Board external connectedness and earnings management

Description
In this study we explore how board external connectedness is related to firm's earnings management.
Board external connectedness is gauged by the percentage of board members with at least one external
board membership, and earnings management is gauged by the level of discretionary accruals from the
modified Jones model. We postulate that externally connected directors could learn something from
their connected firms, and this experience could be translated into monitoring effectiveness which in
turn reduces a firm's level of earnings management. We use the sample consisting of 5940 firm-year
observations from the listed firms in Taiwan in 2007e2011 sampling period to test the learning-effect
hypothesis. The empirical finding supports this hypothesis by showing that board external connectedness
is negatively correlated with the level of earnings management and that the negative relation is
more conspicuous for the same-industry connectedness. However, the negative relation between board
external connectedness and earnings management becomes insignificant when the firm has an external
financing plan. Finally, independent directors' same-industry connectedness does reduce the level of
earnings management when the firm has an external financial plan.

Board external connectedness and earnings management
Pei-Gi Shu
a, *
, Yin-Hua Yeh
b
, Shean-Bii Chiu
c
, Ya-Wei Yang
d
a
Department of Business Administration, Fu Jen Catholic University, Taiwan, ROC
b
Graduate Institute of Finance, National Chiao Tung University, Taiwan, ROC
c
Department of Finance, National Taiwan University, Taiwan, ROC
d
Department of International Business and Finance, Fu Jen Catholic University, Taiwan, ROC
a r t i c l e i n f o
Article history:
Received 26 August 2013
Accepted 3 March 2015
Available online 26 June 2015
Keywords:
Earnings management
Board external connectedness
Learning effect
a b s t r a c t
In this study we explore how board external connectedness is related to ?rm's earnings management.
Board external connectedness is gauged by the percentage of board members with at least one external
board membership, and earnings management is gauged by the level of discretionary accruals from the
modi?ed Jones model. We postulate that externally connected directors could learn something from
their connected ?rms, and this experience could be translated into monitoring effectiveness which in
turn reduces a ?rm's level of earnings management. We use the sample consisting of 5940 ?rm-year
observations from the listed ?rms in Taiwan in 2007e2011 sampling period to test the learning-effect
hypothesis. The empirical ?nding supports this hypothesis by showing that board external connected-
ness is negatively correlated with the level of earnings management and that the negative relation is
more conspicuous for the same-industry connectedness. However, the negative relation between board
external connectedness and earnings management becomes insigni?cant when the ?rm has an external
?nancing plan. Finally, independent directors' same-industry connectedness does reduce the level of
earnings management when the ?rm has an external ?nancial plan.
© 2015 College of Management, National Cheng Kung University. Production and hosting by Elsevier
Taiwan LLC. All rights reserved.
1. Introduction
Prior studies of board interlocking illustrate how it affects
corporate behaviors such as investment choices, mergers and ac-
quisitions, compensation practices, poison pill adoption, stock ex-
change listing decisions, and earnings management (Bizjak,
Lemmon, & Whitby, 2009; Cai & Sevilir, 2012; Cohen, Frazzini, &
Malloy, 2008; Davis, 1991; Fracassi & Tate, 2012; Haunschild,
1993; Hirshleifer & Teoh, 2009; Ishii & Xuan, 2014; Rao, Davis, &
Ward, 2000; Stuart & Yim, 2010). A recent study of Chiu, Teoh,
and Tian (2013) indicates that earnings management is like a vi-
rus that spreads from one ?rm to another via board connections of
shared directorships. These shared directors are like virus carriers
in the sense that the directors of the infected ?rms carry these
earnings management behaviors to susceptible ?rms on whose
boards they also sit on. The study implies that board links are
positively correlated with a ?rm's propensity to engage in earnings
management.
In this study we explore a similar issue to relate board members'
external connectedness to a ?rm's earnings management. In
contrast to the contagion effect illustrated in Chiu et al. (2013), we
develop the learning effect hypothesis indicating that externally
connected board members could learn experiences from their
outside sitting boards. These experiences could be translated into
monitoring effectiveness and therefore result in a lower level of
earnings management. The results from our sample consisting of
5940 ?rm-year observations from the listed ?rms in Taiwan during
2007e2011 verify the learning effect hypothesis. Moreover, we
further look into the attributes of the connectedness and ?nd that
the monitoring effectiveness in reducing level of earnings man-
agement is more conspicuously found for the same-industry
connectedness than for different-industry connectedness.
We further follow prior studies to relate earnings management
to a certain critical corporate event such as seasoned equity offering
(e.g., Dechow, Sloan, & Sweeney, 1996; Rangan, 1998; Teoh, Welch,
* Corresponding author. No. 510, Zhongzheng Rd., Xinzhuang Dist, New Taipei
City 24205, Taiwan, ROC.
E-mail address: [email protected] (P.-G. Shu).
Peer review under responsibility of College of Management, National Cheng
Kung University.
HOSTED BY
Contents lists available at ScienceDirect
Asia Paci?c Management Review
j ournal homepage: www. el sevi er. com/ l ocat e/ apmrvhttp://dx.doi.org/10.1016/j.apmrv.2015.03.003
1029-3132/© 2015 College of Management, National Cheng Kung University. Production and hosting by Elsevier Taiwan LLC. All rights reserved.
Asia Paci?c Management Review 20 (2015) 265e274
& Wong, 1998). We postulate that ?rms with external ?nancing
plans including seasoned equity offerings and debt offerings would
have stronger motives to engage in earnings management so that
their seasoned shares or debt could be sold at a higher price level.
We ?nd that the factor of having an external ?nancing plan is so
strong as to dominate other factors in explaining the level of
earnings management: not only board external connectedness but
also other factors including ?rm size, Tobin's Q, supervisors'
educational level, and ROA lose their explanatory power when
?rms have external ?nancing plans.
We further decompose the connectedness by board members'
status: outside, inside, independent directors, and supervisors. The
results indicate that the external connectedness of outside di-
rectors and supervisors is able to signi?cantly reduce the level of
earnings management. Moreover, if the learning effect could
explain why board external connectedness is negatively correlated
with the level of earnings management, we would expect that for
these connected board members the learning effect fromthe same-
industry exposure is stronger than that from different-industry
exposure. The results support this hypothesis that the same-
industry connectedness of outside directors, inside directors, and
supervisors is able to reduce the level of earnings management
when there is no external ?nancing plan. Finally, we ?nd a special
pattern regarding independent directors: they are passive in
affecting ?rm's level of earnings management when ?rms have no
external ?nancing plan. However, their external connectedness in
the same industry is negatively correlated with ?rm's earnings
management when the ?rm has an external ?nancing plan.
There are several differences between our investigation and
Chiu et al. (2013). First, we investigate the level of earnings man-
agement rather than the inclination of earnings management.
Second, we explore board external connectedness rather than
board interlock.
1
More importantly, we ?nd a signi?cantly negative
relation between board external connectedness and level of earn-
ings management. This is in sharp contrast with the ?nding of Chiu
et al. (2013). The negative correlation is mainly due to the
connectedness of outside directors and supervisors, implying that
these outside directors and supervisors are able to provide a better
monitoring function when they have external connectedness.
However, the monitoring effectiveness due to board external
connectedness simply disappears when the underlying ?rm has an
external ?nancing plan. The positive relation between external
?nancing plan and earnings management is so strong as to domi-
nate the monitoring effectiveness indebted to board external
connectedness. Such relation implies that independent directors
play a good goalkeeper when ?rms have an external ?nancing plan
that involves new stakeholders.
The potential contributions of this paper are multifold. First, we
document the negative relation between board external connect-
edness and ?rm's earnings management and propose the learning
effect to account for the negative relation. The relation might sus-
tain in an emerging market where board interlock creates value in
terms of an enhancement of monitoring effectiveness. Second, we
identify the predominating effect of external ?nancing plans on
?rms highly motivated to engage in earnings management. Third,
we identify board member's status and calculate the external
connectedness for each status. An upclose look allows us to get a
better understanding of how external connectedness of different
board status is related to ?rm's earning management. Finally, we
speci?cally portray the role of independent directors: we ?nd them
passive when there is no external ?nancing plan and active in
reducing ?rm's level of earnings management when there is an
external ?nancing plan. The rest of this paper is organized as fol-
lows. Section 2 reviews literature and develop testing hypotheses.
Section 3 depicts the data and variables. Section 4 reports the
empirical results. Section 5 concludes.
2. Literature review and hypothesis development
2.1. Board external connectedness and earnings management
The existing literature relating to board external connectedness
could be rami?ed into four threads: reputation, information ?ow,
busyness, and contagion. The reputation effect proposed by Fama
(1980) and Fama and Jensen (1983) indicates that multiple di-
rectorships allow directors to develop reputations as monitoring
specialists (Fama, 1980; Fama & Jensen, 1983) and increase their
prestige, visibility, and commercial contact (Mace, 1971). Support-
ing evidence for the positive reputation effect of multiple di-
rectorships was found in various corporate scenarios.
2
Information
?ow emerges as a new thread of studies indicating that with the
conduit of interpersonal and inter-organizational networks, board
external connectedness facilitates information communication,
resources exchange, formation of new relationships, and
enhancement of existing relationships. Connelly and Van Slyke
(2012) indicate that the social context is instrumental for ?rms to
determine what information they must constantly gather and
assess. The information will help them identify and evaluate
emerging opportunities. Unfortunately, many ?rms allow them-
selves to become under-connected and isolate themselves from
new sources of information. Other than the bene?t of accessing
useful information, board external connectedness also mitigates
information asymmetry. For example, Cai and Sevilir (2012) ?nd
that M&A transactions are bene?ted from board external
connectedness which is further rami?ed into ?rst-degree and
second-degree connection. Butler (2008), Botosan (1997) and
Chuluun, Prevost, and Puthenpurackal (2014) ?nd that board
external connectedness is negatively correlated with yield spread.
If reputation effect illustrates motives while information ?ow
illustrates the conduits for these externally connected directors to
have the ?rm run properly, the learning effect is jointly developed
to contain both motives and conduits and indicates that externally
connected directors could learn from their external experiences
from serving multiple directorships. Xie, Davidson, and DaDalt
(2003) indicate that board members' ?nancial sophistication is an
important factor in constraining the propensity of managers to
engage in earnings management. If directors could learn from their
outside directorships and therefore enhance the capability of
monitoring, surveillance, and advising to the incumbent manage-
ment, their acquired knowledge or sophistication from external
connectedness would help detect and reduce the underlying ?rm's
level of earnings management. If the learning effect prevails, we
would expect to ?nd that board external connectedness is nega-
tively correlated with ?rm's level of earnings management.
1
The two terms are similar in meaning. However, there are two subtle differ-
ences between the two terms. First, board external connectedness focuses more on
the connected directors while board interlock focuses more on the connected ?rms.
Second, board external connectedness focuses more on directors' outbound re-
lations while board interlock focuses more on reciprocal relations between two
connected ?rms.
2
The positive reputation effect of multiple directorships, using the proxy of the
number of board seats held by an outside director (Brown & Maloney, 1999;
Shivdasani, 1993; Vafeas, 1999), is applied to antitakeover provisions (Coles &
Hoi, 2003), CEO replacement (Farrell & Whidbee, 2000), CEOs following retire-
ment (Brickley, Linck, & Coles, 1999), ?nancially distressed ?rms (Gilson, 1990),
?rms that cut dividends (Kaplan & Reishus, 1990), and ?rms that are sold (Harford,
2003).
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 266
Moreover, it is reasonable to argue that the learning effect is
more conspicuous for the same-industry connectedness. The
argument is made based on the premise that earnings management
could be industry speci?c. For example, prior studies explore
earnings management for speci?c industries, such as oil industry
(Hall & Stammerjohan, 1997), cable television industry (Key, 1997),
and health-care industry (Mensah, Considine, & Oakes, 1994). The
unique features and characteristics pertaining to speci?c industries
imply that the learning fromcompanies within the same industry is
better than the learning from companies across industries.
Hypothesis 1a. If learning effect prevails, board external
connectedness is negatively correlated with ?rm's level of earnings
management, and the negative correlation is more conspicuous
when the external connectedness is in the same industry.
In contrast, board busyness and contagion effects are two
hypotheses indicating that board external connectedness could
be value detrimental. Board busyness weakens corporate
governance and therefore is detrimental to ?rm value. The
related studies are widely seen in different corporate scenarios.
3
Still, Chiu et al. (2013) follow the studies on social networks
4
and
develop the contagion effect hypothesis indicating that earnings
management is like a virus spreads from ?rm to ?rm via board
connections of shared directors. They ?nd that contagion is
stronger when the shared director has a leadership position or
an accounting-relevant position in the susceptible ?rm. The
contagion hypothesis predicts that board external connectedness
is positively correlated with ?rm's level of earnings
management.
Hypothesis 1b. If contagious effect prevails, board external
connectedness is positively correlated with a ?rm's level of earn-
ings management.
2.2. Earnings management and external ?nancing needs
According to Generally Accepted Accounting Principles (GAAP),
accounting accruals are used in periods when an entity of trans-
actions have cash consequences rather than in periods when cash is
received or paid. This gives managers discretion in determining
which period of actual earnings to be recorded. They can control
over the timing of advertising expenses or R&D expenditure on one
hand, and alter the timing of revenue/expense recognition on the
other (Teoh et al., 1998).
The level of earnings management is affected by managers'
self-interest when their incentives are aligned with ?rms'
?nancial performance. The incentives include salary and bonus
that are directly related, and perquisites, promotions, and job
security that are indirectly related to ?rms' ?nancial perfor-
mance. A potential agency problem arises when managers are
allured to engage in earnings management for their self-interest.
Earnings management results in false information that affects
investors' investment performance and the ef?ciency of capital
markets.
A growing thread of studies explored corporate events around
which earnings management is most severe. For example, earn-
ings management is more conspicuous when ?rms meet company
forecasts (Kasznik, 1999) or analyst forecasts (Burgstahler &
Eames, 2006), are in political, regulatory and legal proceedings
(Hall & Stammerjohan, 1997; Jones, 1991; Key, 1997; Lim &
Matolcsy, 1999; Magnan, Nadeau, & Cormier, 1999; Makar &
Alam, 1998; Mensah et al., 1994), in management buyouts (Perry
& Williams, 1994; Wu, 1997), in takeover or merger settings
(Easterwood, 1998; Erickson & Wang, 1999), and in issuance of
seasoned equity (Dechow et al., 1996; Rangan, 1998; Teoh et al.,
1998). One commonality among these events is the interface
with capital market in the sense that ?rms are more likely to
engage in earnings management when they plan to interact with
the external capital market. In this study we use the event of
external ?nancing which includes seasoned equity offering and
the public placement of company debt. We propose that ?rm's
earnings management is more salient when the ?rm has an
external ?nancing plan.
Hypothesis 2. Firm's earnings management is positively corre-
lated with the external ?nancing plan taking place in the following
period.
3. Data and variables
Our data comprising 5836 ?rm-year observations in the
period of 2007e2010 is collected from the Taiwan Economic
Journal (TEJ), a database company in Taiwan. We exclude ?nancial
?rms from the sample since they are subject to different
regulations.
Table 1 reports the sample distribution with yearly and in-
dustry breakdown. The fact that electronic ?rms comprise more
than half of the sample re?ects its importance in driving the
economic development of Taiwan in the sampling period. The
second large industry is the biomedical industry which includes
the subcategories of biotechnology, pharmaceuticals, and medical
devices.
The ?rst proxy variable, earnings management, aims to detect
the abnormal accruals (or discretionary accruals). Abnormal ac-
cruals are assumed to be the consequence of accounting manipu-
lation. There are several suggested models to estimate non-
discretionary (or “normal”) accruals and discretionary accruals.
We use the more frequently used modi?ed Jones model to measure
accrual earnings management.
5
3
For example, interlocked boards spread the practices of option backdating
(Bizjak et al., 2009), weaken monitoring effectiveness (e.g., Core, Holthausen, &
Larcker, 1999; Shivdasani & Yermack, 1999), stipulate excessive managerial
compensation (Barnea & Guedj, 2009; Core et al., 1999), and are ineffective to
replace an underperforming CEO (Core et al., 1999; Fich & Shivdasani, 2006).
Moreover, Beasley (1996) indicates that the probability of committing accounting
fraud is positively related to the average number of directorships held by outside
directors. Hwang and Kim (2008) indicate that ?rms with audit committees
maintaining greater social ties to the CEO are associated with higher levels of
earnings management. Hwang and Kim (2009) point out that the socially-
independent boards are better in monitoring than conventionally-independent
boards. Fracassi and Tate (2012) indicate that ?rms with powerful CEOs are more
likely to appoint friendly directors leading to weaker board monitoring.
4
A thread of research on social networks indicates that social linkages especially
through the network of interlocking boards might affect several types of corporate
behaviors such as investment choices, mergers and acquisitions, compensation
practices, poison pill adoption, and stock exchange listing decisions (Bizjak et al.,
2009; Cai & Sevilir, 2012; Cohen et al., 2008; Davis, 1991; Fracassi & Tate, 2012;
Haunschild, 1993; Ishii & Xuan, 2014; Rao et al., 2000; Stuart & Yim, 2010).
5
There are numerous models for gauging earnings management, such as the
Jones model (Jones, 1991), the modi?ed Jones model (Dechow, Sloan, & Sweeney,
1995), the modi?ed Jones model with book-to-market ratio and cash ?ows as
additional independent variables (Larcker & Richardson, 2004), the modi?ed Jones
model with either current year or prior year return on assets (ROA) included as an
additional independent variable, performance-matched discretionary accruals
estimated from the modi?ed Jones model (Kothari, Leone, & Wasley, 2005), two
measures of accrual quality (accrual estimation errors) following Dechow and
Dichev (2002) and McNichols (2000), and the Beneish (1997, 1999) unweighted
and weighted probabilities of earnings manipulation. Jones, Krishnan, and
Melendrez (2008) ?nd that all the ten measures are qualitatively similar and
signi?cantly associated with a fraudulent event. In this study we do not speci?cally
cover all alternative measures of earnings management.
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 267
TAC
t
¼ a
t
ð1=TA
tÀ1
Þ þb
t
ðDSALES
t
ÀDAR
t
Þ þg
t
PPE
t
þz
t
ROA
tÀ1
þ?
t
(1)
where TAC
t
denotes total accruals in year t, calculated using the
statement of cash ?ow approach recommended by Hribar and
Collins (2002) as income before discontinued operations and
extraordinary items e (cash from operations e discontinued op-
erations and extraordinary items fromthe statement of cash ?ows);
TA
te1
denotes total assets at the end of year tÀ1; DSALES
t
denotes
the change in sales revenue between year t and year t-1; DAR
t
denotes the change in accounts receivable between year t and year
tÀ1; PPE
t
denotes the gross amount of property, plant and equip-
ment at the end of year t; and ROA
te1
denotes return on assets in
year t-1, calculated as the ratio of income before discontinued op-
erations and extraordinary items to total assets. We note that TAC
t
,
DSALES
t
, DAR
t
, and PPE
t
are scaled by lagged total assets (TA
tÀ1
).
Equation (1) is estimated annually with respect to TEJ industry
classi?cation.
Nondiscretionary accruals are estimated during the event year
as:
NDA
t
¼ atð1=TA
tÀ1
Þ þbtðDSALES
t
ÀDAR
t
Þ þg
t
PPE
t
þz
t
ROA
tÀ1
(2)
Performance-adjusted discretionary accruals are calculated as
the difference between Equations (1) and (2). In Table 2 we report
the summary statistics. The average discretionary accruals are 0%,
implying the listed ?rms on average are not very active in earnings
management.
We assign the dummy of external ?nancing the value of 1 when
the underlying ?rm has an external ?nancial plan and 0 otherwise.
An external ?nancing plan comprises of a seasoned equity offering
and a seasoned debt offering. This variable is included in the model
to capture the possibility that ?rms would have a higher motive to
engage in earnings management so as to have their seasoned eq-
uity/debt offerings tendered at higher selling price. In our ?rm-year
dataset, there are 13.8% of the cases involving external ?nancing
plans. The average ?rm assets are NT 18.771 billion (equivalent to
US 626 million). The average debt ratio is 42.2%.
The most important variable of this paper is board external
connectedness, which is de?ned as the percentage of board
members serving at least one external board membership. Since
prior studies were ambivalent to multiple board membership as it
might be a plus to ?rm value due to the reputation effect
6
or a
minus due to board busyness,
7
we use the neutral term of board
external connectedness to balance the possible pros and cons
associated with it. For all the memberships comprising a board
including directors and supervisors, the average percentage of
external connectedness is 46.4%. We further calculate the
connectedness with respect to status of board members. The board
member's status is decomposed into outside, inside, independent,
and supervisor. The average percentage of member's external
connectedness is 36.3%, 70.8%, 48.6%, and 36.2%, respectively.
Moreover, we expect that the impact of external connectedness on
?rm's earnings management would be more saliently found when
the connection is from the same industry than from different in-
dustries. The average percentage of board member's external
connectedness in the same industry is 33.2%. For each director's
status the average percentage of external connectedness is 33.1%,
24.2%, 36.6% and 26% for the status of outside, inside, independent,
and supervisor, respectively.
Other than the aforementioned control variables, we also
include variables of corporate governance, including board size,
board independence, chairman's tenure, directors' shareholding,
and institutional shareholding. Board size is the number of di-
rectors/supervisors in a board. On average a board is comprised of
10.75 directors/supervisors. Prior studies illustrate two ambivalent
arguments relating board size and corporate governance. On the
one hand, larger boards are less effective in performing their duties
(Dechow et al., 1996; Peasnell, Pope, & Young, 2005). On the other,
increasing the board size increases the possibility of including more
directors on the board with relevant ?nancial reporting knowledge
and experience (Beasley & Salterio, 2001). Board independence is
de?ned as the percentage of independent directors in the board.
Table 2
Summary statistics.
Mean Median S.D. Q1 Q3
DA (%) 0.000 À0.005 0.169 À0.061 0.054
Dummy (external ?nancing) 0.138 0.000 0.345 0.000 0.000
BECall 0.464 0.500 0.307 0.250 0.667
BECoutside 0.363 0.333 0.312 0.000 0.600
BECinside 0.708 0.750 0.313 0.500 1.000
BECindependent 0.486 0.500 0.358 0.000 0.667
BECsupervisor 0.362 0.333 0.331 0.000 0.667
BECall-industry 0.332 0.333 0.305 0.000 0.545
BECout-industry 0.331 0.250 0.357 0.000 0.571
BECinside-industry 0.242 0.000 0.410 0.000 0.500
BECindepdent-industry 0.366 0.000 0.460 0.000 1.000
BECsupervisor-industry 0.260 0.000 0.424 0.000 0.500
Size (NT billion) 18.771 3.699 67.438 1.625 9.853
Debt (%) 0.422 0.421 0.175 0.293 0.544
Board size 10.750 10.000 3.203 9.000 12.000
Board independence (%) 0.164 0.167 0.170 0.000 0.286
Tobin's Q 1.272 1.081 0.769 0.856 1.453
Chairman's tenure (years) 14.710 12.500 11.615 4.500 22.500
Directors' shareholding (%) 23.403 19.610 14.484 12.930 29.990
Institutional shareholding (%) 17.321 11.810 18.071 1.800 27.888
ROA 7.034 7.140 9.302 3.050 11.810
Independent director's
education
1.979 2.000 0.144 2.000 2.000
Supervisor's education 1.926 2.000 0.262 2.000 2.000
Note: Discretionary accruals (DA), estimated by the modi?ed Jones' model, are the
differences between total accruals and non-discretionary accruals. DEF, the dummy
for external ?nancing, is assigned the value of 1 when the underlying ?rmhas a plan
of seasoned equity or debt offerings next year and 0 otherwise. Board external
connectedness (BEC
all
) is the percentage of all directors and supervisors who serve
at least one external board membership. BEC
outside
(BEC
inside
, BEC
independent
, BEC
su-
pervisor
) is the percentage of outside directors (inside directors, independent
directors, supervisors) who serve at least one external board memberships. BEC
all-
industry
(BEC
outside-industry
, BEC
inside-industry
, BEC
independent-industry
, BEC
supervisor-industry
)
is the percentage of all directors (outside directors, inside directors, independent
directors, supervisors) who serve at least one external board memberships in the
same industry. Independent director's (supervisor's) education denotes the di-
rectors' (supervisors') average educational level and is assigned the value of 2 for
receiving the educational level of bachelor degree or above and 1 for receiving the
educational level below bachelor degree.
Table 1
Sample distribution.
2007 2008 2009 2010 2011 Total
Electronics 641 658 684 708 732 3423
Biomedical 70 75 79 82 92 398
Others 400 391 394 405 425 2015
Total 1111 1124 1157 1195 1249 5836
6
For example, Fama (1980) and Fama and Jensen (1983) indicate that the market
for outside directorships allows directors to develop reputations as monitoring
specialists. Mace (1971) suggests that outside directorships provide executives with
prestige, visibility, and commercial contacts and therefore are deemed as valuable
sources of incentives assuming multiple directorships.
7
For example, Fich and Shivdasani (2006) ?nd that board busyness, being
de?ned as boards in which a majority of independent directors within a ?rm hold
three or more directorships, results in lower market-to-book ratio, weaker pro?t-
ability, and lower sensitivity of CEO's turnover to ?rm performance. Moreover, Core
et al. (1999) ?nd that ?rms with busy directors tend to pay their CEOs excessively.
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 268
The average board independence is 16.4%. Prior studies widely
support that earnings management is negatively related to the
proportion of outside directors on the board. In other words,
earnings management is reduced when there is more board inde-
pendence (Beasley, 1996; Dechow et al., 1996; Klein, 2002). Chair-
man's tenure is the number of years of the incumbent CEO in the
company. On average, the incumbent chairman's tenure is 14.71
years. Xie et al. (2003) ?nd that the tenure of outside directors is
positively related to the level of discretionary current accruals. They
interpret that longer tenure as directors may be less effective
monitors and perhaps have been co-opted by management. In this
study we include chairman's tenure. A long-tenured chairman is
more likely to have well control over the board. We therefore
expect a positive relation between chairman's tenure and ?rm's
level of earnings management. Directors' shareholding is the
aggregate shareholding of all directors. On average the aggregate
shareholding of directors is 23.40%. Prior studies indicate that
managerial shareholding acts as a disciplining mechanism (Berle &
Means, 1932; Jensen &Meckling, 1976) and therefore is supposed to
be negatively correlated with the absolute value of abnormal ac-
cruals (War?eld, Wild, & Wild, 1995). However, it could be the case
that managers aiming to maximizing their bonuses (Healy, 1985) or
compensation (Nagar, Nanda, & Wysocki, 2003) would engage in
earnings management. In this study, we focus on the role of di-
rectors and therefore use directors' shareholding instead. The
aforementioned arguments imply that the relation between di-
rectors' shareholding and ?rm's earnings management remains
mixed. In contrast, institutional shareholding serves as a moni-
toring proxy in the sense that institutional investors are able and
are motivated to restrict high levels of earnings management when
they have high stakes in the ?rm. We therefore expect to ?nd a
negative relation between institutional shareholding and levels of
earnings management. The average institutional shareholding is
17.32%. We also include the average educational level of indepen-
dent directors and that of supervisors into analysis.
4. Empirical results
In Table 3 we classify ?rms into large and small discretionary
accruals with the classi?cation scheme based on the sample me-
dian. All variables of interest are conducted the tests in differences
between ?rms with large and small discretionary accruals. The
results indicate that ?rms with large discretionary accruals are
more likely to have an external ?nancing plan (17.07%) than ?rms
with small discretionary accruals (10.49%). Moreover, ?rms with
large discretionary accruals are associated with lower board
external connectedness (45.23%) than ?rms with small discre-
tionary accruals (47.65%). The difference is signi?cant at 1% level.
The result indicates that board external connectedness facilitates
?rm's reduction in earnings management. For the subgroups of
member's status the results indicate that outside directors' and
supervisors' external connectedness is related to ?rm's reduction in
earnings management.
Because the learning effect would be more conspicuous for the
same-industry connectedness than for the different-industry one,
we expect that the relation between board external connectedness
and earnings management would be more signi?cant when the
external connectedness is in the same industry than in different
industries. The results indicate that ?rms with large discretionary
accruals are associated with lower overall external connectedness,
outside directors' external connectedness, inside directors' external
connectedness, and supervisors' external connectedness than ?rms
with small discretionary accruals. This veri?es our postulation that
the same-industry effect accentuates the negative relation between
board external connectedness and earnings management.
For the control variables we ?nd that ?rms with larger discre-
tionary accruals are associated with smaller asset size, a lower debt
ratio, and lower educational levels for independent directors and
for supervisors than ?rms with small discretionary accruals. In
contrast, ?rms with larger discretionary accruals are associated
with higher board independence, higher directors' shareholding,
and higher ROA than ?rms with lower discretionary accruals. The
results seem inclusive. Nevertheless, there are two things to be
sure. First, ?rm's earnings management is probably positively
correlated with ?nancial performance measure as manifested in
ROA. It is unrelated to market performance measure as manifested
in Tobin's Q. Second, the level of earnings management is positively
dictated by directors' shareholding. If directors' aggregate share-
holding is a quali?ed proxy of insiders' incentive, we postulate that
?rm's levels of earnings management are positively correlated with
insiders' incentives.
A ?nal puzzle fromthe ?ndings is why board independence fails
to mitigate ?rm's engagement of earnings management. Not only
the level of board independence but also these independent di-
rectors' external connectedness fails to reduce ?rm's levels of
earnings management. One possible explanation is that most listed
?rms involve resume independent directors. For example, Hwang
and Kim (2009) indicate that social ties do matter and that,
consequently, a considerable percentage of the conventionally in-
dependent boards are substantively not socially independent. The
existence of these independent directors fails to provide moni-
toring function in reducing ?rm's level of earnings management.
Another possibility is that independent directors would more likely
to provide monitoring function when ?rms are in a critical event or
when there is an agency problem (Brickley & James, 1987; Byrd &
Hickman, 1992; Lee, Rosenstein, Rangan, & Davidson, 1992;
Weisbach, 1988). The two possibilities are not mutually exclusive.
That is, independent directors might passively monitor the ?rms in
normal time. However, they become active monitors when ?rms
are in a critical moment, for example, when ?rms have plans of
external ?nancing. An external ?nancing plan involves in new
stakeholders. Independent directors are responsible for protecting
the interests of these stakeholders especially when they are subject
to the harm of information asymmetry. We will revisit this issue
latter.
In Table 4 we report the coef?cients of partial correlation among
variables. The result in Panel A indicates that the level of discre-
tionary accruals is positively correlated with the condition of
external ?nancing with the partial coef?cient of correlation of 0.12,
the highest among all variables relating to discretionary accruals.
This implies that the plan of external ?nancing is very critical to
dictate ?rm's use of discretionary accruals and supports the prior
?nding that earnings management is severe in the issuance of
seasoned equity (Dechow et al., 1996; Rangan, 1998; Teoh et al.,
1998). Moreover, the result indicates that the level of discre-
tionary accruals is negatively correlated with ?rm size, debt ratio,
and the average educational level of independent directors and
supervisors, and is positively correlated with ROA. The negative
correlation between earnings management and ?rm size and the
debt ratio, respectively, could be understood as follows: large ?rms
and high-levered ?rms are subject to public scrutiny and therefore
are left with limited space of earnings management. The negative
correlation with independent directors' and supervisors' education
level indicates that education helps monitoring effectiveness of
board members.
In Panel B we report the partial coef?cients of correlation be-
tween board external connectedness and discretionary accruals.
The result indicates that board external connectedness is negatively
correlated with discretionary accruals. The relation is signi?cant for
overall BEC, outside BEC, and supervisor BEC. When the
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 269
connectedness refers to the same industry connectedness, the
relation is signi?cant for outside and supervisor BEC.
In Table 5 we segregate the sample into two subsamples: ?rms
with an external ?nancing plan versus ?rms without. The segre-
gation is based on two premises. First, the factor of external
?nancing is probably so strong as to dominate other variables in
explaining ?rm's earnings management. Second, the relation be-
tween board external connectedness and the level of discretionary
accruals might be moderately affected by the factor of an external
?nancing plan. The results when ?rms without an external
?nancing plan are summarized in Panel A of Table 5. The result
indicates that overall board external connectedness and the
connectedness with respect to outside directors and with respect to
supervisors is negatively correlated with the level of discretionary
accruals, respectively. This implies that outside directors and su-
pervisors with external connectedness are able to provide moni-
toring function in lowing ?rm's earnings management. This is
consistent with the ?ndings of Peasnell et al. (2005) that the
Table 3
Test in differences.
Large DA Small DA Difference test
Mean Median Mean Median t z
Dummy (external ?nancing) 0.1707 0.0000 0.1049 0.0000 7.324*** 7.292***
BECall (%) 0.4523 0.5000 0.4765 0.5000 À2.997*** À2.610***
BECoutside (%) 0.3492 0.3333 0.3776 0.3333 À3.458*** À3.324***
BECinside (%) 0.7028 0.6667 0.7130 1.0000 À1.178 À1.209
BECindependent (%) 0.4907 0.5000 0.4814 0.5000 0.690 0.765
BECsupervisor (%) 0.3464 0.3333 0.3781 0.3333 À3.602*** À3.240***
BECall-industry 0.3203 0.3333 0.3437 0.3333 À2.892*** À2.722***
BECoutside-industry 0.3149 0.2361 0.3468 0.3333 À3.125*** À3.283***
BECinside-industry 0.2283 0.0000 0.2550 0.0000 À2.308** À2.593**
BECindependent-industry 0.2623 0.000 0.2571 0.0000 0.468 0.603
BECsupervisor-industry 0.3546 0.000 0.3767 0.0000 À1.833* À1.811*
Size 15.1978 15.0251 15.4272 15.2495 À6.071*** À5.688***
Debt (%) 0.4079 0.4070 0.4360 0.4370 À6.146*** À6.101***
Board size 10.7300 10.0000 10.7700 10.0000 À0.447 À0.400
Board independence (%) 0.1693 0.2000 0.1590 0.0000 2.307** 2.224**
Tobin's Q 1.2632 1.0803 1.2802 1.0816 À0.843 À1.016
Chairman's tenure (years) 14.6070 12.2500 14.8074 12.5800 À0.534 À1.012
Directors' shareholding 23.6428 20.0450 23.1639 19.2500 1.259** 2.224**
Institutional shareholding 17.2020 11.5100 17.4397 12.1800 À0.501 À0.702
ROA 7.8750 7.5700 6.1954 6.7300 6.910*** 6.308***
Independent director's education 1.9724 2.000 1.9856 2.000 À2.127** À2.110**
Supervisor's education 1.9158 2.000 1.9353 2.000 À2.073** À2.041**
Note: ***, **, * denote the signi?cance level of 1%, 5%, and 10%, respectively.
Table 4
Coef?cients of correlation.
Panel A: Discretionary accruals and external ?nancing
DA DEF Size Debt BSize BInd Inst BShare Tobin's Q ROA Tenure Ind.-Edu. Sup.-Edu,
DA 1
DEF 0.12
**
1
Size À0.09
**
0.08
**
1
Debt À0.11
**
0.13
**
0.28
**
1
BSize À0.00 0.05
**
0.13
**
À0.01 1
Bind 0.02 0.08
**
À0.24
**
À0.10
**
0.03
*
1
Inst À0.01 À0.03
*
0.12
**
0.05
**
0.08
**
À0.08
**
1
Dshare 0.02 À0.07
**
À0.20
**
À0.01 0.07
**
0.01 0.47
**
1
Tobin's Q À0.01 0.06
**
0.02 À0.13
**
0.02 0.13
**
À0.02 À0.04
**
1
ROA 0.09
**
0.06
**
0.15
**
À0.23
**
0.04
**
0.14
**
0.02 0.06
**
0.35
**
1
Tenure À0.01 À0.00 0.14
**
0.04
*
À0.05
**
À0.20
**
À0.06
**
À0.04
**
À0.01 0.01 1
Ind.-Edu. À0.05
*
0.01 0.03 À0.02 0.05
*
À0.01 À0.00 0.00 0.02 0.01 0.08
**
1
Sup.-Edu. À0.04
*
0.02 0.00 À0.04 0.01 0.03 À0.04
*
À0.02 0.03 0.03 À0.01 0.04 1
Panel B: Discretionary accruals and board external connectedness
DA BEC
all
BEC
outside
BEC
inside
BEC
independent
BEC
supervisor
BEC
all-ind.
BEC
outside-ind.
BEC
inside-ind.
BEC
indep-ind.
BEC
sup-ind.
DA 1
BEC
all
À0.05
**
1
BEC
outside
À0.05
**
0.77
**
1
BEC
inside
À0.02 0.42
**
0.15
**
1
BEC
independent
À0.01 0.19
**
0.24
**
0.06
**
1
BEC
supervisor
À0.06
**
0.36
**
0.39
**
0.12
**
0.23
**
1
BEC
all-ind.
À0.02 0.35
**
0.396
**
0.08
**
0.20
**
0.26
**
1
BEC
outside-ind.
À0.05
**
0.30
**
0.39
**
0.08
**
À0.21
**
0.14
**
0.80
**
1
BEC
inside-ind.
À0.01 0.23
**
0.25
**
0.08
**
0.18
**
0.24
**
0.66
**
0.22
**
1
BEC
indep-ind.
0.00 0.13
**
0.11
**
0.02 0.51
**
0.15
**
0.31
**
À0.11
**
0.08
**
1
BEC
sup-ind.
À0.04
**
0.22
**
0.24
**
0.02 0.17
**
0.54
**
0.42
**
0.32
**
0.27
**
0.26
**
1
Note: ***, **, * denote the signi?cance level of 1%, 5%, and 10%, respectively.
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 270
Table 5
Regression of earnings management conditioned on external ?nancing.
Panel A: No external ?nancing Dependent variable: Earnings management
Variable (1) (2) (3) (4) (5)
Intercept 0.282***
(2.622)
0.284***
(2.656)
0.319**
(2.556)
0.305***
(2.828)
0.321***
(3.680)
Firm size À0.014***
(À3.383)
À0.015***
(À3.690)
À0.014***
(À3.133)
À0.017***
(À4.272)
À0.014***
(À3.893)
Debt À0.063*
(À1.867)
À0.068**
(À1.973)
À0.068*
(À1.839)
À0.063*
(À1.845)
À0.071**
(À2.399)
Board size À0.001
(À0.686)
À0.001
(À0.449)
À0.001
(À0.273)
À0.001
(À0.676)
À0.000
(À0.250)
Board independence À0.008
(À0.130)
À0.011
(À0.179)
0.010
(0.147)
À0.004
(À0.061)
À0.045
(À0.838)
Institutional holding À0.000
(À0.217)
À0.000
(À0.105)
À0.000
(À0.271)
À0.000
(À0.681)
À0.000
(À0.288)
Directors' holding 0.000
(0.507)
0.000
(0.609)
0.000
(0.308)
0.000
(0.634)
0.000
(0.044)
Tobin's Q À0.009*
(À1.737)
À0.009*
(À1.814)
À0.008
(À1.481)
À0.008*
(À1.685)
À0.006
(À1.302)
ROA 0.001**
(2.030)
0.001*
(1.874)
0.001**
(2.075)
0.001**
(2.136)
0.001***
(2.796)
Chairman's tenure À0.000
(À0.515)
À0.000
(À0.558)
À0.000
(À0.351)
À0.000
(À0.349)
À0.001
(À1.509)
Independent director's education À0.004
(À0.107)
À0.004
(À0.128)
À0.023
(À0.531)
À0.006
(À0.164)
À0.013
(À0.407)
Supervisor's education À0.035*
(À1.687)
À0.033*
(À1.600)
À0.036
(À1.550)
À0.037*
(À1.750)
BEC
all
À0.039**
(À2.575)
BEC
outside
À0.049***
(À3.218)
BEC
inside
À0.024
(À1.253)
BEC
independent
0.009
(0.486)
BEC
supervisor
À0.030**
(À2.226)
Adj. R
2
0.070 0.065 0.05 0.061 0.061
Panel B: External ?nancing (1) (2) (3) (4) (5)
Intercept 0.567
(1.571)
0.552
(1.478)
0.724**
(2.148)
0.462
(1.277)
0.535
(2.194)
Firm size À0.007
(À0.649)
À0.003***
(À0.210)
À0.013
(À1.262)
À0.008
(À0.701)
À0.008
(À0.832)
Debt À0.271**
(À2.235)
À0.283**
(À2.291)
À0.243**
(À2.157)
À0.250**
(À2.072)
À0.193**
(À1.999)
Board size À0.009*
(À1.739)
À0.001*
(À1.890)
À0.004
(À0.805)
À0.008
(À1.489)
À0.006
(À1.330)
Board independence À0.160
(À0.889)
À0.179
(À0.953)
À0.161
(À0.871)
À0.068
(À0.363)
À0.121
(À0.818)
Institutional holding 0.001
(0.394)
0.001
(0.375)
0.001
(1.041)
À0.000
(À0.236)
0.000
(0.286)
Directors' holding 0.000
(0.110)
0.000
(0.117)
À0.001
(À0.445)
0.001
(0.522)
0.000
(0.194)
Tobin's Q 0.014
(0.690)
0.014
(0.695)
À0.008
(À0.415)
0.016
(0.832)
0.013
(0.780)
ROA 0.000
(0.024)
À0.000
(À0.045)
0.001
(0.668)
0.000
(0.040)
0.002
(1.367)
Chairman's tenure À0.000
(À0.262)
À0.000
(À0.139)
0.001
(0.743)
À0.001
(À0.417)
À0.000
(À0.337)
Independent director's education À0.148
(À1.434)
À0.156
(À1.502)
À0.163*
(À1.784)
À0.154
(À1.536)
À0.104
(À1.446)
Supervisor's education 0.043
(0.616)
0.039
(0.546)
0.012
(0.173)
0.054
(0.769)
BEC
all
0.007
(0.148)
BEC
outside
0.018
(0.373)
BEC
inside
0.038
(0.770)
BEC
independent
0.095
(1.530)
BEC
supervisor
À0.014
(À0.378)
Adj. R
2
0.049 0.039 0.126 0.066 0.067
Note: The regression coef?cient and the t-statistics in parentheses are reported in the upper and lower case, respectively. ***, **, and * denote the signi?cance level of 1%, 5%,
and 10%, respectively.
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 271
Table 6
Regression of earnings management conditioned on the same-industry connection.
Panel A: No external ?nancing Dependent variable: Earnings management
Variable (1) (2) (3) (4) (5)
Intercept 0.279***
(2.624)
0.308***
(2.903)
0.251**
(1.935)
0.313***
(2.915)
0.319***
(3.640)
Firm size À0.013***
(À3.409)
À0.016***
(À4.066)
À0.011
(À2.267)
À0.017***
(À4.263)
À0.015***
(À4.320)
Debt À0.073**
(À2.153)
À0.065**
(À1.924)
À0.070*
(À1.890)
À0.064*
(À1.870)
À0.087***
(À2.912)
Board size À0.001
(À0.580)
À0.000
(À0.166)
À0.001
(À0.438)
À0.001
(À0.755)
0.000
(0.188)
Board independence À0.018
(À0.291)
À0.044
(À0.704)
À0.002
(À0.035)
À0.009
(À0.138)
À0.030
(À0.544)
Institutional holding À0.000
(À0.567)
À0.000
(À0.454)
À0.000
(À0.279)
À0.000
(À0.676)
À0.000
(À0.417)
Directors' holding 0.000
(0.578)
0.000
(0.592)
0.000
(0.504)
0.000
(0.584)
À0.000
(À0.286)
Tobin's Q À0.009*
(À1.821)
À0.010*
(À1.945)
À0.008
(À1.416)
À0.009*
(À1.710)
À0.009*
(À1.840)
ROA 0.001*
(1.821)
0.001**
(2.104)
0.001*
(1.774)
0.001**
(2.169)
0.001***
(2.714)
Chairman's tenure À0.000
(À0.562)
À0.000
(À0.633)
À0.000
(À0.291)
À0.000
(À0.370)
À0.001
(À1.518)
Independent director's education À0.003
(À0.088)
À0.008
(À0.224)
À0.018
(À0.383)
À0.004
(À0.111)
À0.007
(À0.214)
Supervisor's education À0.035*
(À1.695)
À0.036*
(À1.730)
À0.036
(À1.568)
À0.037*
(À1.762)
BEC
all-industry
À0.099***
(À3.614)
BEC
outside-industry
À0.078***
(À3.560)
BEC
inside-industry
À0.028**
(À2.205)
BEC
independent-industry
À0.004
(À0.321)
BEC
supervisor-industry
À0.021**
(À2.073)
Adj. R
2
0.079 0.079 0.055 0.061 0.057
Panel B: External ?nancing (1) (2) (3) (4) (5)
Intercept 0.563
(1.591)
0.573
(1.605)
0.779**
(2.344)
0.535
(1.508)
0.614***
(2.666)
Firm size À0.004
(À0.349)
À0.007
(À0.674)
À0.013
(À1.255)
À0.006
(À0.541)
À0.014*
(À1.720)
Debt À0.279**
(À2.341)
À0.276**
(À2.296)
À0.245**
(À2.145)
À0.261**
(À2.179)
À0.150
(À1.634)
Board size À0.008
(À1.560)
À0.008
(À1.653)
À0.004
(À0.826)
À0.010*
(À1.955)
À0.005
(À1.134)
Board independence À0.212
(À1.189)
À0.236
(À1.255)
À0.172
(À0.926)
À0.130
(À0.730)
À0.166
(À1.212)
Institutional holding 0.001
(0.494)
0.001
(0.528)
0.002
(1.177)
0.000
(0.301)
0.000
(0.405)
Directors' holding 0.000
(0.169)
0.000
(0.106)
À0.001
(À0.450)
0.000
(0.094)
0.000
(0.034)
Tobin's Q 0.020
(1.104)
0.018
(0.902)
À0.010
(À0.484)
0.014
(0.690)
0.013
(0.781)
ROA À0.001
(À1.014)
À0.000
(À0.106)
0.001
(0.587)
0.000
(0.079)
0.002
(1.393)
Chairman's tenure À0.000
(À0.102)
À0.001
(À0.396)
0.001
(0.785)
À0.000
(À0.258)
À0.000
(À0.317)
Independent director's education À0.158
(À1.587)
À0.131
(À1.297)
À0.174*
(À1.873)
À0.133
(À1.331)
À0.108
(À1.528)
Supervisor's education 0.037
(0.542)
0.035
(0.499)
0.020
(0.293)
0.034
(0.490)
BEC
all-industry
À0.170*
(À1.952)
BEC
outside-industry
À0.081
(À1.205)
BEC
inside-industry
À0.004
(À0.109)
BEC
independent-industry
À0.074*
(À1.763)
BEC
supervisor-industry
À0.021
(À0.794)
Adj. R
2
0.076 0.059 0.118 0.071 0.079
Note: ***, **, and * denote the signi?cance level of 1%, 5%, and 10%, respectively.
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 272
outside board members are able to lower the income-increasing
abnormal accruals when pre-managed earnings are low. The
external exposure enhances their monitoring effectiveness and is
translated into a material reduction of the level of discretionary
accruals. However, board external connectedness loses its effec-
tiveness in reducing level of earnings management when ?rms
have external ?nancing plans (Panel B). Moreover, ?rmsize, Tobin's
Q, and supervisors' educational level, and ROA have explanatory
power on ?rm's earnings management when ?rms have no
external ?nancial plans. However, their explanatory power is gone
when ?rms have an external ?nancing plan. This corroborates with
our postulation that the effect of an external ?nancing plan dom-
inates other factors in explaining ?rm's earnings management. The
effectiveness of board external connectedness in reducing ?rm's
earnings management is also signi?cantly affected by the factor of
an external ?nancing plan.
In Table 6 we regress discretionary accruals on the same-
industry board external connectedness. Again, we segregate the
sample into two subsamples: ?rms without and ?rms with an
external ?nancing plan. The result in Table 6 illustrates an inter-
esting pattern: board external connectedness gauged with respect
to overall directors, outside directors, inside directors, and super-
visors is negatively correlated with the level of discretionary ac-
cruals. If anything, the regression coef?cients are larger in extent
and more signi?cant than those in Table 5. The ?nding implies that
the same-industry connectedness enhances the connected di-
rectors' learning more than different-industry connectedness does.
The result supports the learning effect postulated in Hypothesis 1a.
We note that the external connectedness of independent di-
rectors is insigni?cant in affecting the level of discretionary ac-
cruals when ?rms with an external ?nancing plan. However, the
same-industry external connectedness of independent directors is
negatively correlated with the level of discretionary accruals when
?rms have an external ?nancing plan. The ?nding implies that in-
dependent directors are more likely to provide surveillance func-
tion that in turn reduces ?rm's earnings management when ?rms
have an external ?nancing plan than when ?rms have no external
?nancing plan. The critical issue is that external ?nancing involves
external stakeholders who are more likely to subject to information
asymmetry. Independent directors are supposed to protect the
shareholders' interest and especially for those who are informa-
tionally disadvantaged.
5. Concluding remarks
We explore how board external connectedness is related to
?rm's earnings management which is gauged by the level of
discretionary accruals. Using the dataset of Taiwan listed ?rms in
the sampling period between 2007 and 2011 we ?nd a signi?cantly
negative relation between the two. The ?nding is in sharp contrast
with Chiu et al. (2013) who exemplify board connectedness as virus
contagion. The contagion hypothesis would predict a positive cor-
relation between board external connectedness and earnings
management. In contrast, we ?nd the learning effect embedded in
board external connectedness in the sense that the relation be-
tween board external connectedness and discretionary accruals is
positive and the positive relation is more saliently found in cases of
the same-industry connectedness. This is consistent with the
notion that earnings management could be industry speci?c (Hall &
Stammerjohan, 1997; Key, 1997; Mensah et al., 1994) and therefore
the learning effect is more conspicuous for the same-industry
connectedness. Moreover, we ?nd that the factor of external
?nancing plan is so strong as to dominate other factors in
explaining ?rm's level of earnings management. This is consistent
with previous ?nding that earnings management is more saliently
found when ?rms have an external ?nancing plan (e.g., Dechow
et al., 1996; Rangan, 1998; Teoh et al., 1998). More importantly,
we ?nd that independent directors only provide surveillance
functionwhen ?rms have an external ?nancing planwhich involves
additional external shareholders or debt claimers.
One limitation of this paper is that we are unable to trace the
speci?c time that directors assume outside directorships or to ?nd
out an external shock that could isolate the link between board
external connectedness and ?rm's earnings management. This is
important to address the possible concern that board external
connectedness and earnings management could be endogenously
related. For example, directors when seeking external directorships
would prefer decent ?rms to mitigate the possible litigation cost
when ?rms are in trouble. Moreover, when ?nding outside di-
rectors, ?rms would prefer ones who are socially reputable. These
socially reputable persons happen to assume multiple directorships
in different ?rms. The endogeneity issue and possible spurious
relation are partially ameliorated when ?rm's characteristics have
been controlled in the empirical models.
According to Xie et al. (2003), board members' ?nancial so-
phistication is an important factor in constraining the propensity of
managers to engage in earnings management. Our ?nding echoes
that of Xie et al. (2003) in the sense that externally connected di-
rectors acquire ?nancial sophistication from their external
connectedness and have their learning experiences be transformed
into monitoring effectiveness and therefore a reduction in the level
of earnings management. Nevertheless, in this study we are unable
to meticulously trace the experiences that externally connected
board members learn fromand bene?t. Future studies could further
address these issues.
Con?icts of interest
All contributing authors declare no con?icts of interest.
References
Barnea, A., & Guedj, I. (2009). Director networks (Unpublished paper). University of
Texas at Austin.
Beasley, M. S. (1996). An empirical analysis of the relation between the board of
director composition and ?nancial statement fraud. Accounting Review, 71(4),
443e465.
Beasley, M. S., & Salterio, S. E. (2001). The relationship between board character-
istics and voluntary improvements in audit committee composition and
experience. Contemporary Accounting Research, 18(4), 539e570.
Beneish, M. D. (1997). Detecting GAAP Violation: implications for assessing earnings
management among ?rms with extreme ?nancial performance. Journal of Ac-
counting and Public Policy, 16(3), 271e309.
Beneish, M. D. (1999). The detection of earnings manipulation. Financial Analysts
Journal, 55(5), 24e36.
Berle, A. A., & Means, G. C. (1932). The modern corporation and private property. New
York, NY: MacMillan.
Bizjak, J., Lemmon, M., & Whitby, R. (2009). Option backdating and board interlocks.
Review of Financial Studies, 22(11), 4821e4847.
Botosan, C. A. (1997). Disclosure level and the cost of equity capital. Accounting
Review, 72(3), 323e349.
Brickley, J. A., & James, C. M. (1987). The takeover market, corporate board
composition, and ownership structure: the case of banking. Journal of Law and
Economics, 30(1), 161e181.
Brickley, J. A., Linck, J. S., & Coles, J. L. (1999). What happens to CEOs after they
retire? New evidence on career concerns, horizon problems, and CEO in-
centives. Journal of Financial Economics, 52(3), 341e377.
Brown, W. O., & Maloney, M. T. (1999). Exit, voice, and the role of corporate directors:
Evidence from acquisition performance. Social Science Research Network work-
ing paper series 160308. Retrieved from:http://papers.ssrn.com/sol3/papers.
cfm?abstract_id=160308.
Burgstahler, D., & Eames, M. (2006). Management of earnings and analyst forecasts
to achieve zero and small positive earnings surprises. Journal of Business Finance
and Accounting, 33(5), 633e652.
Butler, A. W. (2008). Distance still matters: evidence from municipal bond under-
writing. Review of Financial Studies, 21(2), 763e784.
Byrd, J. W., & Hickman, K. A. (1992). Do outside directors monitor managers? Evi-
dence from tender offer bids. Journal of Financial Economics, 32(2), 195e221.
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 273
Cai, Y., & Sevilir, M. (2012). Board connections and M&A transactions. Journal of
Financial Economics, 103(2), 327e349.
Chiu, P. C., Teoh, S. H., & Tian, F. (2013). Board interlocks and earnings management
contagion. Accounting Review, 88(3), 915e944.
Chuluun, T., Prevost, A. K., & Puthenpurackal, J. (2014). Board networks and the cost
of corporate debt. Financial Management, 43(3), 533e568.
Cohen, J. L., Frazzini, A., & Malloy, C. (2008). The small world of investing: board
connections and mutual fund returns. Journal of Political Economy, 116(5),
951e979.
Coles, J. L., & Hoi, C. K. (2003). New evidence on the market for directors: board
membership and Pennsylvania Senate Bill 1310. Journal of Finance, 58(7),
197e230.
Connelly, B. L., & Van Slyke, E. J. (2012). The power and peril of board interlocks.
Business Horizons, 55(5), 403e408.
Core, J. E., Holthausen, R. W., & Larcker, D. F. (1999). Corporate governance, chief
executive of?cer compensation, and ?rm performance. Journal of Financial
Economics, 51(3), 371e406.
Davis, G. F. (1991). Agents without principles? The spread of the poison pill through
the intercorporate network. Administrative Science Quarterly, 36(4), 583e613.
Dechow, P. M., & Dichev, I. D. (2002). The quality of accruals and earnings: the role
of accrual estimation errors. Accounting Review, 77(Supplement), 35e59.
Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1995). Detecting earnings manage-
ment. Accounting Review, 70(2), 193e225.
Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1996). Causes and consequences of
earnings manipulation: an analysis of ?rms subject to enforcement actions by
the SEC. Contemporary Accounting Research, 13(1), 1e36.
Easterwood, C. M. (1998). Takeovers and incentives for earnings management: an
empirical analysis. Journal of Applied Business Research, 14(1), 29e48.
Erickson, M., & Wang, S. W. (1999). Earnings management by acquiring ?rms in
stock for stock mergers. Journal of Accounting and Economics, 27(2), 149e176.
Fama, E. F. (1980). Agency problems and the theory of the ?rm. Journal of Political
Economy, 88(2), 288e307.
Fama, E. F., & Jensen, M. C. (1983). Separation of ownership and control. Journal of
Law and Economics, 26(2), 301e325.
Farrell, K. A., & Whidbee, D. A. (2000). The consequences of forced CEO succession
for outside directors. Journal of Business, 73(4), 597e627.
Fich, E. M., & Shivdasani, A. (2006). Are busy boards effective monitors? Journal of
Finance, 61(2), 689e724.
Fracassi, C., & Tate, G. (2012). External networking and internal ?rm governance.
Journal of Finance, 67(1), 153e194.
Gilson, S. C. (1990). Bankruptcy, boards, banks, and blockholders: evidence on
changes in corporate ownership and control when ?rms default. Journal of
Financial Economics, 27(2), 355e387.
Hall, S. C., & Stammerjohan, W. W. (1997). Damage awards and earnings manage-
ment in the oil industry. Accounting Review, 72(1), 47e65.
Harford, J. (2003). Takeover bids and target directors' incentives: the impact on
directors' wealth and board seats. Journal of Financial Economics, 69(1), 51e83.
Haunschild, P. R. (1993). Interorganizational imitation: the impact of interlocks on
corporate acquisition activity. Administrative Science Quarterly, 38(4), 564e592.
Healy, P. M. (1985). The effect of bonus schemes on accounting decisions. Journal of
Accounting and Economics, 7(1), 85e107.
Hirshleifer, D., & Teoh, S. H. (2009). Thought and behavior contagion in capital
markets. In T. Hens, & K. R. Schenk-Hopp e (Eds.), Handbook of ?nancial markets:
Dnamics and evolution (pp. 1e56). Amsterdam, Netherlands: Elsevier.
Hribar, P., & Collins, D. W. (2002). Errors in estimating accruals: implications for
empirical research. Journal of Accounting Research, 40(1), 105e134.
Hwang, B. H., & Kim, S. (2008). Earnings management and social ties (Unpublished
paper). Purdue University, West Lafayette.
Hwang, B. H., & Kim, S. (2009). It pays to have friends. Journal of Financial Economics,
93(1), 138e158.
Ishii, J., & Xuan, Y. (2014). Acquirer-target social ties and merger outcomes. Journal
of Financial Economics, 112(3), 344e363.
Jensen, M. C., & Meckling, W. (1976). Theory of the ?rm: managerial behavior,
agency costs, and ownership structure. Journal of Financial Economics, 3(4),
305e360.
Jones, J. J. (1991). Earnings management during import relief investigations. Journal
of Accounting Research, 29(2), 193e228.
Jones, K. L., Krishnan, G. V., & Melendrez, K. D. (2008). Do models of discretionary
accruals detect actual cases of fraudulent and restated earnings? An empirical
analysis. Contemporary Accounting Research, 25(2), 499e531.
Kaplan, S. N., & Reishus, D. (1990). Outside directorships and corporate perfor-
mance. Journal of Financial Economics, 27(2), 389e410.
Kasznik, R. (1999). On the association between voluntary disclosure and earnings
management. Journal of Accounting Research, 37(1), 57e81.
Key, K. G. (1997). Political cost incentives for earnings management in the cable
television industry. Journal of Accounting and Economics, 23(3), 309e337.
Klein, A. (2002). Audit committee, board of director characteristics and earnings
management. Journal of Accounting and Economics, 33(3), 375e400.
Kothari, S. P., Leone, A. J., & Wasley, C. E. (2005). Performance matched discretionary
accrual measures. Journal of Accounting and Economics, 39(1), 163e197.
Larcker, D. F., & Richardson, S. A. (2004). Fees paid to audit ?rms, accrual choices,
and corporate governance. Journal of Accounting Research, 42(3), 625e656.
Lee, C. I., Rosenstein, S., Rangan, N., & Davidson, W. N. (1992). Board composition
and shareholder wealth. The case of management buyouts. Financial Manage-
ment, 21(1), 58e72.
Lim, S., & Matolcsy, Z. (1999). Earnings management of ?rms subject to product
price controls. Accounting and Finance, 39(2), 131e150.
Mace, M. L. (1971). Directors: Myth and reality. Boston, MA: Harvard Business School
Press.
Magnan, M., Nadeau, C., & Cormier, D. (1999). Earnings management during anti-
dumping investigations: analysis and implications. Canadian Journal of Admin-
istrative Sciences, 16(2), 149e162.
Makar, S. D., & Alam, P. (1998). Earnings management and antitrust investigations:
political costs over business cycles. Journal of Business Finance and Accounting,
25(5), 701e719.
McNichols, M. F. (2000). Research design issues in earnings management studies.
Journal of Accounting and Public Policy, 19(4), 313e345.
Mensah, Y. M., Considine, J. M., & Oakes, L. (1994). Statutory insolvency regulation
and earnings management in the prepaid health-care industry. Accounting Re-
view, 69(1), 70e95.
Nagar, V., Nanda, D., & Wysocki, P. (2003). Discretionary disclosure and stock-based
incentives. Journal of Accounting and Economics, 34(1), 283e309.
Peasnell, K. V., Pope, P. F., & Young, S. (2005). Board monitoring and earnings
management: do outside directors in?uence abnormal accruals? Journal of
Business Finance and Accounting, 32(7), 1311e1346.
Perry, S. E., & Williams, T. H. (1994). Earnings management preceding management
buyout offers. Journal of Accounting and Economics, 18(2), 157e179.
Rangan, S. (1998). Earnings management and the performance of seasoned equity
offerings. Journal of Financial Economics, 50(1), 101e122.
Rao, H., Davis, G. F., & Ward, A. (2000). Embeddedness, social identity and mobility:
why ?rms leave the NASDAQ and join the New York Stock Exchange. Admin-
istrative Science Quarterly, 45(2), 268e292.
Shivdasani, A. (1993). Board composition, ownership structure, and hostile take-
overs. Journal of Accounting and Economics, 16(1), 167e198.
Shivdasani, A., & Yermack, D. (1999). CEO involvement in the selection of new board
members: an empirical analysis. Journal of Finance, 54(5), 1829e1853.
Stuart, T. E., & Yim, S. (2010). Board interlocks and the propensity to be targeted in
private equity transactions. Journal of Financial Economics, 97(1), 174e189.
Teoh, S. H., Welch, I., & Wong, T. J. (1998). Earnings management and the long-run
market performance of initial public offerings. Journal of Finance, 53(6),
1935e1974.
Vafeas, N. (1999). Board meeting frequency and ?rm performance. Journal of
Financial Economics, 53(1), 113e142.
War?eld, T. D., Wild, J. J., & Wild, K. L. (1995). Managerial ownership, accounting
choices, and informativeness of earnings. Journal of Accounting and Economics,
20(1), 61e92.
Weisbach, M. (1988). Outside directors and CEO turnover. Journal of Financial Eco-
nomics, 20(1), 431e460.
Wu, Y. W. (1997). Management buyouts and earnings management. Journal of Ac-
counting, Auditing and Finance, 12(2), 373e389.
Xie, B., Davidson, W. N., & DaDalt, P. J. (2003). Earnings management and corporate
governance: the role of the board and the audit committee. Journal of Corporate
Finance, 9(3), 295e316.
P.-G. Shu et al. / Asia Paci?c Management Review 20 (2015) 265e274 274

doc_415719014.pdf