06) (0.03) (0.17) (0.66) (1.05)
Lgdp95 0.173*** 0.165*** 0.135*** 0.082*** 0.098*** 0.065**
(7.67) (7.40) (7.06) (4.71) (4.77) (2.49)
Inflation -0.001*** -0.001*** -0.001* -0.001** -0.001 -0.001**
(3.10) (3.11) (1.95) (2.01) (1.73) (2.37)
Legor_fr -0.022 -0.033 -0.125** -0.022 -0.027 0.024
(0.39) (0.57) (2.38) (0.45) (0.6) (0.54)
Legor_ge 0.32*** 0.339*** -0.038 0.124 0*** 0***
(2.83) (3.07) (0.54) (1.77) (0.00) (0.00)
Legor_sc -0.108 -0.086 0*** 0*** 0*** 0***
(0.72) (0.57) (0.00) (0.00) (0.00) (0.00)**
Legor_so -0.29*** -0.291*** -0.131* -0.077 -0.281*** -0.17**
(5.60) (6.01) (1.95) (1.45) (4.44) (2.02)
FDI 0.005 -0.002 -0.008
(1.48) (0.34) (1.11)
Trade 0 0.002*** 0.001
(0.32) (3.33) (0.93)
Constant -0.963*** -0.88*** -0.566*** -0.418*** -0.376*** -0.262
(5.18) (4.30) (5.12) (4.18) (3.01) (1.67)
Observations 908 871 629 592 302 260
R-squared 0.57 0.57 0.49 0.53 0.5 0.34
F-statistic
a
0.25 0.47 1.32 1.26 0.84 1.00
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
89
Table 3.9: IV Regression of the Basic Equation and Liberalization
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. Regressions
A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B instruments
Lgdp95 with Distance from the Equator in Degrees (Disteq). Standard errors are clustered at the
country level. The Basic Equation includes the Log of Checks, Polarization, Inflation, and
Lgdp95. Each regression includes a dummy variable for country-year observation after 1995
(GroupB) and interactions terms between GROUPB and Log of Checks (LCHECKSB) and
GROUPB and Polarization (POLARIZB). F-Tests of equality of coefficient between the political
variables and the interaction terms are given below. Absolute value of robust t-statistics are in
parenthesis.*significant at 10%; **significant at 5%; ***significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Log of Checks 0.192** 0.466*** 0.024 0.273** 0.248 0.045
(2.04) (4.45) (0.24) (2.28) (1.17) (0.23)
Polarization 0.049 0.214*** 0.071 0.064 -0.363 -0.528
(0.77) (3.05) (0.48) (0.33) (1.19) (1.37)
LCHECKSB 0.052 0.155 0.089 0.376 0.168 -0.298
(0.35) (0.99) (0.55) (1.76) (0.46) (0.75)
POLARIZB 0.096 0.007 -0.006 -0.118
(0.87) (0.05) (0.03) (0.39)
GROUPB -0.116 -0.296 -0.221* -0.369*** 0.05 0.269***
(0.70) (1.73) (1.88) (2.64) (0.34) (2.30)
Inflation 0.000 -0.001*** 0.001 -0.003* 0.010*** -0.004**
(0.65) (2.06) (1.22) (1.82) (2.88) (2.41)
Instrument -0.790*** 5.527*** -0.661*** 2.813*** -0.284*** 4.875***
(23.51) (16.68) (16.52) (6.13) (4.48) (11.95)
Constant 11.264 6.362*** 10.143*** 6.425 7.421 5.441***
(59.45) (47.89) (47.85) (50.94) (21.09) (54.97)
Observations 491 526 478 600 246 407
R-squared 0.60 0.48 0.43 0.08 0.12 0.29
90
Panel B: Second Stage Regression of the Basic Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit
1 2 3 4 5 6
Lgdp95 0.178*** 0.232*** 0.161*** 0.152*** 0.221* 0.140**
(7.73) (4.86) (4.36) (2.12) (1.83) (2.41)
Log of Checks 0.011 -0.045 0.011 0.029 -0.036 -0.007
(0.28) (0.75) (0.33) (0.88) (0.80) (0.27)
Polarization -0.10 -0.20 -0.085** -0.082** 0.032 0.005
(0.37) (0.52) (2.15) (2.18) (0.80) (0.19)
LCHECKSB -0.06 -0.045 0.056 0.10 0.038 0.015
(1.19) (1.06) (0.92) (0.18) (0.52) (0.37)
POLARIZB 0.024 -0.015 0.004 0.010
(0.57) (0.33) (0.06) (0.19)
GROUPB 0.207*** 0.188*** -0.014 0.015 0.007 -0.010
(3.11) (3.66) (0.37) (0.33) (0.12) (0.29)
Inflation -0.001*** -0.001 -0.001*** -0.001 -0.004** -0.001
(3.34) (3.26) (3.30) (2.85) (2.64) (1.59)
Constant -1.08*** -1.419*** -0.831*** -0.768 1.105 0.632*
(5.75) (4.11) (3.31) (1.53) (1.56) (1.80)
Observations 491 526 478 600 246 407
R-squared 0.55 0.54 0.39 0.42 . .0.27
F-statistic
a
F(2,39)=
0.86
F(2,46)=
0.69
F(2,49)=
1.15
F(2,63)=
0.11
F(1,33)=
0.27
F(1,50)=
0.14
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
91
Table 3.10: IV Regression of the Openness Equation and Liberalization
The Openness Equation includes the Log of Checks, Polarization, Inflation, Lgdp95, Trade, and
FDI. Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Trade is instrumented
with the Log of the Frankel-Romer proxy of natural trade openness (Log FR). Standard errors
are clustered at the country level. The Basic Equation includes the Log of Checks, Polarization,
Inflation, and Lgdp95. Each regression includes a dummy variable for country-year observation
after 1995 (GroupB) and interactions terms between GROUPB and Log of Checks (LCHECKSB)
and GROUPB and Polarization (POLARIZB). F-Tests of equality of coefficient between the
political variables and the interaction terms are given below. Absolute value of robust t-statistics
are in parenthesis. * significant at 10%; ** significant at 5%; *** significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Log of
Checks
-0.033 0.401*** -0.002 0.269** 0.242 0.222
(0.81) (3.82) (0.02) (2.39) (1.09) (1.45)
Polarization 0.003 0.265*** 0.156 -0.156 -0.332 -0.397
(0.11) (3.74) (0.96) (0.90) (0.90) (1.35)
LCHECKSB 0.018 0.207 0.081 0.415** -0.065 -0.109
(0.43) (1.31) (0.43) (2.13) (0.16) (0.36)
POLARIZB 0.021 -0.034 -0.086 -0.207 Dropped Dropped
(0.76) (0.028) (0.36) (0.75)
GROUPB 0.129 -0.398** -0.252** -0.311** 0.018 -0.004
(2.55) (2.23) (2.00) (2.39) (0.12) (0.04)
FDI 0.005 0.030** 0.092*** 0.057*** 0.065* 0.063***
(0.23) (2.02) (4.62) (3.12) (1.90) (2.88)
Inflation 0.001* -0.001 0.002 -0.001 0.009** 0.000
(1.80) (1.53) (1.39) (0.62) (2.54) (0.37)
Instrument -0.820 5.935*** -0.564**** 2.573*** -0.233*** 3.855***
(15.44) (16.92) (11.17) (5.94) (3.42) (12.21)
Log FR 0.195*** 0.114** -0.016 0.147** 0.024* 0.276***
(3.35) (2.34) (1.68) (2.02) (1.73) (4.73)
Constant 10.884*** 5.918*** 9.60 5.838*** 7.014*** 4.574***
(4.02) (28.71) (34.19) (24.02) (18.15) (23.83)
Observations 364 515 397 561 233 354
R-squared 0.48 0.48 0.34 0.11 0.09 0.45
92
Panel B: First Stage Regressions of Trade
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Log of
Checks
7.380*** 7.998*** 5.118 3.943 10.074* 11.538**
(2.68)** (3.08) (1.03) (1.22) (1.82) (2.48)
Polarization -5.049 -7.348*** -10.158 -6.060 14.360 -12.843
(2.55) (4.20) (1.41) (1.19) (1.55) (1.43)
LCHECKSB -0.906 -7.702* -14.319 11.596** 3.148 -1.216
(0.22) (1.97) (1.70) (2.07) (0.31) (0.13)
POLARIZB 2.630 1.769 -1.802 -6.911
(0.77) (0.58) (0.17) (0.87)
GROUPB -1.524 9.094** 1.178 -7.601** 4.54 2.866
(0.31) (2.06) (0.21) (2.04) (1.23) (0.96)
FDI 2.648*** 2.237*** 10.693*** 6.706*** 5.241*** 6.76***
(5.22) (6.00) (12.04) (12.85) (6.16) (10.10)
Inflation -0.017 -0.006 -0.223*** -0.129*** -0.116 -0.001
(1.23) (0.045) (3.93) (2.95) (1.24) (0.03)
Instrument 5.353 15.281 -2.24 26.028** 1.510 39.240***
(1.433) (1.76) (1.00) (2.09) (0.89) (4.10)
Log FR 30.363*** 27.678*** -0.233 34.07*** -0.489 24.108***
(21.19) (22.98) (0.55) (16.35) (1.44) (13.60)
Constant -1.959*** -20.341*** 66.797*** 27.6*** 39.80*** -26.142***
(6.691) (3.99) (5.36) (3.96) (4.13) (4.49)
R-squared 0.64 0.61 0.32 .050 0.16 .
Observations 364 515 397 561 233 354
93
Panel C: Second Stage Regressions of the Openness Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The
dependent
variable is
Private
Credit.
1 2 3 4 5 6
Lgdp95 0.154*** 0.212*** 0.098*** 0.188** 0.270 0.208*
(6.81) (4.94) (2.90) (2.09) (0.63) (1.90)
Trade -0.002** -0.003** 0.002 0.001 0.003 -0.008*
(2.19) (2.54) (0.31) (0.53) (0.12) (1.91)
Log of
Checks
0.033 -0.003 0.006 0.018 -0.073 0.064
(1.48) (0.07) (0.14) (0.52) (0.19) (0.87)
Polarization -0.052** -0.044 -0.053 -0.085* 0.081 -0.126
(2.05) (1.42) (0.72) (1.88) (0.17) (1.44)
LCHECKSB -0.062 -0.053 0.019 -0.002 0.022 -0.032
(1.49) (1.26) (0.18) (0.02) (0.21) (0.26)
POLARIZB 0.049 -0.014 0.057 0.022
(1.12) (0.35) (1.34) (0.38)
GROUPB 0.107* 0.181*** -0.038 0.024 -0.011 0.031
(1.79) (3.16) (0.81) (0.45) (0.07) (0.51)
FDI 0.016** 0.012* 0.000 0.000 -0.031 0.049
(2.25) (1.92) (0) (0.02) (0.18) (1.77)
Inflation -0.001*** -0.001 0.000 -0.001 -0.004** -0.001
(3.73) (3.85) (0.30) (2.31) (2.34) (1.11)
Constant 0.746*** -1.119 -0.551* -1.063 -1.54 -0.665
(4.13) (3.88) (1.74) (1.76) (0.41) (1.23)
Observations 364 515 397 561 233 354
R-squared 0.52 0.57 0.43 0.20 . .
F-statistic
a
F(2,32)=
1.59
F(2,45)=
0.92
F(2,42)=
0.93
F(2,60)=
0.13
F(1,30)=
0.05
F(1,46)=
0.07
a
For Not Free Countries, only the constraint that the Log of Checks is equal to LCHECKSB is
tested.
94
Table 3.11: OLS Regression of the Basic and Openness Equations
The Basic equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Data is averaged over 5-year
intervals. Absolute value of robust t-statistics are in parenthesis. * significant at 10%; **
significant at 5%; *** significant at 1%.
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The Dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of checks -0.0868 -0.089 0.09** 0.098** 0.068 0.083
(1.77) (1.7) (2.12) (2.34) (0.94) (0.9)
Polarization -0.012 -0.012 -0.166*** -0.094** -0.132 -0.113
(0.45) (0.42) (2.64) (2.06) (1.52) (1.25)
Inflation -0.001*** -0.001*** -0.001*** -0.001*** 0 0
(3.87) (3.63) (4.23) (5.28) (1.49) (1.65)
Lgdp95 0.205*** 0.202*** 0.126*** 0.07*** 0.11*** 0.103**
(10.55) (9.47) (8.01) (4.62) (7.08) (2.65)
Trade 0 0.002*** -0.001
(0.8) (4.65) (1.26)
FDI 0.01 -0.004 0.006
(1.7) (0.71) (0.89)
Constant -1.132*** -1.094*** -0.613*** -0.39*** -0.459*** -0.36**
(8.47) (7.01) (6.00) (4.58) (5.03) (2.09)
Observations 187 180 109 102 70 61
R-squared 0.52 0.51 0.51 0.64 0.35 0.2
95
Table 3.12: OLS Regression of the Basic and Openness Equations and Legal Origin
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Data is averaged over 5-year
intervals. Each regression includes dummy variables for French, German, Scandinavian, and
Socialist legal origins. Absolute value of robust t-statistics are in parenthesis. * significant at
10%; ** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
Basic Openness Basic Openness Basic Openness
The Dependent
variable is Private
Credit.
1 2 3 4 5 6
Log of Checks -0.122** -0.127** 0.006 0.087 0.081 0.144
(2.07) (2.22) (0.13) (1.84) (1.08) (1.4)
Polarization -0.003 0.004 -0.099** -0.09 0.283 -0.058
(0.08) (0.13) (2.01) (1.82) (1.8) (0.22)
Lgdp95 0.203*** 0.189*** 0.152*** 0.074*** 0.113*** 0.052*
(8.64) (7.64) (9.65) (3.61) (5.04) (1.84)
Inflation -0.001*** -0.001** -0.001 0 0 0**
(2.80) (2.40) (1.43) (1.23) (1.18) (2.32)
Legor_fr 0.014 -0.015 -0.159*** -0.014 -0.038 0.037
(0.3) (0.32) (3.93) (0.33) (0.69) (0.64)
Legor_ge 0.309*** 0.337*** -0.102 0.128* 0*** 0***
(3.65) (3.94) (1.75) (1.87) (0.00) (0.00)
Legor_sc -0.103 -0.092 0*** 0*** 0*** 0***
(1.08) (0.98) (0.00) (0.00) (0.00) (0.00)
Legor_so -0.154*** -0.197*** -0.159 -0.172 0*** 0***
(2.94) (3.96) (1.21) (1.23) (0.00) (0.00)
FDI 0.013 -0.005 -0.008
(1.66) (1) (0.18)
Trade 0 0.002*** 0.001
(0.36) (3.92) (0.99)
Constant -1.101*** -0.988*** -0.665*** -0.41*** -0.486*** -0.245
(6.04) (4.59) (6.69) (3.91) (3.26) (1.48)
Observations 168 163 92 88 41 35
R-squared 0.58 0.58 0.57 0.64 0.61 0.38
96
Table 3.13: IV Regression of the Basic Equation
The Basic Equation includes the Log of Checks, Polarization, Inflation, and Lgdp95. Regressions
A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B instruments
Lgdp95 with Distance from the Equator in Degrees (Disteq). Data is averaged over 5-year
intervals. Absolute value of robust t-statistics are in parenthesis. * significant at 10%; **
significant at 5%; *** significant at 1%
Panel A: First Stage Regressions of Lgdp95
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Log of
checks
0.499** 1.047*** -0.062 0.400 0.650 -0.291
(2.21) (4.81) (0.22) (1.21) (1.08) (0.45)
Polarization 0.111 0.279** -0.526 -0.278 -0.918 -3.535
(0.72) (1.99) (1.40) (0.51) (0.44) (0.70)
Inflation 0.001 -0.002 -0.002 -0.005 0.008 -0.002
(0.49) (1.08) (0.77) (1.56) (0.61) (0.99)
Instrument -0.733*** 4.123*** -
0.781***
0.126 -0.100 5.25***
(8.13) (6.40) (6.25) (0.10) (0.54) (5.04)
Constant 10.705*** 6.117*** 10.745 6.87*** 6.236*** 5.370***
(21.708) (23.03) (16.32) (19.70) (6.12) (21.49)
Observations 85 93 69 87 33 62
R-squared .56 .54 .42 -0.002 -0.069 .32
Panel B: Second Stage Regressions of the Basic Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
The dependent
variable is
Private Credit.
1 2 3 4 5 6
Lgdp95 0.202*** 0.272*** 0.204*** 1.175 0.554 0.146***
(6.02) (6.62) (5.72) (.10) (.50) (3.01)
Lchecks -0.067 -0.191** 0.059 -0.346 -0.186 0.035
(1.06) (2.22) (1.15) (0.07) (0.23) (.46)
Polarization -0.009 0.023 -0.099 0.162 0.398 0.299
(0.25) (0.63) (1.09) (0.05) (0.35) (1.48)
Inflation -0.002** -0.002** -0.001 0.004 -0.007 -0.000
(2.49) (2.83) (.13) (0.07) (0.72) (1.03)
Constant -1.141*** -1.556*** -1.160*** -7.847 -2.990 -0.672**
(4.69) (5.79) (4.63) (0.09) (0.48) (2.35)
Observations 85 93 69 87 33 62
R-squared .51 .57 .40 . . .2
97
Table 3.14: IV Regression of the Openness Equation
The Openness Equation includes the Log of Checks, Polarization, Inflation, Lgdp95, Trade, and
FDI. Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Trade is instrumented
with the Log of the Frankel-Romer proxy of natural trade openness (Log FR). Data is averaged
over 5 year intervals. Absolute value of robust t-statistics are in parenthesis. * significant at 10%;
** significant at 5%; *** significant at 1%
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Panel A: First Stage Regressions of Lgdp95
Log of
Checks
0.586* 1.092*** -0.124 0.595* 0.445 0.348
(1.93) (4.66) (0.41) (1.84) (0.63) (0.86)
Polarization 0.034 0.289** -0.168 -0.049 -2.341 -2.381
(0.17) (2.00) (0.40) (0.10) (0.48) (0.76)
FDI 0.011 0.009 0.212*** 0.068* 0.208 0.104
(0.22) (0.27) (3.31) (1.73) (1.26) (1.92)
Inflation 0.002 -0.002 -0.001 -0.003 -0.011 -0.000
(0.67) (0.95) (0.33) (1.14) 90.55) (0.31)
Instrument -0.704*** 4.11*** -0.409** 0.076 -0.058 4.239***
(4.07) (5.84) (2.26) (0.06) (0.30) (6.35)
Log FR 0.084 -0.017 -0.015 0.332 0.026 0.287**
(0.56) (0.17) (0.60) (1.66) (0.66) (2.38)
Constant 10.235*** 6.073*** 8.702*** 5.600*** 5.940 4.268***
(12.67) (14.19) (8.87) (8.18) (5.46) (10.54)
Observations 63 91 57 83 31 54
R-squared 0.39 0.53 0.40 0.05 -0.06 0.48
Panel B: First Stage Regressions of Trade
Log of
Checks
3.207 3.92 -9.896 19.684* 17.239 11.386
(0.41) (0.55) (0.73) (1.90) (0.99) (.78)
Polarization -6.642 10.93** -7.365 -12.552 21.460 59.679
(1.27) (2.49) (0.39) (0.78) (0.18) (.53)
FDI 3.167** 2.772** 23.419*** 8.360*** 13.604*** 10.703***
(2.55) (2.71) (8.12) (6.65) (3.36) (5.47)
Inflation -0.014 -0.001 -0.005 -0.128 -0.160 0.023
(0.23) (0.01) (0.05) (1.44) (0.32) (0.48)
Instrument -15.704*** 25.232 9.248 57.455 2.019 25.835
(3.51) (1.18) (1.14) (1.45) (0.43) (1.08)
Log FR 35.30*** 27.922*** -0.705 45.484*** -0.831 20.535***
(9.07) (9.13) (0.64) (7.14) (0.85) (4.75)
Constant 32.91 -16.505 0.018 -61.634*** 31.318 -20.366
(1.57) (-1.27) (0) (2.83) (1.17) (1.40)
Observations 63 91 57 83 31 54
R-squared 0.68 0.060 0.61 0.59 0.36 0.441
98
Panel C: Second Stage Regressions of the Openness Equation
Free Countries Partly Free Countries Not Free Countries
A B A B A B
1 2 3 4 5 6
Lgdp95 0.215*** 0.267*** 0.067 0.694 -6.050 0.180**
(4.57) (5.56) (1.36) (0.45) (0.02) (2.48)
Trade -0.002** -0.002** 0.000 -0.003 -0.197 -0.010***
(2.02) (2.47) (0.32) (0.27) (0.02) (3.23)
Log of Checks -0.039 -0.149 -0.001 -0.187 6.199 0.087
(0.43) (1.40) (0.02) (0.28) (0.02) (0.38)
Polarization -0.067) -0.051 -0.40 -0.130 -9.907 0.720
(1.47) (1.27) (0.73) (0.40) (0.02) (1.32)
FDI 0.028** 0.023*** 0.040 -0.006 3.961 0.107***
(2.48) (2.92) (1.42) (0.15) (0.02) (3.37)
Inflation -0.002** -0.002*** -0.00 0.001 -0.103 -0.000
(2.55) (2.74) (1.58) (0.22) (0.02) (.98)
Constant 1.153*** -1.409*** -0.281 -4.186 42.525 -0.443***
(3.27) (4.59) (1.86) (0.44) (0.02) (11.19)
Observations 63 91 57 83 3131 54
R-squared 0.48 0.60 0.69 . . .
99
Table 3.15: OLS Estimation of Ownership
The dependent variable is the percentage of the publicly listed firms in each country that are
widely held. The Basic equation includes the Log of Checks, Polarization, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Explanatory variables are
averaged over the years 1996-2000. Absolute value of robust t-statistics are in parenthesis. *
significant at 10%; ** significant at 5%; *** significant at 1%
Widely-held10 Widely-held20
Basic Openness Basic Openness
1 2 3 4
Log of Checks 16.076 25.756 10.437 21.396
(0.96) (1.45) (0.56) (1.14)
Polarization -9.407 -11.815 -12.228 -14.42
(1.13) (1.41) (1.37) (1.58)
Democracy -9.053 5.045 14.778 0.539
(0.89) (0.33) (.11) (0.03)
Lgdp95 7.228 8.921* 12.931 14.733*
(1.45) (1.78) (1.67) (1.92)
Trade -0.098 -0.104
(1.20) (0.98)
FDI -0.710* -0.842
(1.78) (1.50)
Constant -48.556 78.850 -74.574 -107.537
(1.31) (1.51) (1.17) (1.26)
Observations 28 28 28
R-squared 0.21 0.31 0.29 0.38
100
Table 3.16: OLS Estimation of Ownership and Legal Origin
The dependent variable is the percentage of the publicly listed firms in each country that are
widely held. The Basic equation includes the Log of Checks, Polarization, and Lgdp95. The
Openness Equation includes the Basic equation plus Trade and FDI. Each regression includes
dummy variables for French, German, Scandinavian, and Socialist legal origins. Explanatory
variables are averaged over the years 1996-2000. Absolute value of robust t-statistics are in
parenthesis. * significant at 10%; ** significant at 5%; *** significant at 1%
Widely-held10 Widely-held20
1 2 3 4
Lchecks -2.816 0.162 -4.273 -0.080
(0.12) (0.01) (0.19) (0.00)
Polarization 1.024 -1.042 -4.419 -6.656
(0.09) (0.11) (0.37) (0.57)
Democracy 76.613 2.250 -21.993 -2.911
(1.14) (0.13) (1.41) (0.13)
Trade Openness -0.182* -0.178
(1.85) (1.56)
FDI 0.70 -0.073
(0.12) (0.10)
Lgdp95 6.885 10.452 10.486 14.115
(0.89) (1.41) (1.15) (1.53)
Legor_fr -23.057 -26.854 -22.144 -25.307
(1.56) (1.65) (1.43) (1.52)
Legor_ge -9.666 -15.588 -2.499 -8.606
(0.53) (0.84) (0.12) (0.43)
Legor_sc -38.691** 42.617** -21.637 -25.260
(2.33) (2.56) (1.27) (1.52)
Constant -8.444 -48.152 -20.735 -62.556
(0.13) (0.66) (0.25) (0.60)
Observations 28 28 28 28
R-squared 0.48 0.53 0.40 0.49
101
Table 3.17: IV Estimation of Ownership
The dependent variable is the percentage of the publicly listed firms in each country that are
widely held. The Basic equation includes the Log of Checks, Polarization, and Lgdp95.
Regressions A instruments Lgdp95 with Log of Settler Mortality (Lsettler). Regressions B
instruments Lgdp95 with Distance from the Equator in Degrees (Disteq). Explanatory variables
are averaged over the years 1996-2000. Absolute value of robust t-statistics are in
parenthesis.*significant at 10%; **significant at 5%; ***significant at 1%
Panel A: First Stage Regressions of Lgdp95
Widely-held10 Widely-held20
1 2 3 4
Lchecks -0.088 2.445** -0.088 2.445**
(0.09) (2.44) (0.09) (2.44)
Polarization 0.308 -0.203 0.308 -0.203
(0.85) (0.44) (0.85) (0.44)
Democracy -0.582 0.784 -0.582 0.784
(1.03) (0.84) (1.03) (0.84)
Lsettler/Disteq -0.697* 1.294 -0.697* 1.294
(2.02) (0.43) (2.02) (0.43)
Constant 12.22*** 5.440** 12.22*** 5.440**
(5.45) (2.69) (5.45) (2.69)
Observations 11 11 11 11
R-squared 0.70 0.52 0.70 0.52
Panel B: Second Stage Regressions of the Basic Equation
Widely-held10 Widely-held20
1 2 3 4
Lgdp95 -4.979 24.764 5.13 21.715
(0.23) (0.32) (0.28) (0.30)
Lchecks 44.393 -33.891 37.589 -26.314
(0.89) (0.16) (1.03) (0.13)
Polarization 7.724 -1.598 8.802 -9.671
(0.44) (0.07) (0.78) (0.49)
Democracy -6.009 -20.508 -0.633 -29.740
(0.19) (0.5) (0.02) (0.56)
Constant 31.800 -144.603 -53.14 -98.926
(0.16) (0.33) (0.27) (0.24)
Observations 11 12 11 12
R-squared 0.53 1.30 0.67 0.43
102
Chapter 4: Securities Laws, Firm Size, and Capital Issuance
1. Introduction
The dictum that institutions matter no longer surprises the vast majority of
economists. Institutional disparities in country organization go a long way in explaining
differences in firm behavior. This chapter seeks to add to the literature investigating the
role of institutions in firm level outcomes by examining the effects of securities laws on
capital issuance, the means by which firm finance their growth.
Several studies have found a role for legal efficiency in the ability of firms to
finance their growth from external sources. Rajan and Zingales (1998) pursue the
relationship between financial development and firm size. The authors find that
“industrial sectors that are relatively more in need of external finance develop
disproportionately faster in countries with more developed financial markets”. Kumar,
Rajan and Zingales (2001) investigate the linkage between firm size and judicial
efficiency. The empirical study mingles institutional and technological explanations for
the determinants of firm size in fifteen European countries. The researchers show that (1)
countries with more efficient judicial systems have larger firms (2) larger firms tend to
operate in capital-intensive industries; however, (3) capital intensive firms in efficient
judicial systems tend to be smaller.
Laeven and Woodruff (2004) focus on Mexican firms, which share the same
national commercial laws, however, are exposed to legal environments with varying
degrees of efficiency and enforcement at the state level. The authors’ data set allows
them to examine firms that are both large (corporations) and small (single person
proprietorships). States with better legal enforcement are found to have larger firms.
103
Laeven and Woodruff hypothesize that “better legal systems increase the investment of
firm owners by reducing the idiosyncratic risk they face.” Reduction of risk is associated
with the benefits of financial development – minority shareholder protection, better
informational exchange between management and investors and improved contracting
among firms and their suppliers. These benefits lead to greater capital accumulation,
higher levels of external finance, and growth in firm size.
Finally, Beck, Demirguc-Kunt, and Levine (2002) explore judicial independence
from the central government and the ability of the legal system to adapt to changing
circumstances. Their paper improves on previous research by incorporating the World
Business Survey (WBES). This data source covers 4000 small, medium and large firms
in 38 countries. The firms are asked about the financing obstacles in general, in meeting
collateral requirements, bureaucratic paperwork and acquiring long-term loans. The
authors show that firms in common law countries are much more able to garner external
finance than those in civil law countries. A limitation of their study is that their results
rely on “perceptions of financing obstacles” from the firm survey as opposed to actual
restrictions. These perceptions may be caused by cultural or environmental factors that
are independent of legal attributes.
Our paper contributes to the literature in legal institutions and firm size by
investigating the role of security laws on the ability of firms to raise external finance by
issuing capital. The paper takes advantage of two unique data sets in order to accomplish
this task. The first data set is compiled by La Porta, Lopez de Silanes, and Shleifer
(2004) and quantifies the different laws regulating the issuance of new equity in 49
countries. We refer to the authors as LLS through out the chapter. Securities regulations
104
are divided into laws governing public and private enforcement. Private security laws aid
in the ability of private agents to contract among themselves, while public security laws
regulate the organization of a public enforcers analogous to the Securities and Exchange
Commission in the US. LLS conclude that while securities laws that enhance private
enforcement of contracts improve stock market development, laws governing public
enforcement have no effect on development in general, or the ability of small firms to
raise external finance in particular.
The second data set features firm accounting information and primary issues and
is compiled by Knill (2004). It includes small, medium, and large firms of 54 countries
from the years 1996-2003. This data set allows us to analyze access to finance at the firm
level and to examine the role of securities legislation on capital issuances of publicly
listed firms of all sizes.
Our first objective is to investigate if a country’s securities laws are a significant
determinant of the probability that the firm issues a security. Second, we asses if
securities laws have asymmetric influence on the issuance of small and large firms and
we determine if the asymmetry is found in both industrialized countries and emerging
markets. Finally, we attempt to confirm at the firm level the conclusion of LLS that only
private enforcement matters in increasing access to capital. We find that the composition
of laws governing security issuance in each country significantly affects the probability
that a firm will issue equity. Furthermore, in contrast to LLS, we show that public
enforcement of security laws is a much more important determinant of firm access to
capital markets than private enforcement. It is our conclusion that aggregate
macroeconomic indicators of stock market size used in the LLS paper belie the
105
importance of a strong regulator in increasing access to capital, particularly for emerging
market firms. Section 2 discusses our economic approach to estimating access to finance.
Section 3 describes the data. Section 4 presents the empirical testing strategy and Section
5 provides the results. Section 6 provides a robustness check of our results and Section 7
concludes.
2. Methodology
A challenge to research examining access to external finance is that firm level
analysis is limited to a few large publicly listed firms per country. Our data set allows us
to investigate the impact of security laws on a wide representation of listed firms and
countries. We investigate whether the probability of capital issuance is dependent on a
country’s security laws, firm characteristics, and macroeconomic factors.
We begin by calculating the need for external funds for firms in our sample in
period t. We believe that firms that need external funds to make investments and do not
issue securities are more financially constrained than the other firms in our data set.
Higgens (1977) presents a financial planning model that calculates a firm’s need for
external funds in period t. This equation relates the firm growth rate to its need for
external funds and derives the external funds necessary from the “percentage of sales”.
Demirguc-Kunt and Maksimovic (2002) also use this financial planning model to identify
firms that require external finance to meet their investment needs. The ‘external funds
necessary’ (or EFN) for firm i in period t is calculated as follows:
) )( ( ) )( / ( ) )( / (
, , , 1 , , , , 1 , , 1 , 1 , , t i t i t i t i t i t i t i t i t i t i t i t i
RR S M S S S L S S S A EFN ? ? ? ? =
? ? ? ?
(1a)
where A
t-1
is the total assets of the firm in time t-1, S
t-1
and S
t
are the sales of the firm in
times t-1 and t respectively, L
t
is the liabilities of the firm in time t, M
t
is the profit
106
margin of the firm as defined by net income divided by sales for time t, RR is the
retention ratio for the firm. As noted by Demirguc-Kunt and Maksimovic (2002) two
simplifying assumptions are made in order for this methodology to be implemented.
First, both the asset utilization (A/S) and the profit margin of the firm must remain
constant per unit of sale. Second, the use of the formula to discern additional funds
necessary depends on true values of assets being reported (relative to their depreciable
basis).
Using firm-specific information to measure S
t
is problematic because Sales in
period t could be determined by weak securities laws that reduce access to external
finance, which, is the very situation that we are testing. Therefore, we approximate
industry wide rates of desired growth in Sales from t-1 to t using US firm level data. We
assume that the US industry growth rates are a good proxy for desired sales growth in
other countries using the logic of Rajan and Zingales (1998). The authors justify the
external dependence of US firms as a proxy for optimal demand for external funds
because the US has the most advanced capital markets in the world. Thus, US firms are
operating in as close to a frictionless financial environment as there is in the world.
Furthermore, Rajan and Zingales reason that industry wide measures of external
dependence are reasonable by stating, “Therefore, much of the demand for external funds
is likely to arise as a result of technological shocks that raise an industry’s investment
opportunities beyond what external funds can support. To the extent these shocks are
worldwide; the need for funds of the US firms is a good proxy.” We assume that since
US firms can access external finance optimally, they are able to increase their sales at an
optimal rate as well. Moreover, optimal rates of growth persist across industries in other
107
countries. For each US firm in our data set, we measure growth in Sales from year t-1 to
year t. Next, we average individual firm growth rates in the same industry and size
classification. The term
US
m j
g
,
is the average growth rate in Sales for firms in industry j of
size m. We measure separate growth rates for each size classification m, where m is
either small or large, because small and large firms in the same industry can grow at
drastically different rates.
21
We use
US
m j
g
,
to approximate desired growth in Sales for a firm outside the US in
industry j and size m. The predicted value of Desired Sales for firm i is equal to
) (
1 , ,
,
*
?
=
t i
US
m j
t i S g S (2)
We substitute t i S ,
*
for actual Sales in period t into equation (1a). Equation (1b) measures
the external funds necessary to increase Sales in year t-1 to Desired Sales in year t.
) )( ( ) )( / ( ) )( / (
,
*
, , 1 ,
*
,
*
, , 1 ,
*
, 1 , 1 ,
*
, t i t i t i t i t i t i t i t i t i t i t i t i
RR S M S S S L S S S A EFN ? ? ? ? =
? ? ? ?
(1b)
We drop all US firms from the sample to avoid bias in our empirical model.
Our econometric model is based on the findings of Korajczyk and Levy who
examine capital structure choice in the presence of financing constraints for US firms
(2003). They find that a firm’s choice of security issuance (either debt or equity) is based
on macroeconomic conditions and firm specific information. We use the independent
variables found in the paper to test the effect of securities laws on access to external
funds. However, our econometric approach differs slightly from Korajczyk and Levy
(2003) because our dependent variable lumps the decision to issue debt or equity into one
action. Thus, we are not modeling capital structure choice and can focus our estimation
21
Size is not a concern for Rajan and Zingales (1998) because they only have data for large
publicly traded firms.
108
entirely on financing constraints that prevent a firm in the panel from acquiring external
finance when it wants to.
We use a probit model to analyze the event that a firm i issues a security in time t.
Y=1 if the firm issues a security, and equals 0 if the firm does not issue.
The model we want to estimate is then
) (
) 1 ( Pr
6 5 4 , 3 1 , 2 1 , 1
,
t j k t i t k t i
t i
T I SECURITIES EFN Z F
Y ob
? ? ? ? ? ? ? + + + + + + ?
= =
? ?
(3)
? represents the standard normal distribution. The constant term is defined by ? .
1 , ? t i
F
is a vector of lagged firm-specific variables such as cash flow, debt/asset level, size,
profitability, risk, uniqueness of assets, trade credit and asset tangibility.
1 , ? t k
Z is a
vector of lagged macroeconomic variables such as GDP, corruption, shareholder rights,
the efficiency of the judiciary and the availability of domestic credit and foreign capital.
*
,t i
EFN , as described above, measures the external funds needed to increase Sales in year
t-1 to Desired Sales in year t. SECURITIES is the primary variable of interest and
represents the composition of securities laws in country k.
j
I represents the industry
indicators and
t
T is time dummies. Equation (3) allows us to investigate if the
composition of securities laws in Country k increases the probability that a firm will issue
a security when it requires external funds to finance its investments.
In addition, to estimating Equation (3) for the entire sample of non-US firms in
the data, we also estimate the equation separately for the economically advanced G 10
countries and emerging markets. We divide the data set into these two sub samples in
order to investigate whether differences in institutional quality of the legal system
109
between G10 and emerging market countries affect the impact of securities laws on
capital issuance.
3. Data
Knill (2004) collects observations for common stock, non-convertible debt,
convertible debt, non-convertible preferred stock and convertible preferred stock issued
domestically. The data covers all domestic issues of securities around the world from the
time period 1/1/1996 through 3/31/2003 and is collected from the SDC Global New
issues database. Table 4.1 shows the amount of each type of security issued by country.
Financials for the companies issuing domestically in a given year are from REUTERS.
This data set enables us to have a much richer sample of global new issues around the
world of both smaller and larger firms than afforded by SDC Platinum alone. REUTERS
provides financial information on all publicly traded firms for the majority of countries in
the world and does not suffer from the bias toward large firms to the extent that other
international databases such as Worldscope/Datastream/Research Insight do. In fact,
REUTERS even covers pink sheets and OTC/Bulletin Board firms whereas the others do
not. As such, the coverage is much more comprehensive and the average firm size is
much smaller.
In order to address sample selection bias from a data set composed entirely of
firms that issue domestic capital in a given year, Knill also collects data on firms not
issuing capital during this period of time to represent those public companies that either
cannot issue capital or are internally ‘wealthy’ to the point where there are financially
unconstrained. For emerging market country firm-year observations of the non-issuing
firms, financials are collected for 1996-2003 for the most exhaustive list of firms for each
110
country as possible from REUTERS, collecting the exact same data utilized for the issuer
dataset. Developed country firm-year observations are collected from Worldscope, due
to the inability of REUTERS to provide such large amounts of data given the
practitioner-oriented setup of this information-rich database. This is not believed to
cause bias due to the careful matching of accounting information and the quality of
information that is provided for the countries. Therefore, our data set includes security
issuance (or non-issuance) and financial accounting information for all listed firms
available in the REUTERS and Worldscope databases.
Following Korajczyk and Levy (2003), financial services are excluded due to the
special circumstances of their asset base and utility firms (Macro Industry: Financial
Services, Real Estate and Energy and Power) due to the abnormal stability and
predictability of cash flow. Knill also excludes those firms that have gone bankrupt due
to the special set of issues that are included in capital structure determination when a
company is failing
22
. This follows the methodology of Asquith et al. (1994) who find
that such situations generally cause a major restructuring of capital structure outside of
the scope of financial constraint relaxation. Lastly, Initial Public Offerings (IPO)s are
excluded. Welch (2002) finds that the firms who undertake IPOs find themselves in
unique environment, similar to those of the other deletions, which hosts a different set of
issues than the post-IPO period. Including these firms would bias the results.
The only firms not covered in REUTERS are those that have gone bankrupt or
have merged with another firm. The first group has deliberately been excluded from the
sample as mentioned above. The second group would only be a problem if the issuing
22
Firms going bankrupt would have additional difficulty obtaining capital, which would
convolute results.
111
company had acquired a firm in the sense that the capital structure may have changed
thus changing the financial environment examined. A cross-examination of the Global
New Issues database with the Mergers & Acquisitions database provides information on
whether any of the firms in the sample have been involved in a merger or acquisition.
Our dependent variable is Capital issuance. Capital Issuance is a dummy
variable, which equals 1 if firm i issued some type of capital (common stock, non-
convertible debt, convertible debt, non-convertible preferred stock and convertible
preferred stock) in time t and 0 otherwise. We explain the probability of capital issuance
by including as independent variables indexes of securities laws, firm specific variables
and macroeconomic indicators. We discuss each type of explanatory variable below. A
detailed list of variables and definitions is given in Appendix D.
Securities laws fall into two broad categories, public enforcement and private
enforcement. Private Enforcement is calculated in the La Porta et al. (2004) data set by
combining indexes of disclosure and liability. The disclosure index covers five distinct
areas (1) insiders compensation; (2) ownership by large shareholders; (3) inside
ownership; (4) contracts outside the normal course of business; and (5) transactions with
related parties. The liability index is the average of four levels of accountability in the
issuance of stock. These levels of accountability summarizes the burden proof for the
issuer of the stock (bdn_iss), the company directory (bdn_dir), the distributor (bdn_dis),
and the accountant (bdn_acc).
Public Enforcement regulates the supervisory behavior of the main regulatory
government authority in charge of the stock market. Public Enforcement is calculated by
the combining the following four sub indexes: (1) supervisory index; (2) investigative
112
powers index; (3) non criminal sanctions index (orders), and (4) criminal sanctions
index. The supervisor index is assessed on its independence from the central government
and that all hirings and firings are given due process (appointment), whether the
supervisor only regulates stock markets and not banks too (focus), and if the supervisor
has the authority to regulate primary offerings and listings on the stock market
(regulatory powers). The investigative index assesses whether the supervisor can
subpoena documents and witnesses. The orders index involves non criminal sanctions
for violations of disclosure standards, such as compensating investors for losses, or
instituting recommendations of the supervisor. The orders can be given to the issuer
(ord_iss), distributor (ord_dis) or accountant (ord_acc). Finally, the criminal index is a
measure the supervisor’s ability to impose criminal sanctions against the director
(crim_dir), distributor (crim_dis), issuer (crim_iss) and accountant (crim_acc).
Along with securities laws, access to capital is also explained by firm specific
variables. REUTERS obtains firm financial statistics for listed firms from country stock
exchanges. Variable definitions vary from country to country and are measured in
different currencies. To make variables comparable across countries, firm variables are
scaled by total assets unless otherwise noted.
As many empiricists have attributed size as a determinant of capital structure, we
assign size categories based on Total Assets. Korajczyk and Levy (2003) and Baker and
Wurgler (2002) find a positive relationship between leverage and size. Titman and
Wessels (1988) find that size influences not only the extent of leverage but also the type.
113
Our proxy for this follows both Titman and Wessels (1988) and Rajan and Zingales
(1995) and is calculated as total assets/GDP
23
.
The ratio of Cash to Total Assets describes the feasibility of using internal funds
to finance growth. According to the pecking order of capital structure choice (Myers,
1984) firms prefer to finance investments from internal funds, then bank loans, and
finally from capital issuance. We expect that the availability of internal funds will
decrease the likelihood that a firm would issue a security in a given year.
Uniqueness of assets is included based on the theory that a high uniqueness
increases the expected costs of bankruptcy. As assets of a company become more
distinctive, the ability to sell those assets when necessary (i.e. the liquidity of those
assets) decreases thus increasing liquidity risk. Within-country industry averages are
used in those cases where there is missing data. This should not be problematic due to
the uniformity of the nature of assets in firms within the same industry.
Profitability of firms would be an obvious influence on firms inasmuch as this
impacts how well a firm can either pay interest and/or dividends. Titman and Wessels
(1988) provide two measurements for this variable that are applicable universally. They
are operating income divided by sales and operating income divided by total assets. We
only provide results for profitability based on sales for brevity. We also include standard
deviation of the firm’s profitability ratio over the three years prior to issue (Risk) as an
independent variable. Firms with riskier profitability are less likely to issue capital.
Also relevant to capital structure determination is Asset Tangibility. This refers to
how palpable the assets of a firm are and relates to capital structure concerns through its
23
This is done annually so that firms may switch size groupings over years. The analyses are also done
using average size of the eight year periods. As results are unchanged, they are omitted for brevity.
114
limitations on debt levels due to the ability to provide collateral. A firm has less
collateral the less tangible its assets are. This, arguably, could be said to increase the
probability of bankruptcy due to the inability to obtain funds when there are especially
needed. This follows logically from the fact that a company without material assets
would not be able to liquidate to obtain the necessary funds to pay off debtors if it were
necessary. This variable is created by calculating fixed assets divide by book value of
assets (following Rajan and Zingales 1995). Once again, within-country industry
averages are used in those cases where there is missing data. For the same reasons given
above justifying the rationale for industry average substitution as proxies for uniqueness
of assets, industry averages are suitable proxies here.
To correct for any additional access a firm might have in other nations which
might affect financial constraints (Lins et al., 1999) it is vital to include an indication of
whether a firm has listings in other countries (i.e. ADR on a U.S. stock exchange). We
include a dummy variable for Cross listing that takes on a value of 1 if a firm is listed on
an exchange outside of its nation and 0 otherwise. Finally, differences in industry
classification are avoided by using as industry indicator the SDC Platinum Macro
industry code as our categorization. The industry dummy is included to account for any
industry fixed effects.
Given the fact that there are over 20,000 firms in our sample, it is not surprising
that the range of firm-level statistics such as cash, uniqueness, profitability and risk span
a range that is considerable in size. Smaller firms seem to have much more leverage than
their larger peers (Rajan and Zingales, 1995). Profitability and risk for the smaller firms
are considerably larger, reflecting the higher growth rate of the smaller firms (and based
115
on the fact that the figure is scaled by assets, controlling for size). The majority of the
sample is not cross listed. Table 4.2 summarizes the firm specific variables and
demonstrates the incredible range of variation in the financial accounting data.
Based on results from such papers as Korajcyk and Levy (2003) and Booth,
Demirguc-Kunt, and Maksimovic (1999), we include macroeconomic factors to capture
their impact on capital structures in different countries. All of our country-level control
variables are averaged over the years t-1 through t-3. Unless otherwise stated,
macroeconomic indicators are obtained via the International Finance Corporation (IFC).
GDP growth is the percentage growth in gross domestic product per capita and is
included to control for business cycle effects on issuance. Securities issuance tends to
occur counter cyclically because in good times firms have enough cash on hand to fund
investments internally, while in bad times they are more likely to raise money for
external capital markets.
We also control for the availability of domestic and foreign capital in each
country in order to isolate the role of securities laws on access to capital from both the
size of a country’s domestic financial markets and the openness of the country to foreign
funds. Domestic capital is the sum of market capital and private domestic credit, scaled
by GDP. Market Cap is equal to the Listed shares * their value and domestic credit is
credit extended to banks and other financial intermediaries scaled. Foreign capital is all
foreign investment – Foreign direct investment, Foreign portfolio investment, and foreign
bank lending and official flows – scaled by GDP.
Additionally, LLS note that better law enforcement could be associated with
increased access to capital regardless of the content of the securities laws. To focus the
116
regression analysis on the relationship between the composition of securities legislation
and capital issuance we control for the quality of the country’s legal institutions. Judicial
Efficiency controls for the efficiency of the legal system in executing the laws.
Corruption controls for the degree to which securities laws are fairly executed. Higher
values of this variable is associated with less corruption. Finally, Anti Director Rights
controls for investor protection given by corporate laws at the firm level as opposed to
national securities laws.
Table 4.3 summarizes the macroeconomic variables. There is considerable
variation in all the macroeconomic controls, especially GDP growth and the availability
of domestic and foreign capital. Table 4.4 presents the correlations between capital
issuance and the firm characteristics. Profitability, Uniqueness, Asset Tangibility, and
Cash on hand are all negatively correlated with our capital issuance indicator. Table 4.5
displays the pair wise correlations between the dependent variable, capital issuance
indicator, and the macro characteristics. The variables governing the legal environment,
Anti Director Rights, Judicial Efficiency, and the Private Enforcement index, are all
negatively correlated with issuance. Only Public Enforcement manifests a positive
correlation with securities issuance, foreshadowing our regression results.
4. Microeconomic Testing Strategy of LLS (2004)
Our objective is to test the accuracy of LLS conclusions at the micro level.
The authors present three different theories of optimal public policy in the regulation of
the private transactions between the issuers of securities (firms) and investors. The first
theory states that law simply codifies existing market arrangements and therefore serves
no purpose in improving securities transactions. The second theory asserts that the
117
codification of arrangements covering disclosure and liability aid in the efficiency of
private transactions by clarifying liability rules and standardizing security contracts.
Both of these theories assert that public enforcement of securities laws by the government
is at best unnecessary and at worse harmful to investors and firms. The last theory
presented by LLS states that government regulation of securities markets corrects market
failure due to information asymmetry by imposing sanctions and securing information by
subpoena. The authors find that law matters in supporting stock market development,
however only security laws enforcing private transactions are important. They argue that
public enforcement of securities laws are irrelevant and have no effect on macro
indicators of stock market development.
In contrast to LLS, we find that the firm level results are more in line with the
theory that government regulation is needed to support trade and to reduce the transaction
cost of issuance. We show that public enforcement of securities is especially important
in improving access to capital for emerging market firms.
We examine the LLS conclusion by estimating the probit model defined in
Equation 3. We regress the probability of capital issuance for each firm-year observation
on Public Enforcement, Private Enforcement, firm characteristics and macro controls.
We use the STATA cluster command in each regression by firm. This command helps us
avoid spurious results caused by omitted variable bias at the firm level. For each
regression, we calculate the marginal coefficients so that each coefficient represents the
increase in probability of capital issuance explained by the independent variable.
Our large sample size allows us to analyze the data in the two different
dimensions of firm size and institutional quality. We show that securities laws have
118
disparate effects on small and large firms and for firms in developed as opposed to
emerging economies. We test the null hypothesis that the coefficients for small firms and
large firms are equal in the whole sample of countries and the G10 and emerging markets
sub samples by computing the likelihood ratio statistic
( )
2 1
ln ln ln 2
U U R
L L L LR ? ? ? = (4)
where ln L
R
is the restricted log-likelihood function of the entire sample of firm year
observations, and ln L
U1
and ln L
U2
are the unrestricted log-likelihood functions for small
firms and large firms, respectively. The likelihood ratio is distributed chi-squared with
the degrees of freedom equal to the number of restrictions. We calculate the test statistic
for each of the three sub samples – all firm year observations, G10 firms, and emerging
market firms.
5. Results
Before looking at the individual regressions, we can make some general
comments about how the firm specific variables affect capital issuance. Cash on hand,
Uniqueness of Assets, and Profit Risk are consistently negative determinants of issuance
over the different sub samples. This is an expected result since the availability of internal
funds reduces the desire for external finance while uniqueness and risk have also been
shown to decrease issuance. Additionally, we find that External Finance Need (EFN) is a
more important determinant of capital issuance for the small firms in our sample than for
large firms. Large firms have more access to capital and can issue either to garner
investment capital or to optimize their existing capital structure. Small firms have less
access and tend only to go to the capital markets when they need funds.
119
There are also differences in the effects of the firm-specific variables between
firm located in a G10 or emerging market country. For instance, Profitability is a
positive variable for large emerging market firms. Profitability determines the ability of
the firm to pay dividends an interest and in countries with weak investor protection, profit
is an important signal of the ability (if not willingness) to pay. Cross listing is only
positive for large G10 firms. There may be benefits to small and EM firms from issuing
capital in foreign as opposed to domestic markets as a signal of financial well-being,
which reduces their domestic issues.
We also note some general trends in the data in regards to the macroeconomic
variables. In almost all of the regressions, the coefficients enter significantly at the 1%
level. Growth in GDP is a negative determinant of capital issuance in G10 countries and
positive in the emerging markets. For the G10 countries, this result is consistent with the
tendency of capital issuance to be counter cyclical. Since GDP growth controls for the
business cycle in good times (GDP growth is high) firms have enough internal funds to
finance investment and may not need to access capital markets. Additionally, the
availability of Domestic Capital often enters the regression equation as a negative
component of issuance. Since this variable is largely composed bank loans, Domestic
Capital maybe proxying for alternative sources of capital other than the securities
markets. Foreign Capital flows are positively related to capital issuance in emerging
market firms and negatively related to issuance for G10 firms. This result is independent
of firm size.
Corruption, Anti Director Rights, and an Efficient Judiciary control for the
fairness and effectiveness of investor protection. The probability of issuance is higher in
120
less corrupt countries. Except for large firms in the G10, Anti Director Rights have a
negative effect on issuance and an Efficient Judiciary has a positive effect. Interestingly,
for the large G10 firms Anti Director Rights are positively related to issuance and an
Efficient Judiciary has a negative relationship.
We use the two aggregate securities law indexes, Public and Private Enforcement,
to summarize effects of securities laws on capital issuance. Recall that the Public
Enforcement index aggregates supervisor independence, investigative powers, orders and
criminal sanction. The Private Enforcement index is composed of disclosure and burden
of proof. Table 4.6A displays the effects of Public and Private Enforcement for all firms
in our data set. The first regression includes Public Enforcement as the main explanatory
variable of security laws in each country. The second regression contains Private
Enforcement only. Finally, the third regression displays a horserace between Public and
Private Enforcement.
Regression 1 generates our primary result; Public Enforcement has a large
positive and significant role in capital issuance. The variable’s coefficient is 0.289 and is
significant at the 1% level. Regression 1 shows that capital issuance is pro cyclical over
the whole sample of firm year observations as GDP_growth is positive and significant.
The availability of domestic capital is a negative determinant of capital issuance and is
significant at the 1% level. Corruption plays a negligible role in issuance however the
other two measures of the fairness and efficiency of the legal system, Anti Director
Rights and Efficient Judiciary are significant components of capital issuance. Protection
of shareholders by corporate laws (anti director rights) is negative, while an efficient
judiciary makes a positive contribution.
121
Regression 2 examines the role of Private Enforcement on our dependent
variable. We find that enforcement of private transactions plays a negligible role in
capital issuance. In the final regression in Table 4.8A, we include both Public and
Private Enforcement in the regression in order to discern the independent contribution of
each variable on capital issuance. The coefficient on Public Enforcement remains
unchanged from Regression 1. It is positive and significant at the 1% level and has a
coefficient of 0.289. Private Enforcement is statistically insignificant.
In the next two tables, Tables 4.7B and 4.7C, we investigate whether securities
laws have disparate effect on small and large firms. A firm is categorized as small (large)
if its Total Assets are below (above) the median for all firms in its country of origin. We
find that for small firms presented in Table 4.6B, Public Enforcement has a larger impact
on capital issuance than that of the whole sample. The coefficient on Public Enforcement
is 0.445 (compared to 0.289 in Table 4.6A). Regression 2 reveals that Private
Enforcement of securities laws is a significant and negative determinant of capital
issuance for small firms. This result is repeated in Regression 3 when both indexes of
enforcement are included. The coefficient on Private Enforcement is -0.301 and is
significant at the 1% level. Public Enforcement on the other hand has a positive
coefficient of 0.453 and is also significant at the 1% level.
The dominance of Public over Private Enforcement does not hold for the large
firms in the sample. In Regression 1 of Table 4.6C, Public Enforcement is shown to have
an insignificant effect on issuance. Regression 2 shows that Private Enforcement is
positive for large firms and significant at the 1% level with a coefficient of 0.198. The
122
horse race between both securities indexes presented in Regression 3 shows that both
variables makes a significant contribution to capital issuance for the large firms overall.
We wrap up the analysis of Tables 4.7A-C by testing whether the differences in
small and large firms are statistically significant. The likelihood ratio test statistic for the
equality of coefficient of small and large firms in Regression 1 of Tables 4.7A-C is given
by Equation 4.
( )
10 . 4150
) 69 . 15060 02 . 6059 76 . 23194 ( 2
ln ln ln 2
2 1
=
+ + ? ? =
? ? ? =
U U R
L L L LR
The test statistic is distributed chi-square with 56 degrees of freedom. At a 1% level of
significance the critical value from of the chi-square distribution is 83.52. The null
hypothesis of equality of coefficients between small and large firms is resoundingly
rejected. We calculate the likelihood ratio statistic for Regressions 2 and 3 at 3333.82
and 2208.68 respectively. The critical values of chi-square at a 1% level of significance
for Regressions 2 and 3 are 83.52 and 85.95, respectively. These test statistics support
our finding that securities laws have disparate effects on small and large firms. Public
Enforcement increases issuance in small firms and Private Enforcement decreases the
probability of issuance. In contrast, both enforcement indexes make a positive
contribution to the probability of large firm issuance overall.
Next, we isolate G10 firms in our sample. The sub sample is composed of all the
countries in the G10 except for the US, which was dropped from the sample to prevent
biasness in the results as described in the methodology section. Overall G10 countries
have better institutional quality than the other countries in the data set, allowing us to
investigate the role of security laws in a good institutional environment. We divide firms
123
into small and large groups as before. In Table 4.7A, we find that Public Enforcement
has a smaller effect on issuance for small G10 firms than for the whole sample. Its
coefficient is -0.190 compared to 0.289 for the entire sample. Regression 2 manifests a
significant negative relationship between Private Enforcement and capital issuance. In
Regression 3, the results indicate that when both indexes are included in the regression
Public Enforcement is a positive factor in the issuance of small G10 firms while Private
Enforcement is a significant deterrent to issuance.
Table 4.6B demonstrates the difference between small and large G10 firms’
response to security laws. In Regression 1, Public Enforcement has a coefficient of
-0.401 and is significant at the 1% level. Private Enforcement is also shown to have a
negative effect on issuance in Regression 2. We also observe in Regression 3 that laws
regulating both private and public enforcement have negative impacts on large firm
issuance. The coefficients on Public and Private Enforcement are -0.422 and -0.340,
respectively. Calculations of the likelihood ratio statistics for all three regressions in
Tables 4.8A & B reject equality of coefficients between small and large G10 firms at the
1% level. The likelihood ratio statistics are 2740.96, 2610.24, and 2736.66 for
Regressions 1, 2 and 3, respectively. The critical values of the test from the chi-square
distribution are 83.52 for Regression 1 and 2 and 85.95 for Regression 3.
Finally, we investigate the role of securities indexes in firms whose country origin
is not in the G10. We define those countries as emerging markets. Table 4.8A displays
the results for all firm year observations in emerging market countries. For small
emerging market firms Public Enforcement is positive and significant at 0.459 as shown
in Regression 1 of Table 4.8A. Private Enforcement enters Regression 2 negatively and
124
significantly at the 5% level. Regression 3 displays the result that the largest impact of
the securities indexes on capital issuance is for small emerging market firms. The
coefficients are 1.001 and -1.054 for Public and Private Enforcement, respectively. Both
variables are significant at the 1% level.
Table 4.8B shows that large emerging market firms behave similarly to small
emerging market firms in their response to the public regulation of securities laws.
However, in contrast to the other sub samples of our data set, Private Enforcement is
positive and significant in Regression 2. When both security laws indexes are included in
Regression 3, only Public Enforcement retains a positive relationship with issuance.
Securities laws have a similar effect on small and large emerging market firms.
However, likelihood ratio tests strongly reject the null hypotheses of equality of
coefficients between all the explanatory variables. The statistics for Regressions 1-3 in
Tables 4.8A and 4.8B are 742.86, 750.64, and 811.08, respectively. The critical values of
the test from the chi-square distribution are 83.52 for Regression 1 and 2 and 85.95 for
Regression 3.
In the last series of regressions, we estimate the impact of each separate sub
index of Public and Private Enforcement in six individual regression equations for each
of the three sub samples. We report the results of these regressions for the securities laws
indexes only in Appendix E. Over the whole sample of firms supervisor characteristics,
and the supervisor’s investigative powers are positive determinants of capital issuance.
However, the significance and sign of the security indexes depend on the sub sample
examined. For instance, the supervisor characteristics, investigative power, criminal and
non-criminal sanctions significantly decrease the probability of issuance for large G10
125
firms. In contrast, all firms in emerging markets increase their securities issuance when
there is a strong supervisor with investigative powers and the authority to impose
sanctions.
6. Robustness Check
A reason for skepticism of the results we present may be found in the large z-
statistics given by the enforcement variables. For example, in the all-firm regressions
presented in Table 4.6A, in Regression 3, Public Enforcement has a z-statistic equal to
21.01. A reader may question the plausibility of these results given the cross-country
nature of the panel. A useful robustness check is to cluster the standard errors by country
in order to account for the cross-country nature of the regression analysis. Table 4.9A
repeats the analysis for all firms in Table 4.6A with the standard errors clustered at the
country level, and Table 4.9B presents these results for the regressions in which both
Public and Private Enforcement are included for each firm category. As Table 4.9A
demonstrates Public Enforcement remains significant at the 1% level in Regressions 1
and 3, while Private Enforcement remains insignificant in Regressions 2 and 3. Though
clustering the standard errors by country leads to a sizable reduction in the z-statistics for
most of the variables, the significant relationship between public enforcement of
securities laws and capital issuance remains unchanged.
Table 4.9 suggests that overall, small firms increase issuance when securities laws
are publicly enforced. There is not significant relationship between securities law
enforcement and capital issuance for large firms overall when standard errors are
clustered by country. Additionally, the significance of Public Enforcement for small G10
firms disappears, however this analysis supports our original conclusion that these firms
126
are harmed by private enforcement of securities laws. Also supported by the robustness
check, is that Public Enforcement is a negative and significant at the 1% level for large
G10 firms and is positive and significant at the 1% level for small and large Emerging
Market firms.
7. Conclusions
The main conclusion of LLS is that private enforcement of securities regulation
increases the level of stock market development while public enforcement has a
negligible effect on stock markets.
Public enforcement plays, at best, a modest role in the development of stock
markets. Specifically, there is no evidence that such factors as Supervisor’s
independence or focus work. Both the Supervisor’s investigative powers and the
strength of both criminal and non-criminal sanctions only matter for a narrow set
of outcomes. In contrast, the development of stock markets is strongly associated
with measures of private enforcement such as extensive disclosure requirements
and a relatively low burden of proof on investors seeking to recover damages
resulting from omissions of material information from the prospectus.
Our paper questions this conclusion by examining the impact of securities
regulation at the firm level. Our data set allows us to investigate effect of securities
regulation on listed small and large firms and in developing and industrialized countries.
We find that securities laws have disparate effects on capital issuance between small and
large firms in G10 and emerging market countries. Private Enforcement of securities
laws is found to be a deterrent to capital issuance for small firms and firms in emerging
127
markets. Public Enforcement significantly increases the probability of issuance for all
emerging market firms.
Our results suggest that private enforcement of securities laws is not the panacea
to stock market development described in LLS. Instead, our results encourage the
establishment of a strong regulatory authority in countries with weak institutions, while
LLS support strengthening laws regulating private transactions. Both courses of actions
have dramatically different effects on capital issuance, therefore securities laws should be
structured in a way that increases the availability of capital for all firms.
128
Table 4.1: Security Issuance by Country
Country Debt Conv. Debt Equity Preferred Conv. Preferred Total
Argentina 29 10 61 2 102
Australia 21 58 8245 48 8372
Austria 2 91 93
Bangladesh 5 5
Belgium 173 173
Bermuda 10 1 11
Bolivia 6 1 7
Brazil 94 25 51 35 205
Canada 26 14 40
Chile 37 160 197
China 7 1291 1298
Colombia 23 32 55
Costa Rica 3 3
Czech Republic 4 4
Denmark 1 192 193
Finland 6 1 224 231
France 48 11 1207 1266
Germany 6 1 585 7 599
Greece 2 133 135
Hong Kong 4 5 900 909
Hungary 16 16
India 125 179 304
Indonesia 40 128 168
Ireland 41 41
Israel 8 8
Italy 3 203 1 207
Japan 2149 239 1951 4339
Luxembourg 7 1 8
Malaysia 64 2 418 1 485
Mexico 91 1 33 125
Netherlands 10 1 136 6 153
New Zealand 2 5 42 3 52
Norway 1 1 102 104
Pakistan 22 22
Papua New Guinea 6 6
Peru 143 3 146
Philippines 18 42 60
Poland 2 32 34
129
Table 4.1: Security Issuance by Country (cont.)
Country Debt Conv. Debt Equity Preferred Conv. Preferred Total
Portugal 46 1 47
Singapore 59 314 373
South Africa 4 4
South Korea 397 9 406
Spain 5 98 103
Sri Lanka 11 11
Sweden 22 236 258
Switzerland 51 7 104 1 163
Taiwan 739 2 316 1057
Thailand 71 2 77 150
Turkey 11 11
US 42 121 3438 3620 17 7238
United Kingdom 7 1855 12 1874
Venezuela 19 38 1 58
Total 3947 497 23692 3764 17 31929
130
Table 4.2: Summary Statistics of Firm Specific Variables
Variable Obs Mean Std. Dev. Min Max
Capital Issuance 108280 0.22 0.41 0.00 1.00
Cash/TA 71742 0.14 0.16 0.00 1.14
Cross listing 108280 0.10 0.30 0.00 1.00
EFN 68977 -1.83E+10 1.96E+12 -4.70E+14 2.37E+12
Risk 103990 -3.36 1.52 -13.82 9.38
Profitability 108280 -0.05 1.31 -165.94 15.48
Uniqueness 71742 0.01 0.41 -76.53 48.80
Asset Tangibility 108280 0.39 0.48 0.00 73.42
Table 4.3: Summary Statistics of Macroeconomic Variables
Variable Obs Mean Std. Dev Min Max
GDP Growth 330 0.030 0.032 -0.131 0.111
Domestic Capital 284 154.943 104.721 28.755 536.873
Foreign Capital 313 0.030 5.719 -27.002 23.839
Corruption 336 3.848 1.296 1 6
Criminal 340 0.502 0.255 0 1
Anti-director 340 3.085 1.346 0 5
Judicial Efficiency 336 7.813 2.143 2.5 10
Private Enforcement Index 340 0.558 0.213 0.11 1
Public Enforcement Index 340 0.501 0.232 0 0.896
131
Table 4.4: Firm Specific Correlations (with Capital Issuance Indicator)
Capital
Issuance Cash/TA Cross listing EFN Risk Profitability Uniqueness
Cash/TA -0.022*** 1.000
Cross listing 0.038*** -0.002 1.000
EFN 0.002 0.003 -0.018*** 1.000
Risk 0.015*** 0.179*** 0.070*** -0.013*** 1.000
Profitability -0.036*** -0.014*** 0.007* 0.000 -0.036*** 1.000
Uniqueness -0.009** -0.006 -0.002 0.000 -0.002 -0.013*** 1.000
Asset Tangibility -0.008** -0.226*** 0.030*** -0.004 -0.086*** 0.008** 0.000
Table 4.5 Macroeconomic Variable Correlations (with Capital Issuance Indicator)
Capital
Issuance
GDP
Growth
Domestic
Capital
Foreign
Capital
Anti-
director
Judicial
Efficiency
Private
Enforcement
GDP Growth 0.072*** 1.000
Domestic Capital -0.135*** 0.041*** 1.000
Foreign Capital 0.029*** 0.157*** 0.106*** 1.000
Anti-director -0.156*** 0.183*** 0.531*** 0.280*** 1.000
Judicial Efficiency -0.044*** -0.010*** 0.534*** 0.060*** 0.580*** 1.000
Private Enforcement -0.171*** 0.213*** 0.488*** 0.325*** 0.825*** 0.462*** 1.000
Public Enforcement 0.002 0.407*** 0.072*** 0.355*** 0.492*** 0.065*** 0.554***
132
Table 4.6A: All Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and standard errors are clustered by firm. All country-level control variables
are averaged over the years t-1 through t-3. Industry and time fixed effects have been suppressed.
Absolute value of z statistics in brackets * significant at 10%; ** significant at 5%; ***
significant at 1%.
1 2 3
Cash_TA -0.039* -0.065** -0.039*
[1.66] [2.57] [1.64]
Uniqueness 0.000** 0.000** 0.000**
[2.08] [2.13] [2.09]
Asset_Tang 0.009 0.003 0.009
[1.49] [0.39] [1.57]
EFN 0.018** 0.020** 0.018**
[2.30] [2.24] [2.30]
Profitability 0.000 0.000 0.000
[1.49] [1.39] [1.49]
LRisk -0.035*** -0.020*** -0.035***
[12.29] [6.39] [12.35]
Cross listing -0.030*** -0.014* -0.03***
[3.81] [1.80] [3.85]
GDP_growth 0.011*** 0.022*** 0.012***
[8.43] [13.73] [8.39]
Domestic Capital -0.001*** -0.001*** -0.001***
[18.70] [13.83] [18.45]
Foreign Capital 0.009*** 0.018*** 0.009***
[12.63] [14.89] [12.58]
Corruption 0.003 0.034*** 0.003
[1.04] [8.24] [0.88]
Anti Director Rights -0.006*** -0.035*** -0.055***
[16.80] [8.24] [12.17]
Judicial Efficiency 0.060*** 0.025*** 0.061***
[17.40] [8.71] [17.48]
Public Enforcement 0.289*** 0.289***
[21.02] [21.01]
Private Enforcement -0.016 -0.027
[0.54] [0.89]
Observations 45179 45179 45179
Log Likelihood -23194.76 -23756.63 -23194.06
Pseudo R-squared 0.08 0.06 0.08
133
Table 4.6B Small Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and standard errors are clustered by firm. Size is determined by Total Assets
divided by GDP. Firms with Total Assets below the median for each country are defined as
small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.043** -0.063** -0.036*
[2.14] [2.34] [1.89]
Uniqueness 0.000* 0.000 0.000**
[1.94] [0.43] [1.98]
Asset_Tang -0.125*** -0.197*** -0.112***
[6.64] [7.58] [6.32]
EFN 0.006* .009* 0.006*
[1.93] [1.81] [1.93]
Profitability -0.001*** -0.001** -0.001***
[3.13] [2.53] [3.25]
LRisk -0.009*** 0.016*** -0.011***
[3.67] [5.34] [4.56]
Cross Listing -0.021** -0.024** -0.019**
[2.21] [2.17] [2.16]
GDP_growth 0.012*** 0.031*** 0.012***
[8.64] [12.89] [8.84]
Domestic Capital 0.000*** 0.000** -0.001***
[9.87] [2.52] [9.11]
Foreign Capital 0.006*** 0.011*** 0.005***
[8.05] [9.86] [7.23]
Corruption 0.016*** 0.055*** 0.009**
[4.24] [11.81] [2.44]
Anti Director Rights -0.088*** -0.040 -0.056***
[23.14] [7.13] [13.35]
Judicial Efficiency 0.040*** 0.005 .043***
[11.14] [1.47] [11.80]
Public Enforcement 0.445*** 0.453***
[26.67] [27.18]
Private Enforcement -0.204*** -0.309***
[5.23] [10.05]
Observations 18708 18708 18708
Log Likelihood -6059.02 -7021.24 -5960.72
Pseudo R-squared 0.33 0.22 0.34
134
Table 4.6C: Large Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Size is
determined by Total Assets divided by GDP. Firms with Total Assets above the median for each
country are defined as large. Probit estimation is used and standard errors are clustered by
firm. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.082** -0.103*** -0.086**
[2.40] [3.07] [2.54]
Uniqueness 0.000** 0.000** 0.000**
[1.99] [1.99] [1.97]
Asset_Tang 0.027 0.018** 0.024
[1.24] [2.38] [1.30]
EFN 0.033* 0.033* 0.033*
[1.83] [1.91] [1.81]
Profitability 0.000 0.000 0.000
[1.41] [1.40] [1.37]
LRisk -0.037*** -0.032*** -0.036***
[1.54] [10.01] [11.37]
Cross Listing -0.014 -0.005 -0.011
[1.45] [0.56] [1.16]
GDP_growth 0.008*** 0.010*** 0.007***
[3.93] [4.99] [3.49]
Domestic Capital -0.001*** -0.001*** -0.001***
[13.60] [13.62] [13.81]
Foreign Capital 0.008*** 0.009*** 0.008***
[8.02] [8.72] [7.95]
Corruption 0.017*** 0.030*** 0.020***
[3.86] [6.67] [4.45]
Anti Director Rights -0.015*** -0.027*** -0.035***
[3.15] [4.63] [5.65]
Judicial Efficiency 0.033*** 0.019*** 0.030***
[7.84] [5.20] [7.51]
Public Enforcement 0.087 0.089***
[1.43] [5.23]
Private Enforcement 0.198*** 0.208***
[4.73] [4.92]
Observations 26295 26295 26295
Log Likelihood -15060.69 -15068.48 -15041.76
Pseudo R-squared 0.05 0.05 0.05
135
Table 4.7A: G10 Small Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Size is
determined by Total Assets divided by GDP. Firms with Total Assets below the median for each
country are defined as small. . Probit estimation is used and standard errors are clustered by
firm. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.009* -0.012*** -0.009*
[1.92] [2.75] [1.87]
Uniqueness -0.009** -0.010** -0.009**
[2.13] [2.27] [2.10]
Asset_Tang -0.009 -0.007* -0.006
[1.62] [1.90] [.1.25]
EFN 0.003*** 0.004*** 0.003***
[3.16] [3.01] [3.20]
Profitability -0.001** -0.001** -0.001**
[2.54] [1.92] [2.47]
LRisk -0.005*** -0.005*** -0.005***
[6.58] [6.65] [6.72]
Cross Listing 0.000 -0.001 -0.001
[0.34] [0.26] [0.42]
GDP_growth -0.657*** -0.340 -0.446***
[4.59] [0.63] [3.81]
Domestic Capital 0.000*** 0.000*** 0.000***
[8.71] [7.63] [8.24]
Foreign Capital -0.002*** -0.002*** -0.002***
[9.01] [9.09] [9.06]
Corruption 0.009*** 0.009*** .006***
[6.99] [6.00] [6.28]
Anti Director Rights -0.005*** 0.002 -0.001
[3.10] [1.34] [1.66]
Judicial Efficiency 0.003 0.001*** 0.005
[0.91] [3.68] [1.20]
Public Enforcement 0.019 0.017***
[7.67] [7.29]
Private Enforcement -0.080** -0.071***
[4.88] [2.77]
Observations 12207 12207 12207
Log Likelihood -1652 -1692.05 -1647.69
Pseudo R-squared 0.39 0.37 0.41
136
Table 4.7B G10 Large Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Size is
determined by Total Assets divided by GDP. Firms with Total Assets above the median for each
country are defined as large. Probit estimation is used and standard errors are clustered by
firm. All country-level control variables are averaged over the years t-1 through t-3. Industry and
time fixed effects have been suppressed. Absolute value of z statistics in brackets * significant at
10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.069 -0.041 -0.067
[1.58] [1.30] [1.58]
Uniqueness -0.116** -0.113** -0.118**
[2.18] [2.02] [2.18]
Asset_Tang 0.109*** 0.118*** 0.113***
[4.49] [4.63] [4.49]
EFN 0.025 0.025 0.025
[0.94] [1.01] [0.94]
Profitability -0.134*** -0.127*** -0.134***
[3.22] [3.23] [3.22]
Lrisk -0.053*** -0.060*** -0.053***
[12.15] [12.65] [12.15]
Cross Listing 0.033** 0.011** 0.031***
[2.49] [2.03] [2.49]
GDP_growth -6.39*** -4.238*** -6.016***
[9.27] [9.74] [9.19]
Domestic Capital 0.001*** 0.000*** -0.001***
[7.93] [10.42] [7.5]
Foreign Capital -0.014*** 0.000*** -0.013***
[8.81] [8.71] [8.80]
Corruption 0.097*** 0.007*** 0.090***
[10.64] [9.75] [10.46]
Anti Director Rights 0.080*** -0.012*** 0.110***
[7.51] [4.13] [6.04]
Judicial Efficiency -0.139*** -0.009*** -0.140***
[7.73] [6.05] [7.67]
Public Enforcement -0.401*** -0.422***
[5.36] [4.75]
Private Enforcement -0.139*** -0.340
[2.61] [0.03]
Observations 18471 18471 18471
Log Likelihood -9894.86 -9913.74 -9894.86
Pseudo R-squared .09 .09 .09
137
Table 4.8A: Small Emerging Market Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and standard errors are clustered by firm. Size is determined by Total Assets
divided by GDP. Firms with Total Assets below the median for each country are defined as
small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.173*** -0.152*** -0.182***
[3.73] [3.33] [3.92]
Uniqueness 0.000* 0.000** 0.000*
[1.92] [2.04] [1.83]
Asset_Tang -0.410*** -0.449*** -0.381***
[8.90] [9.53] [8.41]
EFN 0.007* 0.008* 0.006*
[1.93] [1.92] [1.87]
Profitability -0.009*** -0.009** -0.008**
[2.19] [2.19] [2.14]
Lrisk -0.011* -0.006 -0.017***
[1.98] [1.18] [3.17]
Cross Listing [d) -0.084*** -0.106*** -0.051**
[3.40] [4.45] [2.00]
GDP_growth 0.003 0.004 0.007**
[1.09] [1.76] [2.49]
Domestic Capital 0.000 0.000 0.000
[2.38] [0.31] [0.13]
Foreign Capital 0.005 0.005 0.000
[0.76] [0.76] [0.02]
Corruption 0.127*** 0.118*** .127***
[3.23] [3.54] [0.08]
Anti Director Rights -0.059*** 0.006 -0.049***
[4.58] [0.46] [3.74]
Judicial Efficiency 0.022*** 0.028*** 0.026***
[2.92] [3.62] [3.32]
Public Enforcement 0.459*** 1.001***
[8.49] [12.12]
Private Enforcement -0.166** -1.054***
[2.16] [9.53]
Observations 6501 6501 6501
Log Likelihood 3323.59 3374.17 3234.88
Pseudo R-squared 0.25 0.24 0.27
138
Table 4.8B: Large Emerging Market Firms
The dependent variable is dummy variable for whether the firm issues capital in period t. . Probit
estimation is used and standard errors are clustered by firm. Size is determined by Total Assets
divided by GDP. Firms with Total Assets below the median for each country are defined as
small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
1 2 3
Cash_TA -0.130** -0.127** -0.133**
[2.17] [2.13] [2.24]
Uniqueness 0.000** 0.000** 0.000**
[2.51] [2.35] [2.55]
Asset_Tang -0.087** -0.100*** -0.086**
[2.36] [2.69] [2.34]
EFN 0.005 0.009 0.003
[0.26] [0.45] [0.19]
Profitability 0.000* 0.000* 0.000*
[1.81] [1.85] [1.81]
Lrisk -0.008* -0.004 -0.009**
[1.82] [1.03] [2.01]
Cross Listing (d) -0.049*** -0.054*** -0.047***
[3.21] [3.56] [3.09]
GDP_growth 0.009*** 0.009*** 0.010***
[4.03] [4.02] [2.98]
Domestic Capital -0.001*** -0.001*** -0.001***
[5.79] [5.32] [4.38]
Foreign Capital 0.006** 0.008** 0.005***
[4.79] [5.68] [4.91]
Corruption 0.068*** 0.070*** 0.066***
[5.79] [6.08] [3.74]
Anti Director Rights -0.019 0.002 -0.015
[1.36] [0.13] [1.06]
Judicial Efficiency -0.005 -0.005 -0.006
[0.76] [0.70] [0.78]
Public Enforcement 0.268*** 0.360***
[6.15] [5.78]
Private Enforcement 0.136** -0.194**
[2.22] [2.22]
Observations 7824 7824 7824
Log Likelihood 4546.79 4572.79 4542.18
Pseudo R-squared 0.08 0.07 0.08
139
Table 4.9A: All Firms – Standard Errors Clustered by Country
The dependent variable is dummy variable for whether the firm issues capital in period t. Probit
estimation is used and regressions are clustered by country. All country-level control variables
are averaged over the years t-1 through t-3. Industry and time fixed effects have been suppressed.
Absolute value of z statistics in brackets * significant at 10%; ** significant at 5%; ***
significant at 1%.
1 2 3
Cash_TA -0.039 -0.065** -0.039
[0.54] [0.83] [0.54]
Uniqueness 0.000 0.000** 0.000
[1.27] [1.07] [1.27]
Asset_Tang 0.009 0.003 0.009
[0.29] [0.15] [0.29]
EFN 0.018** 0.020** 0.018***
[2.65] [2.66] [2.65]
Profitability 0.000 0.000 0.000
[1.17] [1.13] [1.16]
LRisk -0.035 -0.020 -0.035*
[1.77] [0.96] [1.76]
Cross listing -0.030 -0.014* -0.03
[1.31] [0.62] [1.31]
GDP_growth 0.011** 0.022*** 0.012**
[2.32] [3.36] [2.44]
Domestic Capital -0.001*** -0.001* -0.001**
[2.62] [1.68] [2.49]
Foreign Capital 0.009*** 0.012 0.009***
[2.27] [1.70] [2.20]
Corruption 0.003 0.034 0.003
[1.17] [1.17] [0.16]
Anti Director Rights -0.006** -0.035 -0.055***
[2.57] [1.07] [1.64]
Judicial Efficiency 0.060*** 0.025* 0.061***
[3.02] [1.81] [3.22]
Public Enforcement 0.289*** 0.289***
[3.35] [3.36]
Private Enforcement -0.016 -0.027
[0.09] [0.16]
N 45179 45179 45179
Log Likelihood -23194.76 -23756.63 -23194.06
Pseudo R-sq 0.08 0.06 0.08
140
Table 4.9B: All Firm Groupings – Standard Errors Clustered by Country
The dependent variable is dummy variable for whether the firm issues capital in period t. . Probit
estimation is used and standard errors are clustered by country. Size is determined by Total
Assets divided by GDP. Firms with Total Assets below the median for each country are defined
as small. All country-level control variables are averaged over the years t-1 through t-3. Industry
and time fixed effects have been suppressed. Absolute value of z statistics in brackets *
significant at 10%; ** significant at 5%; *** significant at 1%.
All Firms G10 Firms Emerging Firms
Small Large Small Large Small Large
Cash_TA -0.036 -0.086* -0.009* -0.067 -0.182** -0.133**
[0.75] [1.68] [1.38] [1.04] [1.99] [2.38]
Uniqueness 0.000 0.000* -0.009** -0.118* 0.000 0.000
[0.80] [1.66] [1.62] [1.89] [0.57] [0.73]
Asset_Tang -0.112* 0.024 -0.006 0.113* -0.381*** -0.086*
[1.94] [0.40] [0.68] [1.69] [4.95] [1.86]
EFN 0.006*** 0.033 0.003** 0.025 0.006*** 0.003
[2.66] [1.38] [2.08] [0.76] [2.73] [0.23]
Profitability -0.001*** 0.000 -0.001** -0.134 -0.008** 0.000
[2.89] [0.91] [2.22] [1.63] [3.82] [1.05]
Lrisk -0.011* -0.036** -0.005* -0.053*** -0.017*** -0.009
[1.85] [2.17] [1.86] [3.75] [2.99] [0.90]
Cross Listing -0.019 -0.011 -0.001 0.031*** -0.051* -0.047***
[1.46] [0.86] [0.25] [3.00] [1.80] [3.49]
GDP_growth 0.012*** 0.007 -0.446** -6.016*** 0.007** 0.010**
[3.84] [1.38] [2.40] [1.76] [1.26] [2.34]
Domestic Capital -0.001 -0.001*** 0.000*** -0.001 0.000 -0.001***
[1.60] [3.08] [2.70] [1.21] [0.05] [2.91]
Foreign Capital 0.005* 0.008** -0.002** -0.013* 0.000 0.005***
[1.78] [2.11] [2.37] [1.65] [0.02] [1.67]
Corruption 0.009** 0.020 0.006** 0.090* 0.127*** 0.066***
[0.43] [1.37] [2.18] [1.82] [3.97] [2.98]
Anti Director Rights -0.056*** -0.035 -0.001 0.110** -0.049 -0.015
[2.90] [1.12] [0.23] [2.41] [1.44] [0.90]
Judicial Efficiency 0.043*** 0.030*** 0.005* -0.140* 0.026 -0.006
[2.66] [2.63] [1.76] [1.70] [0.30] [0.43]
Public Enforcement 0.453*** 0.089 0.017 -0.422*** 1.001*** 0.360***
[4.02] [1.40] [1.43] [3.10] [4.21] [3.90]
Private Enforcement -0.309** 0.208 -0.071** -0.340 -1.054*** -0.194
[2.17] [1.29] [2.42] [1.25] [4.21] [1.16]
Observations 18708 26295 12207 18471 6501 7824
Log Likelihood -5960.72 -15041.76 -1647.69 -9894.86 -3234.88 -4542.18
Pseudo R-squared 0.34 0.05 0.41 0.09 0.27 0.08
141
Appendices
Appendix A: Derivatives of the Constrained Capital Equation
2
2 2 2 2 2 2
) ( 2
) ( 4 ) ( 2
?
? ? ? ? ?
i
wi g g wi g
k k
c
+ + +
= = when <1. (2.3)
A1. Derivative of constrained capital with respect to financial development,
2
2 2 2
2
2 2 2 2
3
2 2 2 2 2 2
) ( 2
) ( 4 2
) ( ' ) ( 8 2 (
) ( ' ) ( 4 ) ( 4 2
) (
) ( 4 ) ( ' ) ( 2
?
? ?
? ? ? ?
? ? ? ? ?
?
? ? ? ? ? ?
?
i
wi g
i wi g g
i wi wi g g g
i
wi g i g wi g k
c
+
+
+ + + +
+
+ + + ?
=
?
?
A2. Derivative of constrained capital with respect to productivity, g
2
2 2 2
2 2 2
2 2
2
) ( 2
) ( 4
) ( 4
2
?
? ? ?
? ?
?
?
i
wi g
wi g
g
g
g
k
c
+ +
+
+
=
?
?
A.3 Derivative of constrained capital with respect to initial wealth, w
2
2 2 2
2
2
) ( 2
) ( 4
) ( 2
) ( 2
?
? ?
? ?
?
i
wi g
i g
i
w
k
c
+
+
=
?
?
142
Appendix B: Derivatives of the Firm Profit Equations with respect to
B.1 Large firm profit
B.2 Derivative of large firm profit with respect to
B.3 Constrained firm capital demand
B.4 Constrained firm profit
143
B.5 Derivative of constrained firm profit with respect to
144
Appendix C: Description of Variables, Chapter 3
Variable Definition Source
Private Credit Private credit by deposit money banks and
other financial institutions to GDP,
calculated using the following deflation
method: {(0.5)*[F
t
/P_e
t
+ F
t-1
/P_e
t-
1
]}/[GDP
t
/P_a
t
] where F is credit to the
private sector, P_e is end-of period CPI, and
P_a is average annual CPI
Beck, Demirguc-Kunt, and
Levine (2003)
Checks The numbers of veto players in the political
system, adjusting for whether these veto
players are independent of each other as
determined by the level of electoral
competitiveness, their respective party
affiliations and the electoral rules.
Database of Political
Institutions
Beck, Clark, Groff, Keefer,
and Walsh. (2001)
Political
Polarization
Maximum polarization between the
executive party and the four principle parties
of the legislature
Database of Political
Institutions
Keefer (2002)
Lgdp95 The natural log of GDP per capita is gross
domestic product divided by midyear
population. GDP is the sum of gross value
added by all resident producers in the
economy plus any product taxes and minus
any subsidies not included in the value of the
products. It is calculated without making
deductions for depreciation of fabricated
assets or for depletion and degradation of
natural resources. Data are in constant U.S.
dollars.
World Bank Indicators (2003)
Trade Trade is the sum of exports and imports of
goods and services measured as a share of
gross domestic product.
World Bank Indicators (2003)
FDI Foreign direct investment is net inflows of
investment to acquire a lasting management
interest (10 percent or more of voting stock)
in an enterprise operating in an economy
other than that of the investor. It is the sum
of equity capital, reinvestment of earnings,
other long-term capital, and short-term
capital as shown in the balance of payments.
World Bank Indicators (2003)
145
Inflation Inflation as measured by the annual growth rate
of the GDP implicit deflator shows the rate of
price change in the economy as a whole.
World Bank Indicators (2003)
Natural
Openness
The Natural logarithm of aggregated fitted
values if bilateral trade equation with geographic
variables
Frankel and Romer (1999)
Lsettler Natural logarithm of estimated European
settlers’ mortality rate.
Acemoglu, Johnson, and
Robinson (2001)
Disteq Distance from Equator of capital city measured
as abs(Latitude)/90.
World Bank (2002)
Engfrac Fraction of Population speaking English as a 1
st
language
Hall and Jones (1999)
Legor_uk British Legal Origin La Porta et al. (1998)
Legor_fr French Legal Origin La Porta et al. (1998)
Legor_ge German Legal Origin La Porta et al. (1998)
Legor_sc Scandanvian Legal Origin La Porta et al. (1998)
Legor_so Socialist Legal Origin La Porta et al. (1998)
GROUPB A dummy variable that equals 1 if the country-
year observation is greater than 1995 and 0
otherwise.
LCHECKSB An interaction term equal to Log of Checks *
GROUPB
POLARIZB An interaction term equal to Polarization *
GROUPB
Widely-Held10 The proportion of firms in a country that are
widely held, where control is inferred at 10%
Morck, Wolfenzon, and Yeung
(2004)
Widely-Held20 The proportion of firms in a country that are
widely held, where control is inferred at 20%
Morck, Wolfenzon, and Yeung
(2004)
146
Appendix D: Description of Variable, Chapter 4
Variable Description
Securities Law, Source and definitions: La Porta et al. (2004)
Disclose The index of disclosure equals the arithmetic mean of: (1) Prospect; (2)
Compensation; (3) Shareholders; (4) Inside ownership; (5) Contracts Irregular; (6)
and Transactions.
Prospectus Equals one if the law prohibits selling securities that are going to be listed on the
largest stock exchange of the country without delivering a prospectus to potential
investors; equals zero otherwise.
Compensa An index of prospectus disclosure requirements regarding the compensation of
directors and key officers. Equals one if the law or the listing rules require that the
compensation of each director and key officer be reported in the prospectus of a
newly-listed firm; equals one-half if only the aggregate compensation of directors
and key officers must be reported in the prospectus of a newly-listed firm; equals
zero when there is no requirement to disclose the compensation of directors and
key officers in the prospectus for a newly-listed firm.
Sharehol An index of disclosure requirements regarding the Issuer’s equity ownership
structure. Equals one if the law or the listing rules require disclosing the name and
ownership stake of each shareholder who, directly or indirectly, controls ten
percent or more of the Issuer’s voting securities; equals one-half if reporting
requirements for the Issuer’s 10% shareholders do not include indirect ownership
or if only their aggregate ownership needs to be disclosed; equals zero when the
law does not require disclosing the name and ownership stake of the Issuer’s 10%
shareholders. No distinction is drawn between large-shareholder reporting
requirements imposed on firms and those imposed on large shareholders
themselves.
Insideow An index of prospectus disclosure requirements regarding the equity ownership of
the Issuer’s shares by its directors and key officers. Equals one if the law or the
listing rules require that the ownership of the Issuer’s shares by each of its director
and key officers be disclosed in the prospectus; equals one-half if only the
aggregate number of the Issuer’s shares owned by its directors and key officers
must be disclosed in the prospectus; equals zero when the ownership of Issuer’s
shares by its directors and key officers need not be disclosed in the prospectus.
Contract An index of prospectus disclosure requirements regarding the Issuer’s contracts
outside the ordinary course of business. Equals one if the law or the listing rules
require that the terms of material contracts made by the Issuer outside the ordinary
course of its business be disclosed in the prospectus; equals one-half if the terms of
only some material contracts made outside the ordinary course of business must be
disclosed; equals zero otherwise.
Transact An index of the prospectus disclosure requirements regarding transaction between
the Issuer and its directors, officers, and/or large shareholders (i.e., “related
parties”). Equals one if the law or the listing rules require that all transactions in
which related parties have, or will have, an interest be disclosed in the prospectus;
equals one-half if only some transactions between the Issuer and related parties
must be disclosed in the prospectus; equals zero if transactions between the Issuer
and related parties need not be disclosed in the prospectus.
bdn_proof The index of burden of proof equals the arithmetic mean of: (1) Burden director;
(2) Burden distributor; and (3) Burden accountant.
147
bdn_dire Index of the procedural difficulty in recovering losses from the Issuer’s directors in
a civil liability case for losses due to misleading statements in the prospectus.
Equals one when investors are only required to prove that the prospectus contains a
misleading statement. Equals two-thirds when investors must also prove that they
relied on the prospectus and/or that their loss was caused by the misleading
statement. Equals one-third when investors prove that the director acted with
negligence and that they either relied on the prospectus or that their loss was
caused by the misleading statement or both. Equals zero if restitution from
directors is unavailable or the liability standard is intent or gross negligence.
bdn_dist Index of the procedural difficulty in recovering losses from the Distributor in a
civil liability case for losses due to misleading statements in the prospectus.
Equals one when investors are only required to prove that the prospectus contains a
misleading statement. Equals two-thirds when investors must also prove that they
relied on the prospectus and/or that their loss was caused by the misleading
statement. Equals one-third when investors prove that the Distributor acted with
negligence and that they either relied on the prospectus or that their loss was
caused by the misleading statement or both. Equals zero if restitution from the
Distributor is unavailable or the liability standard is intent or gross negligence.
bdn_acco Index of the procedural difficulty in recovering losses from the Accountant in a
civil liability case for losses due to misleading statements in the audited financial
information accompanying the prospectus. Equals one when investors are only
required to prove that the audited financial information accompanying the
prospectus contains a misleading statement. Equals two-thirds when investors
must also prove that they relied on the prospectus and/or that their loss was caused
by the misleading accounting information. Equals one-third when investors prove
that the Accountant acted with negligence and that they either relied on the
prospectus or that their loss was caused by the misleading statement or both.
Equals zero if restitution from the Accountant is unavailable or the liability
standard is intent or gross negligence.
Supervisor The index of characteristics of the Supervisor equals the arithmetic mean of: (1)
Appointment; (2) Tenure; (3) Focus; and (4) Rules.
Appoint Equals one if a majority of the members of the Supervisor are unilaterally
appointed by the Executive branch of government; equals zero otherwise.
Tenure Equals one if members of the Supervisor cannot be dismissed at the will of the
appointing authority; equals zero otherwise.
Focus Equals one if separate government agencies or official authorities are in charge of
supervising commercial banks and stock exchanges; equals zero otherwise.
Rules Equals one if the Supervisor can generally issue regulations regarding primary
offerings and/or listing rules on stock exchanges without prior approval of other
governmental authorities. Equals one-half if the Supervisor can generally issue
regulations regarding primary offerings and/or listing rules on stock exchanges
only with the prior approval of other governmental authorities. Equals zero
otherwise.
Investing The index of investigative powers equals the arithmetic mean of: (1) Document;
and (2) Witness.
Document An index of the power of the Supervisor to command documents when
investigating a violation of securities laws. Equals one if the Supervisor can
generally issue an administrative order commanding all persons to turn over
documents; equals one-half if the Supervisor can generally issue an administrative
148
order commanding publicly-traded corporations and/or their directors to turn over
documents; equals zero otherwise.
Witness An index of the power of the Supervisor to subpoena the testimony of witnesses
when investigating a violation of securities laws. Equals one if the Supervisor can
generally subpoena all persons to give testimony; equals one-half if the Supervisor
can generally subpoena the directors of publicly-traded corporations to give
testimony; equals zero otherwise.
Orders The index of orders equals the arithmetic mean of: (1) Orders issuer; (2) Orders
distributor; and (3) Orders accountant.
ord_iss An index aggregating stop and do orders that may be directed at the Issuer in case
of a defective prospectus. The index is formed by averaging the sub-indexes of
orders to stop and to do. The sub-index of orders to stop equals one if the Issuer
may be ordered to refrain from a broad range of actions; equals one-half if the
Issuer may only be ordered to desist from limited actions; equals zero otherwise.
The sub-index of orders to do equals one if the Issuer may be ordered to perform a
broad range of actions to rectify the violation; equals one-half if the Issuer may
only be ordered to perform limited actions; equals zero otherwise.
ord_dis An index aggregating stop and do orders that may be directed at the Distributor in
case of a defective prospectus. The index is formed by averaging the sub-indexes
of orders to stop and to do. The sub-index of orders to stop equals one if the
Distributor may be ordered to refrain from a broad range of actions; equals one-
half if the Distributor may only be ordered to desist from limited actions; equals
zero otherwise. The sub-index of orders to do equals one if the Distributor may be
ordered to perform a broad range of actions to rectify the violation; equals one-half
if the Distributor may only be ordered to perform limited actions; equals zero
otherwise.
ord_acc An index aggregating stop and do orders that may be directed at the Accountant in
case of a defective prospectus. The index is formed by averaging the sub-indexes
of orders to stop and to do. The sub-index of orders to stop equals one if the
Accountant may be ordered to refrain from a broad range of actions; equals one-
half if the Accountant may only be ordered to desist from limited actions; equals
zero otherwise. The sub-index of orders to do equals one if the Accountant may be
ordered to perform a broad range of actions to rectify the violation; equals one-half
if the Accountant may only be ordered to perform limited actions; equals zero
otherwise.
Criminal The index of criminal sanctions equals the arithmetic mean of: (1) Criminal
director; (2) Criminal distributor; and (3) Criminal accountant.
Crim_dir An index of criminal sanctions applicable to the Issuer’s directors and key officers
when the prospectus omits material information. The sub-index for directors
equals zero when directors cannot be held criminally liable when the prospectus is
misleading. Equals one-half if directors can be held criminally liable when aware
that the prospectus is misleading. Equals one if directors can also be held
criminally liable when negligently unaware that the prospectus is misleading. The
sub-index for key officers is constructed analogously.
Crim_dis An index of criminal sanctions applicable to the Distributor (or its officers) when
the prospectus omits material information. Equals zero if the Distributor cannot be
held criminally liable when the prospectus is misleading. Equals one-half if the
Distributor can be held criminally liable when aware that the prospectus is
misleading. Equals one if the Distributor can also be held criminally liable when
149
negligently unaware that the prospectus is misleading.
Crim_acc An index of criminal sanctions applicable to the Accountant (or its officers) when
the financial statements accompanying the prospectus omit material information.
Equals zero if the Accountant cannot be held criminally liable when the financial
statements accompanying the prospectus are misleading. Equals one-half if the
Accountant can be held criminally liable when aware that the financial statement
accompanying the prospectus are misleading. Equals one if the Accountant can
also be held criminally liable when negligently unaware that the financial
statements accompanying the prospectus are misleading.
Private
Enforcement
The index of private enforcement equals the arithmetic mean of: (1) Disclosure
Index; and (2) Burden of proof index.
Public
Enforcement
The index of public enforcement equals the arithmetic mean of: (1) Supervisor
characteristics index; (2) Investigative powers index; (3) Orders index; and (4)
Criminal index.
Firm Characteristics, Source: Knill (2004)
Asset
tangibility
Fixed assets divided by the book value of total assets; industry average is used in
cases of missing data FA/TA
Capital
Issuance
A dummy variable which equals 1 if firm i issued some type of capital (equity,
debt, convertible, etc.) in time t and 0 otherwise
Cross listing A dummy variable which takes on a value of 1 if the firm has stock listed on
additional exchanges and a 0 otherwise
Desired
Growth in
Sales
For each US firm in our data set, we measure growth in Sales from year t-1 to year
t. Next, we average individual firm growth rates in the same industry and size
classification. The term
US
m j
g
,
is the average growth rate in Sales for firms in
industry j of size m. We use
US
m j
g
,
to approximate desired growth in Sales for a
firm outside the US in industry j and size m.
EFN The ‘external funds necessary’ (or EFN) for firm i in period t is calculated as
follows:
) )( ( ) )( / ( ) )( / (
, , , 1 , , , , 1 , , 1 , 1 , , t i t i t i t i t i t i t i t i t i t i t i t i
RR S M S S S L S S S A EFN ? ? ? ? =
? ? ? ?
where A
t-1
is the total assets of the firm in time t-1, S
t-1
and S
t
are the sales of the
firm in times t-1 and t respectively, L
t
is the liabilities of the firm in time t, M
t
is
the profit margin of the firm as defined by net income divided by sales for time t,
RR is the retention ratio for the firm.
Growth in
assets
Growth in total assets (TA
t
– TA
t-1
)/TA
t-1
/(Year
t
-Year
t-1
)
Industry Macro Industry Code from SDC Platinum
Profitability Operating income divided by sales OpInc/Sales
Risk Standard deviation of the firm’s profitability ratio over the three years prior to
issue; industry average is used in cases of missing data SD(ROA
t
, ROA
t-1
, ROA
t-2
)
Uniqueness
of assets
Selling expense divided by sales; industry average is used in cases of missing data
SellExp/Sales
150
Macroeconomic Characteristics
Anti
Director
This index of Anti-director rights is formed by adding one when: (1) the country
allows shareholders to mail their proxy vote; (2) shareholders are not required to
deposit their shares prior to the General Shareholders’ Meeting; (3)cumulative
voting or proportional representation of minorities on the board of directors is
allowed; (4) an oppressed
minorities mechanism is in place; (5) the minimum percentage of share capital that
entitles a shareholder to call for an Extraordinary Shareholders’ Meeting is less
than or equal to ten percent (the sample median); or (6) when shareholders have
preemptive rights that can only be waved by a shareholders meeting. The range for
the index is
from zero to six. Source: La Porta et al. (1998)
Corruption An assigned value from 0 to 6 of perceived Corruption in a country, 0 being the
most Corrupt and 6 the least. The index is based on the likelihood of solicited
bribes from a country in relation to such factors of business as exchange controls,
tax assessment, and loan protection. Source: International Country Risk Guide
Dom
Credit_GDP
Credit provided by monetary authorities and deposit money banks, as well as other
banking institutions (where data is available). It includes all credit to various
sectors on a gross basis, with the exception of credit to the central government,
which is net. Source: WDI
Efficient
Judiciary
Assessment of the “efficiency and integrity of the legal environment as it affects
business, particularly foreign firms” produced by the country risk rating agency
International Country Risk (ICR). Average between 1980 and 1983. Scale from 0
to10, with lower scores representing lower efficiency levels. Source: International
Country Risk
GDP
Growth
GDP per capital growth (%). Source: WDI
Inflation Inflation levels expressed in percent averaged annually over the period 1996-2002.
Source: WDI
MktCap
Percent
Listed shares * their value, scaled by GDP
Domestic
Capital
Sum of all sources of capital in the domestic economy,
equal to MktCap Percent + Domestic Credit
Foreign
Capital
All foreign investment (Foreign direct investment + Foreign portfolio investment +
other (foreign bank lending and official flows)), scaled by GDP
151
Appendix E: Regression of Capital Issuance on Individual Sub Indexes of
Public and Private Enforcement
The dependent variable is dummy variable for whether the firm issues capital in period t.
Variable definitions are given in Appendix 3. Size is determined by Total Assets divided by
GDP. Firms with Total Assets below(above) the median for each country are defined as small
(large). Probit estimation is used and regressions are clustered by country. We estimate
) ( Pr ) 1 ( Pr
, 6 5 4 , 3 1 , 2 1 , 1 , t i t j k t i t k t i t i
T I SECURITIES EFN Z F ob Y ob ? ? ? ? ? ? ? ? + + + + + + + = =
? ?
A,
where SECURITIES is either supervisor, investig, orders, criminal, disclosure, or burden of
proof. Country-level control variables are averaged over the years t-1 through t-3. Only security
law indexes are reported. Absolute value of z statistics are in parenthesis ** significant at 5%;
*** significant at 1%.
Supervisor Investig Orders Criminal Disclosure Burden of
Proof
Whole
Sample
All Firms 0.325*** 0.091*** 0.090 0.147*** 0.80** 0.011
(23.24) (9.55) (10.33) (12.90) (2.26) (0.57)
Small 0.470*** 0.259*** 0.234*** 0.332*** -0.040 -0.116***
(24.76) (21.27) (19.87) (25.61) (0.89) (5.55)
Large 0.120*** 0.050*** -0.037*** -0.045*** 0.198 0.096***
(6.75) (4.02) (3.27) (2.82) (4.47) (3.72)
G10 Firms
Small 0.007 0.014*** 0.012*** 0.018** 0.225*** -0.041***
(1.19) (5.17) (4.00) (2.53) (7.80) (5.98)
Large -0.315*** -0.247*** -0.251*** -0.685*** -0.087 -0.08*
(6.05) (11.26) (9.76) (13.95) (0.42) (1.96)
Emerging
Markets
Small 0.853*** 0.242*** 0.252*** 0.066 -0.214*** -0.023
(15.20) (6.29) (7.30) (0.81) (2.80) (0.13)
Large 0.370*** 0.184*** 0.120*** 0.088*** 0.115*** 0.076
(8.01) (6.61) (5.00) (2.88) (3.08) (1.62)
152
Bibliography
Abiad, Abdul and Ashoka, Mody. “Financial Reform: What Shakes It? What Shapes It?,
IMF Working Paper , No. 03/70, (Apr. 2003).
Acemoglu, Daren, Johnson, Simon, and Robinson, James. “Instituions as the
Fundemental Cause of Long Run Growth”, NBER Working Paper 10481, (May, 2004).
Acemoglu, Daren, Johnson, Simon, Robinson, James, and Yared, Pierre. “Income and
Democracy”, Working Paper, (Feb., 2005)
Aghion, Phillippe, Alesina, Alberto, and Trebbi, Francesco. “Endogenous Political
Institutions”, NBER working Paper Series, No. 9006, (Jun., 2002).
Aivazlan, Varouj, Booth, Lawerence, Demirguc-Kent, Asli, and Maksimovic Vojislav.
“Capital Structure in Developing Countries”. Journal of Finance, Vol. LXI, No. 1, (Feb.,
2001).
Andrews, Donald and Fair, Ray. “Inference in Nonlinear Econometric Models with
Structural Change”, Review of Economic Studies, Vol. 55, No.4, (Oct. 1998), pp. 615-
639.
Andrews, Josephine and Montinola, Grabiella. “Veto Players and the Rule of Law in
Emerging Democracies”, Comparative Political Studies, Vol. 37, No.1, (Feb., 2004),
pp.55-87.
Ayyagari, Meghana, Demirguc-Kunt, Asli, and Maksimovic, Vojislav. “How Well Do
Institutional Theories Explain Firms’ Perceptions of Property Rights” Working Paper
(Feb., 2005).
Bebchuk, Lucian and Roe, Mark. “A Theory of Path Dependence in Corporate
Governance and Ownership”. Stanford Law Review, (Nov 1999).
Beck, Thorsten, Demirguc-Kunt, Asli, and Levine, Ross. “Law and Firms’ Access to
Finance. World Bank Working Paper, (2002).
Beck, T., Demirguc-Kunt, A. and Maksimovic, V. “Financial and Legal constraints to
Firm Growth: Does size matter?” The Journal of Finance, forthcoming.
Beck, Thorsten & Demirguc-Kunt, Asli and Levine, Ross. "Law, Endowments, and
Finance," Journal of Financial Economics, Vol. 70, No. 2, (2003) pp. 137-181.
Beck, Thornsten, Clarke, George, Groff, Alberto, Keefer, Philip, and Walsh, Patrick,
“New Tools in Comparative Political Economy: The Database of Political Institutions”.
World Bank Economic Review, Vol. 15 No.1. (September 2001), pp. 165-176.
153
Becker, Gary S. “A Theory of Competition Among Pressure Groups for Political
Influence”. Quarterly Journal of Economics, Vol. 98 (Aug., 1983) pp. 371 -400.
Beim, David. O and Calomiris, Charles W. Emerging Financial Markets. McGraw-Hill:
New York, (2001).
Bencivenga, Valerie R. and Smith, Bruce D. “Deficits, Inflation and the Banking System
in Developing Countries: The Optimal Degree of Financial Repression”. Oxford
Economics Paper. Vol. 44, Iss.4 (Oct., 1992), pp. 767-790.
Bernheim, B. Douglas and Whinston, Michael D. “Common Agency”, Econometrica
Vol. 54, No. 4 (July 1986) pp. 923-942.
Biasis, Bruno and Mariotti, Thomas “Credit, Wages and Bankruptcy Laws”, Working
Paper (Sep. 2003).
Blackburn, Keith and Hung, Victor. “A Theory of Growth, Financial Development, and
Trade” Economica, New Series, Vol. 65, No. 257, (Feb. 1998), pp. 107-124.
Booth, L. Aivazian, V. Demirguc-Kunt, A. and Maksimovic, V. “Capital Structure in
Developing Countries”, Forthcoming, Journal of Finance.
Buffie, Edward F. “Capital Rationing and Capital Accumulation”, Economica, New
Series, Vol. 58, Iss. 231, (Aug. 1991) pp. 299-316.
Claessens, Stijin, Djankov, Simon, and Lang Larry H.P. “The Separation of Ownership
and control in East Asian Corporations”, Journal of Financial Economic,(2002) Vol. 58
pp. 81-112.
Demirguc-Kunt, Asli, and Maksimovic, Vojislav. “Law, Finance, and Firm Growth”.
Journal of Finance, vol. 63, No. 6, (Dec., 1998).
Demirguc-Kunt, Asli, and Maksimovic, Vojislav. “Stock Market Development and Firm
Finance Choices”, World Bank Policy Research Working Paper 1461, (May 1995).
Desai, Mihir, Dyck, A., and Zingales, Luigi. “Theft and Taxes”, NBER Working Paper
W10978 (Dec. 2004).
Djankov, Simon, La Porta, Raphael, and Shleifer, Andrei. “Regulation of Entry”,
Quarterly Journal of Economics, (2002) Vol. 117 pp. 1 -37.
Djankov, Simeon, McLiesh, Caralee, and Shleifer Andrei. “Private Credit in 129
Countries”, NBER Working Paper No. W11078, (Jan 2005).
154
Djankov, Simon, and Murrell, Peter. “Enterprise Restructuring in Transition: A
Quantitative Survey”. Journal of Economic Literature, (2002).
Drazen, Alan. Political Economy in Macroeconomics, Princeton UP, (2000).
Dyck A. and Zingales L. “Private Benefits of Control: An International Comparison”.
The Journal of Finance, April 2004, vol. 59, no. 2, pp. 537-600.
Easterly, William and Levine, Ross. “Africa’s Growth Tragedy: Policies and Ethnic
Divisions”, The Quarterly Journal of Economics, Vol. 1212, No.4, (Nov., 1997), pp.
1203-1250.
Easterly, William Levine, Ross. “Tropics, Germs, and Crops: How Endowments
Influence Economic Development. Journal of Monetary Economics, Vol. 50, Iss. 1, (Jan.,
2003), pp. 3-39.
Engerman, Stanley, and Kenneth Sokoloff, 2002. “Factor Endowments, Inequality, and
Paths of Development among New World Economies”. NBER Working Paper,No. 9259.
Esfahani, Hadi Saleki. “Institutions and Government Controls”. Journal of Development
Economics. Vol. 63 (2000) pp. 197 -229.
Faccio, Mara. “Politically-Connected Firms: Can They Squeeze the State” Working
Paper 2002.
Fisman, Raymond. “Estimating the Value of Political Connections”, American Economic
Review, Vol. 91, pp. 1095 -1102.
Frankel, Jeffery, and Romer, David. “Does Trade Cause Growth?”, The American
Economic Review, Vol. 89, No. 3, (Jun., 1999), pp. 379-399.
Glen, Jack, and Pinto, Brian. “Debt or Equity? How Firms in Developing Countries
Choose”. IFC Discussion Paper No. 22 (1994).
Gordan, Robert H., and Lei, Wei. “Government as a Discriminating Monopolist in the
Financial Market: The Case of China”. Working Paper (Jun., 2001).
Gradstein, Mark, 2004. “Inequality, Democracy and the Emergence of Institutions”.
CEPR Discussion Paper No. 4187.
Greene, William. Econometric Analysis, Prentice-Hall (1997), Saddle River, NJ.
Greif, Avner, Milgrom, Paul, & Weingast, Barry R. (1994). Coordination, Commitment,
and Enforcement: The case of the merchant guild. Journal of Political Economy, 102(4),
745-777.
155
Grossman, Gene M, and Helpman, Elhanan. Special Interest Politics MIT (2001),
Cambridge, MA, pp. 225-279.
Harris, Milton, and Raviv, Artur. “The Theory of Capital Structure”. Journal of Finance,
Vol. 46, Iss. 1 (Mar., 1991), pp. 297-355.
Hellman, Joel. “Winners Take All: The Politics of Partial Reform in Postcommunist
Transitions”, World Politics, 50, (1998) pp. 203-234.
Henisz, Witold. “The Institutional Environment for Economic Growth”, Economics and
Politics, Vol. 12, No. 1, (Jan., 2002), pp. 1-31.
Henry, Peter Blair. “Do Stock Market Liberalizations Cause Investment Booms”. Journal
of Financial Economics, 58, (2000) pp. 301-334.
Himmelburg, Charles, Hubbard, R. Glen, and Love, Inessa. “Investor Protection,
Ownership and the Cost of Capital”, World Bank Working Paper Series No. 2834.
Islam, Romeen and Montenegro, Claudio. “What Determines the Quality of Institutions”,
Background Paper for the World Develepment Report, 2002: Building Institutions for
Markets, (Jan., 2002).
Jensen, Michael C., and Meckling, William H. “A Theory of the Firm: Managerial
Behavior, Agency Costs, and Ownership Structure”. Journal of Financial Economics.
Vol. 3, No. 4 (Oct., 1976) pp. 305-360.
Johnson, Simon and Todd Mitton. “Cronyism and Capital Controls: Evidence from
Malaysia”. Journal of Finance, (Jan., 2002).
Kaufman, Daniel, Kraay, Aart, and Zoido-Labaton, Pablo, “Governance Matters”, World
Bank Policy Research Working Paper, (1999) No. 2196.
Keefer, Philip, “When Do Special Interests Run Rampant? Disentangling the Role in
Banking Crises Elections, Incomplete Information, and Checks and Balances,” World
Bank Research Working Paper 2543.
Keefer, Philip and Stasavage, David. "The Limits of Delegation: Veto Players, Central
Bank Independence and the Credibility of Monetary Policy." American Political Science
Review (August 2003)
Keefer, Philip and Stasavage, David. “Bureaucratic Delegation and Political Institions:
When Are Independent Central Banks Irrelevant?” (June 2001) World Bank Policy
Working Paper 2356.
Korajczyk, R. and Levy, A. “Capital Structure Choice: Macroeconomic Conditions and
financial constraints”, Journal of Financial Economics Vol. 68, Iss. 1, (2003) pp. 75-109.
156
Kroszner, Randall, “On the Political Economy of Banking and Financial Regulatory
Reform in Emerging Markets”. Working Paper, (Apr., 1998).
Krishnan, Sharma. “The Underlying Constraints of Corporate Bond Market Development
in South East Asia”. DESA Discussion Paper, No.14.
Laeven, L. and Woodruff, C. “The Quality of the Legal System, Firm Ownership, and
Firm Size”, World Bank Working Paper.
La Porta, Raphael, Lopez-de-Silanes, Florencio, Pop_ELeches, Cristian, and Shleifer,
Andrei. “The Guarantees if Freedom”, NBER Working Paper 8157, (Feb. 2002).
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, and Vishny, Robert W.
“Law and Finance”. The Journal of Political Economy, Vol. 106, No. 6. (Dec., 1998), pp.
1113-1155.
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, “Corporate Governance
Around the World”. Journal of Finance, Vol. 54, No.2 April (1999) pp.471-518.
La Porta, Raphael, Lopez-de-Silanes, Florencio, and Shleifer, Andrei “Government
Ownership of Banks”. Journal of Finance. Vol 57, Issue 1 (Jan. 2002).
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, and Vishny, Robert W.
“Investor Protection and Corporate Valuation, Journal of Finance, Vol. 57, No. 3, June
(2002), pp. 1147-1170.
La Porta, Raphael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, and Vishny, Robert W.
“Investor Protection and Corporate Valuation”. Journal of Finance. Vol. LVII, No. 3
(Jun., 2002).
La Porta, Rafael, Lopez-de-Silanes, Florencio, Shleifer, Andrei, & Vishny, Robert.
(1999). The Quality of Government. Journal of Law, Economics, & Organization, 15,
222-279.
Lau, Laurence J., Qian, Yingyi, Gerard Roland. “Reform Without Losers: An
Interpretation of China’s Dual Track Approach to Transition”, Journal of Political
Economy.
Levine, Ross. “Financial Development and Economic Growth: Views and Agenda”.
Journal of Economic Literature, Vol. 35, No. 2. (Jun., 1997), pp. 688-726.
Lederman, Daniel, Loayza, Norman, and Soares, Rodrigo. “Accountability and
Corruption: Political Institutions Matter”. Economics and Politics, Vol. 7, Iss.1, (March,
2005), pp. 1-35.
157
Levine, Ross. “The Legal Environment, Banks, and Long Run Growth”, Journal of
Money, Credit, and Banking, Vol. 30, No.3, Part 2: Comparative Financial Systems
(Aug. 1998), pp. 596-613.
Levine Ross, and Zervos, Sara. “Stock Markets, Banks, and Economic Growth”,
American Economic Review, Vol. 88, Iss. 3, (Jun., 1998), pp. 537-558.
Levine, Ross and Demiriguc-Kunt, Asli. Financial Structure and Economic Growth:
Across Country Comparison of Banks, Markets and Development. MIT, Cambridge
(2001).
Lukauskas, Arvid J. “The Political economy of Financial Restriction: The Case of
Spain”. Comparative Politics. Vol. 27, Iss. 1 (Oct 194), pp. 67-89.
Matsuyama, Kiminori. “Endogeneous Inequality”. Review of Economic Studies, Vol. 67,
pp. 743-759 (2000).
Meyer, Stuart C. “The Capital Structure Puzzle”. The Journal of Finance, Vol. 39, Iss. 3,
(Jul. 1984), pp. 575-592.
Morck, Randall, Yeung, Bernard, and Yu, Wayne. “The Information Content of Stock
Markets: Why do Emerging Markets Have Synchronous Stock Price Movement”.
Journal of Financial Economics, 58, (2000), pp. 215 -260.
Moser, Peter. “Checks and Balances, and the Supply of Central Bank Independence”,
European Economic Review, 43,(1999), 1569-1593.
Nejadmalayeri, A. “Interest Rates, Taxes, Bankruptcy and Financing Decisions”,
Working Paper, University of Nevada.
Olson, Mancur, 1965. The Logic of Collective Action: Public Goods and the Theory of
Groups. Harvard University Press: Cambridge, MA.
Pagano, Marco and Volpin, Paolo F. “The Political Economy of Finance”. CEPR
Discussion Paper, No. 3231 (Feb., 2002).
Perotti, Enrico and Volpin, Paolo. “Lobbying on Entry”. Working Paper, (May 2004).
Persson, Torsten. “Do Political Insitutions Shape Economic Policy?”, Econometrica,
Vol. 70, No. 3 (May 2002), pp. 883-905.
Rajan, Raghuran G. , and Zingales, Luigi. ,Financial Dependence and Growth”. American
Economic Review, Vol. 88, Iss. 3, (Jun., 1998), pp. 559-586.
158
Rajan, Raghuran G. , and Zingales, Luigi. “The Great Reversals: The Political Economy
of Financial Development in the 20
th
Century”. (Jul. 2002) Journal of Financial
Economics, forthcoming.
Rojas_Suarez, Liliana and Wiesbrod, Steven. “Building Stability in Latin American
Financial Markets”, IADB Workin Paper Series 320, (1996).
Roderick, Dani. “Understanding Economic Policy Reform”. Journal of Economic
Literature.Vol. 34, Iss. 1, (Mar., 1996), pp. 9-41.
Rodrik, Dani, Subramanian, Arvind, and Trebbi, Francesco. “Institutions Rule, The
Primacy of Institutions Over Geography and Integration in Economic Development”,
Journal of Economic Growth, Vol. 9, No. 2 (Jun. 2004).
Shleifer, Andrei, and Vishny, Robert W. “A Survey of Corporate Governance”. The
Journal of Finance, Vol. 52, No. 2. (Jun., 1997), pp. 737-783.
Shleifer, Andrei, and Vishny, Robert W. “Large Shareholders and Corporate Control”.
Journal of Political Economy, Vol. 93, Iss. 3. Part 1, (Jun., 1986) pp. 462-488.
Shleifer, A.; Wolfenzon, D. “Investor Protection and Equity Markets”. Journal of
Financial Economics (2002), vol. 66, no. 1, pp. 3-27.
Singh, Ajit. “Corporate Financial Patterns in Industrializing Countries: A Comparative
International Study”, IFC Technical Paper No. 2 (Apr., 1995).
Stultz, Rene and Williamson, Rohan. “Culture, Openness, and Finance” NBER Working
Paper No. 8222, (April 2001)
Tsebelis, George. “Decision Making in Political Systems: Veto Players in
Presidentialism, Parliametarism, Multicameralism and Multipartyism, British Journal of
Political Science, Vol. 25, No. 3. (Jul., 1995), pp. 289-325.
Tsebelis, George. (1999). Veto players and law production in parliamentary democracies:
An Empirical analysis. American Political Science Review, 93(3), 591-608.
Tsebelis, George. (2002).Veto Players: How Political Institutions Work. Princeton, NJ:
Princeton University Press.
Tirole, Jean. “Corporate Governance”, Econometrica, Vol. 60, Iss. 1 (2001), pp. 1-35.
Titman, Sheridan and Roberto Wessels. “The Determinants of Capital Structure Choice”.
Journal of Finance, Vol. 43, Iss. 1 (Mar., 1998) pp. 1-19.
159
Tybout, James R. “Manufacturing Firms in Developing Countries: How Well Do They
Do and Why?”. Journal of Economic Literature. Vol. XXXVIII, (Mar., 2000), pp. 11–
44.
Wurgler, Jeffrey. “Financial Markets and the Allocation of Capital”. Journal of Financial
Economics, Vol. 58, (2000), pp.1
doc_390815765.pdf