Topics in Entrepreneurship and Family Business Management

Description
Topics in Entrepreneurship and Family Business Management



612+%5+0
'064'24'0'745*+2
#0&
(#/+.;$75+0'55/#0#)'/'06











-DQ3KLOLSS$KUHQV
8QLYHUVLW\RI0DQQKHLP






612+%5+0
'064'24'0'745*+2#0&(#/+.;$75+0'55/#0#)'/'06

(+8''55#;5



,1$8*85$/',66(57$7,21


]XU(UODQJXQJGHVDNDGHPLVFKHQ*UDGHV

GRFWRUUHUXPSROLWLFDUXP

'RNWRUGHU:LUWVFKDIWVZLVVHQVFKDIWHQ
GHU8QLYHUVLWlW0DQQKHLP









HLQJHUHLFKWDQGHU
)DNXOWlWIU%HWULHEVZLUWVFKDIWVOHKUH
GHU8QLYHUVLWlW0DQQKHLP

YRQ

'LSORP.DXIPDQQ-DQ3KLOLSS$KUHQV

JHERUHQDPLQ6WXWWJDUW

/



'HNDQGHU)DNXOWlWIU%HWULHEVZLUWVFKDIWVOHKUH
'U-UJHQ06FKQHLGHU

*XWDFKWHU
 3URI'U0LFKDHO:R\ZRGH
 3URI'U+ROJHU'DVNH

7DJGHUPQGOLFKHQ3UIXQJ


//













)



+







%#$

%225

///



!"#$

!%

!%&
'
()
 66666666666666666666666666666666666666666666666666$$$$$$$$$$$$
"


  6666666666666666666666666666666666666666666666666666$
5

-666666666666666666666666666666666666666666666666666666$
#




*!
(

*!
/



*!
5




!

*

$$&
+
,
!
$&
'
-.)
.


7!666666666666666666666666666666666666666666666666$$
&

"$ 8
6666666666666666666666666666666666666666666666666$
"2

%$ )


-



66666666666666666666666666666$
"%

4$ .
(66666666666666666666666666666666666666666666$$
"9

9$ (
(6666666666666666666666666666666666666666666666
%2

3$ 
70666666666666666666666666666666666666666666666666$
%4

5$ :66666666666666666666666666666666666666666666666666
92

)6666666666666666666666666666666666666666666666666666$$
9"

7
;666666666666666666666666666666666666666666666666666666$
95


,
*

$$

01
02

,
,$
!1
$!"

*!"!
3&
'
.-(44
.


7!666666666666666666666666666666666666666666666666$$
9&

"$ 8
6666666666666666666666666666666666666666666666666$
32

%$ )

-666666666666666666666666666666666666666666666$
3%

4$ .
:





+66666666666666666666666$$
33

9$ (
(6666666666666666666666666666666666666666666666
5"

3$ 
70666666666666666666666666666666666666666666666666$
54

5$ :66666666666666666666666666666666666666666666666666
'#

)6666666666666666666666666666666666666666666666666666$$
'&

7
;666666666666666666666666666666666666666666666666666666$
&9

,
$

*

$$

01
02

01
*$!!$
,
6
*!

!
$$&
'
(4((/.

.


7!666666666666666666666666666666666666666666666666$$
"2"
"$ 8
6666666666666666666666666666666666666666666666666$
"2%

%$ )

-666666666666666666666666666666666666666666666$
"24

4$ .
(66666666666666666666666666666666666666666666$$
"23

9$ (
(6666666666666666666666666666666666666666666666
"2#

3$ 
70666666666666666666666666666666666666666666666666$
"2#

5$ :66666666666666666666666666666666666666666666666666
""&

)6666666666666666666666666666666666666666666666666666$$
""&

7
;666666666666666666666666666666666666666666666666666666$
"%"




/s






*!
.



*!
7



0!8!

!$!%&
6
*!&
,
!
$


*

$$

01
0&
'
(/7(7)
.


7!666666666666666666666666666666666666666666666666$$
"%3
"$ 8
6666666666666666666666666666666666666666666666666$
"%5

%$ )

-666666666666666666666666666666666666666666666$
"%'

4$ .
(66666666666666666666666666666666666666666666$$
"%&

9$ (
(6666666666666666666666666666666666666666666666
"44

3$ 
70666666666666666666666666666666666666666666666666$
"43

5$ :66666666666666666666666666666666666666666666666666
"95

)6666666666666666666666666666666666666666666666666666$$
"9#

7
;666666666666666666666666666666666666666666666666666666$
"34


!
,

$!

!!!&
'
(7-(-)
.


7!666666666666666666666666666666666666666666666666$$
"3&
"$ 8
6666666666666666666666666666666666666666666666666$
"52

%$ 7






8
86666$$6666666666666666$
"5%

4$ .


^?

????

????

????

?????

?????

?????

'???????/s?

'???????K>^?

F IGURE 2. C OMPARISON OF OLS (TABLE 6) AND 2SLS-IV (TABLE 7) R ESULTS

As can be seen in figure 2, the 2SLS-IV results suggest that the true effect of predecessor activity is slightly more positive as suggested by OLS results, thus we take this
indication into account when interpreting the OLS results.
Furthermore, we address reservations that a general indicator for prolonged predecessor activity is not sufficient to cover all of the relevant heterogeneity, especially with
respect to the specific role a predecessor can fulfill. However, as highlighted before,
within some of the role categories our observations run pretty low and furthermore, the
indication of multiple roles is possible, which blurs and smudges the potential performance differences of a differentiation of the predecessors’ roles. Nevertheless, we shall
still address this idea in the appendix table A3, but we refrain from putting too much
emphasis on any results we reach due to the limitations highlighted. In order to over-

84

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

come the issues regarding the number of observations and the indication of multiple
roles in table A3, we weighted the predecessor’s roles and by summing up the weighted
roles for each predecessor we create a degree of influence score for each predecessor.
By pursuing this approach, we dramatically increase the number of observations for our
regressions and approximately capture the (formal) intensity of the activity of the predecessor instead of just his activity. The degree of influence score is then employed within
Huber-White robust OLS models similar to columns 1 to 4 of table 6. The regression
results are presented in table 8.
Column 1 of table 8 shows that no significant effect of the degree of influence of a
prolonged predecessor activity is visible without controlling for its highly heterogeneous
and context-dependent nature, which is a very similar result to the general predecessor
activity in table 6. But as soon as we interact the degree of influence with other variables, especially the successor’s human capital score, a positive and significant effect
per degree of influence (coefficient 0.58, significant at the five percent level) becomes
visible. As column 2 of table 8 highlights, this positive effect is contrasted by the negative and significant interaction between human capital score and the degree of influence
score (coefficient -0.22, significant at the one percent level). This interaction term can
quickly gain in magnitude, for example for a degree of influence score of two and a successors human capital score of two the effect would already be a -0.88 percentage points
reduced difference in industry- and performance-adjusted profit margin. Furthermore,
it is interesting to observe that high human capital family successors seem to be capable of shielding parts of the negative effects of the negative interaction between human
capital score and degree of influence, as the triple interaction term family*human capital
score*degree of influence is positive and significant (coefficient is 0.35 at the five percent
level, column 3 of table 8). Interestingly, this “shielding” effect even seems to increase
slightly with the degree of influence of the predecessor, as the positive and significant coefficient (0.07, significant at the one percent level) of family*human capital score*degree
of influence2 in column 5 of table 8 shows. Furthermore, column 4 of table 8 highlights
another feature of the predecessor’s degree of influence in family successions, when we
control for nepotism, the family*degree of influence interaction turns negative and significant (-1.14, significant at the 10% level). This highlights that in family successions the
degree of influence of the predecessor has more positive effects in nepotistic successions
(coefficient of nepotism strong*degree of influence is 1.23 and significant at the five percent level) as compared to family successions without nepotistic traits. This finding is
similar to the findings of column 4 of the table and highlights that once the successor has
reached a sufficient level of human capital to steer the fortunes of the enterprise, the predecessor is best advised to reduce his degree of influence in order to maximize enterprise
performance. This is also mirrored in a statement from another of the CEO successors
(11) who took part in the in-depth interview: “Then you ask yourself: who is calling
the shots here, it is us or him [the predecessor]? This inevitably happens, because you
say A and the senior says B. You have to find a consensus and sometimes you argue out
the differences on the backs of the employees.” We visualize this aspect by depicting the
results of regression two of table 8 in a three-dimensional diagram including differential

CHAPTER 2: CULTIVATING ROSES?

85

TABLE 8—OLS-R EGRESSION - D EGREE OF P REDECESSOR I NFLUENCE AND P ERFORMANCE
Dependent variable:
Variables
Degree of influence
(predecessor)
Degree of influence2
Company age
Company age*Degree of influence
Human capital score (HCS)
Human capital*Degree of influence
Family
External
Family*Degree of influence
External*Degree of influence
Family*HCS*Degree of influence
External*HCS*Degree of influence
Family*HCS*Degree of influence2
External*HCS*Degree of influence2
Nepotism strong
Nepotism weak
Nepotism strong*Degree of influence
Nepotism weak*Degree of influence

(1)
-0.03
(0.115)

1 Industry- & performance-adjusted PM
(2)
(3)
(4)
(5)
0.58**
0.95**
0.97**
0.86*
(0.256)
(0.462)
(0.465)
(0.459)
-0.01
(0.072)
0.00
0.01
0.00
0.00
(0.009)
(0.009)
(0.009)
(0.009)
-0.00
-0.00
-0.00
-0.00
(0.003)
(0.003)
(0.003)
(0.003)
0.81***
0.78***
0.81***
0.88***
(0.261)
(0.290)
(0.310)
(0.313)
-0.22***
-0.49*** -0.50***
-0.41***
(0.083)
(0.155)
(0.158)
(0.130)
-0.35
0.35
0.66
(0.856)
(1.341)
(1.345)
-0.13
-0.13
-0.11
(0.886)
(0.893)
(0.896)
-0.43
-1.14*
-0.96*
(0.434)
(0.602)
(0.516)
-0.30
-0.30
0.17
(0.613)
(0.617)
(0.484)
0.35**
0.46***
(0.150)
(0.166)
0.29
0.29
(0.189)
(0.190)
0.07***
(0.024)
0.02
(0.027)
-3.85**
-3.87**
(1.883)
(1.885)
-0.17
-0.22
(1.438)
(1.430)
1.23**
1.19**
(0.529)
(0.503)
0.53
0.49
(0.435)
(0.419)
?
?
?
?

?
Controls
Observations
391
383
383
383
383
R2
0.04
0.07
0.08
0.11
0.11
Note: The dependent variable is the difference in industry- and performance-adjusted profit margin (PM). The differences
are calculated via: industry- (and performance-) adjusted PM of the year 2009 less industry- (and performance-) adjusted
PM of the succession year. Industry-adjusted PM is PM less the median PM of the accordant year and industry (two-digit
ISIC) of a control group (from the Amadeus database); Performance-adjustments are designed by sorting the industryadjusted values of a control group into deciles and matching the individual industry-adjusted PM values of the sample with
the accordant control group decile in the year of the succession. The median industry-adjusted PM of the relevant control
group decile and year then serves as a control. Independent variables are: Degree of Influence (predecessor) is a score of
the predecessor’s influence consisting of a weighted sum of all predecessor roles; Degree of influence2 is the degree of the
predecessor’s influence squared; Company age is the age of the company in years; Family indicates if a successor is related
by marriage or blood to at least one of the three persons owning more than 50% of the enterprise; External indicates if
a successor had no previous ties to the enterprise; Human capital score (HCS) is derived from the sum of five proxy
elements: (1) age, (2) industry experience, (3) leadership experience, (4) merchant education, and (5) professionalism;
Nepotism strong indicates if the predecessor only considered family members as successors; and Nepotism weak indicates
if the succession is not categorized into the strong nepotism category and a family successor with lower human capital,
compared to the average human capital of non-family successors within the industry of the company, is installed as
successor. Controls are: Ln sales (size) is the natural logarithm of sales in the year of the succession; Industry-adjusted
PM (momentum) is the industry-adjusted ratio earnings before taxes divided by operating revenue in the succession year;
Industry- and performance-adjusted PM (momentum) is industry- and performance-adjusted PM in the succession year;
Ownership is an indicator equal to one if the successor owned a share of the enterprise in the succession year; Default
probability represents probability of default based on the Creditreform solvency-index score of the enterprise in the year
of the succession; Executives displays the number of executives in the succession year; Working relation is an indicator
equal to one if the successor perceives the working relations with the predecessor as good; Investment delay is an indicator
variable equal to one if an investment delay is perceived; Financing requirements is an indicator variable equal to one if
severe unexpected financing requirements were encountered during the succession; and Years is time elapsed since the
succession in years. Interactions between variables are marked via stars *. Significances are displayed within the table
via: * ten percent, ** five percent and *** one percent. The values in parentheses display Huber-White robust standard
errors.

86

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

????
????
????
?
?

?

?

?

?

?

?????

Ž/??

Ž/??

,?????????????Ž??

Ž/??

?????

Ž/??

?????

Ž/??

??????????????? ???????????WD

abnormal performance, the degree of the predecessor’s influence, and the human capital
score of the predecessor in figure 3.

F IGURE 3. S UCCESSOR ’ S H UMAN C APITAL , P REDECESSOR ’ S D EGREE OF I NFLUENCE (D O I) AND P ERFORMANCE

It is interesting to observe that figure 3 reveals the same general pattern as figure 1:
a positive abnormal performance impact of predecessor activity diminishes with the increasing human capital score of the successor and eventually turns negative. Figure 3
now adds another aspect to this finding: the negative relationship between human capital
score and predecessor activity is amplified and increased in magnitude with an increasing
degree of influence of the predecessor. If the successor is equipped with very low level
of human capital, a high degree of influence of the predecessor leads to a higher abnormal performance as compared to a low degree of predecessor influence. In contrast, if
the successor has high levels of human capital, a high degree of predecessor influence
leads to lower (more negative) abnormal performance as compared to a low degree of
predecessor influence.
Overall, the results of this picture support the theoretical arguments for a “phasingin” and “phasing-out” period (also known as the “succession dance”, Handler, 1990),
which allows the successor to gradually catch up with the incumbent’s knowledge while
the predecessor’s activity shields the enterprise from the reducing knowledge and ability
gaps of the successors. Once the successor has acquired a sufficient level of ability, our

CHAPTER 2: CULTIVATING ROSES?

87

results indicate that the predecessor is best advised to minimize his influence and activity
in order to maximize enterprise performance. Again, we quote one interesting passage of
the in-depth interviews with CEO successors (13): “And even if he leaves the enterprise,
I will certainly ask his advice on important decisions: The intuition of an entrepreneur,
who for thirty years has steered his company, is quite different from the normal intuition.
And that is why, thank goodness, he is still on board.” However, the predecessor might
want to avoid becoming an impediment to post-succession enterprise change, or to coin
it in the spirit of Campbell’s book The Hero with a Thousand Faces (1949): the hero of
today has to avoid becoming the tyrant of tomorrow.
VI.

Conclusion

Our results provide evidence that the likelihood of a prolonged predecessor activity
in CEO successions in family firms is significantly driven by family successions and
successions flavored with nepotistic traits. This is possibly because predecessors, in addition to their private benefits from seeing a family heir steering the fortunes of their
former enterprise, derive positive utility from prolonged activity within kin-steered enterprises. Furthermore, predecessors are significantly more likely to stay on board if they
have children and this relationship is most pronounced for daughters. We suspect that
this observation may be explainable by an intrinsic motivation to guide one’s children
into one’s footsteps and a heartfelt wish to ensure a smooth family succession, a possible
need to guide an interim management until the next family generation is old and experienced enough to become the new CEOs, and furthermore an increased need for income
stability when raising children.
Interestingly, in industries with high levels of tacit knowledge (we employ the intangibles-sales ratio as a proxy), we also observe a significantly higher likelihood of activity
by the preceding CEO. We suspect this observation may be explainable by the need for
knowledge transfer between the successor and the predecessor, and furthermore, the need
to keep idiosyncratic knowledge accessible. This observation is also in line with our evidence suggesting that the human capital score of the successor significantly negatively
influences the likelihood of predecessor activity. We interpret this in the following way:
when the ability of the successor is still limited, prolonged activity by the predecessor
may be helpful as a period of adjustment, guidance and mentoring, but when the successor has strong capabilities he will strive for independence and the predecessor’s activity
becomes unnecessary if not an impediment to post-succession change. In addition, we
find that successor ownership significantly decreases the likelihood of predecessor activity. This is in line with arguments which attach an insulating effect to (predecessor)
ownership or, due to the status which ownership entails, a barrier to retirement. Furthermore, in line with intuition we observe that the years passed since the succession and
corporate age significantly negatively affect the likelihood of predecessor activity. The
above results are robust to the inclusion of an array of firm- and industry-level controls.
What might be most relevant to practitioners is the performance impact of the activity
of the preceding CEO. Here we observe a highly ambivalent picture, which reveals a very
clear pattern on a second viewing. Our results indicate that prolonged predecessor activ-

88

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

ity entails a significant positive impact on differential industry- and performance adjusted
profit margins (Barber and Lyon, 1996), but there exists a significant negative interaction
with the level of human capital and the predecessor’s activity with respect to differential abnormal performance. Therefore, the overall performance impact diminishes with
increased successor human capital and eventually turns negative for high levels of successor human capital. The above results of our Huber-White robust OLS regressions are
robust to the inclusion of commonly used controls regarding enterprise size, momentum,
ownership structure, and time and also further indicators of the state of the company and
the working relations between successor and predecessor.
We interpret these findings as evidence that if an information or skill gap between the
successor and predecessor exists, a phase of parallel activity, where management techniques and idiosyncratic knowledge are transferred, is vital for enterprise performance.
This is especially the case if the human capital score of the predecessor is low and a prolonged predecessor activity shields the enterprise from the (arguably reducing) deficits
of the successor. However, for strong and able successors an information or skill gap
is potentially non-existent, while high human capital successors often spark vital organizational change (Miller, 1993, and Ahrens and Woywode, 2012). In such cases the
continuing incumbency of the predecessor may be dysfunctional for the organization,
because openness and responsiveness to change stimuli may be potentially diminished
(Hambrick and Fukutomi, 1991, and Virany et al., 1992) and profound organizational
change may for many reasons spark the resistance of the predecessor and lead to internal
conflicts. Furthermore, if the preceding CEO omits to signal a willingness to clear out
of the way in such conflicts and if the roles are not really defined (Handler, 1990), the
successor might suffer from frustration and lower motivation leading to inferior performance (Dyck et al., 2002, Sharma et al., 2001, Sharma et al., 2003, and Le Breton-Miller
et al., 2004). This is mirrored in the negative and significant performance impact of prolonged predecessor activity when the successor has high levels of human capital. Among
the robustness checks we also employ a 2SLS-IV which reveals that the effect of performance on predecessor activity leads to an underestimation of the effect of predecessor
activity on performance by roughly one percentage point of differential industry- and
performance-adjusted profit margin, while on average the general heterogeneous pattern
described above remains valid.
Furthermore, we find evidence that external and family successors seem to be capable of shielding against some of the negative influence of the successor human capital*predecessor activity interaction on differential abnormal performance as compared
to enterprise successors, which as former veterans of the preceding CEO arguably have
the hardest task in introducing change and avoiding internal conflicts. However, when
we control for nepotism, it becomes visible that for family successions many of the positive influences of predecessor activity are due to “shielding effects” in order to avoid
negative performance effects of successors with growing, but low human capital, while
in non-nepotistic succession with arguably strong successors the predecessor’s activity
entails a significant negative effect on differential abnormal performance.
In addition, we also cast light on the degree of the predecessor’s influence on differ-

CHAPTER 2: CULTIVATING ROSES?

89

ential abnormal performance. Remarkably, we find the same pattern as in the regressions including simply the predecessor’s activity except that it seems to be amplified by
an increasing degree of influence of the predecessor, which is robust to the inclusion
of an array of controls. For low human capital successors a high degree of predecessor influence entails more positive and significant effects on differential industry- and
performance-adjusted profit margins, as compared to a low degree of predecessor influence. By contrast, for successors with strong human capital a high degree of predecessor
influence is related to much worse negative and significant effects on differential abnormal performance as compared to a low degree of predecessor influence.
Overall, our results highlight the potential positive role of a period of parallel activity
with respect to knowledge transfers, especially if the successor’s ability is still limited.
A smooth “phasing-in” and “phasing-out” process, sometimes also referred to as the
“succession dance” (Handler, 1990), which is dynamically guided by and oriented at the
current level of ability of the successor seems to play a key role in unleashing a maximum performance in successions. One of the interviewed CEO successors (12) noted:
“In any case, what my father should be grated as a very positive trait, is that whenever my duties and responsibilities in the company gained weight, he accordingly and
increasingly reduced his [responsibility].” However, if the successor is already equipped
or has reached a higher level of ability, the preceding CEO is perhaps best advised to
minimize his influence and role and to avoid the hubris of king Jemsh´?d.
REFERENCES

Adams, Ren´ee, Heitor Almeida, and Daniel Ferreira. 2009. “Understanding the Relationship between Founder-CEOs and Firm Performance.” Journal of Empirical Finance, 16(1): 136–150.
Ahrens, Jan-Philipp, and Michael Woywode. 2012. “Restructuring, Human Capital,
and Enterprise Performance in CEO Successions in Family Firms.” Dissertation Chapter - Dissertation Jan-Philipp Ahrens, University of Mannheim.
Ahrens, Jan-Philipp, Andreas Landmann, and Michael Woywode. 2012(b). “Gender
Preferences in CEO Successions in Family Firms: Family Characteristics and Human
Capital of the Successor.” Dissertation Chapter - Dissertation Jan-Philipp Ahrens, University of Mannheim.
Ahrens, Jan-Philipp, Sandra Gottschalk, and Michael Woywode. 2012. “Nepotism
- CEO Successions, Ownership and Enterprise Performance.” Dissertation Chapter Dissertation Jan-Philipp Ahrens, University of Mannheim.
Akerlof, George A. 1970. “The Market for “Lemons”: Quality Uncertainty and the
Market Mechanism.” Quarterly Journal of Economics, 84(3): 488–500.
Anderson, Ronald C., and David M. Reeb. 2003. “Founding-Family Ownership and
Firm Performance: Evidence from the S&P 500.” Journal of Finance, 58(3): 1301–
1328.
Arrow, Kenneth J. 1973. “Higher Education as a Filter.” Journal of Public Economics,
2(3): 193–216.

90

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Atchley, Robert C. 1989. “A Continuity Theory of Normal Aging.” The Gerontologist,
29(2): 183–190.
Barach, Jeffrey A., and Jospeh B. Ganitsky. 1995. “Successful Succession in Family
Business.” Family Business Review, 8(2): 131–155.
Barach, Jeffrey A., Joseph Ganitsky, James A. Carson, and Benjamin A. Doochin.
1988. “Entry of the Next Generation: Strategic Challenge for Family Business.” Journal
of Small Business Management, 26(2): 49–56.
Barber, Brad M., and John D. Lyon. 1996. “Detecting Abnormal Operating Performance: The Empirical Power and Specification of Test Statistics.” Journal of Financial
Economics, 41(3): 359–399.
Barnes, Louis B., and Simon A. Hershon. 1976. “Transferring Power in the Family
Business.” Harvard Business Review, 54: 105–114.
Barney, Jay. 1991. “Firm Resources and Sustained Competitive Advantage.” Journal of
Management, 17(1): 99–120.
Beckhard, Richard, and D. Gibb Dyer, Jr. 1983. “Managing Continuity in the Familyowned Business.” Organizational Dynamics, 12(1): 5–12.
Bennedsen, Morten, Kasper Meisner Nielsen, Francisco P´erez-Gonz´alez, and
Daniel Wolfenzon. 2007. “Inside the Family Firm: the Role of Families in Succession Decisions and Performance.” Quarterly Journal of Economics, 122(2): 647–691.
Bertrand, Marianne, and Antoinette Schoar. 2003. “Managing with Style: The Effect
of Managers on Firm Policies.” Quarterly Journal of Economics, 118(4): 1169–1208.
Bigley, Gregory A., and Margarethe F. Wiersema. 2002. “New CEOs and Corporate
Strategic Refocusing: How Experience as Heir Apparent Influences the Use of Power.”
Administrative Science Quarterly, 47(4): 707–727.
Boeker, Warren, and Rushi Karichalil. 2002. “Entrepreneurial Transitions: Factors
Influencing Founder Departure.” The Academy of Management Journal, 45(4): 818–
826.
Burkart, Mike, Fausto Panunzi, and Andrei Shleifer. 2003. “Family Firms.” The Journal of Finance, 58(5): 2167–2202.
Cabrera-Su´arez, Katiuska, Petra De Sa´a-P´erez, and Desiderio Garc´?a-Almeida.
2001. “The Succession Process from a Resource- and Knowledge-Based View of the
Family Firm.” Family Business Review, 14(1): 37–46.
Campbell, Joseph John. 1949. The Hero with a Thousand Faces. New York: Pantheon
Books.
Castanias, Richard P, and Constance E. Helfat. 1991. “Managerial Resources and
Rents.” Journal of Management, 17(1): 155–171.
Chrisman, James J., Jess H. Chua, and Pramodita Sharma. 1998. “Important Attributes of Successors in Family Businesses: An Exploratory Study.” Family Business
Review, 11(1): 19–34.
Churchill, Neil C ., and Virginia L. Lewis. 1983. “The Five Stages of Small Business
Growth.” Harvard Business Review, 61(3): 30–51.

CHAPTER 2: CULTIVATING ROSES?

91

Davis, James H., F. David Schoorman, and Lex Donaldson. 1997. “Towards a Stewardship Theory of Management.” Academy of Management Review, 22(1): 20–47.
Dawson, Alexandra. 2011. “Private Equity Investment Decisions in Family Firms:
The Role of Human Resources and Agency Costs.” Journal of Business Venturing,
26(2): 189–199.
Demsetz, Harold, and Bel´en Villalonga. 2001. “Ownership Structure and Corporate
Performance.” Journal of Corporate Finance, 7(3): 209–233.
Demsetz, Harold, and Kenneth Lehn. 1985. “The Structure of Corporate Ownership:
Causes and Consequences.” Journal of Political Economy, 93(6): 1155–1177.
Dyck, Alexander, and Luigi Zingales. 2004. “Private Benefits of Control: An International Comparison.” Journal of Finance, 59(2): 537–600.
Dyck, Bruno, Michael Mauws, Frederick A Starke, and Gary A Mischke. 2002.
“Passing the Baton: The Importance of Sequence, Timing, Technique and Communication in Executive Succession.” Journal of Business Venturing, 17(2): 143–162.
Dyer, Jr., W. Gibb. 2003. “The Family: the Missing Variable in Organizational Research.” Entrepreneurship Theory and Practice, 27(4): 401–416.
Fama, Eugene F. 1980. “Agency Problems and the Theory of the Firm.” Journal of
Political Economy, 88(2): 288–307.
Firdaus´?. 1832. The Sh´ah N´ameh of the Persian Poet Firdaus´? - translated and abridged
in prose and verse - by James Atkinson. London: The Oriental Translation Fund.
Hambrick, Donald C., and Gregory D. S. Fukutomi. 1991. “The Seasons of a CEO’s
Tenure.” The Academy of Management Review, 16(4): 719–742.
Handler, Wendy C. 1990. “Succession in Family Firms: A Mutual Role Adjustment between Entrepreneur and Next-generation Family Members.” Entrepreneurship Theory
and Practice, 15(1): 37–51.
Heckman, James J. 1978. “Dummy Endogenous Variables in a Simultaneous Equation
System.” Econometrica, 46(4): 931–959.
Helfat, Constance E. 2000. “Guest Editor’s Introduction to the Special Issue: The Evolution of Firm Capabilities.” Strategic Management Journal, 21(10/11): 955–959.
Henderson, Andrew D., Danny Miller, and Donald C. Hambrick. 2006. “How
Quickly Do CEOs Become Obsolete? Industry Dynamism, CEO Tenure, and Company Performance.” Strategic Management Journal, 27(5): 447–460.
Hillier, David, and Patrick McColgan. 2009. “Firm Performance and Managerial Succession in Family Managed Firms.” Journal of Business Finance & Accounting, 36(34): 461–484.
Ibrahim, A. Bakr, and J. R. Goodwin. 1986. “Perceived Causes of Success in Small
Business.” American Journal of Small Business, 11(2): 41–50.
Jensen, Michael C., and William H. Meckling. 1976. “Theory of the Firm: Managerial
Behavior, Agency Costs and Ownership Structure.” Journal of Financial Economics,
3(4): 305–360.

92

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Johnson, Simon, Rafael La Porta, Florencio Lopez-de Silanes, and Andrei Shleifer.
2000. “Tunneling.” The American Economic Review, 90(2): 22–27.
Kerkhoff, Elisabeth, Klaus Ballarini, and Detlef Keese. 2004. Generationswechsel in
mittelst¨andischen Unternehmen. Karlsruhe: Loeper.
Kets de Vries, Manfred F. R. 1985. “The Dark Side of Entrepreneurship.” Harvard
Business Review, 63(6): 160–167.
Lansberg, Ivan. 1988. “The Succession Conspiracy.” Family Business Review,
1(2): 119–143.
Lansberg, Ivan. 1999. Succeeding Generations: Realizing the Dream of Families in
Business. Boston (Mass.): Harvard Business School Press.
La Porta, Rafael, Florencio Lopez de Silanes, and Andrei Shleifer. 1999. “Corporate
Ownership Around the World.” Journal of Finance, 54(2): 471–517.
Lazear, Edward P., and Sherwin Rosen. 1981. “Rank-Order Tournaments as Optimum
Labor Contracts.” Journal of Political Economy, 89(5): 841–864.
Le Breton-Miller, Isabelle, and Danny Miller. 2006. “Why Do Some Family Businesses Out-Compete? Governance, Long-Term Orientations, and Sustainable Capability.” Entrepreneurship Theory and Practice, 30(6): 731–746.
Le Breton-Miller, Isabelle, Danny Miller, and Lloyd P. Steier. 2004. “Toward an Integrative Model of Effective FOB Succession.” Entrepreneurship Theory and Practice,
28(4): 305–328.
Lee, Khai Sheang, Guan Hua Lim, and Wei Shi Lim. 2003. “Family Business Succession: Appropriation Risk and Choice of Successor.” Academy of Management Review,
28(4): 657–666.
Levinson, Harry. 1971. “Conflicts that Plague Family Businesses.” Harvard Business
Review, 49(2): 90–98.
Levinson, Harry. 1974. “Don’t Choose Your Own Successor.” Harvard Business Review, 52(6): 53–62.
Mehrotra, Vikas, Randall Morck, Jungwook Shim, and Yupana Wiwattanakantang. 2011. “Adoptive Expectations: Rising Sons in Japanese Family Firms.” NBER
Working Paper No. 16874.
Miller, Danny. 1990. The Icarus Paradox: How Exceptional Companies Bring About
their own Downfall. New York: HarperBusiness.
Miller, Danny. 1991. “Stale in the Saddle: CEO Tenure and the Match between Organization and Environment.” Management Science, 37(1): 34–52.
Miller, Danny. 1993. “Some Organizational Consequences of CEO Succession.” The
Academy of Management Journal, 36(3): 644–659.
Miller, Danny, and Cornelia Dr¨oge. 1986. “Psychological and Traditional Determinants of Structure.” Administrative Science Quarterly, 31(4): 539–560.
Miller, Danny, Lloyd Steier, and Isabelle Le Breton-Miller. 2003. “Lost in Time: Intergenerational Succession, Change, and Failure in Family Business.” Journal of Business Venturing, 18(4): 513–531.

CHAPTER 2: CULTIVATING ROSES?

93

Mincer, Jacob A. 1974. Schooling, Experience, and Earnings. Columbia University
Press.
Morck, Randall, David Stangeland, and Bernard Yeung. 2000. “Inherited Wealth,
Corporate Control and Economic Growth: The Canadian Disease.” In Concentrated
Corporate Ownership. National Bureau of Economic Research, Chicago: University of
Chicago Press.
Morris, Michael H., Roy O. Williams, Jeffrey A. Allen, and Ramon A. Avila. 1997.
“Correlates of Success in Family Business Transitions.” Journal of Business Venturing,
12(5): 385–401.
Murphy, Kevin J., and J´an Z´abojn´?k. 2004. “CEO Pay and Appointments: A MarketBased Explanation for Recent Trends.” American Economic Review, 94(2): 192–196.
Murphy, Kevin J., and Jerold L. Zimmermann. 1993. “Financial Performance Surrounding CEO Turnover.” Journal of Accounting and Economics, 16(1-3): 273–315.
Osterloh, Margit, and Bruno S. Frey. 2000. “Motivation, Knowledge Transfer, and
Organizational Forms.” Organization Science, 11(5): 538–550.
P´erez-Gonz´alez, Francisco. 2006. “Inherited Control and Firm Performance.” American Economic Review, 96(5): 1559–1588.
Rosen, Sherwin. 1986. “Prizes and Incentives in Elimination Tournaments.” American
Economic Review, 76(4): 701–715.
Schulze, William S., Michael H. Lubatkin, Richard N. Dino, and Ann K. Buchholtz.
2001. “Agency Relationships in Family Firms: Theory and Evidence.” Organization
Science, 12(2): 99–116.
Sharma, Pramodita, James J. Chrisman, and Amy L. Pablo. 2001. “Determinants of
Initial Satisfaction with the Succession Process in Family Firms: a Conceptual Model.”
Entrepreneurship Theory and Practice, 25(3): 17–35.
Sharma, Pramodita, James J. Chrisman, and Jess H. Chua. 2003. “Predictors of Satisfaction with the Succession Process in Family Firms.” Journal of Business Venturing,
18(5): 667–687.
Shleifer, Andrei, and Robert W. Vishny. 1986. “Large Shareholders and Corporate
Control.” Journal of Political Economy, 94(3): 461–488.
Smith, Brian F., and Ben Amoako-Adu. 1999. “Management Succession and Financial
Performance of Family Controlled Firms.” Journal of Corporate Finance, 5(4): 341–
368.
Smith, Mark, and Michael C. White. 1987. “Strategy, CEO Specialization, and Succession.” Administrative Science Quarterly, 32(2): 263–280.
Sonnenfeld, Jeffrey A. 1988. The Hero’s Farewell: What Happens When Chief Executives Retire. New York: Oxford University Press.
Sonnenfeld, Jeffrey A., and Padraic L. Spence. 1989. “The Parting Patriarch of a Family Firm.” Family Business Review, 2(4): 355–375.
Spence, A. Michael. 1973. “Job Market Signaling.” Quarterly Journal of Economics,
87(3): 355–374.

94

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Spence, A. Michael. 1974. “Competitive and Optimal Responses to Signals: An Analysis of Efficiency and Distribution.” Journal of Economic Theory, 7(3): 296–332.
Strober, Myra H. 1990. “Human Capital Theory: Implications for HR Managers.” Industrial Relations, 29(2): 214–239.
Teece, David J., Gary Pisano, and Amy Shuen. 1997. “Dynamic Capabilities and
Strategic Management.” Strategic Management Journal, 18(7): 509–533.
Vancil, Richard F. 1987. Passing the Baton: Managing the Process of CEO Succession.
Boston (Mass.): Harvard Business School Press.
Villalonga, Bel´en, and Raphael Amit. 2006. “How Do Family Ownership, Control and
Management Affect Firm Value?” Journal of Financial Economics, 80(2): 385–417.
Virany, Beverly, Michael L. Tushman, and Elaine Romanelli. 1992. “Executive
Succession and Organization Outcomes in Turbulent Environments: An Organization
Learning Approach.” Organization Science, 3(1): 72–91.
von Hayek, Friedrich A. 1945. “The Use of Knowledge in Society.” American Economic Review, 35(4): 519–530.
Wasserman, Noam. 2003. “Founder-CEO Succession and the Paradox of Entrepreneurial Success.” Organization Science, 14(2): 149–172.
A PPENDIX
A1. Detailed Sample Selection Information

This section is provided for the reader’s convenience and is similar to Ahrens et al.
(2012), Ahrens and Woywode (2012), and Ahrens et al. (2012,b) who employed a very
similar data set. The second filter for the gross sample selection from the Mannheim
Enterprise Panel (MUP) is composed in the following way. We suspect a succession
event to have taken place, if between the years 2002 and 2008:
1) a leading member of the executive board resigned, or
2) a new leading member of the executive board was appointed, or
3) a previous owner reduced his share, or
4) a new or previous owner (a natural person) increased his share, and
5) one of the previous owners and leading members of the executive board was over
55 years old.
The age criterion raises the chance of observing normal successions caused by old
age and the natural person criterion separates out takeovers by other companies (legal
entities) from the sample, such that the observed population of the sample companies
keeps the “concentrated ownership control” attribute. Furthermore, all enterprises which
ceased to exist or went through a re-establishment are unobserved, which introduces
a survivorship bias to our sample. As a consequence, we point out that our findings

CHAPTER 2: CULTIVATING ROSES?

95

are limited to the selected sample of surviving enterprises and are of reduced holistic
representativeness. Nevertheless, these constraints only impose minor restrictions on
this article’s research goals.
Applying both filters to the MUP database yields a total gross sample of 14,250 enterprises, which is categorized according to the International Standard Industry Classification of All Economic Activities (ISIC Rev. 3.1) of the United Nations as serving as
a basis for standardized computer-aided telephone interviews. The ISIC industry sections (A) agriculture, hunting and forestry, (B) fishing, (C) mining, quarrying and (E)
electricity gas and water supply, (L) public administration and defense & compulsory
social security, (P) activities of households, (Q) extra-territorial organizations and bodies
as well as division (91) activities of membership organizations are excluded. In addition,
enterprises for which no telephone number is available in the MUP database are dropped
(less than one percent).
The gross sample of 14,250 enterprises is contacted by the Center for Evaluation and
Methods (CEM) using a standardized computer aided interview (ZEW-Unternehmensbefragung “Generationenwechsel Mittelstand”, 2010).
A2. Additional Tables

In table A1 we present the industry classification key which we employed for the
industry cluster aggregation (Table A1 is taken from Ahrens et al., 2012).

96

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE A1—I NDUSTRY C LASSIFICATION K EY

ZEW industry key
1. Manufacturing
2. Construction
3. Business services

ISIC
Rev. 3.1
code
(1)
D
F
K

ISIC
Rev. 3.1
industry description
(2)

Manufacturing
Construction
Real estate, renting and business activities
(without ISIC 70: real estate activities)
O
Other community, social and personal service activities
(only ISIC 90: sewage and refuse disposal, sanitation and
similar activities)
4. Consumer services
H
Hotels and restaurants
K
Real estate, renting and business activities
(only ISIC 70: real estate activities)
M
Education
N
Health and social work
O
Other community, social and personal service activities
(only ISIC 92: recreational, cultural and sporting activities
and ISIC 93: other service activities)
5. Wholesale & retail
G
Wholesale and retail trade; repair of motor vehicles, motorcycles and personal and household goods
6. Other
I
Transport, storage and communication
J
Financial intermediation
Note: For each aggregated industry cluster the key in form of the ISIC Rev. 3.1 code (column 1) and its description
(column 2) is reported. The ISIC industry sections (A) agriculture, hunting and forestry, (B) fishing, (C) mining, quarrying and (E) electricity gas and water supply, (L) public administration and defense & compulsory social security, (P)
activities of households, (Q) extra-territorial organizations and bodies as well as division (91) activities of membership
organizations are not included. ISIC industry categories with no observations are not displayed.

Employment costs
per employee
B. Controls
?
Successor attributes
?
Company level
?
Industry level
?
Years
Observations
747
Pseudo R2
0.08
Note: A legend is presented separately.

CEO human capital score

Research & development
expenses-sales ratio
Tacitness (intangibles-sales ratio)

?
?
?
?
747
0.08

?
?
?
?
747
0.08

(1)
(2)
(3)
A. Knowledge transfer proxies at the two-digit industry level
Absolute new firms
-0.03
(0.121)
Start-up ratio
0.07
(0.133)
Employee growth
-0.18
(0.111)
Total assets-sales ratio

Panel

747
0.08

?
?
?
?

0.03
(0.126)
0.13
(0.138)
-0.21*
(0.119)

747
0.08

?
?
?
?

-0.15
(0.127)

489
0.11

?
?
?
?

-0.19
(0.136)

587
0.10

?
?
?
?

0.14
(0.126)

747
0.08

?
?
?
?

-0.18
(0.314)

Probit regression dependent variable: active predecessor
(4)
(5)
(6)
(7)
(8)

701
0.09

?
?
?
?

0.08
(0.142)

(9)

TABLE A2—D ISAGGREGATED K NOWLEDGE T RANSFER P ROXIES AND ACTIVITY OF P REDECESSOR

465
0.13

?
?
?
?

0.04
(0.195)
-0.06
(0.155)
0.47**
(0.192)
0.12
(0.552)
0.28
(0.214)

(10)

465
0.13

?
?
?
?

-0.03
(0.164)
0.26
(0.311)
-0.09
(0.181)
0.00
(0.205)
0.00
(0.176)
0.44**
(0.195)
0.25
(0.588)
0.34
(0.223)

(11)

CHAPTER 2: CULTIVATING ROSES?
97

98

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Legend table A2: The dependent variable is an indicator equal to one if the predecessor remains active after the
succession within his company and zero otherwise.
The variables of interest are the disaggregated elements of the knowledge transfer proxies per industry (at the two-digit
ISIC level) used in table 3 and include: Absolute new firms, an indicator equal to one if the absolute number of new firms
in the respective industry and year (using Eurostat data for 2008 and 2009 for Germany) if higher than or equal to the
overall median of new firms per industry; Start-up ratio an indicator equal to one if the ratio of new firms to the total
population of firms per industry category (using Eurostat data for 2008 and 2009 for Germany) is higher or equal to the
overall median of the start-up ratio per industry; Employees growth is an indicator equal to one if the employee growth
(per company) per industry and year (using Amadeus data for 2002 until 2009 for Germany) is higher than the overall
median industry employee growth at the industry level; Total assets-sales ratio is an indicator equal to one if the total
assets over operating revenue ratio per industry (using Amadeus Data from 2002-2009 for Germany) is greater than or
equal to the overall median total assets-sales ratio at the industry level; Research & development expenses-sales ratio is
an indicator equal to one if the research and development expenditure over operating revenue ratio (using COMPUSTAT
data for EU 15 countries from 2002 to 2009) of the industry is higher than or equal to the overall median research &
development expenses-sales ratio at the industry level; Tacitness (intangibles-sales ratio) is an indicator equal to one if
the average intangibles over operating revenue ratio per industry (using COMPUSTAT data for EU 15 countries from
2002 to 2009) is equal to or above the overall median intangibles-sales ratio at the industry level; CEO human capital
score is an indicator equal to one if the average non-family supply of CEO successor human capital per industry (using
the sample data and the ZEW industry classification) is greater than or equal to the median non-family CEO successor
human capital at the industry level; Employment costs per employee is an indicator equal to one if the employee costs per
employee at the industry level (using Eurostat data from 2002-2008 for Germany) is greater than or equal to the median
employee costs per employee at the industry level.
Controls include: Family CEO, for successors related by marriage or blood to at least one of the three persons owning
more than 50% of the enterprise; Enterprise CEO, for unrelated successors who previously worked for the enterprise;
Ownership is an indicator equal to one if the successor owned a share of the enterprise in the succession year; Human
capital (score) is the successor’s human capital score derived from the sum of five proxy elements: (1) age, (2) industry
experience, (3) leadership experience, (4) merchant education, and (5) professionalism; Ln employees refers to the natural logarithm of the number of employees in the succession year; Executives displays the number of executives in the
succession year; Default probability is default probability in the succession year derived from the Creditreform solvency
index; and Company age is the age of the company in the succession year. Industry level controls are indicators equal to
one if the respective category according to the ZEW industry classification (see appendix table A1) is met. Years is the
number of years passed since the succession. All values are displayed in 07/2009 euros. The table presents estimated
changes in probabilities devising a maximum-likelihood probit model. The stars display significances at: * ten percent,
** five percent and *** one percent. The values in parentheses display the standard errors.

Table A2 presents the disaggregated elements of the knowledge intensity proxies “dynamism” and “knowledge intensity” of table 3. An inspection of the results from table A2
yields that the individual proxies are, excluding the proxy for tacitness, all insignificant.
Apart from tacit knowledge, this leads us to reject any proposed relations between the
likelihood of the activity of the predecessor and the devised knowledge transfer proxies.
Table A3 shows a differentiation of the predecessor’s roles with respect to their performance impact. In detail, we employ the same model as in column 4 of table 6, but vary
the roles of the predecessor.
In general table A3 shows some heterogeneity between the predecessor roles with
respect to their performance impact, especially with respect to the interactions with the
successor’s human capital score. However, since the interview allows predecessors to
indicate multiple roles and because the observations in some of the role categories are
rather low, we refrain from emphasizing these results and present them mainly for the
sake of completeness.

CHAPTER 2: CULTIVATING ROSES?

99

TABLE A3—OLS-R EGRESSION A NALYSIS - P REDECESSOR ROLES AND E NTERPRISE P ERFORMANCE

Predecessor
roles [Role]
Variables
[Role]
(predecessor)
Company age
Company age*[Role]
Human capital score (HCS)
Human capital*[Role]
Family
External
Family*[Role]
External*[Role]
Family*HCS*[Role]
External*HCS*[Role]
Nepotism strong
Nepotism weak
Nepotism strong*[Role]
Nepotism weak*[Role]

Owner
(shareholder)
(1)
2.50**
(1.139)
0.01
(0.009)
-0.02*
(0.010)
0.61**
(0.285)
-1.66***
(0.510)
-0.27
(1.190)
-0.81
(0.827)
-3.59
(2.548)
-0.65
(2.052)
1.79***
(0.654)
1.39**
(0.692)
-2.95*
(1.605)
-0.04
(1.278)
3.61
(2.323)
2.06
(2.115)
?

Dependent variable: 1 Ind.- & perf.-adj. PM
Active &
Key
responsible Passive
Board
account
owner
owner
member
holder
(2)
(3)
(4)
(5)
2.86**
1.04
5.97*
2.92
(1.127)
(2.119)
(3.041)
(2.333)
0.00
-0.00
-0.00
-0.01
(0.008)
(0.005)
(0.005)
(0.005)
-0.01
-0.02
-0.02*
-0.01
(0.009)
(0.021)
(0.012)
(0.023)
0.59*
0.30
0.39
0.36
(0.307)
(0.238)
(0.254)
(0.242)
-1.22***
-1.80*
-2.87**
-2.26*
(0.463)
(1.015)
(1.172)
(1.284)
0.25
-0.28
0.21
0.12
(1.080)
(0.980)
(1.117)
(1.059)
0.46
-0.24
0.26
0.15
(0.858)
(0.702)
(0.719)
(0.693)
-3.11
-4.50
-8.71**
-5.84*
(2.815)
(3.698)
(3.847)
(3.443)
-1.10
-3.85
-3.84
col.
(2.137)
(5.232)
(3.480)
col.
1.16*
2.80**
3.54***
2.82*
(0.678)
(1.350)
(1.320)
(1.537)
0.60
3.21*
2.23*
0.93
(0.696)
(1.734)
(1.268)
(0.789)
-2.38*
-1.72
-2.80**
-1.97
(1.400)
(1.221)
(1.321)
(1.307)
-0.01
0.41
-0.05
0.15
(1.119)
(1.043)
(1.194)
(1.129)
2.67
3.09*
6.70***
4.38*
(2.486)
(1.800)
(1.861)
(2.478)
1.87
1.92
3.39*
2.14
(2.301)
(2.666)
(1.970)
(2.030)
?
?
?
?

Special
tasks
(6)
0.20
(1.387)
-0.01
(0.005)
0.02
(0.014)
0.30
(0.260)
-0.94
(0.588)
0.29
(1.061)
0.38
(0.740)
-2.03
(2.671)
-2.02
(2.481)
0.81
(0.988)
1.03
(0.681)
-1.99
(1.349)
0.26
(1.113)
1.42
(1.791)
-0.12
(1.836)
?

Controls
Observations
383
383
383
383
383
383
R2
0.10
0.08
0.11
0.11
0.08
0.07
Note: The dependent variable is the difference in industry- and performance-adjusted profit margin (PM). The differences
are calculated via: industry- (and performance-) adjusted PM of the year 2009 less industry- (and performance-) adjusted
PM of the succession year. Industry-adjusted PM is PM less the median PM of the accordant year and industry (two-digit
ISIC) of a control group (from the Amadeus database); Performance-adjustments are designed by sorting the industryadjusted values of a control group into deciles and matching the individual industry-adjusted PM values of the sample with
the accordant control group decile in the year of the succession. The median industry-adjusted PM of the relevant control
group decile and year then serves as a control. Independent variables are: [Role] (predecessor) is a flexible indicator
equal to one if the role of the predecessor in the respective column is met; Company age is the age of the company in
years; Family indicates if a successor is related by marriage or blood to at least one of the three persons owning more than
50% of the enterprise; External indicates if a successor had no previous ties to the enterprise; Human capital score (HCS)
is derived from the sum of five proxy elements: (1) age, (2) industry experience, (3) leadership experience, (4) merchant
education, and (5) professionalism; Nepotism strong indicates if the predecessor only considered family members as
successors; and Nepotism weak indicates if the succession is not categorized into the strong nepotism category and a
family successor with lower human capital, compared to the average human capital of non-family successors within the
industry of the company, is installed as successor. Controls are: Ln sales (size) is the natural logarithm of sales in the
year of the succession; Industry-adjusted PM (momentum) is the industry-adjusted ratio earnings before taxes divided
by operating revenue in the succession year; Industry- and performance-adjusted PM (momentum) is industry- and
performance-adjusted PM in the succession year; Ownership is an indicator equal to one if the successor owned a share
of the enterprise in the succession year; Default probability represents probability of default based on the Creditreform
solvency-index score of the enterprise in the year of the succession; Executives displays the number of executives in the
succession year; Working relation is an indicator equal to one if the successor perceives the working relations with the
predecessor as good; Investment delay is an indicator variable equal to one if an investment delay is perceived; Financing
requirements is an indicator variable equal to one if severe unexpected financing requirements were encountered during
the succession; and Years is the time elapsed in years since the succession. Interactions between variables are marked via
stars *. Significances are displayed within the table via: * ten percent, ** five percent and *** one percent. The values in
parentheses display Huber-White robust standard errors.

Chapter 3
Gender Preferences in CEO Successions in Family Firms:
Family Characteristics and Human Capital of the
Successor
By JAN -P HILIPP A HRENS , A NDREAS L ANDMANN , AND M ICHAEL W OYWODE ?
We investigate labor market constraints in CEO succession contests devising an unique data set on CEO successions in enterprises with concentrated ownership and control. We find that a preference for male
family heirs limits labor market selectivity: Family successions are significantly more likely to occur when a son is among the predecessor’s
children as compared to daughters. Sons among the children increase
the likelihood of nepotistic successions, while in turn female family successors are equipped with higher human capital due to tougher selectivity criteria. Furthermore, the regional industry supply of CEO resources
influences the observed human capital of installed successors.
(JEL: G30, J13, J24, L26, M51)
Keywords: CEO succession, family firms, family characteristics, human
capital, promotion decisions.

? Ahrens: University of Mannheim, School of Business Studies and Economics, Department of Business Studies,
L9 1-2, D-68161 Mannheim, Germany, Tel: +49-177-656-2031, (e-mail: [email protected]). Landmann:
University of Mannheim, Department of Economics, L7 3-5, D-68161 Mannheim, Germany, Tel: +49-621-181-1842,
(e-mail: [email protected]). Woywode: University of Mannheim, School of Business Studies and
Economics, Department of Business Studies, L9 1-2, D-68161 Mannheim, Germany, Tel: +49-62-1181-2894, (e-mail:
[email protected]). Many thanks to Stefanie Ahrens and Robert and Margret Brownell for helpful
comments. Financial support from the Konrad-Adenauer-Stiftung e.V. is gratefully acknowledged.

101

102

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

I. Introduction

The recent economic literature on ownership and control reports that around the world
most firms are controlled by families, founders or their heirs (La Porta et al., 1999, and
Faccio and Lang, 2002). One of the most perilous moments in course of family- and
in particular owner-controlled firms are CEO successions, as only 30% of these firms
survive past the lifespan of their founders (Sonnenfeld and Spence, 1989).
This circumstance has led to a growing attention to the post-succession period in family firms in the firm performance literature, which highlights imperfections in the CEO
successor labor market contests leading to subsequent performance declines (P´erezGonzal´es, 2006, Bennedsen et al., 2007, and Ahrens et al., 2012). The concentration
of ownership and control opens a leeway for the controlling majority, often the parting
CEO, to pursue private benefit maximization instead of profit maximization (Fama and
Jensen, 1983, and Demsetz 1983). In the case of a CEO succession this pursuit takes
the form of biases towards and favoritism of family members, while such nepotistic tendencies or even primogeniture can have detrimental effects for the family firm and firm
performance (Pollak, 1985, Bloom and Van Reenen, 2007, and Ahrens et al., 2012).
At the core of this imperfection is the preference or wish of the parting CEO to see
a family heir to steer the fortunes of the family enterprise or even of his oeuvre, which
leads to a constraint of the successor pool to family members only (id est persons related
to the parting CEO by either blood or marriage). However, in the light of the recent
literature on U.S. family structures (“The Demand for Sons”, Dahl and Moretti, 2008)
the question remains whether this is the complete story, or phrased differently: Is there a
second constraint within the family successor pool which further reduces the successor
pool towards male-family successors? As a consequence, if nepotistic successions occur
more often among male successors, then the reduced selectivity of these CEO succession
contests should be mirrored in a lower ability of male family CEO successors as compared to female family CEO successors. In this article we try to identify potential forces
among the family, firm and industry characteristics which lead to cases of nepotism in
CEO successions. We harness an unique data set which covers 804 CEO successions
between 2002 and 2008 in German family firms with concentrated ownership and we
differentiate between strong forms and weak forms of nepotism. Our results indicate
that the occurrence of both forms of nepotism is significantly driven by the presence
of male children among the predecessor’s children which is robust to the inclusion of
various controls and is indicative for a gender preference of the predecessor.
The single pieces of evidence taken alone will not allow us to make a point, but bringing them together they deliver a clear picture: Family entrepreneurs seem to prefer male
family successors over female family successors. When the predecessor has at least one
son and one daughter, a male family successor is chosen in 79% of the family successions. In addition, if the predecessor has only sons, a male family successor is chosen
in 94% of the family successions, whereas if the predecessor has only daughters, a male
family successor is still chosen in 21% of the family successions. The presence of sons
among the predecessor’s children significantly (at the one percent level) increases the
probability to observe a family succession. This finding is robust to the inclusion of firm

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

103

and industry level controls. Furthermore, our results on the family structure of family
firm entrepreneurs hint at strategic behavior to increase the number of children until at
least one son is born.1
Interestingly, on average female family successors are equipped with significantly (at
the five percent level) higher levels of human capital as compared to male family successors. For example, female family CEO successors are nearly eleven years older (significant at the five percent level) compared to male family heirs. These differences also
become visible when contrasting post-succession performance developments: the profitability in terms of profit margin reduces by -7.4% in male family successor led enterprises, while it increases by 14% in female family successor steered enterprises. This
difference-in-differences remains robust when controlling for industry and performance
trends. The observed human capital of the successors is significantly negatively related
(at the one percent level) to a family origin of the successor, indicating nepotistic tendencies. In addition, our results show that the observed human capital score of the installed
CEO successor is significantly and positively related to the average regional industry
supply of CEO human capital. These results are robust to the inclusion of various controls.
The main finding of this article is that there seems to be a further constraint of the CEO
labor market in the case of successions in family firms due to the preferences for sons.
Our results highlight the fact that in enterprises with concentrated ownership control,
such as family firms, the preferences of the predecessor play a crucial role in the observed
human capital of the installed CEO successor and the subsequent performance of the
enterprise. The article is designed in the following way: Section II gives a literature
overview and is followed by the theoretical section III. Section IV gives information on
the sample. Section V is dedicated to the data analysis. It begins with a brief descriptive
overview, followed by a proposition testing. Section VI discusses the main results and
offers a conclusion.
II. Related Literature

CEO succession have been studied extensively by various scholars leading to a manychromatic literature strand. This article focuses on potential constraints of the CEO successor labor market with a particular attention on the family structure of the predecessor
and the gender of the successor. Therefore, the main literature strands related to this
article are the theory of the firm, the economic contest literature and the recent economic
literature on family structures.
The question whether gender preferences matter has been raised quite early, as for the
analysis of individual behavior gender preferences might be of importance (Ben-Porath
and Welch, 1976). Two articles of Dahl and Moretti (2004, 2008) find evidence which
supports the notion that parents in the U.S. slightly favor boys over girls. For example,
they report that women with first-born daughters are significantly more likely to become
divorced and significantly less likely to be shotgun married, as compared to women with
1 It might be interesting to validate this finding using another data set.

104

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

first-born sons (Dahl and Morretti, 2008). Furthermore, they find that this gender bias
due to the strong preferences of men to have sons, while women only seem to have slight
preferences of daughters over sons (Dahl and Moretti, 2008). Fathers are reported to
spend more time with their sons (Morgan et al., 1988, and Lamb, 1997). While the labor
market behavior of married men seems to be largely inflexible to family size (Angrist and
Evans, 1998), there is evidence that the labor supply and wage rates of men raise more
with the event of having a son as compared to having a daughter (Lundberg and Rose,
2002). Interestingly, the article of Dahl and Moretti (2008) also highlights that fertility
rates in families with a first-born girl are significantly higher as compared to family with
a first-born sons.2
As many family firms are steered by male CEOs, we speculate that gender preferences
might also find their way into CEO succession contests and decisions. For example,
Bennedsen et al. (2007) report that the decision to appoint a family CEO successor is
significantly higher when the first-born is a boy. There is also evidence in the financial
economics literature. Using data of 90 family firms in Thailand, Bertrand et al. (2008)
show that the sons of founders crowd out the ownership and control right of other family
members, which is most evident when the founder has perished. Furthermore, they highlight that a relative larger number of sons among the children of the founder is negatively
correlated with lower firm-level performance, while no such correlation could be found
for the number of daughters.
It is argued in the literature that among the reasons for such negative performance
outcomes are constraints in the CEO succession contest which curb the natural selectivity
of CEO labor market (P´erez-Gonz´alez, 2006, Bennedsen et al., 2007, and Ahrens et al.,
2012).3 Installing such constraints becomes possible, because in family firms one often
finds unity of ownership and control allowing the departing CEO to enforce his own
preferences and private benefits, while such a pursuit potentially imposes agency costs on
minority shareholders, as highlighted in the respective ownership and control literature
(Jensen and Meckling, 1976, Fama, 1980, Demsetz and Lehn, 1985, Shleifer and Vishny,
1986, Johnson et al., 2000, Demsetz and Villalonga, 2001, Dyck and Zingales, 2004,
and Villalonga and Amit, 2006). In CEO succession contests one possible constraint
is restricting the pool of contestants to the family. Such restrictions adversely affect
the labor market contest leading to a lower ability of the winning candidate compared
to an unrestricted candidate field, as the contest’s structure and the available pool of
contestants are decisive for its outcome (Konrad, 2009). In turn, a lower ability of favored
CEO successors can lead to lower post-succession enterprise performance, as has been
reported by P´erez-Gonz´alez (2006) and Ahrens et al. (2012). The question whether
potential gender preferences also induce observable favoritism (or even nepotism) and
human capital effects in CEO successions has to the best of our knowledge not yet been
addressed directly in the literature.
2 For the United States there is some evidence which suggests that the intensity of these parental gender preferences
declined over the last decades, possibly due to changes in societal gender system (Pollard and Morgan, 2002).
3 CEO successions can also be understood as career ladder contests (Fama, 1980, Lazear and Rosen, 1981, Rosen,
1986, and Ahrens et al., 2012).

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

105

III. Theoretical Section
A. Categories and Definitions

The aim of this article is to explore the influence of gender preferences in CEO successions in enterprises with “concentrated ownership control” (or family firms), which
are defined for the purposes of this article as follows:
DEFINITION 1: An enterprise with concentrated ownership control is present if a maximum of three natural persons own more than 50% of the enterprise and at least one of
these owners is a leading member of the executive board.
Furthermore, in this article the attribute family is attached to successors related by
marriage or blood to at least one of the three persons owning more than 50% of the
enterprise, while the non-family attribute applies to all other successors. To assess the
human capital of the respective successors, we employ the human capital construct previously employed by Ahrens and Woywode (2012). This construct is defined as the sum of
five proxy elements: (1) age above median (proxy for general experience), (2) industry
experience above median (proxy for industry related experience), (3) leadership experience (proxy for practical managerial skills), (4) merchant education, if the successor
holds an university degree in business studies (or strongly related field) or was educated
at an university of cooperative education (proxy for theoretical managerial skills), and (5)
use of a business plan during the succession (proxy for professional managerial skills).
The human capital score mirrors a thought brought forward by Murphy and Z´abojn´?k
(2004) who argue that managerial skills have gained more importance for CEOs over the
last three decades. This is due to the progress in computerization, corporate controlling
and finance, and other arts which, if mastered by the CEO, increase a CEO’s ability to
steer a company and make specific knowledge more accessible than in earlier decades.
In addition, the human capital score also reflects the notion of the human capital theory
that productivity-enhancing investments take place in the education and post-education
phase (Mincer, 1974 and Strober 1990).4
In addition, we address potential biases of the predecessors in favor of specific groups
of CEO successors. Following this idea, we apply a dual approach by distinguishing
between strong nepotism and weak nepotism which we define for purposes of this article
as follows:
DEFINITION 2: Strong nepotism is present if the predecessor only considers family
members as successors.
DEFINITION 3: Weak nepotism is present if the predecessor also considers non-family
members as successors, but decides to install a family member with a human capital
score lower than the average achievable human capital score of managers available in
the respective industry of his enterprise.
4 For a more detailed introduction and explanation of the human capital score we refer to Ahrens and Woywode
(2012).

106

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

In other words, weak nepotism is measured via the amount of the successor’s human
capital which predecessors are willing to give up in order to see a family heir steering
the fortunes of their family business. In order to be able to measure any allowance or
discount in human capital, we have to determine a benchmark for the average achievable
managerial human capital. Our benchmark is designed as follows: By assuming that no
nepotism is at work in succession in which the company was not handed over to a family
CEO, we infer that an able successor was chosen in these cases. Therefore, we take the
mean human capital score of the CEOs per industry of only these cases as a benchmark.5
B. Derived Propositions

Using the definitions of the previous section, we derive propositions which are presented in the following. Apart from strictly excluding non-family successors from the
succession contest (strong nepotism), there may also be weak forms of nepotism, and
the constraints they impose on the succession contest are likely to have an impact on the
contest outcome. Thus, assuming that some family heirs are preferred by their predecessors despite the availability of a better candidate (weak nepotism), then the impact of
such nepotistic tendencies is likely to be mirrored in a lower average human capital of
family successors.
PROPOSITION 1: The successor’s human capital is likely to be lower among family
CEO successors as compared to the average human capital of successors.
In addition to the wish of seeing a family heir to steer the fortunes of ones oeuvre,
we assume that a “demand for sons” (Dahl and Moretti, 2008), thus gender preference
of the predecessor, is likely to find its way into CEO successor decisions, which might
be particularly true for owner controlled family firms which allow the pursuit of private
benefit maximization. Therefore and possibly due to the societal gender system and
general socialization, we suspect that many family firm patriarchs wish their son(s) to
follow in their footsteps.
PROPOSITION 2: Predecessors show a preference of male family heirs as CEO successors.
As a potential consequence of the pursuit of preferences or favoritism in CEO successions are contest constraints, we argue that the predecessors’ wish for sons as successors
fuels tolerance of a lower human capital of the successor (weak nepotism) and might
even lead to an exclusion of non-family successors from the succession contest (strong
nepotism).
PROPOSITION 3: If the predecessor has male children this increases the likelihood of
nepotism in successions.
5 We assume that the mean human capital score of all non-family successors per industry in the sample is a sufficient
proxy on the external supply of CEO human capital and assume that this level of human capital is attainable by all
companies in their respective industry within the boundaries of a standard search for a non-family CEO successor. To
calculate the industry means, we employ the rather broad Centre for European Economic Research industry classification
consisting of six general industry categories, as we have only 804 observations.

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

107

However, we suppose that predecessors might be unwilling to accept any amount of
lower human capital arbitrarily, but that they might compare their heirs to the external
CEO labor market and will not allow an exaggerated mismatch in ability. Thus, we
suspect that the supply of human capital, in terms of the average available managerial
human capital in the respective industry of the succession enterprise, will be positively
related to the human capital of the successor installed.
PROPOSITION 4: The family successors’ human capital is positively related to the
level of average achievable CEO human capital in the respective industry of the enterprise.
IV. Sample Selection

In this article we revisit the data set of Ahrens and Woywode (2012). The data set for
this article relies on the following sources: (a) the Mannheim Enterprise Panel (MUP),
(b) the Bureau van Dyjk Amadeus database (Amadeus), (c) the Hoppenstedt database,
(d) the Creditreform solvency index information, (e) German Bundesbank information,
(f) standardized computer aided telephone interviews (ZEW-Unternehmensbefragung
“Generationenwechsel Mittelstand”, 2010), (g) non-standardized direct interviews and
(h) web-searches. Overall, we observe 804 CEO successions in German family enterprises with concentrated ownership and control with a size of 30 to 1,000 employees,
which took place between 2002 and 2008.6
V. Data Analysis
A. Summary Statistics

A first inspection of the data set shows that female CEO successions occur more often
in family successions, a quarter of the family successions are female family CEO successions, whereas a fifth of all successions are female CEO successions. The fact that
there are much fewer female successions already hints at a potential preference of male
over female successors.7 Overall, female family CEO successions occur relatively often
in the consumer services sector (32.4% of female family successors, and 30.9% of female successors), as opposed to lower relative rates in the manufacturing sector (20.5%
of female family successors, and 16.4% of female successors.).8
We theorize that a wish for a male successor might play a key role in the choice of
a successor. Thus, we start by categorizing the children structure of the predecessors
into: (a) male children only, (b) female children only, (c) mixed children, and (d) no
children. The summary statistics of this categorization which are presented in table 1
already reveal interesting results.
6 For further information on the sample selection, we refer to Ahrens and Woywode (2012).
7 It might also be that female successions occur less often because of gender differences in the CEO labor supply.

However, due to the size of the difference, we believe that it is unlikely that labor supply can exclusively explain the
difference.
8 For additional summary statistics we refer to the appendix A1.

108

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 1—C HILDREN S TRUCTURE OF THE P REDECESSOR AND S UCCESSIONS
Male
Female
No
Gender
Ratio
Ratio (4)
Total
only
only
Mixed
children
unknown (2) to (3) to (2)&(3)
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
A. Type of succession
1. Family
455
141
58
249
2
5
2.43
1.25
(57.81)
(64.68)
(49.15)
(65.70)
(3.45)
(35.71)
[100.00]
[30.99]
[12.75]
[54.73]
[0.44]
[1.10]
2. Non-family
332
77
60
130
56
9
1.28
0.95
(42.19)
(35.32)
(50.85)
(34.30)
(96.55)
(64.29)
[100.00]
[23.19]
[18.07]
[39.16]
[16.87]
[2.71]
Difference
7.8**
-5.3**
15.6***
(1-2)
(3.2)
(2.6)
(3.6)
Total
787
218
118
379
58
14
1.85
1.13
(100.00) (100.00) (100.00) (100.00) (100.00)
(100.00)
[100.00]
[27.70]
[14.99]
[48.16]
[7.37]
[1.78]
B. Gender of family CEO successor
3. Male
336
131
12
191
1
1
10.92
1.34
family CEO
(75.68)
(93.57)
(21.43)
(78.93)
(50.00)
(25.00)
4. Female
108
9
44
51
1
3
0.20
0.96
family CEO
(24.32)
(6.43)
(78.57)
(21.07)
(50.00)
(75.00)
Difference in
87.14
-57.14
57.86
class (3-4)
Total
444
140
56
242
2
4
2.50
(100.00) (100.00) (100.00) (100.00) (100.00)
(100.00)
[100.00]
[31.53]
[12.61]
[54.50]
[0.45]
[0.90]
C. Intensity of nepotism
5. Nepotism
363
112
40
204
2
5
2.80
1.34
(46.12)
(51.38)
(33.90)
(53.83)
(3.45)
(35.71)
[100.00]
[30.85]
[11.02]
[56.20]
[0.55]
[1.38]
5a. Strong
52
17
4
31
0
0
4.25
1.48
nepotism
(6.61)
(7.80)
(3.39)
(8.18)
(0.00)
(0.00)
[100.00]
[32.69]
[7.69]
[59.62]
[0.00]
[0.00]
5b. Weak
311
95
36
173
2
5
2.64
1.32
nepotism
(39.52)
(43.58)
(30.51)
(45.65)
(3.45)
(35.71)
[100.00]
[30.55]
[11.58]
[55.63]
[0.64]
[1.61]
6. Family no
92
29
18
45
0
0
1.61
0.96
nepotism
(11.69)
(13.30)
(15.25)
(11.87)
(0.00)
(0.00)
[100.00]
[31.52]
[19.57]
[48.91]
[0.00]
[0.00]
7. Non-family
332
77
60
130
56
9
1.28
0.95
(42.19)
(35.32)
(50.85)
(34.30)
(96.55)
(64.29)
[100.00]
[23.19]
[18.07]
[39.16]
[16.87]
[2.71]
Difference
-0.6
-8.5**
7.3
(5-6)
(5.4)
(3.9)
(5.8)
Difference
8.8***
-6.2**
18.9***
(5-7)
(3.3)
(2.6)
(3.7)
Total
787
218
118
379
58
14
1.85
1.13
(100.00) (100.00) (100.00) (100.00) (100.00)
(100.00)
[100.00]
[27.70]
[14.99]
[48.16]
[7.37]
[1.78]
Note: The predecessor’s children structure is categorized in columns 2 to 6 in: Male only for predecessors with one or
more sons; Female only for predecessors with one or more daughters; Mixed for predecessors with both one or more
sons and one or more daughters; No children for predecessor without children; and Gender unknown for predecessors
with children whose gender was not indicated in the interview. The successions are clustered according to the following
categories: Family, if a successor is related by marriage or blood to at least one of the three persons owning more than
50% of the enterprise; Non-family, if a successor is not related to one of the three persons owning more than 50% of the
enterprise; Male family CEO is an indicator equal to one if a male family successor was chosen; Female family CEO is
an indicator equal to one if a female family successor was chosen; Nepotism includes all successions which fall either
into the strong or weak nepotism category; Strong nepotism refers to successions in which non-family CEOs were not
considered as successors; Weak nepotism includes non-strong nepotism successions in which a family successor with a
lower human capital than the average of the non-family CEO successors of the respective industry was installed; and
Family no nepotism presents family successions in which no nepotism was observed. The fraction of observations as a
percentage of the absolute amount of the observations per children structure category is displayed in parentheses. The
fraction of observation as a percentage of the absolute amount of observations: (a) per succession type (panel A), (b) per
gender (panel B), or (c) per nepotism category (panel C) is shown in square brackets. Standard errors are reported in
italics.
Variable

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

109

The patterns in table 1 suggest that family entrepreneurs seem to prefer male successors to steer the fortunes of their family enterprise. This becomes clear by inspecting
some basic numbers (panel A): When we compare family to non-family successions, the
“male children only” category has a 7.8 percentage points higher share (significant at the
five percent level) within the family category as compared to the non-family category,
while the “‘female children only” category has -5.3 percentage points lower share (significant at the five percent level) in the family category as compared to the non-family
category. Furthermore, cases where the predecessor has at least one son and one daughter
have a 15.6 percentage points higher share (significant at the one percent level) among
family successions as compared to non-family successions. Interestingly, predecessors
who have exclusively sons hand their enterprises in 64.7% to family heirs, whereas predecessors who have exclusively daughters only do so in 49.2% of the cases.
Column 7 of panel A reveals another striking insight. Seen from a biological point
of view, the ratio “male children only” to “female children only” (column 7) should be
near 1, as the likelihood of giving birth to a girl is (roughly) as high as for a son. Such
a result is in tendency observable for the non-family successions with a ratio of 1.28.
However, the ratio is 2.43 for family successions, a finding which is very unlikely to be
caused by biological randomness. This hints that the predecessors tried to obtain a son
strategically by increasing the number of children until they had a son, which leads to
reduced observations of the “female children only” category. This peculiar result alone
might be interpreted as indication of a preference of sons as successors, as the ratio “male
children only” to “female children only” is 1.85 over the total sample.
Within the family succession category (panel B) another interesting finding underlines
the preference for male successors: If the predecessor has only male children, in 93.6%
of the cases a male family CEO is installed, whereas if the predecessor has only female
children, then the decision in favor of a female family CEO is only observed in 78.6% of
cases. Furthermore, if the children structure is “mixed” a male family CEO is installed
in 78.9% of cases. Column 8 highlights another finding: statistically, the probability of
remaining within the category “female only” or “male only” reduces by each additional
child after the first child, while the chances of following the “mixed” category increase.
Thus, the ratio of “mixed” to “female only” plus “male only” (column 8) is indicative for
the number of total children (and thus the attempts to have a son). When we apply this
intuition, we observe that predecessors who install a male family CEO are more likely
to have more children (ratio 1.34) as compared to cases where a female family CEO is
installed (ratio 0.96). From a more general point of view, bringing the individual pieces
of evidence from panel A and B together, the overall picture is supportive of proposition
2.
In proposition 3 we speculate that the wish of some predecessors to install a male family heir as CEO successor might also boost nepotistic tendencies. In panel C of table 1
we advance this idea. Overall, within the “male children only” category 51.4% of the
successions entailed nepotistic traits, while within the “female children only” category
this was the case in only 33.9% of the successions.9 It is worth noting that the “male9 This finding is mirrored in the significant differences in shares within the children structure categories across the

110

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

children only” to “female children only” ratio (trying to obtain a son) increases with the
intensity of nepotism and is highest for strong nepotism (4.25, column 7, panel C), while
the same is true for the “mixed children” to “male-children only” plus “female children
only” ratio (indicative for the number of children) which is also highest for strong nepotism (1.48, column 8, panel C). We conclude that proposition 3 cannot be rejected on the
basis of these initial observations. Overall, we find that predecessors strategically tried
to obtain a son, while this behavior positively correlates with the intensity of nepotism.
However, if we observe nepotism less often among female family CEOs as table 1
suggests, then this is likely to have a positive impact on the human capital of female
family heirs. Figure 1 presents a visualization of the allowances in family CEO human
capital against the average achievable human capital across the industry dimension.

D????
?????????

Ž???????
??Ž?

???????
????????

Ž??????
????????

t?Ž??????
???????

K????
??????????

F IGURE 1. H UMAN C APITAL OF CEO S UCCESSORS PER I NDUSTRY

As we can see from figure 1, the average human capital of family successors is lower
across the industries as compared to non-family successors, which is indicative for proponepotism dimensions: The “male children only” category has an 8.8 percentage points higher share (significant at the
one percent level) among the nepotistic family succession category (row 5) when compared with the non-family category
(row 7). In contrast, the “female children only” category has a -6.2 percentage points lower share (significant at the
five percent level) among the nepotistic family successions category as compared to the non-family category (and -8.5
percentage points, significant at the five percent level, when nepotistic family successions are compared to non-nepotistic
family successions in row 6 in panel C).

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

111

sition 1.10 In figure 1 it is important to note that the discount of human capital among
female family successors is much lower in many industries. In the business service sector, female family CEO successors are even equipped with a higher endowment of human
capital than non-family successors on average. This indicates that favoritism occurs less
frequently among female CEO successors, which is weakly indicative for proposition 3
and also indicates that the predecessors seem to prefer sons, rather than daughters, as
successors.
Table 2 presents the human capital elements and the succession characteristics in more
detail.
The results in table 2 show that female family successors are in some areas equipped
with significantly higher human capital. Female family successors benefit from significantly higher age (10.9 percentage points, significant at the five percent level) and are
11.5 percentage points more often equipped with a merchant education (significant at
the five percent level) when compared to male family successors. With regard to the
other human capital proxies, the evidence is mixed and insignificant. Furthermore, on
average the predecessor stays active in a leading and shareholding position more often in
female family successions as compared to successions of male family successors, which
potentially mirrors a lack of confidence in the successor. In reading the results from
table 2 it is important to contrast the post-succession performance developments. The
average profitability in terms of profit margin reduces by -7.4% for male family successors, while female family CEO manage to increase profit margin by 14%, even though
they are subject to stronger gravitational forces towards the mean as they on average inherit more profitable enterprises in the successions year as compared to male heirs. This
difference-in-differences remains robust when controlling for industry and performance
trends, however it is not significant.11 Overall, the human capital score reports that the
share of high human capital female family successors is 11.9 percentage points higher
(significant at the five percent level) as compared to male family successors.

10 We excluded cases of strong nepotism in this figure.
11 The performance is measured between the succession year and the year 2009. This entails the advantage to automatically cancel out time-invariant firm characteristics driving performance. Profit margin (PM) might be subject to
industry trends. Hence, we introduce industry adjustments, which are calculated by subtraction of the median PM of the
accordant year and industry (at the two-digit ISIC code level) of a control group (from the Amadeus database) of 187,388
company-year observations. PM is calculated by dividing earnings before taxes (Amadeus item 33) by operating revenue
(Amadeus item 24) and a multiplication by 100. We required all industry categories in the control group to include at least
five observations per year and industry. We employ two-digit industry controls because Richard N. Clarke (1989) shows
that the difference between two-digit and four-digit SIC controls is marginal. Furthermore, we also attempt to eliminate
outliers by winsorizing the unadjusted PM values at the 0.025 level and take into account the influence of performance
trends due to pre-succession performance by introducing performance adjustments employing performance peer groups
(Barber and Lyon, 1996). We design the peer groups for the performance adjustments by dividing the industry-adjusted
values of the control group into deciles for each accounting period. By matching the industry-adjusted profit margin
of each sample firm with the accordant control group decile in the year of the succession, the relevant control group is
identified for each enterprise. The median PM of the relevant control group and year is then used as a control for the PM
observations of the sample group (see also Ahrens and Woywode, 2012).

112

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 2—S UMMARY S TATISTICS OF F EMALE AND M ALE S UCCESSION C HARACTERISTICS
Difference
Succession type
of means
Variable
Family
Female family
Male Family Non-family
(3) and (4)
(1)
(2)
(3)
(4)
(5)
Use of business plan (%)
46.8
43.5
47.3
58.5
-3.8
(2.34)
(4.79)
(2.73)
(2.64)
(5.51)
High experience - age (%)
29.0
38.0
27.1
47.9
10.9**
(2.13)
(4.69)
(2.43)
(2.68)
(5.28)
Merchant education (%)
38.0
46.3
34.8
36.7
11.5**
(2.28)
(4.82)
(2.60)
(2.58)
(5.48)
High experience - industry (%)
35.8
40.7
35.4
47.0
5.3
(2.25)
(4.75)
(2.61)
(2.68)
(5.42)
Leadership experience (%)
9.0
5.6
9.2
38.1
-3.7
(1.34)
(2.21)
(1.58)
(2.6)
(2.72)
High human capital (%)
49.7
59.3
47.3
65.3
11.9**
(2.34)
(4.75)
(2.73)
(2.55)
(5.48)
Predecessor active (%)
47.5
53.7
45.8
43.3
7.9
(shareholder & executive)
(2.34)
(4.82)
(2.72)
(2.66)
(5.54)
Unplanned succession (%)
8.1
12.0
7.2
10.4
4.9
(1.29)
(3.15)
(1.41)
(1.64)
(3.45)
Operating revenue (in mil. euros)
10.5
8.1
11.1
10.5
-3.1*
(succession year)
(0.97)
(0.98)
(1.24)
(0.83)
(1.58)
Employees (number)
78.6
68.3
80.1
86.9
-12.7*
(succession year)
(3.64)
(4.60)
(4.61)
(5.38)
(6.51)
Profit margin (PM) (%)
6.8
7.1
6.8
5.3
0.3
(succession year)
(0.47)
(1.10)
(0.53)
(0.42)
(1.22)
1 Profit margin (PM) (%)
-0.3
1.0
-0.5
0.80
1.5
(0.36)
(1.07)
(0.39)
(0.35)
(1.14)
1 Industry-adjusted PM (%)
-0.1
0.9
-0.2
0.9
1.1
(0.38)
(1.06)
(0.41)
(0.33)
(1.14)
1 Industry- and
1.4
2.9
1.1
1.7
1.8
performance-adjusted PM (%)
(0.33)
(1.02)
(0.36)
(0.32)
(1.08)
Years since succession (years)
3.6
3.8
3.6
3.4
0.2
(0.09)
(0.19)
(0.11)
(0.11)
(0.22)
Note: The table presents succession characteristics in the succession year and the subsequent performance development
between the succession year and the year 2009 (1). Successions are categorized into: Family, for successors who are
related by marriage or blood to at least one of the three persons owning more than 50% of the enterprise (column 1);
Female family, for female family successors (column 2); Male family, for male family successors (column 3); and Nonfamily, for successors who are unrelated to the three persons owning more than 50% of the enterprise (column 4). Use of
business plan is an indicator variable equal to one if a business plan was used during the succession. High experience age is an indicator equal to one if the successor’s age is higher or equal to the median age of the successors of the sample
in the succession year. Merchant education is an indicator variable equal to one if the successor studied business studies
at university or attended an university of cooperative education. High experience - industry is an indicator equal to one if
the successor’s industry experience is higher or equal to the median industry experience of the successors of the sample
in the succession year. Leadership experience is an indicator variable equal to one if the successor benefits from previous
executive board experience. High human capital of the successor is a score derived from the sum of the five elements
age, industry experience, leadership experience, merchant education, and professionalism to serve as a proxy for ability.
Predecessor active (shareholder & executive) refers to an indicator equal to one if the predecessor still holds shares and
remains operatively active in a leading position. Unplanned succession indicates if the succession was unplanned due
to death or disease of the predecessor. Profit margin (PM) is earnings before taxes divided by operating revenue in the
succession year. Operating revenue is operating revenue in the succession year. Employees is the number of employees
in the succession year. Industry-adjusted PM is PM less the median PM of the accordant year and industry (two-digit
ISIC), Industry- and performance-adjusted PM is industry-adjusted PM less the median industry-adjusted PM of the
relevant performance control group. Performance control groups are designed by sorting the industry-adjusted values of
the variables of the control group enterprises into deciles and matching the industry-adjusted values of the variable in the
sample with the accordant control group decile in the year of the succession. The median of the relevant variable in the
respective decile and year is then employed as a control. Years since succession is the time elapsed in years since the
succession. All values are displayed in 07/2009 euros. The stars display significances (Welch-Satterthwaite test) at: * ten
percent, ** five percent, and *** one percent. Standard errors are reported in parentheses.

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

113

B. Proposition Testing

We begin by testing proposition 1 and apply an ordered logistic regression model on
the observed human capital score (HCS) of the successors in order to address concerns
that allowances in successor human capital (weak nepotism) could also be driven by
variables other than family membership.12 Our linear index model reads as follows:

(1)

Y = ? i + ? f i + X i ? + ?i

Ordered logistic model (table 3)

Here, Y denotes the human capital score, ? i is the intercept, ? is the coefficient of
interest (in this case f i is the variable for the succession type, e.g. family versus nonfamily), X i is an array of controls with a vector of coefficients ?, and ?i denotes the
error term.13 In order to take into account that not all firms attract the same number and
quality of potential successors, our controls include company characteristics and industry
controls.
Our results of the ordered logit regressions mirror the intuition derived from figure
1. Overall, they highlight that family heirs are significantly more likely to have a lower
human capital score, which can be seen in columns 1 to 3 of table 3. In detail, column 3
of table 3 entails a highly significant negative coefficient for family CEOs (-0.806) even
after the inclusion of various company and industry controls. We interpret this finding
as evidence for proposition 1 and conclude that proposition 1 cannot be rejected on the
basis of these results.
However, as column 1 shows, the coefficient for the average achievable human capital (industry wide supply of CEO resources) is insignificant. This leads us to refine
proposition 4 towards the idea that successor human capital is positively related to the
average regionally achievable human capital per industry. Pursuing this thought, we split
our benchmark on the average achievable CEO human capital into four geographical regions.14 As can be observed in columns 2 and 3 of table 3, we find support for the refined
version of proposition 4: Higher levels of the successor’s human capital are significantly
more likely (coefficient 1.114, significant at the five percent level) if the average achievable CEO human capital in the enterprise’s region and industry is higher. This finding is
robust to the inclusion of various controls.
It can be observed in column 3 of table 3 that good quality of the predecessor’s succes12 We exclude all cases of strong nepotism from these regressions, as the allowances in human capital are related to the
weak nepotism category, which by definition 3 excludes cases of strong nepotism.
13 We highlight at this point that this is an abbreviation, as both logit and the probit require the effects of x to go
through the index of x?. For logit models this is Pr(Y = 1|X ) = Pr(Y ? > 0) = exp(X ?)/[1 + exp(X ?)] and for probit
models Pr(Y = 1|X ) = Pr(Y ? > 0) = 8(X ?), where 8 is the cumulative distribution function of the standard normal
distribution. In the following we use the abbreviation index model and simply refer to probit and logit.
14 We use four geographical regions as it is reasonable and convenient to split Germany into the clusters north, central
(or west), south and east, as each of these clusters has at least one large metropolitan area while at the same time the
clusters have approximately the same size. In detail, we use the German postal code to distinguish between (1) eastern
Germany with national postal codes starting with 0 and 1, (2) northern Germany with national postal codes starting with
2 and 4, (3) central Germany with national postal codes starting with 3, 5 and 6, and (4) southern Germany with national
postal codes starting with 7, 8, and 9.

114

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 3—W EAK N EPOTISM AND H UMAN C APITAL OF S UCCESSORS
Human capital score of CEO
(1)
(2)
(3)

Family CEO (Artifact test)
(4)
(5)
(6)

A. Variables of interest
Family CEO
-0.880*** -0.883*** -0.806***
(indicator)
(0.142)
(0.145)
(0.190)
Industry-wide supply of
0.644
1.196***
0.938
CEO resources (score)
(0.415)
(0.381)
(1.491)
Regional industry supply of
0.704***
1.114**
0.236
CEO resources (score)
(0.249)
(0.413)
(0.216)
B. Company level variables
Ln operating revenue
-0.018
-0.070
-0.047
-0.186**
(euros)
(0.108)
(0.071)
(0.072)
(0.078)
Employees
0.001
-0.001*
-0.001*
-0.001
(sum)
(0.001)
(0.001)
(0.001)
(0.001)
Default probability
0.153
-0.077
-0.059
-0.085
(percent)
(0.137)
(0.095)
(0.095)
(0.097)
Financing requirements
0.267
-0.392*
-0.400*
-0.366*
(indicator)
(0.284)
(0.204)
(0.207)
(0.208)
Investment delays
0.534**
-0.733***
-0.740*** -0.851***
(indicator)
(0.237)
(0.160)
(0.163)
(0.167)
Good succession planning
-0.538**
0.291*
0.232
0.225
(indicator)
(0.217)
(0.149)
(0.152)
(0.154)
C. Industry level variables
Construction
0.294
0.029
(indicator)
(0.282)
(0.399)
Business services
0.230
-0.915***
(indicator)
(0.280)
(0.302)
Consumer services
0.035
-0.682
(indicator)
(0.389)
(0.495)
Wholesale & retail trade
0.025
-0.025
(indicator)
(0.253)
(0.349)
Number of observations
694
666
469
492
469
492
Pseudo R2
0.0183
0.0210
0.0420
0.0820
0.724
0.1283
Note: The dependent variable is human capital score in columns 1-3 and family CEO successor (an indicator variable
equal to one for family CEO successors) in column 4 to 6. Human capital score is a proxy of ability derived from
the sum of the five elements age, industry experience, leadership experience, merchant education, and professionalism.
The variables of interest are: Family CEO is an indicator equal to one if the successor is related to one of the three
natural persons owning more than 50% of the enterprise by either marriage or blood, Industry-wide supply of CEO
resources is the mean human capital score per industry of non-family successors, and Regional industry supply of CEO
resources is the mean human capital score per industry and regional cluster of non-family successors. Company variables
include: Ln operating revenue is the natural logarithm of operating revenue, Employees is the sum of employees, Default
probability is derived from the Creditreform solvency index, Financing requirements equals one if unexpected financing
requirements occur during a succession, Investment delay equals one if necessary investments were delayed before the
succession, Good succession planning equals one if the predecessor’s succession planning was perceived by the successor
to be good. Industry level variables are indicators equal to one if the respective category according to the ZEW industry
classification is met (see appendix A2). The values of the regressors refer to the succession year. All euro values are
harmonized to 07/2009 euros. The stars display significances at the * ten percent, ** five percent, and *** one percent
level. The table presents estimated changes in probabilities devising a maximum-likelihood ordered logit model (columns
1-3) and a maximum likelihood probit model (columns 4-6). The values in parentheses display the standard errors.

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

115

sion planning seems to be negatively related (the coefficient is -0.538, significant at the
five percent level) to the probability of installing a successor with high human capital.
Interestingly, investment delays are significantly positively related to the successor’s human capital level (0.534, significant at the five percent level). We interpret this finding in
the following way: The leadership ability required for a successor to steer an enterprise
if a succession is well-planned is not as high as if the succession is more complicated
due to cash-in behavior by the predecessor leading to soaring and unsolved issues in the
enterprise’s foresight.
In columns 4 to 6 of table 3 we address concerns that the effect of the family variable
on the human capital score could be affected by artifacts, because the mean external
supply of human capital might reduce the likelihood of seeing a family successor and
increase the human capital of the successor. We investigate this issue using a simple
probit model on the likelihood of observing a family CEO successor including the same
array of controls. By inspecting the results, we reject the articulated concern. We observe
the exact opposite: The likelihood to see a family heir even increases with the mean of
externally available human capital across the industries (in column 4 the coefficient is
1.196, significant at the one percent level).15 From the results of this artifact test we infer
that the discount on human capital score of family heirs is explainable to a great extent
by preferences to install family heirs.16
As a next step, we test proposition 2 using a probit index model, which reads as follows:

(2)

Y = ? i + ?ci + X i ? + ?i

Probit model (table 4)

Here, Y indicates a family succession, ? i is the intercept, ? is the coefficient of interest
(here c covers the children structure of the predecessor, e.g. son(s) vs. no son(s)), X i
is an array of controls with the vector of coefficients ?, and ?i denotes the error term.
The idea is that, given the predecessor has children, the predecessor’s preference for sons
should be mirrored in a significantly higher likelihood to observe a family CEO if a son
is among the children. Thus, the regressions are restricted to predecessors with children.
The results provide a clear picture. The presence of one or more sons significantly
increases the likelihood to observe a family succession as compared to predecessors with
one or more daughters and no sons (the coefficient of the son(s) indicator in column 6
is 0.51, significant at the one percent level). This result remains robust to inclusion of
company and industry characteristics and is indicative for predecessors’ preference for
male family heirs as CEO successors which is in line with proposition 2. Furthermore, in
columns 4 to 6 of table 4 we observe that family successors are less likely in some industries. Especially in the business services sector we find a highly significant and negative
15 This significance vanishes when introducing industry controls in column 6 of table 3.
16 As another robustness check, we included profit margin as a direct performance measure into our regressions. It
turned out to have no significant effect, but introduced a selection bias due to missing observations. Thus, we excluded
this measure from our regressions and refer to default probability, investment delay and financing requirements as proxies
for firm performance and state of the firm.

116

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 4—P ROBIT: FAMILY S UCCESSION D ECISIONS OF P REDECESSORS WITH C HILDREN

Variable
A. Variables of interest
Son(s) only

(1)
0.40***
(0.14)

Daughter(s) only
Mixed children
Son(s) among children

Dependent variable: Family Succession Indicator
(2)
(3)
(4)
(5)

0.43***
(0.13)

(6)

0.52***
(0.19)
-0.40***
(0.14)
0.03
(0.11)

0.50***
(0.17)
0.42***
(0.13)

-0.52***
(0.19)
-0.02
(0.14)
0.51***
(0.16)

B. Company level variables
Ln sales
-0.21***
-0.20***
-0.21***
(Euro)
(0.08)
(0.08)
(0.08)
Employees
-0.00
-0.00
-0.00
(number)
(0.00)
(0.00)
(0.00)
Default probability
0.02
0.03
0.03
(percent)
(0.08)
(0.08)
(0.08)
Financing requirements
-0.57***
-0.57***
-0.57***
(indicator)
(0.21)
(0.21)
(0.21)
Investment delay
-0.91*** -0.92***
-0.92***
(indicator)
(0.15)
(0.15)
(0.15)
Predecessor active
-0.19
-0.19
-0.19
(indicator)
(0.13)
(0.13)
(0.13)
Unplanned succession
-0.26
-0.26
-0.26
(indicator)
(0.25)
(0.25)
(0.25)
Avg. regional human capital
-0.20
-0.20
-0.20
supply (score)
(0.29)
(0.29)
(0.29)
C. Industry level variables
Construction
-0.28
-0.28
-0.28
(indicator)
(0.20)
(0.20)
(0.20)
Business services
-0.98***
-0.98*** -0.98***
(indicator)
(0.19)
(0.19)
(0.19)
Consumer services
-0.32
-0.32
-0.33
(indicator)
(0.27)
(0.27)
(0.27)
Wholesale & retail trade
0.31
0.31
0.31
(indicator)
(0.19)
(0.19)
(0.19)
Number of observations
715
715
715
521
521
521
Pseudo R2
0.01
0.01
0.01
0.14
0.14
0.14
Note: The dependent variable is equal to one if the successor is related by marriage or blood to at least one of the
three persons owning more than 50% percent of the enterprise. Variables of interest include: Son(s) only is an indicator
for predecessors with one or more sons, Daughter(s) only is an indicator for predecessors with one or more daughters,
Mixed is an indicator for predecessors with both one or more sons and one or more daughters, and Son(s) there is an
indicator equal to one if among the predecessor’s children there is at least one son. Controls include: Ln sales, the
natural logarithm of operating revenue, Employees, is the number of employees, Default probability is derived from
the Creditreform solvency index, Financing requirements equals one if unexpected financing requirements occur during
a succession, Investment delay equals one if necessary investments were delayed before the succession, Predecessor
active is an indicator variable equal to one if the predecessor still holds shares of the enterprise and remains operatively
active in a leading position, Unplanned succession is an indicator equal to one if the successions is unplanned due to
death or heavy disease of the predecessor, Regional industry supply of CEO resources is the mean non-family human
capital score per industry and regional cluster (derived from the proxies age, industry experience, leadership experience,
merchant education, and professionalism). Industry indicators according to the ZEW industry classification are included
(see appendix A2). The values of the regressors refer to the succession year. All values are deflated to 07/2009 euros.
The stars display significances at the * ten percent, ** five percent, and *** one percent level. The table present estimated
changes in probabilities devising a probit (maximum-likelihood) model. Standard errors are reported in parentheses.

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

117

coefficient -0.98. We also find evidence that enterprise size in terms of sales reduces the
likelihood of family CEO successors (-0.21, significant at the one percent level). In addition, we observe negative coefficients of investment delay (-0.92, significant at the one
percent level) and financing requirements (-0.57, significant at the 10% level) which mirrors a “cuckoo’s egg behavior” in the sense that companies with a questionable outlook
are family-outsourced like a cuckoo’s egg by choosing non-family successors.
As a last step it remains to search for the drivers of nepotism and to address proposition
3 which postulates that the presence of male heirs plays a major role in this respect.

(3)

Y = ? i + ?ci + X i ? + ?i

Probit models (table 5)

Within the two probit index models Y indicates weak nepotism (or strong nepotism
respectively), ? i is the intercept, ? is the coefficient of interest (here ci circulates around
the children structure of the predecessor), X i is an array of controls with a vector of
coefficients ?, and ?i denotes the error term. The array of controls includes company
characteristics and industry controls. All regressions are limited to successions where
the family structure permitted a potential succession by close family heirs. The results
are presented in table 5.
The results of the direct probit regressions of table 5 are striking. They offer strong
evidence in favor of proposition 3. The presence of a son among the children of the
predecessor significantly increases the probability of weak nepotism (0.65, significant at
the one percent level, column 2) as compared to predecessors with daughters only. The
same relationship is observable for strong nepotism (0.51, significant at the ten percent
level in column 4). We conclude that proposition 3 cannot be rejected on the basis of
these results.17
From these findings we conclude that in enterprises with concentrated ownership and
control the installation of a successor is strongly affected by the preferences and behavior
of the predecessor.
C. Discussion

We point out that the regression results we found may be subject to omitted variable
issues. In particular, it would be very interesting to evaluate how the gender of the pre17 Furthermore, as can be seen from the regression in columns 1 to 4, on average we find a negative marginal effect of
size on the likelihood to observe nepotistic successions for small companies (until roughly three million euros), which is
mirrored in the strongly positive, but insignificant coefficient of “ln sales(below 15)”. This effect is in line with intuition:
When companies become too small, they might be unattractive to externals. Above three million euros, this effect vanishes
and size tends to have a null impact on the likelihood of strong nepotism. With respect to weak nepotism, columns 1 and
2 show that weak nepotism is significantly less likely (at the one percent level) in the construction (-0.51) and business
services (-0.87) sector, as compared to the manufacturing sector. Interestingly, there are no significant deviations towards
strong nepotism across the industries as all industry indicators remain insignificant and of low magnitude. Sudden and
unplanned successions, which occur due to disease or death of the predecessor, increase the likelihood of succession in
accordance with our definition of strong nepotism (0.56, significant at the five percent level). We suspect that this result
is not due to favoritism or biased succession contests, but rather due to time constraints in cases of unplanned situations,
which potentially render non-family successions void.

118

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 5—D RIVERS OF N EPOTISM - P ROBIT REGRESSIONS
Weak nepotism
(1)
(2)
A. Variables of interest
Son(s) only (indicator)
Daughter(s) and son(s) (indicator)

0.59***
(0.167)
0.69***
(0.153)

Son(s) amongst children (indicator)
B. Company level variables
Ln operating revenue (euros)
Ln sales (below 15) (indicator)
Ln sales (above 15) (euros)
Employees (sum)
Employees squared (sum)
Default probability (percent)
Unplanned (indicator)
C. Industry level variables
Construction (indicator)

Strong nepotism
(3)
(4)
0.52*
(0.298)
0.50*
(0.282)

0.65***
(0.146)
-0.10
(0.103)
3.44
(4.653)
-0.24
(0.317)
0.00
(0.002)
1.12e-6
(2.08e-6)
0.00
(0.083)
-0.31
(0.227)

-0.10
(0.103)
3.29
(4.657)
-0.23
(0.317)
0.00
(0.002)
9.98e-7
(2.08e-6)
-0.01
(0.083)
-0.30
(0.227)

0.51*
(0.274)
-0.11
(0.159)
4.50
(5.844)
-0.31
(0.399)
0.01**
(0.003)
-8.09e-6
(5.46e-6)
0.11
(0.070)
0.56**
(0.256)

-0.11
(0.159)
4.50
(5.844)
-0.31
(0.399)
0.01**
(0.003)
-8.08e-6
(5.45e-6)
0.11
(0.070)
0.56**
(0.256)

-0.51***
-0.50***
0.18
0.18
(0.167)
(0.166)
(0.250)
(0.249)
Business services (indicator)
-0.87***
-0.86***
0.02
0.02
(0.176)
(0.176)
(0.254)
(0.253)
Consumer services (indicator)
-0.33
-0.33
0.02
0.02
(0.243)
(0.244)
(0.351)
(0.351)
Wholesale & retail trade (indicator)
-0.11
-0.11
0.17
0.17
(0.161)
(0.161)
(0.240)
(0.240)
Other (indicator)
-0.34
-0.34
0.19
0.18
(0.280)
(0.280)
(0.412)
(0.412)
Observations
557
557
597
597
Pseudo R2
0.0717
0.0709
0.0673
0.0673
Note: The dependent variable is weak nepotism in columns 1 and 2 (an indicator variable equal to one if a family CEO
with human capital lower than the average human capital of non-family successors in the industry of his enterprise was
installed), and strong nepotism in columns 3 and 4 (an indicator variable equal to one if exclusively family members were
considered as successors). The variables of interest are: Son(s) only is an indicator equal to one if all the predecessor’s
children are male, Daughter(s) and son(s) is an indicator equal to one if the predecessor’s children are mixed, and Son(s)
amongst children is an indicator equal to one if among the predecessor’s children there is at least one son. Company
variables include: Ln operating revenue is the natural logarithm of operating revenue of the succession year, Ln sales
(below 15) is an indicator equal to one if sales is smaller than fifteen million euros, Ln sales (above 15) displays the effect
of Ln sales above fifteen million euros, Employees is the sum of employees, Employees squared is the sum of employees
squared, Default probability is derived from the Creditreform solvency index, and Unplanned is an indicator equal to one
if the succession was due to unplanned disease or death of the predecessor. Industry level variables are indicators equal
to one if the respective category according to the ZEW industry classification (see appendix A2) is met. The values of the
regressors refer to the succession year. All values are displayed in 07/2009 euros. The stars display significances at the *
ten percent, ** five percent, and *** one percent level. The table presents estimated changes in probabilities devising a
maximum likelihood probit model. The values in parentheses display the standard errors.

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

119

decessor affects the observed patterns.18 Furthermore, the question remains how much
of the observed pattern is due to gender preferences of the predecessor as compared to
gender differences in CEO labor supply. However, the fact that in unplanned family successions the female succession rate jumps to 35.1% is indicative that the predecessors’
preferences play a major role in the otherwise lower female succession rate.19 We believe that in particular the family patterns observed in table 1 are indicative for a gender
preference of the predecessor. Generally, it would be of value to test our propositions
using another succession sample to exclude a potential sample bias.
VI.

Conclusion

Our results provide evidence that the labor market selectivity in CEO successions in
family firms with concentrated ownership control is constrained due to gender preferences of the preceding CEO. We differentiate between strong nepotism and weak nepotism in successions and find that the occurrence of both forms is significantly driven by
the presence of sons among the predecessor’s children, which is robust to the inclusion
of various controls. Interestingly, the predecessors’ preference of male family successors
leads in turn to the observation that female family successors are equipped with better
levels of human capital as compared to male family successors. Furthermore, successions of female family successors are less often biased by nepotism as compared to male
family successions.
These results highlight that in enterprises with concentrated ownership control such as
family firms, the preferences of the predecessor, exemplia gratia preferences for family
heirs or for male successors, play a crucial role in the successor selection process. They
influence the succession contest’s outcome with respect to the human capital level of
the successor and subsequent enterprise performance. Consequently, it seems crucial to
sensitize the predecessor on his key role and responsibility regarding this aspect.
REFERENCES

Ahrens, Jan-Philipp, and Michael Woywode. 2012. “Restructuring, Human Capital,
and Enterprise Performance in CEO Successions in Family Firms.” Dissertation Chapter - Dissertation Jan-Philipp Ahrens, University of Mannheim.
Ahrens, Jan-Philipp, Sandra Gottschalk, and Michael Woywode. 2012. “Nepotism CEO Succession, Ownership and Enterprise Performance.” Dissertation Chapter - Dissertation Jan-Philipp Ahrens, University of Mannheim.
Angrist, Joshua D., and William N. Evans. 1998. “Children and Their Parents’ Labor Supply: Evidence from Exogenous Variation in Family Size.” American Economic
Review, 88(3): 450–477.
18 Unfortunately, with the data at hand we have no option to overcome these issues.
19 We argue that in unplanned successions, which occur due to the death or disease of the predecessor, the predecessor’s
preferences play a reduced role.

120

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Barber, Brad M., and John D. Lyon. 1996. “Detecting Abnormal Operating Performance: The Empirical Power and Specification of Test Statistics.” Journal of Financial
Economics, 41(3): 359–399.
Bennedsen, Morten, Kasper Meisner Nielsen, Francisco P´erez-Gonz´alez, and
Daniel Wolfenzon. 2007. “Inside the Family Firm: the Role of Families in Succession Decisions and Performance.” Quarterly Journal of Economics, 122(2): 647–691.
Ben-Porath, Yoram, and Finis Welch. 1976. “Do Sex Preferences Really Matter?”
Quarterly Journal of Economics, 90(2): 285–307.
Bertrand, Marianne, Simon Johnson, Krislert Samphantharak, and Antoinette
Schoar. 2008. “Mixing Family with Business: A Study of Thai Business Groups and
the Families Behind them.” Journal of Financial Economics, 88(3): 466–498.
Bloom, Nicholas, and John Van Reenen. 2007. “Measuring and Explaining Management Practices Across Firms and Countries.” The Quarterly Journal of Economics,
122(4): 1351–1408.
Clarke, Richard N. 1989. “SICs as Delineators of Economic Markets.” The Journal of
Business, 62(1): 17–31.
Dahl, Gordon B., and Enrico Moretti. 2004. “The Demand for Sons: Evidence from
Divorce, Fertility, and Shotgun Marriage.” NBER Working Paper Series, Working Paper
10281: 1–69.
Dahl, Gordon B., and Enrico Moretti. 2008. “The Demand for Sons.” The Review of
Economic Studies, 75(4): 1085–1120.
Demsetz, Harold. 1983. “The Structure of Ownership and the Theory of the Firm.”
Journal of Law and Economics, 26(2): 375–390.
Demsetz, Harold, and Bel´en Villalonga. 2001. “Ownership Structure and Corporate
Performance.” Journal of Corporate Finance, 7(3): 209–233.
Demsetz, Harold, and Kenneth Lehn. 1985. “The Structure of Corporate Ownership:
Causes and Consequences.” Journal of Political Economy, 93(6): 1155–1177.
Dyck, Alexander, and Luigi Zingales. 2004. “Private Benefits of Control: An International Comparison.” Journal of Finance, 59(2): 537–600.
Faccio, Mara, and Larry H.P. Lang. 2002. “The Ultimate Ownership of Western European Corporations.” Journal of Financial Economics, 65(3): 365–395.
Fama, Eugene F. 1980. “Agency Problems and the Theory of the Firm.” Journal of
Political Economy, 88(2): 288–307.
Fama, Eugene F., and Michael C. Jensen. 1983. “Separation of Ownership and Control.” Journal of Law and Economics, 26(2): 301–325.
Jensen, Michael C., and William H. Meckling. 1976. “Theory of the Firm: Managerial
Behavior, Agency Costs and Ownership Structure.” Journal of Financial Economics,
3(4): 305–360.
Johnson, Simon, Rafael La Porta, Florencio Lopez de Silanes, and Andrei Shleifer.
2000. “Tunneling.” American Economic Review, 90(2): 22–27.

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

121

Konrad, Kai A. 2009. Strategy and Dynamics in Contests. Oxford: Oxford University
Press.
Lamb, Michael E. 1997. “The Role of the Father in Child Development.” , ed.
Michael E. Lamb, Chapter The Role of the Father: An Overview, 1–36. Hoboken, NJ:
John Wiley & Sons Inc.
La Porta, Rafael, Florencio Lopez de Silanes, and Andrei Shleifer. 1999. “Corporate
Ownership Around the World.” Journal of Finance, 54(2): 471–517.
Lazear, Edward P., and Sherwin Rosen. 1981. “Rank-Order Tournaments as Optimum
Labor Contracts.” Journal of Political Economy, 89(5): 841–864.
Lundberg, Shelly, and Elaina Rose. 2002. “The Effects of Sons and Daughters on
Men’s Labor Supply and Wages.” Review of Economics and Statistics, 84(2): 251–268.
Mincer, Jacob A. 1974. Schooling, Experience, and Earnings. Columbia University
Press.
Morgan, S. Philip, Diane N. Lye, and Gretchen A. Condran. 1988. “Sons, Daughters,
and the Risk of Marital Disruption.” American Journal of Sociology, 94(1): 110–129.
Murphy, Kevin J., and J´an Z´abojn´?k. 2004. “CEO Pay and Appointments: A MarketBased Explanation for Recent Trends.” American Economic Review, 94(2): 192–196.
P´erez-Gonz´alez, Francisco. 2006. “Inherited Control and Firm Performance.” American Economic Review, 96(5): 1559–1588.
Pollak, Robert A. 1985. “A Transaction Cost Approach to Families and Households.”
Journal of Economic Literature, 23(2): 581–608.
Pollard, Michael S., and S. Philip Morgan. 2002. “Emerging Parental Gender Indifference? Sex Composition of Children and the Third Birth.” American Sociological
Review, 67(4): 600–613.
Rosen, Sherwin. 1986. “Prizes and Incentives in Elimination Tournaments.” American
Economic Review, 76(4): 701–715.
Shleifer, Andrei, and Robert W. Vishny. 1986. “Large Shareholders and Corporate
Control.” Journal of Political Economy, 94(3): 461–488.
Sonnenfeld, Jeffrey A., and Padraic L. Spence. 1989. “The Parting Patriarch of a Family Firm.” Family Business Review, 2(4): 355–375.
Strober, Myra H. 1990. “Human Capital Theory: Implications for HR Managers.” Industrial Relations, 29(2): 214–239.
Villalonga, Bel´en, and Raphael Amit. 2006. “How Do Family Ownership, Control and
Management Affect Firm Value?” Journal of Financial Economics, 80(2): 385–417.
A PPENDIX
A1. Additional Summary Statistics and Additional Tables

Table A1 presents additional summary statistics of the classification and distribution
of the observed successions.

122

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE A1—C LASSIFICATION AND D ISTRIBUTION OF O BSERVED S UCCESSIONS
ManuBusiness
Consumer Wholesale
Total
facturing Construction
services
services
& retail
Other
(1)
(2)
(3)
(4)
(5)
(6)
(7)
1. Male
625
260
88
112
47
96
22
successor
(80.7)
(83.6)
(80.0)
(84.2)
(69.1)
(78.1)
(75.9)
[100.0]
[41.6]
[14.1]
[17.9]
[7.5]
[15.4]
[3.5]
2. Female
149
51
22
21
21
7
15
successor
(19.3)
(16.4)
(20.0)
(15.8)
(30.9)
(22.0)
(24.1)
[100.0]
[34.2]
[14.8]
[14.1]
[14.1]
[18.1]
[4.7]
Total
774
311
110
133
68
123
29
(100.0)
(100.0)
(100.0)
(100.0)
(100.0)
(100.0)
(100.0)
[100.0]
[40.2]
[14.2]
[17.2]
[8.8]
[15.9]
[3.7]
3. Male family
336
155
48
32
23
65
13
successor
(75.7)
(79.5)
(69.6)
(74.4)
(67.7)
(76.5)
(72.2)
[100.0]
[46.1]
[14.3]
[9.5]
[6.9]
[19.4]
[3.8]
4. Female family
108
40
21
11
11
20
5
successor
(24.3)
(20.5)
(30.4)
(25.6)
(32.4)
(23.5)
(27.8)
[100.0]
[37.0]
[19.4]
[10.2]
[10.2]
[18.5]
[4.6]
Total
444
195
69
43
34
85
18
(100.0)
(100.0)
(100.0)
(100.0)
(100.0)
(100.0)
(100.0)
[100.0]
[43.9]
[15.5]
[9.7]
[7.7]
[19.1]
[4.0]
Note: Successions are categorized into: Male successor, for male successors, Female successor, for female successors;
Male family successor, for male successors who are related by marriage or blood to at least one of the three persons
owning more than 50% of the enterprise, and Female family successor, for female successors who are related by marriage
or blood to at least one of the three persons owning more than 50% of the enterprise. The industries are clustered using
aggregated ISIC classifications, see appendix A2. The fraction of successions as a percentage of the absolute amount
of the gross sample (column 1) or of the absolute amount of observed successions per industry cluster (columns 2-7) is
displayed in parentheses. The fraction of successions as a percentage of the absolute amount of observed successions per
succession type is shown in square brackets.
Industry

The industry classification key for the industry aggregation employed (using aggregations of the International Standard Industry Classification of All Economic Activities of
the United Nations, ISIC Rev. 3.1) is shown in table A2.

CHAPTER 3: FAMILY CHARACTERISTICS AND HUMAN CAPITAL

123

TABLE A2—I NDUSTRY C LASSIFICATION K EY

ZEW industry key
1. Manufacturing
2. Construction
3. Business services

ISIC
Rev. 3.1
code
(1)
D
F
K

ISIC
Rev. 3.1
industry description
(2)

Manufacturing
Construction
Real estate, renting and business activities
(without ISIC 70: real estate activities)
O
Other community, social and personal service activities
(only ISIC 90: sewage and refuse disposal, sanitation and
similar activities)
4. Consumer services
H
Hotels and restaurants
K
Real estate, renting and business activities
(only ISIC 70: real estate activities)
M
Education
N
Health and social work
O
Other community, social and personal service activities
(only ISIC 92: recreational, cultural and sporting activities
and ISIC 93: other service activities)
5. Wholesale & retail
G
Wholesale and retail trade; repair of motor vehicles, motorcycles and personal and household goods
6. Other
I
Transport, storage and communication
J
Financial intermediation
Note: For each aggregated industry cluster the key in form of the ISIC Rev. 3.1 code (column 1) and its description
(column 2) is reported. The ISIC industry sections (A) agriculture, hunting and forestry, (B) fishing, (C) mining, quarrying and (E) electricity gas and water supply, (L) public administration and defense & compulsory social security, (P)
activities of households, (Q) extra-territorial organizations and bodies as well as division (91) activities of membership
organizations are not included. ISIC industry categories with no observations are not displayed.

Chapter 4
Restructuring, Human Capital, and Enterprise
Performance in CEO Successions in Family Firms
By JAN -P HILIPP A HRENS AND M ICHAEL W OYWODE ?
Devising an unique data set we analyze managerial actions and their
performance impact during CEO successions in family firms. We find
that corporate change unleashes additional performance due to accumulated improvement potentials from the pre-succession period. High
human capital successors implement more changes and perform significantly better when compared to low human capital successors. Furthermore, the amount of observed changes is subject to the economic contingency and is highest in CEO successions in turnaround situations. In
particular reviews of the supplier relations, the product portfolio, and
the compensation scheme were found to significantly enhance performance.
(JEL: G30, G34, L25, M10, M51)
Keywords: CEO succession, family firms, organizational restructuring,
turnaround management, human capital, firm performance.

? Ahrens: University of Mannheim, School of Business Studies and Economics, Department of Business Studies,
L9 1-2, D-68161 Mannheim, Germany, Tel: +49-177-656-2031, (e-mail: [email protected]). Woywode:
University of Mannheim, School of Business Studies and Economics, Department of Business Studies, L9 1-2, D-68161
Mannheim, Germany, Tel: +49-62-1181-2894, (e-mail: [email protected]). Many thanks to Stefanie Ahrens, Andreas Landmann, and Robert and Margret Brownell for helpful comments. Financial support from the
Konrad-Adenauer-Stiftung e.V. is gratefully acknowledged.

125

126

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

I. Introduction

Changes in organizational leadership are pivotal moments. But in enterprises where
concentrated ownership meets ownership control, CEO changes are regularly intertwined
with ownership changes and become decisive in a company’s fate, as the CEO is often
the heart of the organization he is conducting. Provided the leadership of the departing
CEO has lasted long enough, both an internal equilibrium of well-established rules or
norms, roles, and power distribution, but also of processes and routines, as well as an
external equilibrium of economic and sociological embeddedness have developed (Cyert
and March, 1963, Nelson and Winter, 1982, and Granovetter, 1985).
By relinquishing the focus on the departing CEO and by the enthronement of a new
CEO, a company may become subject to profound reorganization and substantial change
(Miller, 1993). Furthermore, in many circumstances this change proves to be vital and
advantageous, because experience often makes organizations and leaders cling to old
routines of behavior whilst the optimal solutions to organizational challenges might have
changed over time (Miller, 1993, Beck et al., 2008). In this context the new leader can act
as a reset: He assesses the contemporary fit of the organizational equilibria and general
business model and introduces changes.
Consequently, the recent succession literature discusses the ties of successions with
change management (Pardo-del-Val, 2009). Managing momentous change is complex:
Adapting the stabilizing institutionalized fabric possibly fuels opposition from organizational members (Coch and French Jr., 1948) and inertial forces stemming from structural
and environmental embeddedness arise which hinder intended adaptations to a dynamic
environment (Hannan and Freeman, 1977). To make things worse, older organizations,
for example family firms with a long tradition, might be subject to particularly strong
inertial forces and their reorganizations are especially complex and hazardous implying a high default probability (Hannan and Freeman, 1984, Amburgey et al., 1993, and
Haveman, 1993). The work of P´erez-Gonz´alez (2006) and Ahrens et al. (2012) hints
that CEO successors with high levels of human capital seem to cope better with these
challenges. However, despite the growing literature on successions, it is surprising how
little is known in detail about the performance impact of such restructuring decisions
during successions (Miller, 1993). After all and ceteribus paribus, managerial actions
should play a key role in explaining firm performance, which is perhaps most true for
leader-centered family firms.1
This article addresses this gap and visualizes the post-succession managerial actions
and evaluates their impact on performance. For this purpose we harness an unique data
set which tracks 25 management action types and enterprise performance in 804 leadership changes in enterprises with high ownership concentration and ownership control. As the CEO successors’ ability and propensity to implement changes vary significantly across their origins (family, enterprise or external) (Helmich and Brown, 1972,
1 Family firms are possibly the most prominent example of constellations of a large shareholder with some satellite
minority shareholders, as modeled by Shleifer and Vishny (1986). In addition, ownership concentration and family
influence in medium-sized publicly traded companies is a substantial phenomenon across the world. La Porta et al.
(1999) report that 45 percent of the observed enterprises are controlled by families.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

127

and Ahrens et al., 2012), we develop a CEO successor typology with respect to origin and
ability and combine it with a contingency approach which controls for the economic conditions of the succession enterprise (normal, low relative profit margin, industry downturn, and turnaround situation). By relying on accounting-based performance indicators,
we carry out a differential performance analysis. A central advantage of the differential
performance approach is that it controls by construction for time-invariant firm characteristics which might drive firm performance. Following the financial economics literature strand on firm performance, our performance measures include industry- and
performance-adjustments of comparison groups in order to control for the effects of time
and industry trends as well as possible mean reversion effects (Barber and Lyon, 1996).
In addition and to account for further heterogeneity of the successions, we incorporate
an array of commonly used controls for firm characteristics into our regressions. These
include measures for firm size, momentum (starting profitability), ownership, default
probability as well as further indicators on the state of the firm in the succession year
including possible investment delays and financing requirements, but also sociological
controls such as the ability (human capital) and origin of the CEO and the activity of the
predecessor to control for influences of knowledge transfers.
The results of our analysis offer empirical evidence that in many family firms which
experience a CEO succession, a potential for vital corporate amendments has accumulated, which is unleashed by intensive post-succession corporate change. This change
entails a positive and significant impact on enterprise performance even after controlling
for industry and performance trends and including an array common company controls.
Successors with high human capital are capable of detecting and initiating significantly
(at the one percent level) more corporate change and achieve a 1.15 percentage points
higher (significant at the one percent level) industry- and performance-adjusted profit
margin in the post succession period as compared to successors with low human capital.
In addition, the total amount of change is subject to the origin of the successor and the
economic contingency in which the succession occurs. The highest total change is observable in the turnaround category, followed by the low relative profit margin category
of the economic contingencies.
With regard to the individual actions’ performance impact, we find that a moderate
review of the existing product portfolio seems to lead to a 1.23 percentage points higher
industry- and performance-adjusted profit margin (significant at the five percent level and
robust to the inclusion of various controls). In addition, a review of the existing suppliers (0.92, significant at the five percent level) and a review of the compensation scheme
(0.82, significant at the five percent level) seem to drive abnormal post-succession performance. Furthermore, these effects gain magnitude if the successor manages to conquer
new customer groups. Interestingly, internationalization strategies during successions
are accompanied by significantly reduced (-1.70, at the five percent level) abnormal enterprise performance.
This article is designed in the following way: Section II gives a literature overview
and is followed by the theoretical section III. Section IV gives detailed information on
the sample. Section V is dedicated to the data analysis. The final section VI discusses

128

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

the main results and concludes.
II. Related Literature

The determining factors of success in successions in family firms are considered in a
variety of intertwined literature strands. However, the perspective taken by this article
mainly relates to three disciplines: the theory of the firm, the firm performance literature,
and the change and business restructuring literature.
As successions in family firms often include the replacement of an old unity of ownership and control by a new unity of ownership and control, the topic is closely related to
the theory of the firm literature. Several ownership structure and agency issues (Jensen
and Meckling, 1976, Fama, 1980, Demsetz and Lehn, 1985, Demsetz and Villalonga,
2001, and Schulze et al., 2001) are touched upon. As concentrated ownership and ownership control structures in family firms possibly curb the surveillance mechanisms of the
old CEO’s actions to zero or only internal checks, a possible pursuit of private benefit
maximization could potentially endanger firm performance around CEO and ownership
changes (Johnson et al., 2000, Dyck and Zingales, 2004, and Villalonga and Amit, 2006).
For example, firm performance might be adversely affected because of favoritism, exempli gratia through the appointment of family CEO successors with inadequate abilities
(P´erez-Gonz´alez, 2006).
Furthermore, the classical agency costs issue (Arrow, 1963, Holmstr¨om, 1979, and
Arrow, 1985) becomes relevant: With respect to firm performance, is it better to install
an external or a family member as CEO successor? From a theoretical point of view, over
the generations the endowment of talents, including leadership and managerial talents, is
likely to return to the population’s mean (Galton, 1886, Galton, 1890, Heckman, 1995,
Mulligan, 1999, and Mehrotra et al., 2011), while inheriting great amounts of wealth
might possibly induce lethargic and contemplative behavior (Carnegie, 1889, Carnegie,
1891/1933(reprint), and Holtz-Eakin et al., 1993). Nevertheless, it is argued that the
installation of family managers is vital due to the reduced agency costs and improved
monitoring (Fama and Jensen, 1983), stronger family peer pressure (Kandel and Lazear,
1992), pro-organizational behavior (Davis et al., 1997) and long-term orientation (Le
Breton-Miller and Miller, 2006, and Block and Thams, 2007).
In line with the above divergent considerations, the general empirical evaluation of
family influence and firm performance is mixed.2 However, with respect to firm performance in the succession phase, the literature is fairly unidirectional and observes a
negative impact of family CEO successors on firm performance, which has so far been
explained by a lower human capital of family heirs (possibly due to nepotism), a lower
propensity to implement changes of family heirs and the effects of the “Carnegie Conjecture” (Smith and Amoako-Adu, 1999, P´erez-Gonz´alez, 2006, Bennedsen et al., 2007,
2 Positive relationships using various measures are reported by McConaughy et al. (1998), Anderson and Reeb (2003),
Maury (2006), Adams et al. (2009), and Mehrotra et al. (2011). Mixed evidence is provided by Morck et al. (1988),
Villalonga and Amit (2006), and Miller et al. (2007). Furthermore, negative relationships were observed by Yermack
(1996), Morck et al. (2000), Hillier and McColgan (2009). For a detailed literature overview and descriptions of the
measures employed we refer to P´erez-Gonz´alez (2006).

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

129

and Ahrens et al., 2012). In addition, many succession conditions are complex. They
often constitute a phase of organizational change or (r)evolution and sometimes even
organizational crisis, thus the decision making in such situations (Smart and Vertinsky,
1977, and Gladstein and Reilly, 1985) is non-trivial and the quality of the new CEOs’
decisions may depend on their experience and human capital (Canella Jr. and Rowe,
1995, Adner and Helfat, 2003, Holcomb et al., 2009).
In the spirit of Hannan and Freeman (1984) inertial forces due to established power
structures and procedures, or due to a company’s history and experiences or even anxious
ignorance (Hedberg et al., 1976, Nystrom and Starbuck, 1984, Miller, 1990 and 1991,
Schein, 1993, and Miller, 1994), could have potentially desensitized a company to necessary adaptations. Just as tents are sometimes better than palaces, a new CEO who has
fewer links to the corporate’s past experience evaluates the firm’s issues from a different
perspective. He or she has fewer commitments to past policies and organizational rules
and is thus in a position to act as a catalyst for productive and adaptive change (Grinyer
and Spender, 1979, Hofer, 1980, Tushman and Romanelli, 1985, Miller, 1991 and 1993,
and Barker III and Duhaime, 1997) and may spark a new phase of organizational evolution, if not revolution (Greiner, 1972, and Miller, 1980). The amount of change in
successions is often substantial (Miller, 1993), sometimes even comparable to profound
restructuring situations. However, the succession literature has so far (to the best of our
knowledge) mainly ignored addressing this post-succession change and its performance
impact.
Similar to restructuring and turnaround activities, the change which unfolds in the postsuccession renewal processes can be categorized into strategic adaptations (Barker III
and Duhaime, 1997) and efficiency-enhancing changes (Hambrick and Schecter, 1983,
and Robbins and Pearce II, 1992).3 As a consequence, Barker III and Duhaime (1997)
suggest the enrichment of financial data with detailed field data on managerial actions,
as the potential for inference from financial data on managerial decision making is very
limited.4 Furthermore, findings of the responsible restructuring literature (Cascio, 2005)
play a key role because the small and medium sized enterprises observed here cannot
restructure at any rate without touching critical resources or limits due to their limited
size. In line with this argument, the literature hints that strategies including a successful
new usage of existing company resources strongly contribute to organizational recovery
(Morrow et al., 2007). On such a firm size level a successful re-adaptation may involve,
in the spirit of Hammer and Champy (2003), the combination of several jobs into one
without damaging the functional core of the organization.
III. Theoretical Section
A. Categories and Definitions

The aim of this article is to investigate the managerial actions and their performance
impact following CEO successions in enterprises with “concentrated ownership control”.
3 For a literature overview including early works on turnaround situations we refer to Pearce II and Robbins (1993).
4 In this article, we track 25 different managerial actions in order to address this issue.

130

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

For the purposes of this article we begin by defining the “concentrated ownership control” attribute:
DEFINITION 1: An enterprise with concentrated ownership control is present if a maximum of three natural persons own more than 50% of the enterprise and at least one of
these owners is a leading member of the executive board.
Furthermore, we classify the successions along the dimensions “origin” of the successor, “human capital” of the successor, and “economic contingency” of the enterprise in
the succession year, because we suspect that these dimensions influence the pattern of
the post-succession managerial actions and thereby post-succession performance.
The origin dimension distinguishes between the following origins: family, for successors related by marriage or blood to at least one of the three persons owning more than
50% of the enterprise, enterprise, for unrelated successors who were previously employees of the enterprise, external, for successors with no previous ties to the enterprise, and
hybrid, for multiple successors with differing backgrounds.
We assess the human capital dimension by creating a human capital score as a proxy.
This proxy is composed of the sum of five elements which include (1) age above median
(proxy for general experience), (2) industry experience above median (proxy for industry
related experience), (3) leadership experience (proxy for practical managerial skills), (4)
merchant education, if the successor holds an university degree in business studies (or
strongly related field) or was educated at an university of cooperative education (proxy
for theoretical managerial skills), and (5) use of a business plan during the succession
(proxy for professional managerial skills).
In the fashion of Ahrens et al. (2012), the selection of the human capital proxies
mirrors an argument brought forward by Murphy and Z´abojn´?k (2004). Murphy and
Z´abojni´?k (2004) argue that general management skills have gained more importance
for CEOs over the last 30 years due to the advances in management science, corporate
controlling and finance and other arts, which, if mastered by the CEO, enhance the CEO’s
ability to lead an enterprise.5 Thus, we try to capture general managerial skills and
argue that leadership experience, a degree in business studies and the ability to apply
this knowledge using advanced professional instruments, such as a business plan (which
includes a stragic plan, but also a finance- and liquidity-plan and an earnings forecast),
are satisfactory proxies. The role of the successor’s education is also emphasized in the
succession literature (Morris et al., 1997 and Le Breton-Miller et al., 2004). Further,
the small business literature emphasizes the beneficial aspects of education in business
studies and managerial skills (exemplia gratia cash-flow management) for small business
success (Ibrahim and Goodwin, 1986). In addition, education may also be interpreted as
a signal of ability, since the investment “degree in business studies” is easier to obtain
for more gifted individuals (Spence, 1973, Arrow, 1973, and Spence 1974).6
5 In addition Murhpy and Z´abojn´?k (2004) highlight that the advances in computerization of the digital age over the
last 30 years have made company specific knowledge more accessible to managers as compared to earlier times.
6 In a way, the business plan indicator variable also monitors whether the general managerial skills acquired are in
active use as opposed to being only a signal.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

131

However, the human capital theory highlights that productivity-augmenting investments in human capital might also occur in the post-education period (Mincer, 1974
and Strober 1990), which we try to capture by incorporating experience related human
capital proxies. First of all, the succession literature casts light on the fact that proven
skills, such as previous leadership experience, are vital and generate credibility in successions (Barach et al., 1988, Barach and Ganitsky, 1995, Chrisman et al., 1998, and
Le Breton-Miller et al., 2004). For this reason we include leadership experience as an
investment with respect to building up CEO relevant human capital. Furthermore, Mincer (1974) shows that the time distribution of the investments in human capital leads to
an age variation in earnings resulting in a positive correlation between age and earnings
(sometimes called the age profile). As this higher payment might be seen as remuneration for higher ability due to experience, we use age as a proxy for general experience,
which might also be relevant for the ability to lead a company as a CEO.7 Finally, by
employing industry experience, we also add one company specific knowledge proxy to
the four general proxies, which mirrors potential advantages due to mastering the tricks
of the trade of the specific industry of the accordant company.8 Finally, we decided to
work with a sum of these five proxies as many combinations of the above named elements might reflect a augmented ability to succeed as a CEO. The attribute HHC for
high human capital is assigned to the successors if the overall score of the respective
successor is above or equal to the median score of the sample, and LHC for low human
capital otherwise.
The economic contingency dimension is designed to reflect insights from the restructuring and crisis literature and contains four categories which are normal successions,
low relative profit margin cases, successions during industry downturns, and turnaround
cases. Inspired by Miller (1993) and Barker III and Duhaime (1997), we employ relative
performance in the succession year as an indicator of the respective strategic alignment
of the company. More precisely, we attach the low relative profit margin attribute to a
company if it performs -1.0 percentage points in profit margin below its respective industry mean in the succession year and if it does not fall into the turnaround category.
In such circumstances, the respective company’s profit margin situation is likely to stem
from firm-based actions or inactions leading to a need for strategic adaptation (Cameron
et al., 1988).
Furthermore, companies may be forced to adapt, if their respective industry suffers
from a downturn or environmental decline due to shifts in demand size or demand shape
(Harrigan, 1980; Zammuto and Cameron, 1985; and Whetten, 1987). Such industry
contractions often lead to pressure at the firm level and sometimes also to organizational
decline, as the general profitability level declines due to survival contests in industries
which support only a reduced number of firms. We attach the industry downturn attribute
7 We have to add that Mincer (1974) highlights that this relation between age and earnings decreases over time and
eventually turns negative at old age. In addition, Miller (1991) argues that seasoned CEOs might become “stale in the
saddle” and less effective. We argue that we can ignore these diminishing or even negative effects in our study, as the
average age of the CEO successors falls into the increasing part of the relationship.
8 One could also measure the experience within the respective enterprise. However, this would exclude external CEO
successors by definition.

132

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

to a company if the following conditions are met: The average profit margin of their
respective industry is below 2.0% in the succession year and if it does not earn more
than 5.0% profit margin in the succession year and if it does not fall into the turnaround
category.
As a third category we identify companies which are either suffering from both low
relative profit margin and industry downturn attributes or which earn less than 0.5% profit
margin in the succession year. We attach the turnaround attribute to these companies as
they are likely to need major operative and strategic adaptations. All companies which
do not meet these three categories are put into the normal succession category for healthy
enterprises.9
B. Derived Propositions

Using the categories defined above, we derive propositions which are presented in the
following. To start with, many succession conditions are complex and often go along
with a phase of organizational change (Miller, 1993), sometimes even a revolution or
organizational crisis. It follows that the decision making in such crisis situations (Smart
and Vertinsky, 1977, and Gladstein and Reilly, 1985) is demanding.10 In general, successors with a higher human capital should be better at detecting the need for and implementing change and thereby outperform less experienced or less able successors (Trow,
1961, Canella Jr. and Rowe, 1995, Adner and Helfat, 2003, P´erez-Gonz´alez, 2006, and
Holcomb et al., 2009).
PROPOSITION 1: Successors with high human capital implement more changes as
compared to successors with low human capital.
We also argue that the economic contingency dimension affects the extent and type of
change. For example, companies with the low relative profit margin attribute are more
likely to have failed to adapt strategically as compared to firms in the normal succession category. However, the current fit of the firm’s strategy is likely to have an impact
on the choice of implemented changes. Companies with weak strategic positions have,
ceteribus paribus, a higher need for strategic change as compared to companies with a
strong strategic position (Barker III and Duhaime, 1997).11
Furthermore, if reductions in demand size or demand shape due to an industry downturn occur during a succession, then such industry contractions may pressurize firms
9 Furthermore, it would be interesting to capture the social contingencies such as the social system in the company
which each successor inherited upon succession to office, as this is likely to have an effect on the actions (and performance
of the actions) taken by the successor (Guest, 1962). However, this approach is unavailable with the data we have at hand.
In addition, please note that some of the definitions and categories applied in this article follow Ahrens et al. (2012),
while the human capital score employed in this article is reapplied in Ahrens et al. (2012,b) and Ahrens et al. (2012,c).
10 This should especially be the case for successions in enterprises which suffer from turnaround conditions prior to a
succession. While the literature agrees that a top management change (a succession) in such situations is often desirable,
the typical following steps (Bibeault, 1982) for a turnaround such as “evaluation” of the situation and of cash flows, the
“emergency” stage (immediate measures) and the restructuring and rebuilding of the organization (“stabilization”) require
advanced management skills.
11 Apart from the higher likelihood of strategic adaptations, a low relative profit margin might also stem from operative
deficits, thus we also expect operative changes to occur more frequently as compared to the normal successions category.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

133

into striving for efficiency enhancing changes (Harrigan, 1980, Zammuto and Cameron,
1985, and Whetten, 1987), but it also increases incentives to evade the downturn by
following a strategic reorientation (Barker III and Duhaime, 1997). However, industry
downturns are an external source of pressure, while low relative profit margin situations
hint at internal issues and are thus more likely to be amendable by actions by the management. We thus expect that the overall change in low relative profit margin successions
is higher compared to successions in the industry downturn category. In turnaround situations we expect an urgent and important need for change and that the overall extent of
changes in the turnaround category is unsurpassed by the changes in the other categories.
PROPOSITION 2: The number of changes during successions increases over the economic contingencies in the following order: normal (lowest), industry downturn, low
relative profit margin, and turnaround (highest).
We argue that in many successions a potential for improvements and adaptions has
accumulated due to the presence of organizational inertia which may be fueled by established power structures, rules and procedures, a company’s history and experiences,
and anxious ignorance (Hedberg et al., 1976, Hannan and Freeman ,1984, Nystrom and
Starbuck, 1984, Miller, 1990 and 1991, Schein, 1993, and Miller, 1994). By overcoming these barriers, a successor or new CEO may spark productive and adaptive change
(Grinyer and Spender, 1979, Hofer, 1980, Tushman and Romanelli, 1985, Miller, 1991
and 1993, and Barker III and Duhaime, 1997), which taps a company’s potential for improvements and translates it into boosted enterprise performance. However, we believe
that once the major improvement potentials have been salvaged, additional change is unlikely to entail large positive effects. Assuming that the most obvious and vital changes
are the first to be introduced, it might be possible to speak of diminishing marginal returns to additional changes.12
PROPOSITION 3: There exists a previously accumulated and limited performance improvement potential in CEO successions which is salvageable by post-succession change.
IV. Sample Selection

In this article we revisit the data set of Ahrens et al. (2012). The data set for this article relies on the following sources: (a) the Mannheim Enterprise Panel (MUP), (b) the
Bureau van Dyjk Amadeus database (Amadeus), (c) the Hoppenstedt database, (d) the
Creditreform solvency index information, (e) German Bundesbank information, (f) standardized computer-aided telephone interviews, (g) non-standardized direct interviews,
and (h) web-searches.
We begin by extracting a gross sample of owner-controlled enterprises from the MUP
database filtering for the following settings for the years 2002 to 2008:13
12 Additional change might even be disruptive only. Furthermore, it is possible that the amount of change an organization can bear at a time without causing additional negative effects is limited.
13 We chose this time horizon as MUP data from earlier than 2002 are less complete and reliable.

134

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

1) 30 to 1,000 employees, and
2) going concern, and
3) possesses the concentrated ownership control attribute.
Employing a second filter, we single out 14,250 companies which experienced a succession within this time horizon.14
We collect financial data and impute missing values by employing the following hierarchy: 1. MUP information, 2. Amadeus information, 3. Creditreform solvency index
information, 4. Hoppenstedt information, and 5. web-searches. Apart from the sample group, a control group of 187,388 company-year observations is drawn from the
Amadeus database.15 Employing Bundesbank price index information, all values are
reported in 07/2009 euros.
Devising standardized computer-aided telephone interviews, all gross sample enterprises are contacted. In addition, all interviewees suit the following criteria:
1) interviewee is a successor, and
2) interviewee is a leading member of the executive board, and
3) interviewee holds an ownership fraction of the enterprise, and
4) the succession took place between the years 2002 and 2008.16
We arrive at a net sample of 804 CEO successions. The interview spans four key
nexi which are succession type, human capital of the successor, post-succession managerial decision pattern of the successor, and enterprise performance. The first interview
nexus is designed to allow a fine distinction between succession types and includes information on the successor’s origin (e.g. family, enterprise or external). The second
information nexus (human capital) delivers observations on variables such as the successor’s age, leadership experience, industry experience and his or her highest educational
degree or qualification. Nexus three collects data on the post-succession decision pattern
by tracking the management actions executed between the succession year and the year
2009. Enterprise performance is measured using accounting variables, for example profit
margin, number of employees, and credit rating score measured in the succession year
and in 2009. In addition, indicators of the state of the company, such as unexpected postsuccession financing requirements and perceived pre-succession investment delays, flank
the set of traditional accounting variables. In addition to the 804 standardized interviews,
we record 20 successors and two experts in non-standardized in-depth interviews, which
serve as an additional, but unrepresentative, source for the interpretation of the empirical
results, especially with regard to questions which are difficult to capture in standardized
interviews.
14 Appendix A1 presents further information on the second filter.
15 The control group enterprises are required to be of a size between 30 up 1,000 employees and their accounting
data to cover the same time horizon as the sample group (2002 to 2009). Furthermore, we exclude unconsolidated sister
statements (Amadeus consolidation code U2) as well as duplicates.
16 We ask for the year when the successor became a leading member of the executive board.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

135

V. Data Analysis
A. Summary Statistics

We begin by presenting the summary statistics of the changes implemented across the
dimensions “economic contingency” and “human capital” in table 1.
It is most striking to observe the high general intensity of change during successions.17
Successions seem to unleash a great wave of corporate change, which underlines ties to
reorganization and restructuring research strands. One CEO in-depth interview partner
(18) remarked on this: “...the whole structure was in a sort of daily grind, ...now that
we were initiating changes and new approaches, ...it was like a wake-up call which went
through the enterprise.” In addition, table 1 offers a very clear and intuitive pattern: high
human capital successors implement significantly more changes than low human capital
successors (1.70, significant at the one percent level, column 7).18 It is remarkable that
all elements of the organizational structure are changed significantly more often by high
human capital successors as compared to low human capital successors. Enterprises in
the turnaround category experience the highest amount of total change, followed by the
categories low relative profit margin and industry downturn, whereas the lowest change
intensity can be found under normal conditions (columns 1 to 4). In general, the evidence
of table 1 supports the ideas advanced in propositions 1 and 2.
Within the normal category we also observe a relatively high percentage of product
innovation (39.4%) and improvement in the production methods (54.1%). This may
be explained by the circumstance that these enterprises can concentrate resources on
innovation and research & development due to their relative strength. In addition, we
find that in normal succession cases, companies successfully tried to push into national
and international markets and reduced the importance of regional markets.
Interestingly, in successions in the industry downturn category, we measure the lowest rate of dismissals of execute directors (20.9%). This may be explained by the fact
that an industry downturn is not something the management can be blamed for. Furthermore, attempts to strategically evade a shrinking general demand in the downturn
become visible: CEO successors try hardest, compared with the other categories, to find
new customers (92.5%) and also have the highest rate of product portfolio diversification
(additional products: 66.7%). This goes along with an eleven percentage points higher
rate of change within marketing and sales as compared with normal successions. In addition and compared with normal successions, higher rates of operative and efficiencyrelated changes occur in the industry downturn category, possibly in a push to become
more efficient in the face of a declining and embattled market. Arguably, this explains
and is mirrored in the higher rates of change compared to normal successions in the categories working time policy (9.0 percentage points higher), compensation scheme (6.7
percentage points higher), production (5.0 percentage points higher), and new suppliers
(9.1 percentage points higher).
17 This is inline with observations by Miller (1993).
18 Product innovation is not included in the total change score due to low observations.

136

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 1—E CONOMIC C ONTINGENCY, H UMAN C APITAL AND I MPLEMENTED C HANGES

Variable

Economic contingencies
Low rel.
Industry
Normal
PM
downturn
(1)
(2)
(3)
8.4
9.7
9.2

Turnaround
(4)
10.6

Human
capital
LHC
HHC
(5)
(6)
7.41
9.11

Diff.
LHC HHC
(7)
-1.70***

Total change
A. Labor organization (%)
New executive directors
50.2
69.5
49.2
56.9
44.4
53.2
-8.8**
Dropped executive directors
26.0
40.7
20.9
43.2
20.0
28.0
-8.0***
Flattened hierarchy
27.6
42.4
30.3
27.3
23.9
32.1
-8.2**
Steepened hierarchy
12.5
6.8
13.6
13.6
12.7
14.1
-1.4
Working time policy
29.8
45.8
38.8
45.5
28.1
39.7
-11.7***
Compensation scheme
33.6
35.6
40.3
43.2
26.4
41.3
-14.9***
B. Organizational structure (%)
Purchasing
50.0
52.5
50.7
70.5
42.0
53.1
-11.1***
Production
51.1
60.3
56.1
67.4
43.8
50.5
-6.6**
Marketing and sales
65.1
78.0
76.1
81.8
57.6
71.2
-13.5***
Personnel
51.5
64.4
55.2
55.3
44.0
55.6
-11.6***
Corporate finance & controlling
57.0
61.0
58.2
63.6
48.9
57.5
-8.7**
C. Products and innovations (%)
Product innovation
39.4
50.0
20.5
33.3
40.0
41.0
-1.0
Additional products
56.2
64.4
66.7
61.4
57.3
60.5
-3.2
Additional methods of production
54.1
50.8
55.2
61.4
47.7
53.1
-5.4
Sorting out of products (moderate)
30.5
29.3
28.4
43.2
23.6
35.0
-11.4***
Sorting out of products (heavy)
3.0
6.9
1.5
6.8
1.4
4.0
-2.6**
D. Business relations (%)
New customers
83.8
84.5
92.5
88.4
81.5
87.4
-5.9**
Loss of old customers
24.4
25.4
26.9
20.5
20.2
27.6
-7.4**
New suppliers
53.6
54.2
62.7
65.9
49.9
58.5
-8.7**
Dismissal of old suppliers
36.1
42.4
40.3
52.3
27.4
39.5
-12.1***
New bank relations
16.2
15.2
16.4
22.7
13.8
17.7
-3.9
New financiers
18.3
28.8
25.4
25.0
12.4
20.4
-8.0***
E. Geographical activity (%)
1 Regional markets
-3.8
0.0
-3.0
-6.8
-2.0
-3.5
1.5
1 National markets
3.8
-1.7
7.5
2.3
2.9
2.6
0.2
1 International markets
6.8
6.8
3.0
13.6
2.9
7.9
-5.1***
Note: The table presents changes implemented between the succession year and the year 2009. Successions are categorized into Normal, for successions in healthy enterprises which do not fall into the categories low relative profit margin
(PM), industry downturn or turnaround in the succession year; Low rel. PM, for successions in companies which earn
lower profit margins compared to their industry peers in the succession year; Industry downturn, for successions in enterprises which experience an industry downturn (low profit margins across the industry) and which do not earn more
than five percent profit margin in the succession year; and Turnaround, for successions in companies which face both low
relative profit margin and an industry downturn or which earn less than 0.5% profit margin in the succession year. Furthermore, successors are distinguished by high human capital (HHC), or low human capital (LHC) otherwise, depending
on the successor’s score on a human capital proxy (HCS) including (1) age, (2) industry experience, (3) leadership experience, (4) merchant education, and (5) professionalism. Interviewees were asked to indicate post-succession changes of
their enterprise in each of the subcategories from A to D. Each of the change indicator variables records 1 if a change is
indicated during the interview and 0 otherwise. The geographical activity of the enterprise in the year of the succession
and in 2009 is recorded in the interview. Category E presents the change in activity per geographical subcategory. The
stars display statistical significance of differences at the * ten percent, ** five percent, and *** one percent level.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

137

In contrast to the low dismissal rates in the industry downturn category, the low relative
profit margin category is accompanied by high executive staff dismissal rates (40.7%)
and the highest rate of additional appointments of executive staff (69.5%). Arguably, this
is because low relative profit margins are highly likely to stem from internal deficits of
operative or strategic nature for which the management may be held responsible. Within
the low relative profit margin category we also argue to observe stronger strategic reconfigurations, which we believe are mirrored in the highest observed rate of product
innovation (50.0%), a 12.9 percentage points higher change intensity of in the marketing
and sales function and an 8.2 percentage point higher rate of product portfolio diversification as compared with normal successions, whilst the complete product portfolio is
subject to critical assessment (heavy sorting out of products in 6.9% of the cases). We
also find a stronger commitment to close possible gaps in operative efficiency within the
low relative profit margin category, which is reflected in a higher intensity of changes in
the areas of production (8.9 percentage points higher), personnel (12.9 percentage points
higher), working time policy (16 percentage points higher) and dismissal of old suppliers
(6.3 percentage points higher) as compared to the normal category.
It is not really surprising that the turnaround category entails the highest rate of dismissal of executives (43.2%). Major attempts to reconfigure the organization and to
push for organizational efficiency become visible and are mirrored in the highest rates of
change indications, compared to all other categories, in compensation scheme (43.2%),
production (67.4%), corporate finance & controlling (63.6%), purchasing (70.5%), new
suppliers (65.9%), and a dismissal of old suppliers (52.3%). Furthermore, the product
portfolio is often subject to a thorough review (moderate/heavy sorting out of products
indicated in 43.2% / 6.8% of the turnaround successions), while product innovation is
often pursued in a more conservative way as compared to normal successions (6.1 percentage points less often). Actions to stabilize revenues are evident: 81.8% of the surviving turnaround companies indicate changes within marketing and sales, which is 16.7
percentage points more compared with normal successions. At the same time, 88.4% of
the surviving turnaround companies indicate having gained new customers and in only
20.5% of the cases losses of old customers, which is the lowest rate across the “economic
contingency” dimension.19
Table 2 offers an overview using two basic differential performance indicators which
are profit margin and number of employees. Profit margin (PM) is a straightforward
measure of operational efficiency and can serve as an intuitive benchmark for company
performance. The differential performance is measured between the succession year and
the year 2009. Although the measure is a simple difference, it entails the advantage that
it automatically cancels out time-invariant firm characteristics which might drive firm
performance. As the PM values might be subject to industry trends, we introduce industry adjustments. We also take into account the influence of potential mean reversion due
to transitory accounting components, but also the influence of performance trends due to
pre-succession performance by introducing performance adjustments using performance
19 We also perform a factor analysis to visualize structures across the managerial changes. We present the results in
appendix A2.

138

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 2—H UMAN C APITAL , E CONOMIC C ONTINGENCY AND D IFFERENTIAL P ERFORMANCE

Variable
PM (succession year)
1 PM
1 Industry-adjusted PM
1 Ind.- and performanceadjusted PM
1 Employees (%)
1 Industry-adjusted
employees (%)

All
(1)
6.12
0.21
(0.25)
0.39
(0.25)
1.54
(0.23)
20.7
(3.27)
36.7
(3.42)

Normal
(2)
9.12
-0.69
(0.37)
-0.28
(0.38)
1.86
(0.35)
17.5
(3.77)
33.9
(4.05)

Economic contingency
Low rel.
Industry
PM
downturn
(3)
(4)
1.84
3.15
1.28
0.69
(0.48)
(0.30)
2.13
0.11
(0.51)
(0.33)
1.21
0.98
(0.52)
(0.31)
16.4
25.9
(5.42)
(5.63)
19.0
45.0
(5.27)
(6.26)

Turnaround
(5)
0.02
2.90
(0.76)
2.36
(0.55)
1.08
(0.53)
9.0
(3.58)
21.4
(4.34)

Human capital
LHC
HHC
successor successor
(6)
(7)
7.55
5.17
-0.74
0.84
(0.44)
(0.29)
-0.66
1.09
(0.42)
(0.30)
0.96
1.92
(0.36)
(0.29)
15.1
24.9
(1.75)
(5.62)
31.1
41.0
(2.09)
(5.85)

1
LHCHHC
(8)
2.39***
-1.58***
(0.53)
-1.75***
(0.52)
-0.96**
(0.47)
-9.8*
(5.88)
-9.9
(6.21)

Note: The table presents the mean performance development between the succession year and the year 2009. Successions
are categorized into: Normal, for successions in healthy enterprises which do not fall into the categories low rel. profit
margin, industry downturn or turnaround in the succession year; Low rel. PM, for successions in companies which have
a lower profit margin (PM) compared to their industry peers in the succession year; Industry downturn, for successions
in enterprises which experience an industry downturn (low profit margin across the industry) and which do not earn more
than five percent profit margin in the succession year; and Turnaround, for successions in companies which face both low
relative profit margin and an industry downturn or which earn less than 0.5% profit margin in the succession year. The
human capital of successors is distinguished by Low human capital (LHC) and High human capital (HHC) depending on
the successors’ score on a human capital proxy score including (1) age, (2) industry experience, (3) leadership experience,
(4) merchant education, and (5) professionalism. The performance indicators presented are calculated via: (a) profit
margin (PM): earnings before taxes divided by operating revenue, (b) industry-adjusted variables: the subtraction of the
control group median of the respective variable in the according year and industry (two-digit ISIC) from the sample
variable, (c) industry- and performance-adjusted variables: industry-adjusted variables less the median industry-adjusted
variable of the relevant performance control group, (d) number of employees: the growth of the firm’s employees in
percent. Control groups for performance are designed by sorting the industry-adjusted values of the variables of control
group enterprises (drawn from the Amadeus database) into deciles and matching the industry-adjusted values of the
variable in the sample with the accordant Amadeus decile in the year of the succession. The median of the relevant
variable in the respective decile and year is then employed as a control. The stars display (Welch-Satterthwaite test)
significances at: * ten percent, ** five percent and *** one percent. Standard errors are reported in parentheses.

peer groups (Barber and Lyon, 1996).20
20 PM is calculated by dividing earnings before taxes (Amadeus item 33) by operating revenue (Amadeus item 24)
and a multiplication by 100. PM relies on accruals of the same accounting period, which might be seen as an advantage
compared to employing return on assets (ROA) instead. Assets might go into the ROA indicator at historic costs, while
the earnings before taxes is measured in current dollars. ROA is subject to idle or recently acquired assets, while the
downside of profit margin or return on sales is that it does not measure the efficiency of the assets (Barber and Lyon, 1996).
Furthermore, PM and number of employees are accounting-based indicators and thus mainly mirror past, but not future,
performance. The literature points out that accounting values might be influenced by over- and understatement (Barber
and Lyon, 1996). Therefore, a combination of market-based data, for example market-to-book ratios, and accountingbased data might be vital (for examples, see Villalonga and Amit, 2006, and P´erez-Gonz´alez, 2006). Ideally, one would
create a difference over a three year average before and after a succession. However, in this study these approaches are
unavailable with the data at hand because many small and medium sized enterprises are not publicly traded. The industryadjusted values are calculated by subtraction of the median PM of the accordant year and industry (at the two-digit ISIC
code level) of a control group of 187,388 company-year observations taken from Amadeus. All industry categories in the
control group are required to include at least five observations per year and industry (at the two-digit ISIC code level).
We use two-digit industry controls because Richard N. Clarke (1989) shows that the difference between two-digit and
four-digit SIC controls is marginal. In addition, we try to eliminate the influence of extremes and outliers by winsorizing
the unadjusted PM values at the 0.025 level. We design the peer groups for the performance adjustments by dividing the
industry-adjusted values of the control group (the Amadeus database enterprises) into deciles for each accounting period.
By matching the industry-adjusted variable (e.g. PM) of each sample firm with the accordant Amadeus decile in the year

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

139

To begin with, we only have limited performance observations which reduces the possibilities of performance analyses within the economic contingency dimension. We count
235 enterprises in the normal category, 59 with low relative profit margin, 67 in the
industry downturn category, and 44 turnaround enterprises. We find that the average
succession case in our sample achieves a profit margin of 6.12% in the succession year.
Performance in the succession year decreases as intended along the categories normal
(9.12%), industry downturn (3.15%), low relative profit margin (1.84%), and turnaround
(0.02%). We draw attention to the fact that these numbers are subject to a survivorship
bias, because we do not include companies which went out of business. When comparing
low human capital (LHC) and high human capital (HHC) successors, it is not surprising
to observe that on average high human capital successors accept slightly less profitable
enterprises (mean PM of 5.17% in the succession year). Enterprises with a lower performance tend also to be more difficult and challenging to manage, and might thus be
avoided by low human capital successors who on average bought or inherited enterprises
with an PM of 7.55% in the succession year. After controlling for performance and
industry trends as well as time-invariant firm effects, we find that HHC successors performed significantly better with respect to the profit margin development as compared to
LHC successors (difference-in-differences is 0.96 percentage points profit margin). Furthermore, it is interesting to see that the surviving companies in the turnaround category
(column 5) manage the restructuring without reductions in the overall workforce. We
even measure an average increase of 9.0% in workforce. This hints that the small and
medium-sized companies observed seem to value responsible restructuring techniques
(Cascio, 2005) and try to protect their workforce from layoffs, possibly because their
workforce constitutes an investment and vital asset for them which is more difficult to
replace as compared to large enterprises. However, from our data we cannot see possible
shifts from permanent workforce towards a more flexible and contract based workforce.21
B. Proposition Testing

We begin by testing propositions 1 and 2 using a simple ordinary least squares regression with Huber-White robust standard errors and with the sum of the reported management actions as the dependent variable. Results are presented in table 3.
High human capital successors carry into effect significantly more changes as compared to successors with low human capital (column 1 of table 3). In addition, the sum
of the changes is related to the economic contingency in the manner that was postulated
by proposition 2. Turnaround successions spark significantly more change as compared
to normal successions. In the reading of table 3 it is interesting to note the external CEO
seems to spark significantly more change as compared to non-external CEOs, which
of the succession, the relevant control group is identified for each enterprise. The median value of the relevant control
group and year is then used as a control for the performance observations of the sample group.
21 In addition we find that the enterprises of HHC successors also experience a 9.8 percentage points higher increase in
workforce (significant at the 10% level) as compared to LHC successors (column 8 of table 6). However, the significance
of the difference-in-differences in workforce vanishes when introducing industry adjustments. This is not astonishing at
a second glance, as one might expect that HHC successors might be expected to be potentially capable of deploying their
existing workforce more efficiently, which may lead to a lower demand for workforce in HHC successor-led companies.

140

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 3—OLS R EGRESSIONS - OVERALL CHANGE

Variable
A. Variables of interest
High human capital (HHC)

(1)

Dependent variable: sum of reported changes
(2)
(3)
(4)
(5)
(6)

1.64***
(0.43)

External CEO

Turnaround
Male successor

Default probability
Further controls
Construction (industry)
Further industry controls
Years since succession

1.56***
(0.44)
0.93**
(0.51)
1.29**
(0.52)
1.73**
(0.67)

1.37***
(0.43)
1.14**
(0.50)
1.39***
(0.52)
-

0.45
(0.59)
1.09*
(0.64)
2.05***
(0.55)

Low rel. PM

B. Control variables
Industry- and performanceadjusted PM
Industry-adjusted PM

1.71**
(0.67)

1.59***
(0.44)
0.94*
(0.52)
1.71**
(0.67)

1.51***
(0.50)

Industry downturn

(7)

0.38***
(0.11)
-0.12***
(0.04)
-0.30
(0.19)
?

0.38***
(0.11)
-0.13***
(0.04)
-0.27
(0.18)
?

-0.28
(0.17)
?

0.37***
(0.11)
-0.13***
(0.04)
-0.30
(0.19)
?

0.35***
(0.11)
-0.10**
(0.04)
-0.33*
(0.18)
?

-0.34*
(0.17)
?

-0.31*
(0.17)
?

-0.87
(0.57)
?

-1.06*
(0.58)
?

-1.04*
(0.59)
?

-0.89
(0.60)
?

-0.79
(0.57)
?

-0.87
(0.58)
?

-0.97*
(0.57)
?

0.29***
0.29***
0.25**
0.32***
0.30***
0.27***
0.25**
(0.11)
(0.10)
(0.10)
(0.10)
(0.10)
(0.10)
(0.10)
Observations
350
350
353
339
339
341
353
R-squared
0.12
0.11
0.08
0.10
0.15
0.14
0.12
Note: The dependent variable is the sum of the reported management actions in the post-successions period. Independent
variables are: High human capital (HHC) indicates if the successor’s human capital score (derived from the proxies
age, industry experience, leadership experience, merchant education, and professionalism) is greater or equal than the
median score; External CEO is an indicator for successors without previous ties to the enterprise; Low relative PM,
for successions in companies which earn lower profit margins compared to their industry peers in the succession year;
Industry downturn, for successions in enterprises which experience an industry downturn (low profit margins across the
industry) in the succession year; Turnaround, for successions in companies which face both low relative profit margin
and an industry downturn or which earn less than 0.5% profit margin in the succession year; and Male successor indicates
one if the successor is male. Controls include: Industry- and performance-adjusted PM (momentum) is the industryadjusted profit margin (PM) in the succession year less the median industry-adjusted PM of the relevant control group;
Industry-adjusted PM (momentum) is PM in the succession year less the median PM of the according year and industry
of a control group. Industry-adjusted PM values are calculated by the subtraction of a control group median (drawn from
the Amadeus database) of the accordant year and industry (two-digit ISIC) from the sample variable. Control groups for
performance are designed by sorting the industry-adjusted values of the Amadeus database enterprises (control group)
into deciles and matching the industry-adjusted PM values of the sample group with the according control group decile
in the year of the succession. The median industry-adjusted PM of the relevant control group decile and year then serves
as a control; Default probability is based on the Creditreform solvency index score of the enterprise in the year of the
succession; Ln sales (size) is the natural logarithm of operating revenue in the year of the succession; Ownership is
an indicator equal to one if the successor owned a share of the enterprise in the succession year; Predecessor active
(shareholder & executive) is an indicator variable equal to one if the predecessor still holds shares of the enterprise and
remains operatively active in a leading position; Unplanned successions is an indicator for unplanned successions due to
heavy disease or death of the predecessor; Construction (industry) and the further industry indicator variables are equal
to one if the respective category according to the ZEW industry classification (see appendix) is met; and Years since
succession reports time elapsed in years since the succession. All values are displayed in 07/2009 euros. The stars within
the table display significances at the * ten percent, ** five percent, and *** one percent level. The values in parentheses
display Huber-White robust standard errors.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

141

supports a finding discussed by Ahrens et al. (2012), while there also seems to be a
gender component: Male successors introduce significantly more changes as compared
to female successors. These results are robust to the inclusion of an array of company,
industry and time controls. Moreover, as columns 5 to 7 show, these effects are to a great
degree independent from each other. Propositions 1 and 2 are supported in a remarkable
way by the evidence visualized in table 3 and we conclude that we cannot reject them on
the basis of our results.
As a next step, we investigate the performance impact of this change (proposition 3)
employing a Huber-White robust ordinary least squares regression. In detail our model
reads as follows:

(1)

Y = ? i + ?ci + X i ? + ?i

Ordinary least squares model (table 4)

Here, Y denotes differential industry-adjusted PM (for columns 1 to 3) or differential
industry- and performance-adjusted PM (for columns 4 to 6), ? i is the intercept, ? is the
coefficient of interest (in this case of the variable sum of the changes), X i is an array
of controls with a vector of coefficients ?, and ?i denotes the error term. In line with
the recent literature on firm performance and successions we include controls for typical firm characteristics (Villalonga and Amit, 2006, P´erez-Gonz´alez, 2006, and Ahrens
et al., 2012, who use a similar model). In detail, our controls include size (operative
revenue) in the succession year, momentum (e.g. industry-adjusted and industry- and
performance-adjusted profit margin in the succession year), ownership structure in the
succession year, default probability in the succession year, and year controls. In addition, we include further controls to address the effect of family successors, investment
delays and unexpected financing requirements, and the activity of the predecessor for
possible effects of the predecessor’s tacit knowledge (Berman et al., 2002) to increase
the robustness of our results. Furthermore, we exclude four observations (?1.25%) due
to extreme values when inspecting studentized residuals, Cook’s D, DFITS and leverage
to residuals squared plots. The results are visualized in table 4.
Robust to the inclusion of various controls, the natural logarithm of the cumulative
change has a positive and significant impact on post-succession enterprise performance
(coefficient 0.93, significant at the five percent level; 1.16 for industry-adjusted), which is
observable in columns 1 and 4 of table 4. This result is indicative that in many succession
cases a potential for improvement has accumulated and that, once it is salvaged by postsuccession management changes, it unleashes additional enterprise performance. Furthermore, the logarithmic form of the relationship already hints at decreasing marginal
returns from additional changes, as the additional performance potential becomes increasingly salvaged. What this translates into on the operative level, becomes clear when
we cite some of the in-depth CEO successor interviewees, for example (1): “... since last
year, we have not even had a sales and operation plan”; (4): “If you ask me, what was
missing in the company was controlling.”; (10): “... a lot had settled around it, we had
75 employees, but we figured out that we could do the same job with 50”; “We used to
have three tax and accounting consultancies, one for our assets and the wage payments,

142

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

TABLE 4—OLS R EGRESSIONS - H UMAN C APITAL , C HANGES AND E NTERPRISE P ERFORMANCE

Variable
Ln changes
High human
capital (HHC)
Ln changes *
HHC
Industry- and performanceadjusted PM
Industry-adjusted PM

Dependent variable: differential adjusted profit margin (PM)
Ind.-adj.1PM
Ind.& perf.-adj.1PM
(1)
(2)
(3)
(4)
(5)
(6)
1.16***
1.44**
0.93**
1.14*
(0.42)
(0.61)
(0.42)
(0.61)
1.17***
3.10*
1.15***
2.78
(0.43)
(1.69)
(0.44)
(1.73)
-0.97
-0.82
(0.77)
(0.78)
-0.31**
-0.27*
-0.32**
(0.15)
(0.14)
(0.15)
-0.30*** -0.27***
-0.28***
0.05
0.06
0.06
(0.04)
(0.04)
(0.04)
(0.06)
(0.06)
(0.06)

Family-, company-,
?
?
?
?
?
?
industry- and time-controls
Observations
290
309
290
290
309
290
R-squared
0.2285
0.2224
0.2452
0.0448
0.0508
0.0642
Note: The dependent variables are the difference in industry-adjusted profit margin (PM) in columns 1 to 3 and difference in industry- and performance-adjusted PM in columns 4 to 6. The differences are calculated via: industry(and performance-) adjusted PM of the year 2009 less industry- (and performance-) adjusted PM of the succession year.
Industry-adjusted PM values are calculated by the subtraction of a control group median (drawn from the Amadeus
database) of the accordant year and industry (two-digit ISIC) from the sample variable. Control groups for performance
are designed by sorting the industry-adjusted values of the Amadeus database enterprises (control group) into deciles
and matching the industry-adjusted PM values of the sample group with the according control group decile in the year
of the succession. The median industry-adjusted PM of the relevant control group decile and year then serves as a control. The variables of interest include: Ln changes, is the natural logarithm of the sum of observed changes during the
succession; and High human capital (HHC) indicates if the successor’s human capital score (derived from the proxies
age, industry experience, leadership experience, merchant education, and professionalism) is greater or equal than the
median score. Controls are: Ln sales (size) is the natural logarithm of operating revenue in the year of the succession;
Industry- and performance-adjusted PM (momentum) is the industry-adjusted PM in the succession year less the median
industry-adjusted PM of the relevant control group; Industry-adjusted PM (momentum) is PM in the succession year less
the median PM of the according year and industry of a control group; Ownership is an indicator equal to one if the successor owned a share of the enterprise in the succession year; Default probability is based on the Creditreform solvency
index score of the enterprise in the year of the succession; Family CEO indicates if a successor is related by marriage or
blood to at least one of the three persons owning more than 50% of the enterprise; Financing requirements is an indicator
variable equal to one if severe unexpected financing requirements were encountered during the succession; Investment
delay is an indicator variable equal to one if an investment delay is observed; Predecessor active (shareholder & executive) is an indicator variable equal to one if the predecessor still holds shares of the enterprise and remains operatively
active in a leading position; and Years is the number of years since the succession until 2009. Interaction terms between
variables are displayed as variable I * variable II. The stars within the table display significances at the * ten percent, **
five percent, and *** one percent level. The values in parentheses display Huber-White robust standard errors.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

143

the next for the taxes and the third for public accounting. And still we factorized our
bills while the payroll accounting was calculated by hand.”; “...the internal organization
was really 30 to 40 years old, and it was about time to overhaul it.”; (13): “...the goods
received were booked by the incoming goods department by hand, every day hundreds
of book entries... . We have completely new logistic systems. That had an immense
impact.”; and (16): “...we experienced some skidding,... because we were clinging to
long to formerly successful structures, but whose success was individual-related.” Overall, this is strong evidence in favor of proposition 3 and we conclude that proposition 3
should not be rejected on the basis of these results.
Furthermore, table 4 shows that HHC successors generate a 1.15 percentage points
higher industry- and performance-adjusted post-succession profit margin (significant at
the one percent level, column 5 of table 4) as compared to LHC successors, which underlines results observed by Ahrens et al. (2012). This indicates that HHC successors are
potentially better at seeing and executing the potential for improvements and highlights
the crucial role of CEO human capital for enterprise performance in successions. Arguably, management actions can be interpreted as human capital “put into practice”. In
this respect it seems that there is a performance impact of managerial actions during successions. The idea that the management’s actions might be a ritual of symbolic meaning
which entails no effect on firm performance (Gamson and Scotch, 1964, Pfeffer, 1981,
and Meindl and Ehrlich, 1987) seems not to fit in this setting. However, there seem to be
more benefits from CEO human capital to enterprise performance than just the management actions we track, as can be seen in column 3 of table 4 where both the coefficient
of the changes (1.44) and the coefficient of the high human capital indicator (3.10) are
positive and significant.
In the reading of table 4 it is interesting to observe that the coefficient of the interaction term between the natural logarithm of the sum of the changes implemented and
high human capital is insignificantly negative. The finding is difficult to interpret, as
the coefficient of high human capital jumps from 1.17 (column 2) to 3.10 (column 3)
when including the interaction term. However, from this finding we suspect that HHC
successors also tackle changes which are more difficult to implement, and furthermore,
as HHC successors also implement more changes (which might suffer from decreasing
marginal returns), that there seems to be a “speed-limit” for the amount of organizational
change an organization can absorb without jeopardizing enterprise performance. One
of the CEO successors (18) of the in-depth interviews noted on this: “I said to myself:
“Man, a change here should really go faster!” But there are certain operations whose
size you have to obey or which simply take their time.”
With the results of the tables 3 and 4 in mind, one question, which is perhaps most
relevant for practitioners, is which management changes are winners with respect to
enterprise performance. Or put differently, in terms of a search for best practice, is there
something we can learn from the performance impact of the individual management
actions? In order to investigate this question, we track abnormal differential performance
due to execution of a specific managerial action using the same regression model as in
table 4, besides the variable of interest now being the respective managerial action. Due

144

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

to the great amount of changes, we restrict the presentation to significant results, which
are shown in table 5.

TABLE 5—OLS R EGRESSIONS - P ERFORMANCE I MPACT OF C HANGES

Variables
Compensation scheme
Moderate sorting
Sorting out of products
Sorting out of products2
Dismissal of
old suppliers
Internationalization
New customers

Full sample
(1)
(2)
0.82*
0.81*
(0.46)
(0.47)
1.23**
(0.50)
2.59**
(1.01)
-1.38**
(0.57)
0.92**
0.93**
(0.45)
(0.46)
-1.70**
-1.72**
(0.73)
(0.74)

Dependent variable: 1 Ind.& perf.-adj. PM
Long-run
Family=1
Interactions
(3)
(4)
(5)
(6)
(7)
0.97
1.26*
0.15
(0.63)
(0.68)
(0.50)
0.44
-1.60
(0.60)
(1.03)
3.48**
3.60**
(1.46)
(1.53)
-1.77*
-1.82**
(0.89)
(0.87)
1.54**
0.39
-1.33
(0.69)
(0.73)
(1.15)
-2.02*
-2.73***
(1.03)
(0.83)
-0.01
-0.11
(0.65)
(0.69)
2.03**
(1.00)
3.40***
(1.16)
2.92**
(1.26)
?
?
?
?
?
?
?
?
?
?

Moderate storing*
comp. scheme
New customers*
moderate sorting
New customers*
dism. of old suppliers
?
?
Controls
?
?
Years since succession
Observations
305
305
146
164
307
305
305
R-squared
0.10
0.10
0.18
0.12
0.09
0.09
0.08
Note: The dependent variable is the difference in industry- and performance-adjusted profit margin (PM). The differences
are calculated via: industry- (and performance-) adjusted PM of the year 2009 less industry- (and performance-) adjusted
PM of the succession year. Industry-adjusted PM values are calculated by the subtraction of a control group median
(drawn from the Amadeus database) of the accordant year and industry (two-digit ISIC) from the sample variable. Control
groups for performance are designed by sorting the industry-adjusted values of the Amadeus database enterprises (control
group) into deciles and matching the industry-adjusted PM values of the sample group with the according control group
decile in the year of the succession. The median industry-adjusted PM of the relevant control group decile and year then
serves as a control. The variables of interest are [Management action indicators] (for example Compensation scheme)
which are equal to one if a change was indicated in the accordant action category. Control variables include: Ln sales
(size) is the natural logarithm of operating revenue in the year of the succession; Industry- and performance-adjusted
PM (momentum) is the industry-adjusted PM in the succession year less the median industry-adjusted PM of the relevant
control group; Industry-adjusted PM (momentum) is PM in the succession year less the median PM of the according year
and industry of a control group; Ownership is an indicator equal to one if the successor owned a share of the enterprise
in the succession year; Default probability is based on the Creditreform solvency index score of the enterprise in the
year of the succession; Family indicates if a successor is related by marriage or blood to at least one of the three persons
owning more than 50% of the enterprise; Financing requirements is an indicator variable equal to one if severe unexpected
financing requirements were encountered during the succession; Investment delay is an indicator variable equal to one
if an investment delay is observed; Predecessor active (shareholder & executive) is an indicator variable equal to one if
the predecessor still holds shares of the enterprise and remains operatively active in a leading position; and Years since
succession reports the time in years since the succession. Long-run refers to successions which took place between
2002 and 2005. Interaction terms between variables are displayed as variable I * variable II. The stars within the table
display significances at the * ten percent, ** five percent, and *** one percent level. The values in parentheses display
Huber-White robust standard errors.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

145

To start with, what seems to be most fruitful with respect to inducing abnormal performance is a moderate review of the product portfolio. Enterprises in which a moderate
sorting out of products is indicated benefit from a 1.23 percentage point higher industryand performance-adjusted profit margin as compared to enterprises which do not perform this action (column 1 of table 5). Possibly an honest due diligence of a company’s
product portfolio during successions reveals pet products or products which are mainly
offered due to a company’s history rather than for performance reasons. In addition, a
positive impact on performance for enterprises which follow a selective product or market pruning strategy and refocus on their strengths is in line with the restructuring literature (Hambrick and Schecter, 1983). However, a heavy sorting out of products is rather
indicative of severe difficulties and likely to break through the speed limit of organizational change, sparking organizational resistance and being very difficult to implement.
Therefore, we would expect a heavy sorting out of products to have a negative impact
on performance and the relationship of the intensity of product sorting out and enterprise
performance to be described by an inverted u-shape. Interestingly, this is exactly what
we observe in column 2 of table 5 as can be seen in the positive coefficient of sorting out
of products (2.59, significant at the five percent level) and the negative and significant
coefficient of sorting out of products squared (-1.38, significant at the five percent level).
In addition, reviewing relations with old suppliers seems to be a worthwhile idea with
respect to inducing abnormal performance in successions. A dismissal of old suppliers
seems to be related to a significant (at the five percent level) increase of 0.92 percentage
points in industry- and performance-adjusted profit margin (column 1 of table 5). Moreover, we find that a review of the existing compensation scheme is significantly positively
(0.82, significant at the ten percent level) related to abnormal enterprise performance in
successions. It is worth noting that for companies which have to manage a succession
process it seems definitely not the time to start to pursue internationalization projects
(coefficient -1.70, significant at the five percent level). As internationalization strategies
are risky and require profound market and industry knowledge, a relatively new successor may not yet be equipped for such ventures and is perhaps best advised to focus on
national and regional markets. The above effects are robust to the inclusion of various
controls and seem to be robust in nearly all cases when subjecting them to sub-samples
such as a long-run view or restricting the sample to family successions (column 3 and
4).22
We draw attention to the fact that the positive impact of the review of the product portfolio and of previous supplier relations is amplified when the the new CEO successfully
gains access to new customer groups, as is indicated by the positive and highly significant
interaction terms new customers*moderate sorting (3.40, significant at the one percent
level) and new customers*dismissal old suppliers (2.92, significant at the five percent
level) in column 6 and 7 of table 5. Besides the large amount of change, this highlights
the vital and important aspect of keeping and acquiring new customers in the aftermath
of CEO successions in family firms.
22 Furthermore, it would be interesting to have the possibility to verify our findings using a separate data set on CEO
successions.

146

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

In general, our findings give an indication of the average performance impact of managerial actions in successions, such as a beneficial role for a review of the product portfolio, the compensation scheme and old supplier relations. At the same time, we would like
to emphasize that these findings should not be confused with rules. There are no “cooking recipes” or “blueprint solutions” available with regard to the managerial actions: a
thorough understanding of the company’s specific situation is vital for identifying the
appropriate changes.
VI.

Conclusion

Our results provide evidence that in many family firms (companies with concentrated
ownership control), which experience a CEO succession, a potential for vital corporate
amendments has accumulated which is unleashed by intensive post-succession corporate
change. The overall impact of this change on enterprise performance was found to be
positive and significant, even after controlling for industry and performance trends and
including an array of common company controls.
In line with this evidence, the crucial role of the CEO successor’s human capital for
post-succession enterprise performance becomes visible. We find that successors with
high human capital are capable of detecting and initiating significantly more corporate
change and achieve a 1.15 percentage points higher industry- and performance-adjusted
profit margin in the post succession period when compared with successors with low
human capital. These findings are robust to the inclusion of an array of controls. Furthermore, we observe that the total amount of change is subject to the origin of the
successor and is influenced by the economic contingency in which the succession occurs. The lowest total change is found in normal successions followed by successions
during an industry downturn and successions in the low relative profit margin category.
The highest amount of change is observable in the turnaround category.
We also present a performance analysis regarding the impact of individual management actions. We find that a moderate review of the existing product portfolio seems
to lead to a significant increase in differential industry- and performance-adjusted profit
margin (plus 1.23 percentage points, significant at the five percent level and robust to the
inclusion of various controls). In addition, a review of the compensation scheme (0.82,
significant at the ten percent level) and a review of the existing suppliers (0.92, significant at the five percent level) seem to drive abnormal post-succession performance. In
addition, we observe that some of these effects amplify if the successors successfully
acquires new customer groups. Interestingly, internationalization strategies during successions are observed to occur along with significantly reduced (-1.70, at the five percent
level) abnormal enterprise performance.
We emphasize that these general findings should not be confused with “blueprint rules”
and highlight the role of understanding the individual company’s situation before initiating change. When following a predecessor’s footsteps, the new CEO not only inherits
a formal position, but also finds a current internal and external equilibrium of organizational rules, norms, power structures and procedures, which are, in the abstract, nothing
other than the result of the organization’s past experiences and behavior. The question

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

147

is this: are the reasons for these organizational forms still extant, and if not, are there
better forms achievable within the limits of reasonable effort? The new CEO often has
to combine the organization’s past experience with new, possibly external, insights to
unleash the additional performance potential of his company. This approach is perhaps
best explained with an example: In 1911 Roald Amundsen and Robert Scott started rival expeditions in a race to be the first explorer and pioneer to reach the Antarctic South
Pole. As a seasoned polar veteran, Amundsen made a huge effort in his native Norway to
plan his journey and continued to subject his equipment to rigorous scrutiny and refinement during the time in his winter base camp in Antarctica’s Bay of Whales. Amundsen
employed old knowledge and experiences, exemplia gratia employing traditional Nordic
skiing and Eskimo Husky dog-sledge techniques, harnessing the sledges to Huskies used
to the vicious conditions, crafting and adapting Eskimo clothing made of thick leather
and fur capable of withstanding the lowest temperatures, and combining this knowledge
with modern, contemporary equipment, such as the 1892 invention of the Primus stove,
which was the first pressurized-burner kerosene (paraffin) stove and which had a reputation of being reliable under adverse conditions.
The British Antarctic expedition under the command of Captain Scott was equipped
with very modern motor sledges and ponies, and relied techniques which were formerly
employed by the British navy. This gallant British company of scientists, explorers and
adventurers had high morale, grit, valor and courage. When their motor sledges and
ponies failed, Scott and his men pulled the sledge themselves on their frostbitten journey
of more than 800 miles. Fighting for every mile, Scott and his company reached the pole,
but perished tragically in the unforgiving winter only a few miles away from the safety
of their base camp. In contrast, Amundsen’s team reached the pole, had no casualties
and after being en route for 99 days en route returned safely to their base camp.23
In this vein, our advice to CEO successors when searching for the “right” changes is:
adopt the techniques of the Amundsen approach, but in the spirit and the manner of the
Scott approach.
REFERENCES

Adams, Ren´ee, Heitor Almeida, and Daniel Ferreira. 2009. “Understanding the Relationship between Founder-CEOs and Firm Performance.” Journal of Empirical Finance, 16(1): 136–150.
Adner, Ron, and Constance E. Helfat. 2003. “Corporate Effects and Dynamic Managerial Capabilities.” Strategic Management Journal, 24(10): 1011–1025.
Ahrens, Jan-Philipp, Andreas Landmann, and Michael Woywode. 2012(c). “Gender Preferences in CEO Successions in Family Firms: Family Characteristics and Human Capital of their Successor.” Dissertation Chapter - Dissertation Jan-Philipp Ahrens,
University of Mannheim.
23 Today, the southernmost place on earth is named the Amundsen-Scott South Pole Scientific Station to honor both
expeditions.

148

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Ahrens, Jan-Philipp, Michael Woywode, and Jan Zybura. 2012(b). “Inside CEO Successions in Family Firms: Should predecessors stay active or cultivate roses?” Dissertation Chapter - Dissertation Jan-Philipp Ahrens, University of Mannheim.
Ahrens, Jan-Philipp, Sandra Gottschalk, and Michael Woywode. 2012. “Nepotism CEO Succession, Ownership and Enterprise Performance.” Dissertation Chapter - Dissertation Jan-Philipp Ahrens, University of Mannheim.
Amburgey, Terry L., Dawn Kelly, and William P. Barnett. 1993. “Resetting the
Clock: The Dynamics of Organizational Change and Failure.” Administrative Science
Quarterly, 38(1): 51–73.
Anderson, Ronald C., and David M. Reeb. 2003. “Founding-Family Ownership and
Firm Performance: Evidence from the S&P 500.” Journal of Finance, 58(3): 1301–
1328.
Arrow, Kenneth J. 1963. “Uncertainty and the Welfare Economics of Medical Care.”
American Economic Review, 53(5): 941–973.
Arrow, Kenneth J. 1973. “Higher Education as a Filter.” Journal of Public Economics,
2(3): 193–216.
Arrow, Kenneth J. 1985. “The Potentials and Limits of the Market in Resource Allocation.” In Issues in Contemporary Microeconomics and Welfare. ed. George R. Feiwel,
107–124. London: Macmillan.
Barach, Jeffrey A., and Jospeh B. Ganitsky. 1995. “Successful Succession in Family
Business.” Family Business Review, 8(2): 131–155.
Barach, Jeffrey A., Joseph Ganitsky, James A. Carson, and Benjamin A. Doochin.
1988. “Entry of the Next Generation: Strategic Challenge for Family Business.” Journal
of Small Business Management, 26(2): 49–56.
Barber, Brad M., and John D. Lyon. 1996. “Detecting Abnormal Operating Performance: The Empirical Power and Specification of Test Statistics.” Journal of Financial
Economics, 41(3): 359–399.
Barker III, Vincent L., and Irene M. Duhaime. 1997. “Strategic Change in the
Turnaround Process: Theory and Empirical Evidence.” Strategic Management Journal,
18(1): 13–38.
¨
Beck, Nikolaus, Josef Bruderl,
and Michael Woywode. 2008. “Momentum or deceleration? Theoretical and methodological reflections on the analysis of organizational
change.” Academy of Management Journal, 51(3): 413–435.
Bennedsen, Morten, Kasper Meisner Nielsen, Francisco P´erez-Gonz´alez, and
Daniel Wolfenzon. 2007. “Inside the Family Firm: the Role of Families in Succession Decisions and Performance.” Quarterly Journal of Economics, 122(2): 647–691.
Berman, Shawn L., Jonathan Down, and Charles W. L. Hill. 2002. “Tacit Knowledge as a Source of Competitive Advantage in the National Basketball Association.”
Academy of Management Journal, 45(1): 13–31.
Bibeault, Donald B. 1982. Corporate Turnaround: How Managers Turn Losers Into
Winners! New York: MacGraw-Hill.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

149

Block, J¨orn H., and Andreas Thams. 2007. “Long-term Orientation in Family and
Non-family Firms: A Bayesian Analysis.” SFB 649 discussion paper 2007, 059.
Cameron, Kim S., David A. Whetten, and Robert I. Sutton. 1988. “Issues in Organizational Decline.” In Readings in Organizational Decline: Frameworks, Research and
Prescriptions. ed. Kim S. Cameron and David A. Whetten and Robert I. Sutton, 3–19.
Boston: Ballinger Books.
Carnegie, Andrew. 1889. “Wealth.” North American Review, 148(391): 653–664.
Carnegie, Andrew. 1933. “The Advantages of Poverty.” In The Gospel of Wealth and
Other Timely Essays. New York: Doubleday, Doran & Co.
Cascio, Wayne F. 2005. “Strategies for Responsible Restructuring.” The Academy of
Management Executive, 19(4): 39–50.
Chrisman, James J., Jess H. Chua, and Pramodita Sharma. 1998. “Important Attributes of Successors in Family Businesses: An Exploratory Study.” Family Business
Review, 11(1): 19–34.
Clarke, Richard N. 1989. “SICs as Delineators of Economic Markets.” The Journal of
Business, 62(1): 17–31.
Coch, Lester, and John R. P. French, Jr. 1948. “Overcoming Resistance to Change.”
Human Relations, 1(4): 512–532.
Cyert, Richard Michael, and James Gardner March. 1963. A Behavioral Theory of
the Firm. New Jersey: Prentice-Hall.
Davis, James H., F. David Schoorman, and Lex Donaldson. 1997. “Towards a Stewardship Theory of Management.” Academy of Management Review, 22(1): 20–47.
Demsetz, Harold, and Bel´en Villalonga. 2001. “Ownership Structure and Corporate
Performance.” Journal of Corporate Finance, 7(3): 209–233.
Demsetz, Harold, and Kenneth Lehn. 1985. “The Structure of Corporate Ownership:
Causes and Consequences.” Journal of Political Economy, 93(6): 1155–1177.
Dyck, Alexander, and Luigi Zingales. 2004. “Private Benefits of Control: An International Comparison.” Journal of Finance, 59(2): 537–600.
Fama, Eugene F. 1980. “Agency Problems and the Theory of the Firm.” Journal of
Political Economy, 88(2): 288–307.
Fama, Eugene F., and Michael C. Jensen. 1983. “Separation of Ownership and Control.” Journal of Law and Economics, 26(2): 301–325.
Galton, Francis. 1886. “Regression Towards Mediocrity in Hereditary Stature.” Journal
of the Anthropological Institute of Great Britain and Ireland, 15: 246–263.
Galton, Francis. 1890. “Kinship and Correlation.” North American Review,
150(401): 419–431.
Gamson, William A., and Norman A. Scotch. 1964. “Scapegoating in Baseball.” American Journal of Sociology, 70(1): 69–72.
Gladstein, Deborah L., and Nora P. Reilly. 1985. “Group Decision Making under
Threat: The Tycoon Game.” The Academy of Management Journal, 28(3): 613–627.

150

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Granovetter, Mark. 1985. “Economic Action and Social Structure: The Problem of
Embeddedness.” American Journal of Sociology, 91(3): 481–510.
Greiner, Larry E. 1972. “Evolution and Revolution as Organizations Grow.” Harvard
Business Review, 50(4): 37–46.
Grinyer, Peter H., and John-Christopher Spender. 1979. “Recipes, Crises, and Adaptation in Mature Businesses.” International Studies of Management & Organization,
9(3): 113–133.
Guest, Robert H. 1962. “Managerial Succession in Complex Organizations.” American
Journal of Sociology, 68(1): 47–56.
Hambrick, Donald C., and Steven M. Schecter. 1983. “Turnaround Strategies for
Mature Industrial-Product Business Units.” The Academy of Management Journal,
26(2): 231–248.
Hammer, Michael, and James Champy. 2003. Reengineering the Corporation: A Manifesto for Business Revolution. New York: HarperBusiness Essentials.
Hannan, Michael T., and John Freeman. 1977. “The Population Ecology of Organizations.” American Journal of Sociology, 82(5): 929–964.
Hannan, Michael T., and John Freeman. 1984. “Structural Inertia and Organizational
Change.” American Sociological Review, 49(2): 149–164.
Harrigan, Kathryn R. 1980. Strategies for Declining Businesses. Lexington(Mass.):
Lexington Press.
Haveman, Heather A. 1993. “Ghosts of Managers Past: Managerial Succession and
Organizational Mortality.” Academy of Management Journal, 36(4): 864–881.
Heckman, James J. 1995. “Lessons from the Bell Curve.” Journal of Political Economy,
103(5): 1091–1120.
Hedberg, Bo L. T., Paul C. Bystrom, and William H. Starbuck. 1976. “Camping
on Seesaws: Prescriptions for a Self-Designing Organization.” Administrative Science
Quarterly, 21(1): 41–65.
Helmich, Donald L., and Warren B. Brown. 1972. “Successor Type and Organizational Change in the Corporate Enterprise.” Administrative Science Quarterly,
17(3): 371–381.
Hillier, David, and Patrick McColgan. 2009. “Firm Performance and Managerial Succession in Family Managed Firms.” Journal of Business Finance & Accounting, 36(34): 461–484.
Hofer, Charles W. 1980. “Turnaround Strategies.” Journal of Business Strategy,
1(1): 19–31.
Holcomb, Tim R., R. Michael Holmes Jr., and Brian L. Connelly. 2009. “Making the
Most of What You Have: Managerial Ability as a Source of Resource Value Creation.”
Strategic Management Journal, 30(5): 457–485.
Holmstr¨om, Bengt Robert. 1979. “Moral Hazard and Observability.” The Bell Journal
of Economics, 10(1): 74–91.
Holtz-Eakin, Douglas, David Joulfaian, and Harvey S. Rosen. 1993. “The Carnegie

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

151

Conjecture: Some Empirical Evidence.” Quarterly Journal of Economics, 108(2): 413–
435.
Ibrahim, A. Bakr, and J. R. Goodwin. 1986. “Perceived Causes of Success in Small
Business.” American Journal of Small Business, 11(2): 41–50.
Jennrich, Robert I. 2004. “Rotation to Simple Loadings Using Component Loss Functions: The Orthogonal Case.” Psychometrika, 69(2): 257–273.
Jensen, Michael C., and William H. Meckling. 1976. “Theory of the Firm: Managerial
Behavior, Agency Costs and Ownership Structure.” Journal of Financial Economics,
3(4): 305–360.
Johnson, Simon, Rafael La Porta, Florencio Lopez de Silanes, and Andrei Shleifer.
2000. “Tunneling.” American Economic Review, 90(2): 22–27.
Kandel, Eugene, and Edward P. Lazear. 1992. “Peer Pressure and Partnerships.” Journal of Political Economy, 100(4): 801–817.
La Porta, Rafael, Florencio Lopez de Silanes, and Andrei Shleifer. 1999. “Corporate
Ownership Around the World.” Journal of Finance, 54(2): 471–517.
Le Breton-Miller, Isabelle, and Danny Miller. 2006. “Why Do Some Family Businesses Out-Compete? Governance, Long-Term Orientations, and Sustainable Capability.” Entrepreneurship Theory and Practice, 30(6): 731–746.
Le Breton-Miller, Isabelle, Danny Miller, and Lloyd P. Steier. 2004. “Toward an Integrative Model of Effective FOB Succession.” Entrepreneurship Theory and Practice,
28(4): 305–328.
Levine, Charles H. 1979. “More on Cutback Management: Hard Questions for Hard
Times.” Public Administration Review, 39(2): 179–183.
Maury, Benjamin. 2006. “Family Ownership and Firm Performance: Empirical
Evidence from Western European Corporations.” Journal of Corporate Finance,
12(2): 321–341.
McConaughy, Daniel L., Michael C. Walker, Glenn V. Henderson Jr., and Chandra S. Mishra. 1998. “Founding Family Controlled Firms: Efficiency and Value.” Review of Financial Economics, 7(1): 1–19.
Mehrotra, Vikas, Randall Morck, Jungwook Shim, and Yupana Wiwattanakantang. 2011. “Adoptive Expectations: Rising Sons in Japanese Family Firms.” NBER
Working Paper No. 16874.
Meindl, James R., and Sanford B. Ehrlich. 1987. “The Romance of Leadership and
the Evaluation of Organizational Performance.” Academy of Management Journal,
30(1): 91–109.
Miller, Danny. 1990. The Icarus Paradox: How Exceptional Companies Bring About
their own Downfall. New York: HarperBusiness.
Miller, Danny. 1991. “Stale in the Saddle: CEO Tenure and the Match between Organization and Environment.” Management Science, 37(1): 34–52.
Miller, Danny. 1993. “Some Organizational Consequences of CEO Succession.” The
Academy of Management Journal, 36(3): 644–659.

152

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Miller, Danny. 1994. “What Happens after Success: The Perils of Excellence.” Journal
of Management Studies, 31(3): 325–358.
Miller, Danny, and Peter H. Friesen. 1980. “Momentum and Revolution in Organizational Adaptation.” The Academy of Management Journal, 23(4): 591–614.
Miller, Danny, Isabelle Le Breton-Miller, Richard H. Lester, and Albert A.
Cannella, Jr. 2007. “Are Family Firms Really Superior Performers?” Journal of Corporate Finance, 13(5): 829–858.
Mincer, Jacob A. 1974. Schooling, Experience, and Earnings. Columbia University
Press.
Morck, Randall, Andrei Shleifer, and Robert W. Vishny. 1988. “Management Ownership and Market Valuation: An Empirical Analysis.” Journal of Financial Economics,
20: 293–315.
Morck, Randall, David Stangeland, and Bernard Yeung. 2000. “Inherited Wealth,
Corporate Control and Economic Growth: The Canadian Disease.” In Concentrated
Corporate Ownership. National Bureau of Economic Research, Chicago: University of
Chicago Press.
Morris, Michael H., Roy O. Williams, Jeffrey A. Allen, and Ramon A. Avila. 1997.
“Correlates of Success in Family Business Transitions.” Journal of Business Venturing,
12(5): 385–401.
Morrow Jr., J. L., David G. Sirmon, Michael A. Hitt, and Tim R. Holcomb. 2007.
“Creating Value in the Face of Declining Performance: Firm Strategies and Organizational Recovery.” Strategic Management Journal, 28(3): 271–283.
Mulligan, Casey B. 1999. “Galton versus the Human Capital Approach to Inheritance.”
Journal of Political Economy, 107(S6): S184–S224.
Murphy, Kevin J., and J´an Z´abojn´?k. 2004. “CEO Pay and Appointments: A MarketBased Explanation for Recent Trends.” American Economic Review, 94(2): 192–196.
Nelson, Richard Robinson, and Sidney Graham Winter. 1982. An Evolutionary Theory of Economic Change. Cambridge (Mass.), London: Belknap Press.
Nystrom, Paul C., and William H. Starbuck. 1984. “To Avoid Organizational Crises,
Unlearn.” Organizational Dynamics, 12(4): 53–65.
Pardo-del-Val, Manuela. 2009. “Succession in Family Firms from a Multistaged Perspective.” International Entrepreneurship and Management Journal, 5(2): 165–179.
Pearce II, John A., and Keith Robbins. 1993. “Toward Improved Theory and Research
on Business Turnaround.” Journal of Management, 19(3): 613–636.
P´erez-Gonz´alez, Francisco. 2006. “Inherited Control and Firm Performance.” American Economic Review, 96(5): 1559–1588.
Pfeffer, Jeffrey. 1981. “Management as Symbolic Action: The Creation and Maintenance of Organizational Paradigms.” In Research in Organizational Behavior. Vol. 3,
ed. Barry M. Staw and Larry L. Cummings, 1–52. Greenwich(Conn.): JAI Press.
Robbins, D. Keith, and John A. Pearce II. 1992. “Turnaround: Retrenchment and Recovery.” Strategic Management Journal, 13(4): 287–309.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

153

Schein, Edgar H. 1993. “How Can Organizations Learn Faster? The Challenge of Entering the Green Room.” Sloan Management Review, 34(2): 85–92.
Schulze, William S., Michael H. Lubatkin, Richard N. Dino, and Ann K. Buchholtz.
2001. “Agency Relationships in Family Firms: Theory and Evidence.” Organization
Science, 12(2): 99–116.
Shleifer, Andrei, and Robert W. Vishny. 1986. “Large Shareholders and Corporate
Control.” Journal of Political Economy, 94(3): 461–488.
Smart, Carolyne, and Ilan Vertinsky. 1977. “Designs for Crisis Decision Units.” Administrative Science Quarterly, 22(4): 640–657.
Smith, Brian F., and Ben Amoako-Adu. 1999. “Management Succession and Financial
Performance of Family Controlled Firms.” Journal of Corporate Finance, 5(4): 341–
368.
Spence, A. Michael. 1973. “Job Market Signaling.” Quarterly Journal of Economics,
87(3): 355–374.
Spence, A. Michael. 1974. “Competitive and Optimal Responses to Signals: An Analysis of Efficiency and Distribution.” Journal of Economic Theory, 7(3): 296–332.
Strober, Myra H. 1990. “Human Capital Theory: Implications for HR Managers.” Industrial Relations, 29(2): 214–239.
Trow, Donald B. 1961. “Executive Succession in Small Companies.” Administrative
Science Quarterly, 6(2): 228–239.
Tushman, Michael L., and Elaine Romanelli. 1985. “Organizational Evolution: A
Metamorphosis Model of Convergence and Reorientation.” In Research in Organizational Behavior. Vol. 7, ed. Barry M. Staw and Larry L. Cummings, 171–222. Greenwich(Conn.): JAI Press.
Villalonga, Bel´en, and Raphael Amit. 2006. “How Do Family Ownership, Control and
Management Affect Firm Value?” Journal of Financial Economics, 80(2): 385–417.
Whetten, David A. 1987. “Organizational Growth and Decline Processes.” Annual Review of Sociology, 13: 335–358.
Yermack, David. 1996. “Higher Market Valuation of Companies with a Small Board of
Directors.” Journal of Financial Economics, 40(2): 185–211.
Zammuto, Raymond F., and Kim S. Cameron. 1985. “Environmental Decline and
Organizational Response.” In Research in Organizational Behavior, 7. Vol. 7, ed. Barry
M. Staw and Larry L. Cummings, 232–262. Greenwich(Conn.): JAI Press.
A PPENDIX
A1. Further Sample Selection Information

The second filter is designed to find enterprises where a succession event was likely to
have occurred. The filter required the following arguments to be true between the years
2002 and 2008:
1) a leading member of the executive board resigned, or

154

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

2) a new leading member of the executive board was appointed, or
3) a previous owner reduced his share, or
4) a new or previous owner (a natural person) increased his share, and
5) one of the previous owners and leading members of the executive board was at
least 55 or older.
The age requirement increases the likelihood to observe normal successions caused
by old age. The natural person requirement ensures that the observed companies keep
the “concentrated ownership control” (family firm) attribute, which excludes takeovers
by other enterprises (legal entities) from the sample. We draw attention to the fact that
we do not observe enterprises which went out of business during a succession. This
creates a survivorship bias and thus we emphasize that any results reached are of reduced holistic representativeness and limited to the selected sample. However, we believe that this bias is only a minor caveat with respect to this article’s research aims. In
addition to the second filter, the ISIC industry sections (International Standard Industry
Classification of All Economic Activities (ISIC Rev. 3.1) of the United Nations) (A)
agriculture, hunting and forestry, (B) fishing, (C) mining, quarrying and (E) electricity gas and water supply, (L) public administration and defense & compulsory social
security, (P) activities of households, (Q) extra-territorial organizations and bodies as
well as division (91) activities of membership organizations are excluded. Also, companies for which no telephone number is available in the MUP database are dropped (less
than one percent). The gross sample of 14,250 enterprises is contacted by the Center
for Evaluation and Methods (CEM) employing a standardized computer aided interview
(ZEW-Unternehmensbefragung “Generationenwechsel Mittelstand”, 2010).
A2.

Additional Tables

The industry classification key for the industry aggregation employed is shown in table A1. We categorize the successions using aggregations of the International Standard
Industry Classification of All Economic Activities (ISIC Rev. 3.1) of the United Nations.
Table A2 shows the results of a factor analysis of the management actions. Employing an unrotated plain-vanilla version (estimates of the communality are computed using
squared multiple correlates), we find one factor with an Eigenvalue greater or equal to
1.0, which we name ORGRES as it seems to represent mostly internal organizational
restructuring. This factor seems to be more prominent with high human capital successors and in turnaround situations, mirroring our previous findings. Using the principal
component factor method and employing varimax or minimum entropy rotation (Jennrich, 2004), we find a very stable pattern which includes 9 factors with an Eigenvalue
greater or equal to 1.0. These factors seem to represent: internal organization restructuring, supply-restructuring, portfolio review, heavy portfolio review, management change,
hierarchy flattening, financial restructuring, national focus, and international focus.

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE

155

As can be observed in table A3, the prominence of the factors across the various dimensions is highly heterogeneous, which is analyzed using ordinary least squared regressions
with Huber-White robust standard errors. In detail our model reads as follows:

(A1)

Y = ? i + origin ci + economic contingency di + human capital f i + ?i
Ordinary least squares model (table A3)

Here, Y denotes the respective factor, ? i is the intercept, and the independent variables include the dimensions “origin”, “economic contingency” and “human capital”. ?i
denotes the error term.
TABLE A1—I NDUSTRY C LASSIFICATION K EY

ZEW industry key
1. Manufacturing
2. Construction
3. Business services

ISIC
Rev. 3.1
code
(1)
D
F
K

ISIC
Rev. 3.1
industry description
(2)

Manufacturing
Construction
Real estate, renting and business activities
(without ISIC 70: real estate activities)
O
Other community, social and personal service activities
(only ISIC 90: sewage and refuse disposal, sanitation and
similar activities)
4. Consumer services
H
Hotels and restaurants
K
Real estate, renting and business activities
(only ISIC 70: real estate activities)
M
Education
N
Health and social work
O
Other community, social and personal service activities
(only ISIC 92: recreational, cultural and sporting activities
and ISIC 93: other service activities)
5. Wholesale & retail
G
Wholesale and retail trade; repair of motor vehicles, motorcycles and personal and household goods
6. Other
I
Transport, storage and communication
J
Financial intermediation
Note: For each aggregated industry cluster the key in form of the ISIC Rev. 3.1 code (column 1) and its description
(column 2) is reported. The ISIC industry sections (A) agriculture, hunting and forestry, (B) fishing, (C) mining, quarrying and (E) electricity gas and water supply, (L) public administration and defense & compulsory social security, (P)
activities of households, (Q) extra-territorial organizations and bodies as well as division (91) activities of membership
organizations are not included. ISIC industry categories with no observations are not displayed.

ORGRES
(1)

Plain vanilla
& unrotated
IntOrgRes
(2)

A. Labor organization
New executive directors
0.36
Dropped executive directors
0.36
Flattened hierarchy
0.32
Steepened hierarchy
0.09
Working time policy
0.39
0.43
Compensation scheme
0.36
0.40
B. Organizational structure
Purchasing
0.50
0.50
Production
0.39
Marketing and sales
0.50
0.54
Personnel
0.46
0.64
Corporate finance & controlling
0.35
0.63
C. Products and innovations
Additional methods of production
0.46
0.40
Sorting out of products (moderate)
0.29
Sorting out of products (heavy)
0.08
D. Business relations
New customers
0.39
Loss of old customers
0.25
New suppliers
0.49
Dismissal of old suppliers
0.52
New bank relations
0.22
New financiers
0.25
E. Geographical activity
1 Regional markets
-0.20
1 National markets
0.15
1 International markets
0.18
Observations
747
Sampling adequacy (KMO)
0.72
Note: Factor loadings below 0.4 are not shown in columns 2-10.

Factorname

Method

0.80
0.78

0.49
0.78

0.42

-0.49
0.80

0.74
0.74
0.73
-0.79

0.73
0.71
-0.88
0.87

Principal-component factors - orthogonal minimum entropy rotation
Supply- Portf- HPortf- Mgmt- HierFinNatRes
Rev
Rev
Cha
Flat
Res
Focus
(3)
(4)
(5)
(6)
(7)
(8)
(9)

TABLE A2—FACTOR A NALYSIS : FACTOR L OADINGS - P OST-S UCCESSION M ANAGEMENT ACTIONS

0.85

IntFocus
(10)

156
DISSERTATION JAN-PHILIPP AHRENS
DECEMBER 2012

Dependent variable
ORGRES IntOrgRes
SupplyRes
PortfRev
HPortfRev
MgmtCha
HierFlat
FinRes
NatFocus
IntFocus
Variable
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
Family
0.05
0.10
-0.08
0.16
-0.07
-0.06
0.30**
-0.28**
0.01
0.13
(0.11)
(0.12)
(0.13)
(0.13)
(0.14)
(0.13)
(0.13)
(0.14)
(0.11)
(0.12)
External
0.33***
0.24*
0.15
-0.04
0.01
-0.01
0.51***
0.01
0.54***
-0.01
(0.12)
(0.13)
(0.15)
(0.14)
(0.15)
(0.16)
(0.15)
(0.17)
(0.18)
(0.19)
High human capital
0.25***
0.19*
0.14
0.03
0.10
0.19*
-0.03
0.08
-0.17
0.22*
(0.09)
(0.10)
(0.11)
(0.11)
(0.11)
(0.11)
(0.11)
(0.11)
(0.11)
(0.12)
Industry downturn
0.07
0.07
0.10
0.13
-0.02
-0.16
-0.10
0.08
-0.04
-0.11
(0.12)
(0.15)
(0.15)
(0.15)
(0.12)
(0.15)
(0.15)
(0.16)
(0.15)
(0.15)
Low rel. PM
0.22
0.21
0.01
-0.08
0.24
0.32*
0.22
0.22
-0.23**
-0.01
(0.14)
(0.14)
(0.16)
(0.15)
(0.20)
(0.17)
(0.15)
(0.17)
(0.09)
(0.15)
Turnaround
0.38***
0.18
0.33*
0.06
0.09
0.13
0.01
0.19
-0.01
0.29
(0.13)
(0.12)
(0.18)
(0.17)
(0.21)
(0.19)
(0.19)
(0.19)
(0.21)
(0.26)
Ln sales
0.03
0.02
0.01
0.00
-0.03
0.07
0.08
-0.02
-0.07
-0.02
(0.05)
(0.05)
(0.06)
(0.06)
(0.06)
(0.06)
(0.06)
(0.06)
(0.07)
(0.06)
Construction
-0.16
0.21
-0.20
-0.37***
-0.24*
-0.12
-0.01
-0.41***
0.12
-0.12
(0.12)
(0.14)
(0.14)
(0.14)
(0.13)
(0.16)
(0.14)
(0.15)
(0.17)
(0.16)
Business services
-0.02
0.22
-0.37**
0.30*
-0.47***
-0.02
-0.03
-0.43***
0.06
0.04
(0.13)
(0.14)
(0.15)
(0.15)
(0.16)
(0.16)
(0.16)
(0.16)
(0.14)
(0.17)
Consumer services
0.41*
0.30
0.44**
-0.13
-0.44**
0.11
0.40
0.12
0.24
-0.23
(0.23)
(0.22)
(0.22)
(0.23)
(0.19)
(0.22)
(0.24)
(0.28)
(0.29)
(0.15)
Wholesale & retail
0.06
0.11
0.06
-0.01
-0.43**
-0.21
0.15
-0.06
0.18
0.00
(0.13)
(0.15)
(0.17)
(0.18)
(0.18)
(0.18)
(0.16)
(0.16)
(0.14)
(0.17)
Other
0.17
0.25
0.04
-0.02
0.11
-0.38
0.23
0.28
0.33
-0.27
(0.22)
(0.22)
(0.24)
(0.24)
(0.42)
(0.26)
(0.24)
(0.29)
(0.29)
(0.20)
Years
0.04*
0.04
-0.02
0.05*
-0.01
0.01
0.02
0.01
0.01
0.04
(0.02)
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
Observations
373
373
373
373
373
373
373
373
373
373
R-squared
0.11
0.05
0.06
0.06
0.05
0.05
0.06
0.06
0.07
0.04
Note: The table presents the prominence of the action factors across various subgroups. The values in parentheses display Huber-White robust standard errors.

TABLE A3—OLS A NALYSIS OF M ANAGEMENT ACTION FACTORS

CHAPTER 4: RESTRUCTURING, HUMAN CAPITAL, AND ENTERPRISE PERFORMANCE
157

Chapter 5
Entrepreneurship under Imperfect Institutions
By JAN -P HILIPP A HRENS AND M ICHAEL W OYWODE ?
We consider the effects of imperfect institutions on entrepreneurial
activity. Employing a microeconomic model, we argue abnormal uncertainty and transactions costs reduce profitability and volume of entrepreneurial activity while companies complying with good practices
enjoy rewards in the optimal contracts. However, applied to an international context, counterintuitive niche cases where imperfect institutions
enhance the wealth of nations occur, mirroring seemingly contrary findings in the development literature. Policies and business strategies to
escape imperfect institutions are discussed, casting light on medieval
merchant guilds or township and village enterprises and arguing for a
holistic approach when introducing policies for institutional change.
(JEL: D02, D21, D82, D86, F23, L26, O34, O43, R00)
Keywords: Entrepreneurship, Principal-Agent Model, Institutions, Institutional Change, Development Economics

? Ahrens: University of Mannheim, School of Business Studies and Economics, Department of Business Studies,
L9 1-2, D-68161 Mannheim, Germany, Tel: +49-177-656-2031, (e-mail: [email protected]). Woywode:
University of Mannheim, School of Business Studies and Economics, Department of Business Studies, L9 1-2, D-68161
Mannheim, Germany, Tel: +49-62-1181-2894, (e-mail: [email protected]). Many thanks to Stefanie Ahrens, Robert and Margret Brownell, Andreas Landmann, Niclas R¨uffer, and Robert Strohmeyer for helpful comments. Financial support from the Konrad-Adenauer-Stiftung e.V. is gratefully acknowledged. Financial support from
the Konrad-Adenauer-Stiftung e.V. and the Excellence Initiative of the German federal state government is gratefully
acknowledged.

159

160

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Commerce and manufactures gradually introduced order and good government, and
with them, the liberty and security of individuals, [...]. This, though it has been the least
observed, is by far the most important of all their effects. (Adam Smith (1776), Book III,
Chapter IV, 3.4.4)

I. Introduction

This article explores the effects of imperfect institutions on entrepreneurial activity by
considering its microfoundations using a principal-agent framework. By “microfoundations” we mean the underlying economic mechanics hidden below the surface of the observed contracting behavior and profits of a market participant. “Imperfect institutions”
are characterised by inefficiency and an incomplete supply of Dixit’s (2009) “good governance”: (1) security of property rights, (2) enforcement of contracts and (3) collective
action to remedy the effects of externalities.
Studying entrepreneurial activity under imperfect institutions is important because
markets tend to develop failures in the absence of sound institutional economic governance. Good governance and thus sound institutions are essential to Smithian capitalism
(Dixit, 2009).1 Moreover and adding to the confusion, the development literature offers
contrary findings on the role of imperfect institutions. There is some literature where imperfect institutions “sand” and others where they “grease” the wheels of growth.2 To
practitioners, these opposing observations give economists the flavour of ancient astronomers, who had observed and predicted the night sky, but could not provide reasonable explanations for their observations. However, billions of people and millions of
companies trade under imperfect institutions every day, a state which calls for advances
and further understanding in the field.
Applying results and a model of the principal-agent theory, this article offers an applied perspective arguing that possibly both views are right under certain contingencies
and discusses practical policy and management implications. Its results suggest that,
contingent on the assumptions introduced below, overall entrepreneurial activity is reduced by the various effects of an imperfect institutional design. It argues that com1 Smith (1776) writes of the effect of secured property rights: “But men in [...] defenceless state naturally content
themselves with their necessary subsistence, because to acquire more might only tempt the injustice of their oppressors.
On the contrary, when they are secure of enjoying the fruits of their industry, they naturally exert it to better their
condition, and to acquire not only the necessaries, but the conveniences and elegancies of life.” (Adam Smith (1776),
Wealth of Nations, Book III, Chapter III, 3.3.12). Whilst he writes on contracts and conflicting objective functions of two
parties: “What are the common wages of labour, depends, everywhere upon the contract [...] made between [...] two
parties, whose interest are not the same.” (Adam Smith (1776), Book I, Chapter VIII, 1.8.11). However, both property
and contracts are subject to institutions, as Hobbes (1651) notes: “If a Covenant be made, wherein neither of the parties
performe presently, but trust one another; in the condition of meer Nature, (which is a condition of Warre of every man
against every man,) upon any reasonable suspition, it is Voyd: But if there be a common Power set over them both, with
right and force sufficient to compell performance; it is not Voyd. For he that performeth first, has no assurance the other
will performe after; because the bonds of words are too weak to bridle mens ambition, avarice, anger, and other Passions,
without the feare of some coerceive Power; [...]” (Thomas Hobbes (1651), Leviathan, Chapter XIV).
2 For some different positions, see Leff (1964), Lui (1985), and M´eon and Sekkat (2005). An overview is given by
Aidt (2009).

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

161

panies complying with fair practices will gain partnership rents for not exploiting the
sub-optimal institutional set-up. Contingent on the specific parameterization, the results
of the employed principal-agent model also allow for counter-intuitive niche cases where
the wealth of a nation is enhanced by an imperfect institutional design.
For entrepreneurs, inefficient institutions are often associated with a costly hurdle,
thereby tempering entrepreneurial activity as a direct effect. But aside from this obvious
effect, for companies well acquainted with their underdeveloped local institutional setup, a leeway of opportunities to gain advantages through exploitative behavior unfolds.
This is particularly the case, when these companies are dealing with uninformed third
(foreign) companies. Examples of such exploitations include favouritism, bribery, minor
breaching of contractual agreements and moderate violations of property rights such as
patent right infringements, which are not punished by a domestic court of law. Accordingly, this article explores the impact of scarcity of “sound economic governance” with
respect to the entrepreneurial activity and profits of the parties involved.
In this article entrepreneurial activity is modeled by a principal-agent model in order
to investigate a simple trade or venture between two entrepreneurs. In order to model
the government and market failures, this article applies a private values framework.3
The model developed is best suited to situations, where a non-informed entrepreneur
or company wishes to venture in a foreign or alien market and is obliged to establish
a venture (or to trade) with a specific local partner due to his lack of choice, lack of
information or even by force of authorities. With respect to these spatial contingencies,
the real world examples of such a setting include cases of western enterprises wishing
to venture in emerging economies, interregional trade in less developed (or developing)
economies and medieval economies operating under imperfect institutions.4 Clearly,
such a contracting situation occurs countless times in many economies each day and is
therefore also relevant from a macroeconomic perspective.
The perspective of this article is of applied and policy oriented nature and constitutes
an interdisciplinary synthesis. As far as the model is concerned, it builds on the work
of numerous researchers of the principal-agent and contract theory realm, especially the
fundamental and technical work of Akerlof (1970) and Laffont and Martimort (2002).
However, its applied perspective builds a bridge to findings and observations made by
scholars of the realms of institutional economics, entrepreneurship and development economics literature, especially the work of Rodrik (2008) and Dixit (2009).
In section I the specific terminology and the model assumptions are introduced. Section II discusses the model and derives its respective optima, while section III discusses
its implications for policy advisers and for strategic management in the light of development, entrepreneurship and business policy literature. The conclusions of this article are
briefly summarized in section IV.
3 A survey and treatment of similar classes of such theoretical problems is given by Guesnerie and Laffont (1984).
For a technical exposition of the general principal-agent realm we refer to Laffont and Martimort (2002).
4 The context of Greif’s (1993) Maghribi Traders’ Coalition is a good example.

162

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

II. A Model of Entrepreneurial Activity under Imperfect Institutions
A. Terminology

The following classifications of institutions are derived for the purpose of this article
and to serve within the limits of this article. We believe that the distinctions allow us to
carve out the different economic effects more clearly.
To start with, this article’s notion of a sound institution is a “second-best institution”.
Note, that a second-best institution may still not be perfect, but as perfect as a policy
maker is currently able to design it. This embraces the idea that appropriate and sound
institutions may still diverge greatly from the ultimate best-practice economic governance. Indeed the economist’s first best state may be unattainable, since context-specific
market and government failures, especially in less developed and developing economies,
will always constrain the set of feasible institutional designs. Realizing this general principle, Rodrik (2008) coined the term “second-best” institution.
Installing optimal second-best institutions is complex and various authors have contributed research to this field. Dixit (2004) argues that self-enforcing governance arrangements are often more efficient than formal institutions in less developed countries.
Laffont and Martimort (1999) highlight the benefits of institutional separation. Furthermore, the role of informal institutions replacing formal ones for entrepreneurship is emphasized by Tonoyan et al. (2010). Acemoglu et al. (2001) point out that institutions
for the security of property rights could, from a historical perspective, be the single most
important factor for the wealth of nations. Summarizing, Rodrik (2006) drives home
the point that there are no blueprints for optimal institutions, but that country specific
economic constraints should be taken into account.
Taking these ongoing discussions into account, we refrain from specifying the constructs (and variables) which constitute an optimal institutional set-up contingent on its
environment. We assume instead that a governance optimization has already been performed by a sophisticated policy maker and refer to the result as a “second-best institution”. Furthermore, if an institution is not classifiable as “second-best institution”, it
is referred to as an “imperfect institution” within this article. In order to distinguish
within the principal-agent model between different economic effects of institutional imperfections, the class of imperfect institutions is further divided into two sub-categories:
“inefficient” and “inferior” institutions.
“Inefficient institutions” include institutions which are superfluously costly. For example, an inefficient institution may be a bureaucratic process which imposes unnecessary
transaction costs on its customers (e.g. an entrepreneur wishing to acquire a license).
The term “superfluous” indicates that a sophisticated policy maker could suppress the
inefficiencies in the given context. The costs may also be induced by externalities which
the inefficient institutions fail to internalize. Examples include insufficient organization
of “collective action” to incorporate external social costs induced by individuals and the
under-supply of public goods or social-security-nets. Broadly understood, inefficient institutions coincide with the scarcity of the third element of Dixit’s (2009) definition of
functions of good economic governance: (3) collective action.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

163

The second subcategory of imperfect institutions is that of “inferior institutions”, which
refers to institutions which superfluously forbear exploitative behavior. An example of
an inferior institution is a court of law which does not punish a minor breach of contractual agreements or violations of property rights. Further examples of inferior institutions
are institutions which allow favoritism or bribery. To a certain extent, this represents a
scarcity of the first two elements of Dixit’s (2009) definition of functions of good economic governance: (1) security of property rights and (2) enforcement of contracts. Note
that according to these definitions an institution may be inefficient as well as inferior.
B. Model Specification

The model applies the Gauß-notation for sums and the Leibniz-notation for calculus.
In detail, the model incorporates the following assumptions:
A1. Players, Behavior and Risk. There exists an entrepreneur (principal), referred to
in a Schumpeterian sense as the “initiator” (or sometimes initiator-company), and
a partner-trader (agent), referred to as the “partner” (or partner-company), both of
them being risk-neutral and profit maximizers.
A2. Initiator’s objective. The initiator wishes to venture in a market currently being
regulated by an imperfect institution. For this purpose, he considers a venture
with a partner who is active in the market being considered.5 Let 5 denote the
initiator’s profit of the venture. Furthermore, assume that ? denotes the revenue
of the venture, depending on the volume of professional entrepreneurial activity e
that is created by it.6
(1)

?(e).

Assume the revenue function of the venture to be defined as

(2)

??(e)
> 0,
?e
? ??(e)
?e
< 0,
?e
?(0) = 0,
??(0)
? +? as e ? 0.
?e

In addition, the initiator is obliged to contribute a certain amount of resources to
the venture, in order to assure the partner of the attractiveness of a co-operation
agreement. Such an input can take various forms: direct payments or investments,
5 Alternatively, it is also possible to think of a trade between a local and a stranger.
6 We shall add that we employ e as a construct for simplicity reasons. The construct e may be composed of numerous
variables, possibly in a macroeconomic context of the number of small and medium sized enterprises and their respective
turnover.

164

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

delegation of human resources or sharing of sensible knowledge. Hence, for simplicity’s sake, we model all benefits of the partner abstractly as a transfer t in
monetary units from the initiator to the partner. Hence, the objective function of
the initiator reads
(3)

5 = ?(e) ? t.

A3. Information structure. The partner has private information ? on the inferior institutions in his market. This private information allows the partner to exploit the
venture to an extent depending on its knowledge of the inferiority. For example,
the partner has private information with regard to which person can be bribed in
order to commit a non-punishable larceny of knowledge within a venture. It is
public knowledge, that ? is randomly drawn from the interval
(4)

9 = {?, ..., ?}.

with probability density function h(?), which is identically for each partner. H (?)
denotes the respective cumulative distribution function. H (?) is absolutely continuous and differentiable.
A4. Exploitative behavior and preferences. To the extent of the partner’s possibilities due to his information draw ?, the partner executes a bribe (for instance to
be privileged in a contract distribution by the institution or to commit a larceny
of the initiator’s knowledge without punishment) which will result in costs for the
initiator. A well informed partner ? can carry out a substantial bribe, while an
uninformed partner ? cannot bribe. The costs of bribing or exploiting directly
correspond to the amount of the entrepreneurial activity which the bribe needs to
cover.7 However, for simplicity it is assumed in this model that the partner always
bribes the institution and furthermore that the partner’s benefits from bribing equal
its bribing costs.8 Covered by the bribe any exploitation of the contract is observable by the initiator ex post, but cannot be punished judicially due to the inferior
institutions.
A5. Partner’s objective. Taking into account the venture transfers t from the initiator,
the partner has the following objective function
(5)

? = t ? C(e, ? ).

The partner incurs costs of the function
(6)

C(e, ? ) = e? .

7 This corresponds to a variable bribe. A simple, fixed bribe would also be thinkable, but is not introduced here.
Simple fixed bribes can be considered as a costly hurdle for investment.
8 This assumption implies that this model does not cover moral hazard profit maximizations, which is not the focus of
this article and would be a distraction from the main arguments. We suggest it as a topic for future research.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

165

The marginal costs ? of the partner include his variable production costs u =
?, which are normalized to be equal for all partners, and also his bribing costs.
Directly corresponding to the draw of the partner’s private information 9 on the
institutional inferiority, the realization of the marginal costs of the partner belongs
to the set
(7)

2 = {? (?), ..., ? (?)}.

As an example, ? describes an informed partner type with high exploitation ability and mathematically with high marginal costs, while ? describes an uninformed
reliable partner with low marginal costs. It is public knowledge, that the corresponding continuous probability density function g(? ) is distributed identically
for each partner. Let G(? ) denote the corresponding cumulative distribution function. G(? ) is commonly known, absolutely continuous and differentiable. For
simplicity’s sake, we furthermore assume a monotone hazard rate of ?


? G(? )
(8)
? 0 ? (e, t, ? ) ? A × 2.
?? g(? )
A6. Contracting variables. A denotes the set of all feasible contracts. Both parties can
contract on the level of entrepreneurial activity e and on the transfer t, which are
both observable and verifiable. A is defined as
(9)

A = {(e, t), where e, t ? R+ }.

A7. Imperfect institution. The institution may be imperfect in two ways. Firstly, due
to institutional inefficiencies, employing the rule of the institution inflicts costs
of 8 on the initiator (This may be due to an inefficient routine or due to a bribe).
Secondly, the institutional inferiority is modeled as follows: The institution detects
any exploitation by the partner greater than his exploitation ability according to his
information draw ?. Thus, the institution imperfectly supervises the contracts and
property rights of the initiator which inflict variable costs on the initiator.
A8. Timing. Contracting takes place at an interim stage. At time T = 0 the partner
draws on his level of private information of the inferior institutions in his market
and bribes. At the successive stage T = 1, the initiator offers a menu of venture
contracts. Accordingly, the contracts are accepted or refused by the partner in
T = 2. At stage T = 3 the chosen contract is executed.9
A9. Outside options. There are no outside options available to both initiator and partner.
9 Clearly, it is perfectly feasible for contracting to take place at an ex-ante stage T = 0. However, then we also would
need to control for risk-attributes of the parties, which would be distracting from the main argument.

166

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

A10. Initial endowment. Any fixed costs of the partner are considered as sunk costs.
The initial endowment of wealth of the partner ? is zero. In the end the partner
will accept the venture only if at least his status-quo level of wealth is covered.
III. The Model - Discussion and Optimal Contracts

Employing the principal-agent model described above, how much entrepreneurial activity will unfold? Who benefits from such an environment and what are the wealth levels
of the parties? What is the optimal menu of contracts for the initiator? In what ways will
contracts and entrepreneurial activity diverge compared to a second-best institutional surrounding? The core idea is that imperfect institutions can both directly increase the costs
of entrepreneurial activity and create abnormal uncertainty on the quality and trustworthiness of trading partners. The resulting reallocations of wealth leave the relationship
between imperfect institutions and wealth in most cases negative, but are subject to contextual parameters which also allow niche cases with positive results.
These questions will be the driving forces of the following discussion, which is of technical nature and follows findings of the principal-agent theory, which is a well researched
theoretical framework. In detail, the model described above belongs to the class of adverse selection with incentive and participation constraints under a continuum of types.10
Thus, with great respect for its historical development, we refer to the synthesis of Laffont and Martimort (2002), for theoretical and earlier versions of the mathematics, proofs
and concepts applied within this section. We must not be credited for the model’s mathematical derivation, as we rely on their towering achievement and adopt their notation in
the following mathematical section.11 We highlight that our contribution in the following
mathematical section is strictly limited to the application and transfer of the Laffont and
Martimort (2002) theory to an entrepreneurship under imperfect institutions setting in
order to offer a model based on logic to the “sand” or “grease” debate around imperfect
institutions, which is capable to unify seeming contrary positions and from which we
then derive management and policy implications.
A.

Second-Best Institutions - A Benchmark

Following Laffont and Martimont (2002), pp. 28-81, we comment briefly on the optimal contract under second-best institutions to establish a benchmark as a starting point of
the discussion. We assume an utopian world where properly and efficiently functioning
institutions prohibit any exploitations (or bribes) and that the initiator is correctly informed about the level of costs of his venture partners u = ? . Accordingly, the initiator’s
maximization problem reads as follows:
maximize 5 = ?(e(? )) ? t (? )
(10)

(e,t)

subject to t (? ) ? C(e, ? ) ? 0.
10 See Laffont and Martimort (2002), appendix of chapter 3, pp. 134-140.
11 See Laffont and Martimont (2002), chapters 2 and 3, pp. 28-140.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

167

Clearly, the participation constraint must bind, since the proper institutional surrounding
does not allow exploitations
t (? ) = e? .

(11)

A substitution and maximization leads to the rearranged first-order condition
SOLUTION (PROPOSITION) 1:
??(e? (? ))
= ?.
?e

(12)

The initiator contributes t ? = e? (? )? to the venture and accesses the market and the
partner’s resources at a fair price (i.e. he has to pay for no additional costs). If we assume
that there were two types of partners: an inefficient partner with marginal costs u = ?
and an efficient partner u = ? , the optimal contracts read
(e? , t ? ) iff ? = ? ,

(13)

?

(e? , t ) iff ? = ? .

It is possible to depict these contracts on the (e, t) space by representing the initiator’s
profit 5 and the partner’s wealth level ? in the form of isoprofit functions. The isoprofit
functions of the partners will follow the indifference path t ? e? = ? from the origin,
according to their respective definition in assumption (A5.).12 In addition, the isoprofit
function of the initiator is by definition in assumption (A2.) a strictly concave function in
e. Considering both isoprofit functions, the optimal contract is described by the tangential point between them, when the marginal costs equal the marginal benefits. Two main
observations can be drawn from this figure. Firstly, the initiator maximizes his profits by
expanding entrepreneurial activity e until his marginal benefit equals his marginal costs.
Since in the optimal contracts the marginal benefit from the venture for the initiator is
higher when venturing with type ? , this in turn implies the volume of the entrepreneurial
activity e with this type is lower e? < e? .13 Secondly, the objective of the initiator is to
accrue e and to avoid t, his profit 5 increases when moving to the south-east direction
on the (e, t) space. For the partner the opposite is true. Therefore, since ? > ? , the
tangential point when dealing with the partner of type ? will allow the initiator to reach a
higher profit level, than when venturing with the partner type ? (Laffont and Martimont,
2002). In a nutshell, these results are well in line with economic intuition. Companies
12 Since the gradient of these functions is determined by ? , two different isoprofit functions will only cross once in
the (e, t) space. Hence, the partners’ isoprofit functions in this model fulfill the Spence-Mirrlees property, which assures
that a monotone ranking between the types ? on the marginal rate of substitution between entrepreneurial volume e and
transfer
 t ispossible. Or mathematically stated
?
??

??
?e
??
?t

?0

? (e, t, ? ) ? A × 2.

See Mirrlees (1971) and Spence (1974).
13 Since ? > ?, it follows that the marginal benefits are ordered (??(e? (? )))/(?e) > (??(e? (? )))/(?e). Note that
this fact holds due to the spatial boundaries of ? defined in assumption (A2.) which postulate positive and diminishing
marginal revenues in e.

168

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

t

? = t (? ) ? e (? )?
t

t

? = ? (( e (? ) ) ? t (? )

*

*

? = t (? ) ? e (? )?

0

e

*

e

*

e

F IGURE 1. T HE BENCHMARK CONTRACT (e? , t ? ) ON THE ( E , T )-S PACE

venture more profitably and intensely with more efficient partners.
B. Inefficient Institutions

We start with the easiest case of an imperfect institution: assume an institution is
inefficient, but not inferior. Its inefficiency inflicts costs 8 on the entrepreneurial activity
(for example the initiator has to purchase an expensive license). Exploitation is not
tolerated by the institution. The initiator is correctly informed about the type of partner
he is venturing with u = ? . Furthermore assume again, that there are only two types
of partners: an inefficient partner u = ? and an efficient partner u = ?. With the
assumptions of the model above, the following three cases are possible.
?
As a first case, if 8 ? 5 there will be no effect on the volume of entrepreneurial ac?
tivity. Secondly, once the threshold value 5 is reached, projects with sub-optimal partners will begin to be located outside the profitability horizon. Projects, which otherwise
would have been viable, are now subject to shut-down due to the inefficient institutional
environment. Thus, the reduced range of projects executed will only include (e;
˜ e? ).
?
Thirdly, if 8 > 5 , entrepreneurial activity breaks down completely since no project is
located inside the profitability horizon. These direct effects of institutional inefficiency
can be visualized in the (e, t) space, as displayed in the figure below.
Although this effect may seem very intuitive, it entails a big effect for economies, because, similar to a black hole, all entrepreneurial projects beyond the profitability horizon

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

169

t

? (? )

?

?>?

Not profitable

t

???

*

~
t
*
t

?

Profitability horizon

Reduced
activity range
0

?

e

*

e~

e

*

e

F IGURE 2. T HE PROFITABILITY HORIZON AND CONTRACTS UNDER INEFFICIENT INSTITUTIONS

will not be realized.
C. Discussion of Inferior Institutions - Trading with Thieves

Assume now that the institution is inferior and thus tolerates exploitations or bribes,
but is not inefficient. In accordance with the assumptions, the marginal costs ? of possible partners include their variable production costs u, which are now normalized to be
equal for all partners u = ?, but also the costs of their bribery. The partner’s bribing
ability directly corresponds with the draw of the partner’s private information 9 on the
institutional inferiority, the realisation of the marginal costs of the partner thus belongs
to the set 2 = {?(?), ? (?)}. Furthermore, in order to clarify the mechanics, assume for
the moment that there are still only two levels of information about the inferiority.
The initiator must now budget for being exploited, since he knows that he is now
trading with thieves whose abilities he does not know. In addition, for the non-informed
partner of type ? it is more profitable to select the contract designed for the exploiting
partner ? when offered a menu of proposals during a contract negotiation.14 To see this
numerically, we compute the profits of a partner of type ? when mimicking a partner of
14 The contract (t ? , e? ) is located westwards of (t ? , e? ) on the (e, t) space, choosing the (t ? , e? ) contract will enable
the non-exploiting partner of type ? to reach a higher iso-profit function ? .

170

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

type ?
?

t ? e? ? ,
?

? t ? e? ? + e? ? ? e? ? ,

(14)

?

? (? ) + e? (? ? ? ).

Thus, the partner ? is in a position to barter for e? (? ? ?).15 Either he pretends to
have knowledge about the inferior institutions or he barters for an additional rent as a
compensation. Hence, the contracts tailored for a normal second-best institutional set-up
normal are no longer optimal as the partners no longer self-select the contract designed
for them. In order to avoid such pretensions in negotiations, the initiator can only identify
the true level of information ? , if it is in the best interest of the partner to reveal his level.
Therefore, an optimal contract must take into account, that being offered a choice of
contracts in A, truth-telling is weakly preferred by the partner for any possible pair of
? ? 2 × 2. Following Laffont and Martimont (2002), this puts additional truth-telling
constraints on the problem, which read
t (? ) ? e(? )? ? t (? ) ? e(? )? ,

(15)

t (? ) ? e(? )? ? t (? ) ? e(? )? .

Moreover, a simple addition yields an additional insight:
t (? ) ? t (?) + t (? ) ? t (? ) + e(? )? ? e(?)? ? e(? )? + e(?)? ? 0,
(16)

?

e(? )(? ? ?) ? e(? )(? ? ? ) ? 0,

?

e(?) ? e(? ).

Therefore, the truth-telling constraints force the initiator to limit the volume of his entrepreneurial activity with the exploiting partner of type ? at least to that of the nonexploiting type ? . In other words, the level of entrepreneurial activity must be ranked
monotonically according to the level of exploitation of the inferior institution by the
partner types. It is worth noting, that this peculiar structure arises regardless of the objective function of the initiator. The initiator’s objective function is now weighted by the
likelihood ? to meet the ? partner type, since the trustworthiness of the venture partner
is blurred by a veil of uncertainty. Following Laffont and Martimont (2002), pp. 28-81,

15 The partner of type ? already manifests the most informed type and hence it cannot avoid the initiator’s discrimination, thus ? (? ) = 0.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

171

the objective function and the initiator’s problem read
maximize 5 = ?(?(e(? )) ? t (?)) + (1 ? ?)(?(e(? )) ? t (? ))
(e(? ),t (?));(e(? ),t (? ))
subject to
t (?) ? e(?)? ? 0,
(17)

t (? ) ? e(? )? ? 0,

(I)
(II)

t (? ) ? e(?)? ? t (? ) ? e(? )? ,

(III)

t (? ) ? e(? )? ? t (? ) ? e(?)? .

(IV)

The solution to this problem yields the optimal contracts with all possible venture partners. After some calculations one arrives at the following first order conditions:16
SOLUTION (PROPOSITION) 2:
??(e? (?))
?e(?)
??(e? (? ))

(18)

?e(? )

= ?,
= ?+

?
(? ? ? ).
(1 ? ?)

By observing the resulting contracts, we note that overall entrepreneurial activity is
greatly reduced, as compared to the contract (e? , t ? ) that would have been implied under
second-best institutions. The first order conditions yield the same result for the partner of
type ?. However, the first order conditions indicate that the marginal benefits with partner
type ? have increased, reflecting a reduced entrepreneurial activity with this type.17
?
The transfers of the optimal contracts t ? = e? ? + e? (? ? ? ) and t = e? ? , as well
?
?
?
as the partners’ resulting wealth levels ? = e (? ? ?) and ? = 0, indicate another
economic link: the partner type ? accrues a positive partnership rent for not exploiting.
Due to this rent, the wealth level of the initiator will also be lower, 5? > 5? .18
D.

Optimal Contracts under Inferior Institutions

Let us now allow for the more realistic case of a continuum of information levels ? ?
2 = {? , ..., ? }, with a density function g(? ) and a cumulative distribution function G(? )
behaving in accordance to assumption (A5.). This problem discussed in Laffont and
Martimont (2002) pp. 134-140 and we follow their solution. In line with the approach
in the discussion above, we now search for the optimal contract (e? (? ), t ? (? )) in order
to compare it with the benchmark case. As a starting point, the objective function of the
16 For the reader’s convenience, we omitted the calculations and refer to the appendix.
17 This is due to the assumed shape of the ?-function, (??(e))/(?e) > 0 and ((??(e))/(?e))/(?e) < 0, which leads to
e? < e? and e? = e? . Also, the results follow the monotone ranking demanded by equation (16): e? = e? > e? > e?
(Laffont and Martimont, 2002).
18 See Laffont and Martimont (2002).

172

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

initiator under these conditions reads19
Z ?
(19)
maximize 5 =
[?(e(? )) ? t (? )] g(? ) ?? .
{(e(? ),t (? ))}

?

The objective function is now subject to an infinite amount of constraints due to the
infinity of possible types of ? . As the problem is designed to be well behaved in order
to carve out the possible economic effects of the inferiority of an institution, it turns out
that the infinite amount of constraints in our case collapses to an individual rationality
constraint ? (? ) ? 0, a second constraint which imposes a monotone ranking on the
entrepreneurial activity depending on the information level (?e(? ))/(?? ) ? 0 and a truthtelling constraint (?? (? ))/(?? ) = ?e(? ).20 This makes the program of the initiator much
handier and allows us to calculate the optimal contracts. The program of the initiator, in
terms of profit levels, now reads
maximize 5 =
(e(·),? (·))

(20)

subject to

Z

?

(?(e(? )) ? e(? )? ? ? (? )) g(? )?? ,
?

? (? ) ? 0,
?e(? )
? 0,
??
?? (? )
= ?e(? ).
??

(I)
(II)
(III)

In line with Laffont and Martimont (2002), we try to reduce the variables in the objective
?
function. For this purpose, we employ the truth-telling constraint (?? (?? ))/(? ?? ) = ?e(?)
to enlarge ? (? ) = ? (? ).

(21)

? (? ) = ? (? ) + ? (? ) ? ? (? ),
Z ?
?? (? )
?? ,
? ? (? ) ?
??
?
Z ?
e(? )?? + ? (? ).
?
?

However, the most informed type ? of the partners is not capable to pretend to be even
more informed, since the cumulative distribution function G(? ) is common knowledge.
It follows immediately that he will be discriminated perfectly. As a consequence, the
individual rationality constraint of the most informed type ? will be binding: ? (? ) = 0.
19 The integral is used, since the initiator can only form an expected value. This will be the sum of all possible (thus
type dependent) profit situations times their respective likelihood in the form of the density function.
20 The derivation and the calculations belong to the standard repertoire of the principal-agent realm and would be
distracting from the main results. They are provided in the appendix for the reader’s convenience.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

173

Thus, we can write
(22)

? (? ) =

Z

?

e(? )?? .

?

This is a remarkable feature of the optimal contract. It dictates that partners will receive
R?
a partnership rent ? e(? )?? for compliance with good practices.21 Observe that we can
plug this result into the objective function of the initiator, which also implies that the
constraint ? (? ) ? 0 is fulfilled due to the partnership rent. As Laffont and Martimont
(2002) point out the objective function of the initiator now reads
maximize 5 =
e(·)

Z
Z ?
?(e(? )) ? e(? )? ?
?

(23)

?
?


e(? )?? g(? )?? ,

?e(? )
? 0.
??

subject to

(II)

For the moment, we will neglect the remaining monotonicity constraint and follow a
more practical approach by checking whether the omitted constraint is fulfilled afterwards. We directly proceed by maximizing the objective function of the initiator. The
double integral exudes an air of difficulty, but it turns out that it can be tackled by inteR?
gration by parts and a substitution of ? e(? )?? .22
Z ?

?(e(? )) ? e(? )? ?

?

?

Z

|?

?


e(? )?? g(? ) ?? ,
|{z}
{z } g(x)
f (x)

Z ?

?(e(? )) ? e(? )?

?

"Z
?

(24)

?
?

|

#?

?

e(? )?? G(? )
?

?

{z
=0

}

Z ? R ?
 
? ? e(? )??
?
G(? ) ?? g(? )?? ,
??
?

Z ?
Z ?
 
G(? )
g(? ) ?? g(? )?? ,
?e(? )
?(e(? )) ? e(? )? +
g(? )
?
?

Z ?
G(? )
g(? )?? .
?(e(? )) ? e(? )? ? e(? )
g(? )
?

21 Note that this rent must be paid independently of the initiator’s objective function. See also Laffont and Martimont
(2002).
22 We employ the extension of Leibniz’s law for integration by parts:
Rb
R



b R b R
g(x)?x ? f?x(x) ?x.
a f (x)g(x)?x = f (x) g(x)?x a ? a

174

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Now we are in a position to optimize this expression point-wise which yields the following results:
SOLUTION (PROPOSITION) 3:
??(e? (? ))
G(? )
?? ?
= 0,
?e(? )
g(? )
??(e? (? ))
G(? )
=?+
.
?e(? )
g(? )

(25)
?

It is remarkable, that for the partner of type ?, this result shows the following attributes
??(e? (?))
?e(? )

(26)

=?+

G(? )
= ?.
g(?)
| {z }
=0

This indicates that non-exploiting partners will enjoy the same extent of co-operation as
under second-best institutions, as shown by Laffont and Martimont (2002). Let us summarize the situation derived for the optimal contracts under inferior institutions. For this
purpose, we expose the results in the (e, t) space. We note that overall entrepreneurial

t
G (? )
?? ( e (? ))
=? +
? e (? )
g (? )

? =0

?? ( e (? ))
=?
? e (? )

? >0
?*= 0

?

Optimization

t

?

³? e (? )??

*

t

*

t

Partnership-rent

?

t
Reduction
0

e

e

*

e

*

= e

e

F IGURE 3. V ISUALISATION OF THE CONTRACTS UNDER INFERIOR INSTITUTIONS

activity is harshly reduced, since the level of entrepreneurial activity with bribing part-

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

175

ners (or companies) is reduced according to the extent of their exploitations and their
position in the population’s general compliance with good practices. Vice versa, partnerR?
companies with a reputation for good practices enjoy partnership rents ? e(? )?? , depending on their degree of compliance. Remarkably, non-exploiting partners of type ?
will enjoy the same volume of trade as under second-best institutions, plus being rewarded for their good practices due to the partnership rent. Overall, it is more profitable
for the initiator to venture with non-exploiting partners than with exploiters. However,
the general level of profitability will be lower as compared to second-best institutions
5? > 5? . This is due to the fact that the initiator has to reward good compliance with
partnership rents to remedy the inferior institutional environment. Thinking beyond the
assumptions of the model, we note that these effects are sensitive to improvements in
competition. If a competitive situation develops, for example two partners starting a bidding race for the contract with the initiator, the veil of uncertainty on the partners’ costs
is lifted considerably, since the bargaining position of the initiator improves dramatically.
IV. Derived Propositions and Policy and Management Implications
A.

Propositions and Boundaries

The insights gained from applying the Laffont and Martimont (2002) model above will
be cumulatively exposed in the form of propositions of a theorem.
THEOREM 1:
I. Inefficient institutions:
PROPOSITION 1: An inefficient institution has no impact on the volume of entrepreneurial activity if the cost inflicted by its inefficiency is outside the profitability horizon.
PROPOSITION 2: An inefficient institution diminishes the volume of entrepreneurial
activity if the cost inflicted by its inefficiency exceeds the profits generated by some
projects. All projects outside the shifted profitability horizon will be shut down.
PROPOSITION 3: An inefficient institution prevents all entrepreneurial activity if the
cost inflicted by its inefficiency exceeds the profits generated by all projects.
II. Inferior institutions:
PROPOSITION 4: An inferior institution has no impact on the volume of entrepreneurial activity for partners complying with good practices.
PROPOSITION 5: An inferior institution diminishes the volume of entrepreneurial activity by an extent according to the general compliance of partners with good practices
(G(? ))/(g(? )) and the extent of their individual exploitational abilities ? (?).
PROPOSITION 6: An inferior institution diminishes the profitability of all entrepreneurial activity for the initiator.

176

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

PROPOSITION 7: The presence of an inferior institution forces the initiator to award
R?
partnership rents ? e(? )?? to its partners.

PROPOSITION 8: Competition dampens the effects of inferior institutions on entrepreneurial activity.
We stress that these propositions should be read from an agnostic perspective, as they
are subject to the spatial boundaries defined in the assumptions of the model, as well
as the implicit assumptions of the model. One such implicit assumptions is that the
construct “entrepreneurial activity” is modeled sufficiently in the form of a simple trade
or venture and a production function. We are aware that it is influenced by many more
variables in practice.23 We purposely neglect and fade out other variables in order to
carve out the economic effects of imperfect institutions on entrepreneurial activity.
Our approach is driven by the profit maximization objective and views the entrepreneur
as an autonomous decision making entity. Other authors, for example Cyert and March
(1963) emphasize that companies are a collection of individuals. In addition, in the
literature strand on organizational theory, personality traits of persons in charge, political processes within a company and generally broader goal systems of the company
are considered as determinants of the actual behavior of firms. These sociological and
psychological dimensions are neglected in our approach, since the perspective taken is
related more to industrial organization, microeconomics and contract theory.
Furthermore, we would like to stress that the approach taken implicitly assumes a
deterministic framework. We argue that exogenous changes in the structural elements
of the economy, e.g. in our article the institutional design, induce changes in the entrepreneurial activity. We are aware of the fact that other authors, for example Brandenburger and Nalebuff (1996), highlight that companies can gain advantages by changing
the structural elements in the market themselves. According to our approach, this corresponds to an institutional amendment, be it informal or formal. Such considerations are
however beyond the scope of this article.
B. Policy Implications

Reflecting the recognition that markets are unlikely to operate well in the absence of
proper institutions which provide them with security through predicable rules, much of
the reform focus in the developing world has shifted towards improving and reforming
institutions and governance. But at what institutions should governance reformers aim?
Some insights for practical policy making can be drawn from the propositions of the
theorem. Due to the generality of the model, they are applicable to a wide range of settings, ranging from a regional, microcosmic trade situation to an international venture.
However, and in line with contemporary development economics research, see Rodrik
23 For example one could postulate that variables like “literacy rate” influence the “entrepreneurial activity” construct.
But the decision to form a venture itself is also influenced by numerous variables. A large body of literature has attempted
to explain the portfolio of venture decisions by firms and has identified numerous reasons for inter-firm collaboration, thus
influencing venture activity. See Hitt et al. (2011), or for an agency theoretical point of view, see Reuer and Ragozzino
(2006).

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

177

(2007), the contextuality of these policy implications cannot be overstated. Proper institutions may be necessary for economic growth, but they are not a blueprint for achieving
economic growth. Thus, the policy implications given here are formulated with the emphasis on humility, they should not be confused with rules.
I. The Partnership Rent: A Diagnostic Signal. - The propositions predict that once a
country allows for inferior institutions, the entrepreneurs and companies will isomorphically introduce partnership rents for good compliance as a rational response.24 In turn,
high partnership rents relative to second-best levels indicate the need for lifelines outside
the formal institutional environment. Thus, they might be taken as a diagnostic signal
of high institutional inferiority for policy makers (in parts, as there may also be other,
e.g. cultural, reasons). In practice, such partnership rents may take various forms such
as special conditions and discounts in relational contracting situations with the goal of
building up a long-term relationship of mutual trust between trading partners to allow
successive projects (i.e. repeated partnerships).25
II. Reforming Institutions: i. Inferior Institutions. - Following the principal-agent
framework applied here, there are two main approaches open to the government reformer.
The first approach is to enhance the institutional surrounding in which the exploitation
occurred (e.g. improve law enforcement and rule of law). However, in practice this
relationship is non-trivial since the model predicts that as a consequence the enhancement
will undermine the lifelines of partnership rents which formerly saved companies from
exploits. Therefore, one has to consider whether the informal partnership rents c pr or
the formal law enforcement routines cinst due to exploitations impose more transaction
costs on firms. In line with Dixit (2003) as well as Li (2003), we argue that trading based
on partnership rents requires no fixed costs but high marginal costs. In contrast, law
enforcement is associated with high fixed costs for its provision but lower marginal costs.
Maintaining and enforcing an institution may be only become optimal for entrepreneurs
when a certain amount of trading contacts is exceeded, which may not be the case for
early development stages. Thus, the answer obviously depends on the country specific
development situation: how interwoven is the economy and how many ties or business
contacts have to be maintained by the individual entrepreneurs? Is there (or will there
be) a demand for an institution such that c pr > cinst ?
The second approach is to tackle the exploitation at its roots. One reason for the
exploit’s magnitude is asymmetric information. Entrepreneurs foreign to a specific market possessed inefficient screening opportunities and the quality of its trading partners
remained opaque. Thus, increasing competition, helping companies to gather information about market participants and disseminating or even ostracizing transgressors will
remedy the institutional situation. Supply of information may be provided from official
institutions or on a private profit-maximizing basis, as Dixit (2003) shows in a theoretical
24 In this context, we draw the attention of the interested reader to a neo-institutional perspective outlined from DiMaggio and Powell (1983) which consider isomorphic institutional change in organizational fields as a response to uncertainty
(amongst other reasons).
25 Work from Abreu et al. (1991) on repeated partnerships shows that under imperfect conditions long-term agreements
may foster compliance.

178

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

context. However, supplying market information is costly and may be difficult to achieve
in practice.
Besides the above economic suggestions, sociological and psychological aspects shall
not be forgotten. Even if the exploitation or bribe was possible, it would be the individual behavior and action which made it real. In the manner in which B´enabou and
Tirole (2003) argued, intrinsic incentives will shape and interact with the outcomes of
extrinsic incentives. If the internal value system of the individual had attributed utility
or happiness to behavior in accordance with social norms of honesty, righteousness and
fairness, and if the internal values system reacted to a violation of these norms by increased disutility in the form of a feeling of guilt, then less exploitations might be carried
out. The prevalence of egoistic and transgressive behavior is thus broken and the need
for governance or partnership rents is reduced. However the installation of such so called
“pro-social preferences” is a long-winded process, which will take hold only gradually as
individuals become educated in and familiar with social norms, as the work of Camerer
(2003) on behavioral economics indicates. Thus, fostering “pro-social preferences” by
direct or indirect education is considerable as a worthwhile, but long-run policy.
The above political advice is not new. An early form of these insights appears in
Hobbes’s Leviathan (1651): “The force of Words, being [...] too weak to hold men to the
performance of their Covenants; there are in mans nature, but two imaginable helps to
strengthen it. And those are either a Feare of the consequence of breaking their word;
or a Glory, or Pride in appearing not to need to breake it. This later is a Generosity
too rarely found to be presumed on, especially in the pursuers of Wealth, Command, or
sensuall Pleasure; which are the greatest part of Mankind.” (Thomas Hobbes (1651),
Leviathan, Chapter XIV)
Reforming Institutions: ii. Inefficient Institutions. - If an institution is found to be
inefficient, it is crucial to determine what caused the inefficiencies and to determine the
related costs 8. For example, if the costs 8 are caused by externalities, a reform of
an institutions must take into account the incentives of the individuals causing them or
force them to compensate their victims, which will induce resistance by those who must
pay for it. Regarding processual inefficiencies it is perhaps best to consider benchmark
institutions, which offer more effective routines. As Rodrik (2008) points out, this may
be a second-best institution with respect to the country specific appropriateness.
III. Economic Desirability of Reforms: i. Inefficient Institutions. - Regardless of the
context, institutional inefficiencies should be avoided since their effects are purely of
negative nature. To make things even worse, highly inefficient institutions may even induce institutional inferiority. High costs related to institutional inefficiency, for example
extensive licensing or entry costs (de Soto, 1989, and Djankov et al., 2002), can create
˜ by bribing.
incentives for market participants to bypass them at lower costs 8
Niche Case: Interestingly, in the niche case of a given strong institutional inefficiency,
an induced institutional inferiority may even entail positive effects here. In this niche
case, as the costs for the bribing bypass (exploit) are lower than the original costs induced

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

179

t

? (? )

?

t

*

~
t
*
t

???
Profitability horizon
inefficient institution

~
?
0

~
???
Shifted
profitability horizon
by bypassing

Added
activity range

?

e

*

e~

e

*

e

F IGURE 4. I NSTITUTIONAL INEFFICIENCY INDUCING INSTITUTIONAL INFERIORITY

˜ < 8, this expands the profitability horizon and more projects
by the high inefficiency, 8
are being realized because entrepreneurial activity is more profitable. Furthermore, the
increased entrepreneurial activity e may also have positive effects on a country’s wealth
5c (e), which arguably may surpass the potential negative effects of the institutional inferiority 8in f on the country’s wealth 5c (8in f ), leading to positive net wealth effect
5c ?. The above case also suggests that there may, under certain parameter values,
exist an optimal level of inferiority, if the inefficiency cannot be avoided. Such a correlation was empirically observed by M´endez and S´epulveda (2006) and M´eon and Weill
(2010), while Levy (2007) offers historical evidence. These efficiency gains relative
to an inefficient institution are better understood if one compares bribing to competitive bidding procedures (Beck and Maher, 1986) which reveals similarities between both
mechanisms.
However, let us be very clear: the model strongly suggests that institutional inefficiencies should be avoided if possible.
Economic Desirability of Reforms: ii. Inferior Institutions. In general, it cannot be
overstated that the mathematical results of this article strongly indicate that institutional
inferiority is likely to be undesirable due to the partnership rents and the reduced entrepreneurial activity, which arguably reduces a countries wealth 5c (e) ?.
Niche Case: However, the model also allows for parameters values for which this is
not the case: the relationship is non-monotonic from a national perspective. This id-

180

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

iosyncratic effect is best visualized with an example: A country hosts institutions of
inferior quality, which favor domestic entrepreneurs and allocate foreign entrepreneurs
to specific domestic venture partners. Subsequently, the model suggests that the foreign
entrepreneurs will adjust their contract offerings and reduce their entrepreneurial activity. After the contract negotiations the contracts are executed and the domestic players
(partner companies and institutions) reap bribing profits ? b or enjoy partnership rents
? pr .
Given that the reduction of overall entrepreneurial activity e ? generates less welfare
loss than the extra profits enjoyed by the domestic players, thus 5c (e) ?< 5c (? b +
? pr ) ?, inferior institutions are desirable for a domestic policy maker. Overall the welfare of the country rises 5c ?, while foreign entrepreneurs suffer a decreased level of
profits 5 ?, which may be regarded as irrelevant from a national perspective. In addition, privileging incumbent firms over entrants may in certain contexts be desirable.
In addition the capital accumulation generated domestically may be necessary to open
a leeway for the next development stage and induce growth. For instance Acemoglu et
al. (2006) argue that rents generated by incumbents stimulate necessary investment for
countries lagging behind the global technology frontier. In line with this argument, Qian
(2003) argues that China’s rise is due to the proper application of imperfect institutions
as stepping stones.
Here, the question remains whether the government would not better of extracting
a surplus from a foreign entrepreneur by imposing a simple lump-sum tax, rather than
using the indirect approach of creating additional uncertainty for the foreign entrepreneur
through an institutional inferiority. However, some capital might be inaccessible by taxes
and but accessible by inferior institutions, for example through a tolerance of patent right
infringements and accesses to firm specific knowledge or technologies. Furthermore, in
practice a tax may be a less feasible instrument, because it potentially may create high
transaction costs and because multi-national organizations are often very professional in
reducing their tax burdens. Such (peculiarly) beneficiary effects of inferior institutions
were highlighted by scholars quite early, for instance by Leff (1964). However, we
emphasize that this is a niche case, which may only be relevant as a stepping stone, and
that we refrain from putting to much emphasis on it.
On the contrary, if the institutional inferiority is located on a domestic or regional basis
(i.e. a domestic entrepreneur trading with an domestic entrepreneur), then the impact of
an institutional inferiority is very likely to have only strongly negative, reallocative and
wasteful welfare effects. Entrepreneurial activity and, assuming beneficiary effects from
entrepreneurial activity, welfare will be reduced. Furthermore, if the costs of partnership
rents can be passed to the consumers of the entrepreneurial activity, consumers will suffer
a loss in consumer surplus. Given such a context, inferior institutions should obviously
be avoided and amended. For such negative effects, there is large body of supportive,
empirical, and argumentative evidence in the corruption and growth literature (Shleifer
and Vishny, 1993, Mauro, 1995, Rose-Ackerman, 1999, M´eon and Sekkat, 2005, and
Aidt, 2009). Dixit (2009) points out that a failure to protect (and a violation of) property
rights by the government and its agents is a major cause of poor economic performance

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

181

in many countries. Furthermore, there is the danger that the institutional inferiority induces damaging lock-in effects: considering the benefits some players enjoy, it will be
hard to introduce incentives in favor of a policy reform towards a second-best institution
(Coate and Morris, 1999, and Blackburn et al., 2006). This may lead to complication in
economic transitions and be a source of unsustainable development (Paldam, 2002, and
Aidt, 2009).
IV. Imperfections as Barriers. - Furthermore, imperfect institutions create barriers to
entering the domestic market and reduce entrepreneurial activity. This result is in line
with current findings in the literature. For instance Klapper et al. (2006) consider entry
barriers reducing entrepreneurial activity. Campos et al. (2010) show that corruption
is an important barrier to entry and thereby inflicts social costs. In addition, according
to Ricardo (1817) a liberal perspective would suggest that such barriers are undesirable
with regard to the benefits of free trade.26 Barriers to entry also curb entrepreneurial
selection and thus fuel inefficiencies.
But on the other side of the coin, entry barriers may be vindicated from a protectionistnational perspective, since protected higher profits of incumbents create higher incentives
for domestic entrepreneurial activity and thereby fuel growth. Hausmann and Rodrik
(2003) argue that modest entry barriers shield incumbents from imitators, thus providing them with rents in equilibrium. These rents subsidize the risk-taking which goes
along with entrepreneurial activity, and thus promote domestic entrepreneurial activity.
Also, according to Bliss and Di Tella (1997), a profit-maximizing bureaucrat could wish
to reduce the number of firms, since bribe-profits of the remaining firms are potentially
higher. Thus, regarding the desirability of these entry barriers, the objective function of
the governance reformer plays an important role.
V. Addressing the Right Constraints. - We shall draw attention to the fact that many
developing countries face constraints of various natures in their ability to tackle and
implement reforms. Reforms and amendments are thus scare resources. Consequently
the above has to be judged in a holistic framework. The reforms which create higher
marginal benefits should be realized first. If for instance the main constraint for entrepreneurial activity is the literacy rate or an inefficient financial sector, then a reform
of the rule of law will clearly not be advisable and will only waste scarce resources. Inspired by a framework proposed by Rodrik (2006), the figure below attempts to provide
a holistic framework for constraints of entrepreneurial activity and highlights the area
in which our model grips. Obviously, the figure below is necessarily imperfect and we
apologize for any omissions.
C. Strategic Management Implications

To formulate a necessarily inchoate collection of strategies based on the insights of
the principal-agent model of this article, it is worthwhile to expand the scope of this
industrial-organizational related framework and, loosely restating Brandenburger and
26 See Ricardo (1817), pp. 146-185.

182

DISSERTATION JAN-PHILIPP AHRENS

? 6RFLDOVHFXULW\QHW
? ,QIRUPDOVROLGDULW\
? &XOWXUDOEDUULHUV
? HWF

? ,QWHUPHGLDWLRQ
? &RXQWU\UDWLQJ
? )RUHLJQH[FKDQJH ? 'RPHVWLFVDYLQJV
? HWF
? HWF
,QWHUQDWLRQDO
'RPHVWLF
ILQDQFH
ILQDQFH

5LVNDYHUVLRQ

&RVWVRIILQDQFH

DECEMBER 2012

? &DSLWDOVWRFN
? .QRZOHGJHEDUULHUV
? ,PLWDWLRQ
? HWF
&RVWVRIHQWU\H[LW

(QWUHSUHQHXULDODFWLYLW\
5HWXUQV
'HPDQG
VLGH
? *'3SHU
FDSLWD
? ,QFRPH
GLVWULEXWLRQ
? HWF

6XSSO\
VLGH
? &RPSOHPHQWRUV
? +XPDQFDSLWDO
? ,QIUDVWUXFWXUH
? 7HFKQRORJLFDO
DFFHVV
? HWF

*RYHUQPHQW
IDLOXUHV
0DFUR
SHUVSHFWLYH
? )LQDQFLDO
SROLF\
? )LVFDOSROLF\
? 0RQHWDU\
SROLF\
? 6WDELOLW\
? HWF

0LFUR
SHUVSHFWLYH
? &RQWUDFW
HQIRUFHPHQW
? 3URSHUW\ULJKWV
? &RUUXSWLRQ
? 7D[DWLRQ
? HWF

0DUNHW
IDLOXUHV
? ,QIRUPDWLRQ
? &RRUGLQDWLRQ
? ([WHUQDOLWLHV
? &RPSHWLWLRQ
? HWF
$UHDVWRXFKHG
E\WKHPRGHO

F IGURE 5. H OLISTIC FRAMEWORK FOR CONSTRAINTS ON ENTREPRENEURIAL ACTIVITY

Nalebuff (1996), to allow the companies to change the structure of the game itself. The
basic question is: What determines the magnitude of the exploitation (of the foreigner)
and how can companies reduce it?
I. Corporate Citizenship. - In the strategic management literature Hymer (1976), Zaheer and Mosakowski (1995), and Mezias (2002) argue that there exists a set of liabilities due to foreignness which inflict additional costs on foreign players in local or
domestic markets. According to Zaheer and Mosakowski (1995), Vernon (2001), and
Gardberg and Fombrun (2006) these liabilities are composed of transaction costs, information costs, lack of familiarity with and awareness of the domestic institutional environment, lack of legitimacy, and local biases. If one assumes, that the magnitude of the
exploitation or bribe is influenced by these liabilities of foreignness, then entrepreneurs
or companies should invest in measures to increase their local embeddedness.
Gardberg and Fombrun (2006) argue that corporate citizenship, understood as the socioeconomic and philanthropic activities which companies undertake to fulfill perceived
duties as members of society, creates intangible assets such as reputational and social
capital as well as legitimacy. These intangible assets help companies to integrate into the
sociocognitive fabric of domestic and local communities, thus buffering the negative impacts of the foreignness attribute. Indeed, research suggests that companies can leapfrog
such a nationalistic or regionalistic barrier by enhancing the perceived legitimacy to oper-

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

183

ate in the considered domestic market (Oliver, 1991, and Gardberg and Fombrun, 2006).
Hence, if foreignness is identified as the reason for the exploitative behavior of partnercompanies, then investing in corporate citizenship might be one strategy for the initiator
to dampen the magnitude of exploits.
II. Internalization. - Another strategy might be to incorporate the institutions which
forbear the exploitation into the venture. Following this approach, the initiator partially
sells or shares the assets of the venture with the institution. The rationale behind this
approach is that once the institution becomes a claimant of residual cash flows generated
by the venture, it is incited not to tolerate any exploits. Thereby the initiator actively
changes the “rules of the game”: The institution and the initiator streamline their objective functions towards parallelism and the internalization transforms the exploitation into
an agency problem of the institution. Such an approach can be observed in China, where
community enterprises, so called “township and village enterprises”, are a common business model. Such an approach may furthermore save the company transaction costs and
enhance a company’s position in an environment of imperfect institutions. However, a
joint venture with authorities can also backfire, because the company may actually lose
more bargaining power to the authority than they have sold in terms of shares. Thus to a
certain extent, according to von Hayek (1944), the positive effects of such a strategy are
accompanied by a loss of freedom, the company has made a few steps on “the road to
serfdom”.
The analogous approach of integration of the partner-company into the initiator’s company, in accordance with Williamson’s (1985) transaction cost theory, is also possible.
In a similar way, this transforms the economic governance problem into a corporate
governance problem. The vitality of this approach depends on whether internal or external governance is more costly to the initiator. Dixit (2009) points out that examples of
such behavior can be found in family-owned conglomerates in less-developed countries,
which are integrated mainly to escape weak external governance and tolerate the inefficiencies caused, since exposure to external governance would be even costlier.
III. Competition. - As a next strategy, since the effects of the model are limited to the
case of non-competitive interaction between two parties, the initiator’s strategy might include fuelling competition between several partners or institutions across regions. Both
partnership rents and institutional inferiorities are sensitive to competition if the partners or the institution are interested in trading with the initiator. The initiator might,
for instance, consider another foreign or domestic market for his activities and thereby
enhance his bargaining position.
However, such a strategy may have limited effects, since increasing competition alone
may not suffice to fully extinguish exploitations. For example Ades and Di Tella (1999)
point out corruption may be fuelled not only by rents from imperfect competition, but
also natural rents (e.g. revenue from rich natural resources). Furthermore, Bliss and Di
Tella (1997) argue that competition may not be an exogenous parameter, but could also
be endogenously determined by the objective function of a profit-maximizing bureau-

184

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

crat.27
IV. Strategic Governance Alliances. - Another approach for companies wishing to
escape the “world of thieves” setting is to form “strategic governance alliances” which
shield individual companies from exploitations by threatening transgressors with sanctions by all alliance members. When calculating the pay-off from exploiting, any transgressor would then have to incorporate the profits forgone with other alliance members
into his objective function. If the profits lost are higher than the immediate gains from the
exploitation, then compliance with fair practices is ensured. However, the punishment
of the transgressor equals a private provision of a public good for all alliance members
except for the one who has been cheated. Thus, in order to overcome free-riding behavior a second layer of punishment directed at free-riders within the alliance must be
introduced. Such strategies and codices were successfully employed by medieval merchant guilds in Europe, as Greif et al. (1994) show. Another work by Greif (1993) on the
11th century Maghribi traders’ coalition in the Mediterranean area shows that such coalitions worked extraordinarily well in very hostile institutional surroundings. A review of
ancient reports of the time reveals that only a handful of reports contain allegations of
misconduct. We wonder whether there are insights to be gained from the knowledge of
these guilds for modern alliances too. Such strategies may have become viable again,
especially since information on transgressors is distributable via modern communication
technologies very easily and transparently.
V. Private Enforcement. - Private enforcement of contracts and property rights under
the shadow law can be observed in practice. For example, before the modern Italian
state was installed, formal institutions in Italy could not perfectly fulfill parts of their
functions, especially in the poorer south of Italy. The civil society and merchants thus
reacted by hiring guards who would enforce their masters’ claims. For potential transgressors this guardianship acted as a strong deterrence and credible threat (Gambetta,
1993). Another example of private enforcement is private collecting agencies which can
also be understood as private strategies to escape institutional imperfections.
VI. Embracing Imperfections. - From a corporate perspective institutional imperfections might also be a chance to fortify a corporate strategic alignment. The astute
initiator could thus bear the institutional imperfections, since they constitute an entry
barrier which defends his strategic position once he has entered the market. Companies
adopting this strategy will invest in human resources capable of navigating in imperfect
institutional surroundings, such as experts in relational contracting.
The model we applied is actually good news for all companies complying with good
practices. It predicts that honest behavior pays, especially if competitors are known
for not tolerating the laws of good business conduct. Adopting an honest strategy is
remunerated with the first-best volume of entrepreneurial activity, an interesting point
27 But this is only the case if the bureaucrat is a monopolist. Fuelling competition between corrupt bureaucrats might
actually reduce bribes, for example by installing competing official institutions offering identical services.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

185

with regard to scale effects, and on top of this is rewarded with partnership rents. This is
clearly one of the most encouraging results of the model.
V. Conclusion

The propositions we derived from applying a standard principal-agent model to the
context of international or interregional entrepreneurial activity suggest that imperfect
institutions reduce the volume and the profits of entrepreneurial activity for the initiating entrepreneur. Analyzing the incentives structures in partnerships reveals that it is
optimal for entrepreneurs venturing under imperfect trading conditions to distribute and
pay partnership rents, as the imperfect institutions may induce abnormal uncertainty due
to the hidden information on the trustworthiness of the trading partner. Applied on a
national or regional basis the model mainly suggests redistributive and negative effects
of imperfect institutions on the wealth of a nation. However, applying the model to a
holistic international macroeconomic context or to countries with feeble institutions unveils niche-cases, where the wealth of a nation may (given specific parameter values) be
increased through an imperfect institution, which is mirrored in contemporary findings
and debates in the development literature.
TABLE 1—E FFECTS OF I MPERFECT I NSTITUTIONS

Case
General case:
1. Imperfect institution
Niche cases:
1. Inferior institution &
foreign initiator
2. Inefficient institution &
exploit bypass

Level of
entrepreneurial
activity
(1)

Level of
initiator’s
wealth
(2)

Level of
host-country’s
wealth
(3)

?

?

?

?

?

?, ?, ?

?

?

?, ?, ?

Note: The level of entrepreneurial activity is measured in e and the profit of the initiator in 5. The level of the host
country’s wealth incorporates the profit of domestic initiators 5, the profit of partners ? , the payments to the institution
(? ? u)e and possible macroeconomic wealth effects of the level of e. In the general case the effects of imperfect
institutions (both inefficient and inferior) are described in a standard domestic setting. Niche case one assumes that the
initiator and his profit levels are accounted within the realm of a foreign economy. Niche case two describes a standard
domestic setting with strong institutional inefficiencies, where inferior institutions allow an exploit bypass. The arrows
describe the direction of possible level changes compared to the levels under second-best institutions. For niche case two,
the arrows describe possible level changes compared to the levels of inefficient institutions without the exploit bypass.

We highlighted several strategies (e.g. corporate citizenship, internalization, strategic
governance alliances, and private enforcement) which allow entrepreneurs to navigate in
and to escape from an imperfect institutional setting. Furthermore, a discussion of the
driving forces and the effects of the imperfections as well as the policies to remedy them
highlight the need for a holistic approach. Due to the existence of several trade-offs, no

186

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

deterministic policy advice can be given on the basis of the derived propositions without
knowing the country-specific parameters necessary to judge the policy’s appropriateness.
The importance of the context rather reminds economists and policy advisers of the “art
of humility” in policy making. Since although institutions may be imperfect, they may
be so for a very good reason. A current institutional “species” of a country may be the
outcome of an experimental evolutionary process which has settled in a current equilibrium. The famous author Herbert G. Wells noted on this: “Biologically the species is
the accumulation of the experiments of all its successful individuals since the beginning,
and the World State of the Modern Utopist will, in its economic aspect, be a compendium
of established economic experience [...]” (Herbert G. Wells (1905), A modern Utopia,
Chapter III, 4). But at the same time they may be an accident or relict of history which
has erroneously been defended and survived previous suggestions of change: a winner of
the Nobel Prize in Literature once wrote: “[...] that all evolution in thought and conduct
must at first appear as heresy and misconduct.” (George B. Shaw (1923), Saint Joan,
Preface, The Law of Change is the Law of God.). So the questions are: Is the reason for
the current equilibrium still intact? And, since resources for reform are scarce, is there a
better equilibrium available at a reasonable price?
All this indicates that before advising institutional change, economists should inquire
very carefully into the context of the institutional setting: How do the domestic companies of the considered country trade? Have they many partners or just a few? Do they
trade a lot with foreign countries? Is the binding constraint on growth of this specific
country really the institutional sub-optimality, or is it from a holistic point of view another issue? A good economic “country due diligence” is needed, otherwise any advice,
with our model in mind, is of uncertain outcome. For what Leonardo da Vinci once noted
on the art of painters, may also be true for the art of economists:
“Those who are in love with practice without knowledge are like the sailor who gets
into a ship without rudder or compass and who never can be certain whether he is going.”
(Leonardo da Vinci, Book I, 19.)
REFERENCES

Abreu, Dilip, Paul Milgrom, and David Pearce. 1991. “Information and Timing in
Repeated Partnerships.” Econometrica, 59(6): 1713–1733.
Acemoglu, Daron, Philippe Aghion, and Fabrizio Zilibotti. 2006. “Distance to Frontier, Selection, and Economic Growth.” Journal of the European Economic Association,
4(1): 37–74.
Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2001. “The Colonial Origins of Comparative Development: An Empirical Investigation.” American Economic
Review, 91(5): 1369–1401.
Ades, Alberto, and Rafael Di Tella. 1999. “Rents, Competition and Corruption.” American Economic Review, 89(4): 982–994.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

187

Aghion, Philippe, and Jean Tirole. 1997. “Formal and Real Authority in Organizations.” Journal of Political Economy, 105(1): 1–29.
Aidt, Toke S. 2009. “Corruption, Institutions, and Economic Development.” Oxford Review of Economic Policy, 25(2): 271–291.
Akerlof, George A. 1970. “The Market for “Lemons”: Quality Uncertainty and the
Market Mechanism.” Quarterly Journal of Economics, 84(3): 488–500.
Beck, Paul J., and Michael W. Maher. 1986. “A Comparison of Bribery and Bidding
in Thin Markets.” Economics Letters, 20(1): 1–5.
B´enabou, Roland, and Jean Tirole. 2003. “Intrinsic and Extrinsic Motivation.” Quarterly Journal of Economics, 70(3): 489–520.
Blackburn, Keith, Niloy Bose, and M. Emranul Haque. 2006. “The Incidence and
Persistence of Corruption in Economic Development.” Journal of Economic Dynamics
and Control, 30(12): 2447–2467.
Bliss, Christopher, and Rafael Di Tella. 1997. “Does Competition Kill Corruption?”
Journal of Political Economy, 105(5): 1001–1023.
Brandenburger, Adam M., and Barry J. Nalebuff. 1996. Co-opetition. New York:
Doubleday.
Camerer, Colin F. 2003. Behavioral Game Theory: Experiments in Strategic Interaction. New York: Russell Sage Foundation; Princeton: Princeton University Press.
Campos, Nauro F., Saul Estrin, and Eugenio Proto. 2010. “Corruption as a Barrier to
Entry: Theory and Evidence.” IZA Discussion Paper No. 5243, 1–23.
Coate, Stephen, and Stephen Morris. 1999. “Policy Persistence.” American Economic
Review, 89(5): 1327–1336.
Cyert, Richard M., and James G. March. 1963. A Behavioural Theory of the Firm.
New York: Prentice-Hall.
da Vinci, Leonardo, and Jean P. Richter. 1883. The literary works of Leonardo da
Vinci: compiled and edited from the original manuscripts by Jean Paul Richter. London:
Sampson Low, Marston and Searle & Rivington.
de Soto, Hernando. 1989. The Other Path - The Invisible Revolution in the Third World.
New York: Harper & Row.
DiMaggio, Paul J., and Walter W. Powell. 1983. “The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields.” American
Sociological Review, 48(2): 147–160.
Dixit, Avinash K. 2003a. “On Modes of Economic Governance.” Econometrica,
71(2): 449–481.
Dixit, Avinash K. 2003b. “Trade Expansion and Contract Enforcement.” Journal of Political Economy, 111(6): 1293–1317.
Dixit, Avinash K. 2004. Lawlessness and Economics: Alternative Modes of Governance.
Princeton: Princeton University Press.
Dixit, Avinash K. 2009. “Governance Institutions and Economic Activity.” American
Economic Review, 99(1): 5–24.

188

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Djankov, Simeon, Rafael La Porta, Florencio Lopez de Silanes, and Andrei Shleifer.
2002. “The Regulation of Entry.” Quarterly Journal of Economics, 117(1): 1–37.
Gambetta, Diego. 1993. The Sicilian Mafia: The Business of Private Protection. Cambridge(Mass.): Harvard University Press.
Gardberg, Naomi A., and Charles J. Fombrun. 2006. “Corporate Citizenship: Creating Intangible Assets across Institutional Environments.” Academy of Management
Review, 31(2): 329–346.
Greif, Avner. 1993. “Contract Enforceability and Economic Institutions in Early Trade:
The Maghribi Traders’ Coalition.” American Economic Review, 83(3): 525–548.
Greif, Avner, Paul Milgrom, and Barry R. Weingast. 1994. “Coordination, Commitment, and Enforcement: The Case of the Merchant Guild.” Journal of Political Economy, 102(4): 745–776.
Guesnerie, Roger, and Jean-Jacques Laffont. 1984. “A Complete Solution to a Class
of Principal-Agent Problems with an Application to the Control of a Self-Managed
Firm.” Journal of Public Economics, 25(3): 329–369.
Hausmann, Ricardo, and Dani Rodrik. 2003. “Economic Development as SelfDiscovery.” Journal of Development Economics, 72(2): 603–633.
Hitt, Michael A., R. Duane Ireland, and Robert E. Hoskisson. 2011. Strategic Management: Competitiveness and Globalization. 9th ed. Cincinnati: South-Western College Publishing.
Hobbes, Thomas. 1651. Leviathan or the Matter, Forme and Power of a Common-wealth
Ecclesiasticall and Civil. London: Andrew Crooke.
Hymer, Stephen H. 1976. The International Operations of National Firms: A Study of
Foreign Direct Investment. Cambridge(Mass.): MIT Press.
Klapper, Leora, Luc Laeven, and Raghuram Rajan. 2006. “Entry Regulation as a
Barrier to Entrepreneurship.” Journal of Financial Economics, 82(3): 591–629.
Laffont, Jean-Jacques, and David Martimort. 1999. “Separation of Regulators against
Collusive Behavior.” Rand Journal of Economics, 30(2): 232–262.
Laffont, Jean-Jacques, and David Martimort. 2002. The Theory of Incentives: The
Principal-Agent Model. Princeton: Princeton University Press.
Laffont, Jean-Jacques, and Jean Tirole. 1988. “The Dynamics of Incentive Contracts.”
Econometrica, 56(5): 1153–1175.
Leff, Nathaniel H. 1964. “Economic Development through Bureaucratic Corruption.”
American Behavioral Scientist, 8(3): 8–14.
Levy, Daniel. 2007. “Price Adjustment under the Table: Evidence on EfficiencyEnhancing Corruption.” European Journal of Political Economy, 23(2): 423–447.
Li, John S. 2003. “Relation-Based versus Rule-Based Governance: An Explanation
of the East Asian Miracle and Asian Crisis.” Review of International Economics,
11(4): 651–673.
Lui, Francis T. 1985. “An Equilibrium Queuing Model of Bribery.” Journal of Political
Economy, 93(4): 760–781.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

189

Mauro, Paolo. 1995. “Corruption and Growth.” Quarterly Journal of Economics,
110(3): 681–712.
´
M´endez, Fabio, and Facundo Sepulveda.
2006. “Corruption, Growth and Political Regimes: Cross Country Evidence.” European Journal of Political Economy,
22(1): 82–98.
M´eon, Pierre-Guillaume, and Khalid Sekkat. 2005. “Does Corruption Grease or Sand
the Wheels of Growth?” Public Choice, 122(1/2): 69–97.
M´eon, Pierre-Guillaume, and Laurent Weill. 2010. “Is Corruption an Efficient
Grease?” World Development, 38(3): 244–259.
Mezias, John M. 2002. “Identifying Liabilities of Foreignness and Strategies to Minimize their Effects: The Case of Labor Lawsuit Judgments in the United States.” Strategic Management Journal, 23(3): 229–244.
Mirrlees, James A. 1971. “An Exploration in the Theory of Optimum Income Taxation.”
Review of Economic Studies, 38(2): 175–208.
Myerson, Roger B. 1979. “Incentive Compatibility and the Bargaining Problem.”
Econometrica, 47(1): 61–73.
Myerson, Roger B. 1981. “Optimal Auction Design.” Mathematics of Operations Research, 6(1): 58–73.
Oliver, Christine. 1991. “Strategic Responses to Institutional Processes.” Academy of
Management Review, 16(1): 145–179.
Paldam, Martin. 2002. “The Cross-Country Pattern of Corruption: Economies, Culture
and the Seesaw Dynamics.” European Journal of Political Economy, 18(2): 215–240.
Qian, Yingyi. 2003. “How Reform Worked in China.” In Search of Prosperity: Analytic Narratives on Economic Growth, ed. Dani Rodrik, 297–333. Princeton: Princeton
University Press.
Reuer, Jeffrey J., and Roberto Ragozzino. 2006. “Agency Hazards and Alliance Portfolios.” Strategic Management Journal, 27(1): 27–43.
Ricardo, David. 1817. On the Principles of Political Economy and Taxation. London:
John Murray.
Rodrik, Dani. 2006. “Goodbye Washington Consensus, Hello Washington Confusion?
A Review of the World Bank’s “Economic Growth in the 1990s - Learning from a
Decade of Reform”.” Journal of Economic Literature, 44(4): 973–987.
Rodrik, Dani. 2007. One Economics, Many Recipes: Globalization, Institutions, and
Economic Growth. Princeton: Princeton University Press.
Rodrik, Dani. 2008. “Second-Best Institutions.” American Economic Review,
98(2): 100–104.
Rose-Ackerman, Susan. 1999. Corruption and Government: Causes, Consequences,
and Reform. Cambridge: Cambridge University Press.
Shaw, George B. 1924. Saint Joan - A Chronicle Play in Six Scenes and an Epilogue.
London: Constable.

190

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

Shleifer, Andrei, and Robert W. Vishny. 1993. “Corruption.” Quarterly Journal of
Economics, 108(3): 599–617.
Smith, Adam. 1776. An Inquiry into the Nature and Causes of the Wealth of Nations.
5th ed., London: W. Strahan and T. Cadell.
Spence, A. Michael. 1974. “Competitive and Optimal Responses to Signals: An Analysis of Efficiency and Distribution.” Journal of Economic Theory, 7(3): 296–332.
Tonoyan, Vartuhi, Robert Strohmeyer, Mohsin Habib, and Manfred Perlitz. 2010.
“Corruption and Entrepreneurship: How Formal and Informal Institutions Shape Small
Firm Behaviour in Transition and Mature Market Economies.” Entrepreneurship Theory and Practice, 34(5): 803–831.
Vernon, Raymond. 2001. Sovereignty at Bay: The Multinational Spread of U.S. Enterprises. New York: Routledge.
von Hayek, Friedrich A. 1944. The Road to Serfdom. London: Routledge.
Wells, Herbert G. 1905. A Modern Utopia. London: Chapman & Hall.
Williamson, Oliver E. 1985. The Economic Institutions of Capitalism. New York: The
Free Press.
Zaheer, Srilata, and Elaine Mosakowski. 1997. “The Dynamics of the Liability of
Foreignness: A Global Study of Survival in Financial Services.” Strategic Management
Journal, 18(6): 439–464.
M ATHEMATICAL A PPENDIX - O MITTED C ALCULATIONS
A1. Preface to the Mathematical Appendix

The version of the principal-agent model, which we apply for the imperfect institution
setting in this article, is a well researched theoretical model. The derivations and calculations presented below are given for the reader’s convenience and belong to the standard
repertoire of the principal-agent realm. They build on the work of numerous researchers
of the principal-agent realm and the discovery of the revelation principle. With great respect for the field’s historical development, we predominantly rely on the approaches and
terminology used in the book of Laffont and Martimont: The Theory of Incentives - The
Principal-Agent Model, 2002, which is one of the towering standard technical references
of this field of research.
A2.

Equations (17)

In order to solve the problem of the equations numbered with (17), it is necessary to
determine which constraints are the relevant ones.
LEMMA 1: ( A .) Constraint (II) is active and equal to zero. ( B .) Therefore, constraint
(III) is also active and binding.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

191

L EMMA 1 ( A .) We begin by establishing that constraint (II) is binding. Recall that
? > ?. Observe that constraint (III) unveils us that
(A1)

t (?) ? e(?)? ? t (? ) ? e(? )? ? t (? ) ? e(? )? .

This implies constraint (II) must be active and equal to zero. Imagine constraint (II) is not
equal to zero, which would imply accordingly constraint (I) also not equal to zero, since
t (?) ? e(?)? ? t (? ) ? e(? )? . This cannot be optimal, since the initiator could profitably
lower both t (? ) and t (?) by an amount 1t without any changes in the incentive structure
and generate profits 1t. Hence, the second constraint must bind in the optimum (Laffont
and Martimont, 2002).
L EMMA 1 ( B .) From Lemma 1 ( A .) we know that t (? ) ? e(? )? = 0. Combining this
insight again with constraint (III) yields
(A2)

t (? ) ? e(? )? ? t (? ) ? e(? )? ? t (? ) ? e(? )? = 0.

Now assume for a moment that constraint (III) would not be binding. Then, by inspection
of the equation above, the initiator could decrease his payments t (? ) by an amount 1t (?)
and thereby improve his profits by ?1t (?). Clearly, this cannot be true in the optimum,
therefore the opposite must be true and constraint (III) must bind (Laffont and Martimont,
2002).
LEMMA 2: ( A .) Constraint (I) and ( B .) constraint (IV) are not relevant to the optimization problem.
L EMMA 2 ( A .) Note that t (? ) ? e(? )? ? t (? ) ? e(? )? can be expanded by +e(? )? ?
e(? )? without defects. A simple rearrangement yields
(A3)

t (? ) ? e(? )? ? t (? ) ? e(? )? + e(? )(? ? ?).

We see that constraint (II) and (III) directly imply constraint (I).28 What remains is to
validate its slackness property. Again, from Lemma 1 ( A .) we are assured that t (?) ?
e(? )? = 0. Furthermore, since ? ? ? > 0 and if we assume e(? ) to be positive, the
slackness property is fulfilled.29 Constraint (I) can therefore be omitted (Laffont and
Martimont (2002).
L EMMA 2 ( B .) Applying Lemma 1 ( B .) yields
(A4)

t (? ) ? e(?)? = t (? ) ? e(? )? ,
? t (?) ? t (? ) = (e(?) ? e(? ))?.

28 Intuitively, this is in line with our argumentation above, we argued that partner ? must be given e? (? ? ? ) to induce
truthful behavior, hence he will enjoy a positive wealth level.
29 We assume here for convenience, that entrepreneurial activity is desirable and e > 0. We shall prove this later.

192

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

However, from equation (16) we already know that e(?) ? e(? ). Combining this knowledge with ? > ? and rearranging again shows
t (?) ? t (? ) = (e(? ) ? e(? ))? ? (e(? ) ? e(? ))? ,

(A5)

?

t (? ) ? e(? )? ? t (? ) ? e(?)? .

Thereby we have shown that if constraints (II) and (III) are active and fulfilled, constraint
(IV) will hold implicitly also. Hence, we can neglect constraint (IV) (Laffont and Martimont, 2002). Applying Lemma 1 and Lemma 2 this program is solvable. However,
before determining the optimum, we will rewrite the problem in terms of wealth levels ?
and of entrepreneurial activity e. The rearranged problem of the principal reads (Laffont
and Martimont, 2002):
maximize 5 = ?(?(e(?)) ? e(?)?) + (1 ? ?)(?(e(? )) ? e(? )? )
(e(? ),? (?));(e(? ),? (? ))
? (?? (? ) + (1 ? ?)? (? ))
(A6)

subject to

? (? ) ? 0,
? (? ) ? 0,
? (? ) ? ? (? ) + e(? )(? ? ?),
? (? ) ? ? (?) ? e(?)(? ? ?).

Applying Lemma 1 and Lemma 2 this problem will shrink to
maximize 5 = ?(?(e(?)) ? e(?)?) + (1 ? ?)(?(e(? )) ? e(? )? )
(e(? ),? (?));(e(? ),? (? ))
? (?? (? ) + (1 ? ?)? (? ))
(A7)
subject to

? (? ) = 0,
? (? ) = ? (? ) + e(? )(? ? ? ).

A substitution yields
(A8)
maximize 5 = ?(?(e(? )) ? e(? )?) + (1 ? ?)(?(e(? )) ? e(? )? ) ? ?e(? )(? ? ?).
(e(? ),e(? ))

However, we assumed implicitly that it is profitable for the initiator not to exclude exploiting types from the venture.

LEMMA 3: The initiator maximizes his profits when venturing with all possible companies.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

193

L EMMA 3. Following Laffon and Martimont (2002), for Lemma 3 to hold, the profit
of venturing with both possible partner-types must exceed the profits of venturing with
the more profitable partner alone. Or expressed in other words, the difference in both
profit levels yields a positive value. This we will now demonstrate. To start with, recall
that e? = e? . Observe that venturing with both partners is optimal, as compared to with
only one partner, when
(A9)






? ?(e? ) ? e? ? ? e? (? ? ?) + (1 ? ?) ?(e? ) ? e? ? > ? ?(e? ) ? e? ? ,
??e? (? ? ? ) + (1 ? ?)(?(e? ) ? e? ? ) > 0,


?
?
?
?
?(e ) ? e ? ?
e (? ? ?) > 0,
(1 ? ?)


?
?
?
(? ? ?) > 0.
?(e ) ? e ? +
(1 ? ?)

?

?
?

Applying equation (18) to this system yields
Ã
!
?
??(e
(?
))
(A10)
?(e? ) ? e?
> 0,
?e(? )
From this rearranged equation it is possible to show that the above is true. Since
??(e)
>0
?e

(A11)
?

??(e)
?e

?e

< 0,

it follows that ?(e) ? e ((??(e(? )))/(?e(? ))) is strictly increasing in e. Since e cannot be
?
negative, and e? must have

a positive value due to equation (18), it follows that ?(e ) ?
?
?
e (??(e (? )))/(?e(? )) must be positive and it is optimal to venture with both partners
(Laffont and Martimont, 2002).30
A3. Equations (19)

In the following we will descriptively categorize the infinity of constraints of equations
(19) in order to reach a calculable version of the problem. The following considerations
build on earlier discoveries such as the revelation principle, see Myerson (1979) and
(1981), and follow an approach used in Laffont and Martimont (2002).
Participation. To start with, it is individually rational for a partner to participate in a
30 This result is not generalizable, since a shut-down may occur if the model is changed marginally, for instance if one
allows for positive levels of initial endowments of wealth ? (or positive fixed costs).

194

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

contractual agreement only if he reaches at least his status-quo level of wealth
t (? ) ? e(? )? ? 0,
?
? (? ) ? 0.

(A12)

Monotonicity. Secondly, in order to avoid that the partner pretends to be another type,
the initiator has to take into account the truth-telling constraints, such that it is optimal
for each possible partner-type to self-select the venture contract designed for its type ?
(Laffont and Martimont, 2002). Let ?ˆ ? 2 be a pretended value, then the following must
hold
(A13)

t (? ) ? e(? )? ? t (?ˆ ) ? e(?ˆ )?

? (? , ?ˆ ) ? 2 × 2.

However, if the above holds globally for all types of partners ? , then it must also hold
locally for particular pretended values ?? ? 2 such that
(A14)

t (? ) ? e(? )? ? t (?? ) ? e(?? )?
t (?? ) ? e(?? )?? ? t (? ) ? e(? )??

? pairs (? , ?? ) ? 2 × 2.

In a rearranged and added form, this system yields an additional insight
(A15)

t (? ) ? e(? )? + t (?? ) ? e(?? )?? ? t (?? ) ? e(?? )? + t (? ) ? e(? )??


?
(? ? ?? ) e(?? ) ? e(? ) ? 0.

Assuming ?? > ?, we can see


(A16)

 

? ? ??
e(?? ) ? e(? ) ? 0.
| {z } |
{z
}
?

must be ?

Therefore, the truth-telling constraints dictate that the entrepreneurial volume e is decreasing in ? (Laffont and Martimont, 2002).31 This is in line with our discussion results
in equation (16), however this condition requires all levels of entrepreneurial activity to
be ranked in a monotonic way (Laffont and Martimont, 2002)
(A17)

?e(? )
? 0.
??

Small pretensions. Following Laffont and Martimont (2002), a third point to note is
that the local truth-telling constraint t (? ) ? e(? )? ? t (?? ) ? e(?? )? is expandable by

31 While, of course, the reverse case ?? < ? yields a parallel result. Furthermore, we assume that e is differentiable
almost everywhere.

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

195

+e(?? )?? ? e(?? )?? . Rewriting this extended constraint in terms of wealth levels reveals
(A18)

? (? ) ? ? (?? ) + e(?? )(?? ? ? ).

From the point of view of the partner, this is an optimization problem with two steps, due
to the two variables. If the partner wants to maximize his profit from pretending, he will
calculate the following
(A19)

argument of the maximum ? = ? (?? ) + e(?? )(?? ? ? ).
?? ? 2

This will be maximized if ? = ?? . Therefore, the partner proceeds and derives the first
order condition to determine his optimal behaviour32
(A20)

?e(?? )
?? (?? )
+ e(?? ) +
(?? ? ? ) = 0.
? ??
? ??

If we insert ? = ?? into the above first order condition we arrive at
(A21)

?? (? )
= ?e(? ),
??

which shows us what the constraints impose on small changes (Laffont and Martimont,
2002).
Large pretensions. To be complete, we need to check whether the optimum is indeed
also a global optimum. In others words, now we have set a condition for the partner not
to pretend slightly, but we also need to exclude his ability to pretend greatly. Returning
to the global truth-telling constraint
(A22)

t (? ) ? e(? )? ? t (?ˆ ) ? e(?ˆ )?

? (? , ?ˆ ) ? 2 × 2,

we simulate a large pretension ?ˆ . In order to prove that such a large pretension is economically uninteresting for the partner, it must be that there is a positive amount of profits
not realized by the partner, an opportunity loss ? due to the pretension. Therefore we
need to prove the following (Laffont and Martimont, 2002):
(A23)

t (? ) ? e(? )? = t (?ˆ ) ? e(?ˆ )? + ?

? (? , ?ˆ ) ? 2 × 2;

? ? R+ .

LEMMA 4: There exists a ? ? R+ , when a partner is not telling the truth. Hence, the
global truth-telling constraint is fulfilled.
L EMMA 4. The initiator will offer different transfers for different types, as we know
32 Here we employ Leibniz’s law: ?ab = a?b + b?a .
?x
?x
?x

196

DISSERTATION JAN-PHILIPP AHRENS

DECEMBER 2012

from the local truth-telling constraints. Therefore a difference in transfers to the partner
can be expressed in the following way33
t (? ) ? t (?ˆ ) =

(A24)

Z

?ˆ

?

?

?e(? )
?? .
??

An integration by parts and a rearrangement of this term yields34
t (? ) ? t (?ˆ ) =

(A25)

? t (? ) ? e(? )? =
? t (? ) ? e(? )? =
{z
}
|
Truth

e(? )? ? e(?ˆ )?ˆ ?
t (?ˆ ) ? e(?ˆ )?ˆ ?
t (?ˆ ) ? e(?ˆ )? + e(?ˆ )(? ? ?ˆ ) ?
{z
}
|
|
{z
Pretension

Z

?

e(? )?? ,
?ˆ

Z

Z

?

e(? )?? ,
?ˆ

?ˆ

?

e(? )?? .
}

Loss due to pretension

R?
It remains to establish that ? = e(?ˆ )(? ? ?ˆ ) ? ?ˆ e(? )?? belongs to the realm of R+ ,
so we can indeed speak of losses instead of profits.RHowever, since e is not increasing in
?
either ? nor ? , it follows that ? = e(?ˆ )(? ? ?ˆ ) ? ?ˆ e(? )?? ? 0. Therefore, the global
truth-telling constraint for the partner is fulfilled by our approach and we can neglect it
in the following (Laffont and Martimont, 2002).
We are now in a position to collect the above results on the structure of the infinite
amount of constraints and summarize them. However, any results are only valid, if the
neglected monotonicity constraint (?e(? ))/(?? ) ? 0 holds.

LEMMA 5: The monotonicity constraint (?e(? ))/(?? ) ? 0 holds, the contract is thus
optimal.
L EMMA 5. We start by checking if e? (? ) is decreasing in ? . The derivative at e?
reads:35

 2 ?
? G(? )
? ?(e (? )) ?e? (? )
(A26)
=1+
.
2
?e (? )
??
?? g(? )
In assumption (A2.) we assumed diminishing marginal returns from entrepreneurial activity, and in assumption (A5.) we postulated a monotone hazard rate. Therefore we can
33 ? replaces ? here.
34 The version of the rule employed here:

R b ? f (x)
R b ? f (x)
?g(x)
b
a f (x) ?x ?x = [ f (x)g(x)]a ? a ?x g(x)?x = f (b)g(b) ? f (a)g(a) ? a ?x g(x)?x.
35 We apply the chain rule here.

Rb

CHAPTER 5: ENTREPRENEURSHIP UNDER IMPERFECT INSTITUTIONS

197

write
?
(A27)

?

? ? 2 ?(e? (? )) ? ?e? (? )
? G(? )
?
?
? ?e2 (? ) ? ?? = 1 + ?? g(? ) ,
| {z }
{z
}
|
?

?

+

?e? (? )
? 0,
??

which is the desired inequality for our approach to be valid (Laffont and Martimont,
2002). It is worth mentioning that our results are thus limited to these assumptions,
especially the monotone hazard rate.36

36 It turns out that, if one does not assume this, this article would plunge into fancy mathematics, which clearly would
not be suited to carve out and model the economic effects.





(UNOlUXQJJHPl??GHU3URPRWLRQVRUGQXQJGHU8QLYHUVLWlW0DQQKHLPYRP0lU]

,FK-DQ3KLOLSS$KUHQVJHERUHQDPKDEHELVODQJNHLQHQ3URPRWLRQVDQWUDJDQGHUZHLWLJJHVWHOOWE]ZKDEHPHLQH
'LVVHUWDWLRQQLFKWDOV*DQ]HVRGHU]XZHVHQWOLFKHQ7HLOHQHLQHUDQGHUHQ3UIXQJVEHK|UGHYRUJHOHJW

_________________________________


(LGHVVWDWWOLFKH(UNOlUXQJJHPl??GHU3URPRWLRQVRUGQXQJGHU8QLYHUVLWlW0DQQKHLPYRP0lU]
,FK -DQ3KLOLSS $KUHQV JHERUHQ DP  KDEH GLH YRUJHOHJWH 'LVVHUWDWLRQ VHOEVWlQGLJ DQJHIHUWLJW XQG PLFK QLFKW
DQGHUHUDOVGHULQLKUDQJHJHEHQHQ+LOIVPLWWHOEHGLHQW(QWOHKQXQJHQDXVDQGHUHQ6FKULIWHQVRZHLWVLHLQGHU'LVVHUWDWLRQ
QLFKW DXVGUFNOLFK DOV VROFKH JHNHQQ]HLFKQHW XQG PLW 4XHOOHQDQJDEHQ YHUVHKHQ VLQG KDEHQ QLFKW VWDWWJHIXQGHQ ,FK
YHUVLFKHUHGDVVLFKQLFKWGLH+LOIHHLQHUNRPPHU]LHOOHQ3URPRWLRQVYHUPLWWOXQJEHUDWXQJLQ$QVSUXFKJHQRPPHQKDEH



_________________________________



1DFKZHLVEHUGLH7HLOQDKPHDP*UDGXLHUWHQVWXGLXPJHPl??GHU3URPRWLRQVRUGQXQJGHU8QLYHUVLWlW0DQQKHLP
YRP0lU]

'HU1DFKZHLVEHUGLH7HLOQDKPHDP*UDGXLHUWHQVWXGLXPZLUGDXIGHUIROJHQGHQ6HLWHEHUGDV7UDQVFULSWRI*UDGHV GHU
JUDGXDWHVFKRRORIHFRQRPLF VRFLDOVFLHQFHVGHU8QLYHUVLWlW0DQQKHLP&HQWHUIRU'RFWRUDO6WXGLHVLQ%XVLQHVVYRP
$SULOJHIKUW



_________________________________






%
%
%
%
%
YghZij[\
Z]%^_]%`_Y\_]ab%c\d`eZc%e[%fdce[Zcc%
klmino%pq%rshhtkiu%%
v%wx%y%z{ppu%|}~y%
????%rshhtkiu%
??m???hi?ushhtkiu}?k%
tnn??????m?}?hi?ushhtkiu}?k%
%

%a????%`???????%

%Y???%£?¤??%

?????%`??%`???%c????%
a¤?¥%¦§????¥§??%
?tphk?%|w????????????% ?tphk?%|w?????????~??%
 s¡?% |w???????????¢%  s¡?% |w?????????~|?%

%

rshhtkiux%¨?li©%|x%?~??%

\??????ª?%??%«???%
¬tim%im%np%?klniqo%ntsn%­?®?¥?§?ªª%a¥???%npp¯%?sln%ih%ntk%qp©©p°ih±%?p?lmkm?%
³p?lmk%
¸hnlp???niph%np%ntk%¹?phpui?m%pq%
³phnkmnm%zs?¯hp°©k?±k?%qlpu%º?u?p©?n?
ghijklmino%¶kl©ih%
¬tkplo%pq%¸h?khnijkm%zs?¯hp°©k?±k?%qlpu%
º?u?p©?n?ghijklmino%¶kl©ih%
r¨½~?%¨?jsh?km%ih%µnlsnk±i?%
rshs±kukhn%
r¨½~?%¨??©ik?%¹?phpuknli?m%ih%
rshs±kukhn%sh?%¹hnlk?lkhk?lmti?%
{kmksl?t%

vk?n?lkl%
·ik?k%

%
µnls?m¼%
µ?uukl% ?xy% %
%
?~?~%
%
»po°p?k% º»µ%?~?~% ?x?% %
%
%
»po°p?k%  µµ%?~??% ?x?% %
%
%
%

%
%
%
¨±sntk%²©pmnklts©qkh%
z³khnkl%rshs±kl%³´µ¶%

¬klu% ·ls?k% %
%
»ihnkl% ?x?% %
?~~w??~?~%

%

01234%67%89*32+8%,-,64.%
/0123%455%67789%:%;4>4+6?@%A%;8773?@%B%;,*62,C*=679-?@%D%;C*2>?E%FG4%C7>>7H2+8%32CC494+62*627+%2,%I7,,2J>49%:K@%
AL@%AK@%BLE%B719,4,%H26G716%*+-%89*34,%*94%*,,4,,43%J-9%M%;I*,,?%79%D%;C*2>?%
4NOP%QR5%67789%S@T%;4>4+6?%67%U@T%;C*29VI*,,?E%W@T%;C*2>?E%:%32CC494+62*627+%2+%T@X%,64I,%2,%I7,,2J>4E%
%

SVS%

CURRICULUM VITAE
Jan-Philipp Ahrens

..........
Geburtstag:
Geburtsort:
Familienstand:
Staatsangehörigkeit:

17.12.1982
Stuttgart
Verheiratet
Deutsch

Akademischer Lebenslauf

..........

2010 – 2013
Mannheim

Universität Mannheim
Studiengang: Promotion Betriebswirtschaftslehre
Lehrstuhl für Mittelstandsforschung und Entrepreneurship
(Prof. Dr. Michael Woywode)

2011 - 2013
München

Ludwig-Maximilians-Universität München
Visiting Student
Courses in Econometrics (Prof. Dr. Winter)

2010 - 2011
Berlin

Humboldt-Universität zu Berlin
Visiting Student: Berlin Doctoral Program in Economics & Management
Science
Courses in Economics (Prof. Dr. Strausz & Prof. Dr. Wolfstetter)

2002 – 2008
Mannheim

Universität Mannheim
Studiengang: Diplomstudium Betriebswirtschaftslehre mit interkultureller
Qualifikation Englisch
Schwerpunkte
Organisation (Prof. Dr. Dr. h.c. Kieser)
Internationales Management (Prof. Dr. Perlitz)
Anglistik & Anglo-Amerikanischer Kulturraum (Prof. Dr. Reichardt)

2000 - 2002
England

Haileybury Imperial College
International Baccalaureate & A-Level



doc_976220409.pdf