Corporate Governance and Incentive Contracts:
Historical Evidence from a Legal Reform
Christian Bayer
Carsten Burhopyz
Abstract
First Version: June 2006
This Version: August 14, 2006
This paper proposes to employ a major shift in the legal and institutional environment to identify contractual incentives from the correlation of executive pay and
…rm performance. We use the reform of the German joint-stock companies act in 1884
as such a major shift and estimate the correlation of pay and performance between
1870 and 1910 for executives of 39 manufacturing …rms and nine banks. The reform
substantially enhanced corporate control and strengthened monitoring incentives of
shareholders. Accordingly the pay-performance correlation decreased signi…cantly after the reform. While executives received a bonus of about 31 Mark per 1,000 Mark
increase in pro…ts before 1884, the reform decreased this correlation coe¢ cient by
about two thirds for manufacturing …rms and by about one …fth for banks. This part
of the pay-performance correlation must have been an incentive component before the
reform, which is identi…ed by our di¤erence-in-di¤erence estimation approach.
JEL-Classi…cation: G30, J33, N23
Keywords: pay-performance sensitivity, natural experiment, legal reform, corporate governance
Universität Dortmund, Wirtschafts- und Sozialwissenschaftliche Fakultät, D-44221 Dortmund, Germany, e-mail: [email protected]
y
corresponding author: University of Muenster, Institute for Economic and Social History, Domplatz
20-22, D-48134 Münster, Germany, e-mail: [email protected]
z
We would like to thank Kornelius Kraft, Michael Roos, participants of the Royal Economic Society
Meeting 2004, and of the EEA-Meeting 2004, and seminar participants at the University of Dortmund for
helpfull comments and suggestions. The authors would like to thank the Deutsche Forschungsgemeinschaft
for …nancial support. An earlier version of this paper was written while Christian Bayer was Jean-Monnet
Fellow at the European University Institute and while Carsten Burhop was a research fellow at the Center
for Development Research, Bonn University. Financial support of these institutions is gratefully acknowledged.
1
1
Introduction
Incentives are at the core of economic analysis, so that a large body of theoretical and
empirical literature is devoted to their analysis. In particular, working relations have been
intensively investigated (see Prendergast, 1999, for an overview). Within working relations
individuals respond to the incentives set by their labor contracts in a rational way and
adjust their e¤orts accordingly. In turn, contracts are written with this reaction in mind
and will therefore reward performance.
This means that most empirical analysis of contractual incentives in labor relations
starts with a correlation of pay and performance, since actual labor contracts are typically
not observable to the econometrician. Though, it is not clear that incentive pay is the only
reason for the correlation, notwithstanding it being an important measure. We look at
the manager-shareholder relationship in particular. Here, a statistical correlation of pay
and performance may result from several factors, e.g. more able managers get paid more.
However, this type of correlation re‡ects no incentive phenomenon, but is present also
in a world with complete and perfect information.1 In other words, we have to separate
contractual correlation of pay and performance from pay and performance being correlated
ex post as a result of market outcome. We propose a di¤erence-in-di¤erence estimation
technique as a solution to this identi…cation problem and suggest to use a major shift in
the legal and institutional environment as an instrument for identi…cation.
Corporate and private law builds the basis on which economic agents form optimal
contracts. It de…nes the set of enforceable agreements, it sets the rules of corporate
control and hence determines a principal’s information, and last but not least, it restricts
an agent’s set of possible and legal actions. Legal rules de…ne the parameters under which
economic agents form contracts. Consequently, reforms of legal rules will in‡uence the
contracts written.2
Of particular interest for our research strategy are reforms that alter the contractual
correlation between pay and performance, but have no signi…cant e¤ect on the correlations
of pay and performance that are result of other market forces. Such reforms should
not a¤ect the abilities or other …xed characteristics of a manager which are performance
relevant and observable to the principal but not to the econometrician. This makes changes
in legal (minimum) standards of corporate governance particularly promising. These, legal
reforms have clear-cut implications on managerial contracts from a theoretical point of
1
For example, Himmelberg and Hubbard (2000) or Grossmann (2003) argue that general-equilibrium
e¤ects already induce some correlation between performance and pay even without any strategic considerations when able managers are scarce. Additionally, better quali…ed managers will work for …rms that
highly reward performance. Thus, incentive contracts become a selection device and increase performance
but do not induce managers to exert more e¤ort (Lazear 1986, 2003).
2
For example, Conyon (1997) and Core et al. (1999) …nd that changes in corporate governance signi…cantly in‡uence the level of managerial compensation.
2
view and as such can serve as a test for the incentive theory of contracts itself.
Yet, it is rare to observe a major change of legal institutions qualifying for this testing
strategy. First of all, the legal reform must lead to unambigous predictions for parameters
of contracts between principals and agents. Typically, only few and substantial refoms
thus qualify. Secondly, we need to observe the contractual relationship for a prolonged
period of time before and after the reform. For example, the Sarbanes-Oxley-Act of 2002
may well qualify as a major reform, but for our testing strategy the time period after the
reform has been too short to understand its long term implications as of now. In short, our
testing strategy needs a historical perspective. This however brings up a third restriction
in the choice of a legal reform as an identi…action instrument. Both, before and after
the reform the data base of the analysis has to be sound and will typically be based on
accounting data.
This leads us to consider data from the German Empire before World War I. Legal
minimum standards on accounting and publication of company accounts were introduced
in Germany relatively early (in 1861) compared to their introduction in Britain (1900/06),
in Massachusetts (1903), and in the US in general (1933).3 Furthermore, there was a
legal reform in Germany in 1884 that primarily aimed at a new regulation of IPOs but
substantially reshaped corporate law. This reform can be expected to have a substantial
impact on contractual relationships between managers and owners of a …rm. Because of
this, we have sampled a data base that allows us to analyze the impact of the 1884 reform
on managerial contracts and therefore allows us to test for the incentive character of the
performance and pay correlation. Our data embraces information on …rm-level accounting,
stock market, and management compensation for 38 manufacturing companies and nine
banks for the time period between 1871 and 1911.
Our major …ndings are the following: First, we observe a substantial correlation between …rm performance and executive compensation for German manufacturing …rms and
banks for the period 1871-1913. Second, a substantial part of this correlation can be
interpreted as an incentive component. The 1884 legal reform reduced the correlation
coe¢ cient between pay and performance of executives by about two thirds for manufacturing …rms and by about one …fth for banks. Finally, this result is stable with respect to
di¤erent estimation techniques, sample splits, and control variables.
The remainder of the paper is organized as follows: Section 2 reviews the related
literature and discusses the theoretical ideas upon which we base our empirical analysis.
3
Already, in 1869 the NYSE Committee on Stock Lists requires disclosure of …nancial conditions. Until
1880 the NYSE Committee on Stock Lists introduced a requirement to publish Statement of Conditions
and List of corporate o¢ cers. However, until 1910 there is a signi…cant number of unlisted companies that
are traded to which these conditions do not apply. In 1910 the NYSE closes its unlisted department and
most …rms apply for listing (Simon, 1989, p 289)
3
Section 3 gives a summary over the historical setting from which we take our data. It
particularly focuses on the legal reform that forms the key ingredient to our identi…cation
strategy. Section 4 describes our data, Section 5 presents the results of our empirical
analysis, and …nally Section 6 concludes.
2
Literature, Theory, and Econometric Background
First empirical studies on managerial incentive pay concentrated on the reduced form
relation of pay and performance. Particularly, they focus on the remuneration of top
executives of publicly traded companies. Jensen and Murphy (1990) report in their seminal
article that compensation of CEOs increases by 3.25$ per 1000$ increase in shareholders’
wealth, which has been perceived as overly weak incentives for executives. Since many
factors shape the contracts, it is hard to argue what would be a good or optimal payperformance sensitivity. In any case, Hall and Liebmann (1998) …nd an increase in the
pay-performance sensitivity for the late 1990s. In between these two seminal papers, a
large strand of research got triggered by the …ndings of Jensen and Murphy (1990).4
More recently, a number of authors has emphasized potential problems of identi…cation that may strain this …rst generation of empirical articles on pay for performance. A
correlation of pay and performance can result from a number of other factors besides managerial contracts that reward performance to induce e¤ort. For example good managers
may simply be paid better since they deterministically perform better. In a similar fashion, Himmelberg and Hubbard (2000) argue that business or sectoral cycles will induce a
pay and performance correlation as demand for managerial services rises during booms.
To achieve identi…cation nonetheless, a typical estimation strategy has become the use
of di¤erence-in-di¤erence methods. For instance, Baker and Hall (2004) look at the di¤erence in compensation of manangers of large and small …rms. Elston and Goldberg (2003)
analyze the in‡uence of ownership structure on the pay performance relation. Cunat and
Guadalupe (2005) use a shift in exchange rates as a quasi-natural experiment to analyze
the impact of an increase in competitiveness of a market on managerial compensation
schemes. Most closely related to our approach is the paper by Talley and Johnsen (2004)
which analyzes the interaction of corporate governance, executive compensation and securities litigation. They provide a detailed model of this interaction and …nd that corporate
governance serves as a substitute for incentive pay, both, empirically end theoretically. But
the changes in corporate governance that Talley and Johnsen (2004) analyze are mostly
4
Following Jensen and Murphy’s approach, Schwalbach and Graß
ho¤ (1997), and Kraft and
Niederprüm(1999a) and (1999b) investigate the link of managerial remuneration and company performance
for Germany. Like Jensen and Murphy …nd for the US, they also …nd a rather weak pay-performance sensitivity.General overviews of the executive remuneration and corporate governance literature are given by
Core et al. (2003), Murphy (1999), Tirole (2001), and Shleifer and Vishny (1997).
4
Figure 1: In‡uences of corporate law on managerial contracts
endogenously adopted, which makes identi…cation more di¢ cult.5
We put the identi…cation problem back to the focus follow the di¤erence-in-di¤erence
approach, too. We employ a substantial reform of the legal and institutional environment
as an instrument for identi…cation. We investigate how the legal reform that we look at
changes the pay performance correlation in particular.
Corporate and private law can be understood as the ground upon which principals and
agents base their contracts. It de…nes the set of enforceable agreements, it sets the rules of
corporate control and hence determines a principal’s information, and last but not least,
it restricts an agent’s set of possible and legal actions, see Figure 1.
Consequently, any shift in the legal rules that de…ne corporate governance can be
expected to in‡uence the agent’s contract. At the same time, legal reforms of coporate
governance will typically not a¤ect the abilities or other …xed characteristics of a manager
which are performance relevant and observable to the principal but not to the econometrician. Therefore, we can exploit a shift in legal rules for identi…cation. The legal reform
that we look at may be summarized as one that gave more sticks to the hand of principals,
so that we can expect the use of incentive pay to be decreased by the reform (see next
section).
This identi…cation strategy may be expressed in the following formal way. Suppose
that
wit =
0
+
it
5
+
i
+ uit
(1)
Also the article of Joskow et al. (1996) analyzes the e¤ects of regulation on incentive pay. Yet its focus
is rather on political in‡uence on compensation packages than on structural identi…cation.
5
de…nes the employment contract of a manager, where wit denotes the compensation of
manager i at time t: The compensation package is linear and depends on the published
pro…t
it
of the …rm where manager i is employed in period t with the pay-performance
sensitivity
plus the random components
and uit : Here
i
i
captures the unobserved
characteristics of the manager.
The most straightforward empirical approach would be a direct estimation of (1) by
means of some least-squares estimator. However, this approach will not necessarily identify incentives
, because pro…ts
it
outlined before. Nonetheless, if
identify at least the change in
and the error-term (
i
+ uit ) may be correlated, as
varies with changes in legal institutions, we can still
using a di¤erence-in-di¤erence approach.
Consider two sub-periods, period 0 and 1 ; for which
the simple OLS estimator ^ j asymptotically converges to
^
j
=
j
+
cov ( it ; uit +
var ( it )
The di¤erence of the estimates ^
^
0
1
i)
=
j
+ corr (
is only biased if corr (
di¤ers. For each period
it ; uit
+
it ; uit
+
j
i)
i)
is not constant
for both sub-periods. In fact, we can expect the correlation of the error terms and pro…ts
corr (
it ; uit
+
i)
not to be in‡uenced by legal reforms most likely. This correlation
captures the relationship of individual characteristics and performance or business cycle
e¤ects for example. As a result, the unbiased estimate of the di¤erence ^ 0 ^ 1 can serve
as a lower bound estimate for the sensitivity of pay with respect to performance that is
meant to induce incentives.
To put this identi…cation strategy to work, we estimate
wit =
0
+^
thus. In this equation It2
the reform (t 2
1)
1
0
it
+ ^
^
1
0
it It2
1
+ xit +
i
+ uit
(2)
denotes an indicator function that takes the value 1 after
and zero otherwise. The vector xit represents other covariates that
could in‡uence bonus payments. On a more technical side, censoring does complicate the
analysis somewhat. The managerial compensation wit is always positive, so that we have
to estimate equation (2) by a random e¤ects tobit estimator.
3
Historical background
Germany’s …nancial history of the 1870s and 1880s is shaped by the repercussions of the
so-called Gründerzeit (foundation or promotion period). Between 1871 and 1873, the time
immediately after the foundation of the German Empire in 1871, 843 joint-stock companies
were founded in Germany. Out of these, at least 442 were quoted at the stock market. This
6
period was followed by the so-called Gründerkrise (foundation crisis) of 1873-79, during
which several hundreds of the new joint-stock companies failed and 225 companies were
de-listed from the stock market. Classical accounts by contemporaries as well as recent
scholarship relate the boom of 1871-73 to the introduction of free incorporation of jointstock companies in Germany during 1870, and to a number of accompanying economic and
psychological factors, e.g. the optimistic public mood after the foundation of the German
Empire and the substantial indemnity paid by France after the German-French War of
1870-71 (see e.g. Henning 1995, pp 205; Kiesewetter 2004, chapter 5; Berghahn 1994, pp
11-17).
Already contemporary observers emphasised that the hot market for IPOs had negative
long-term consequences for the German economy. Several hundreds joint-stock companies
were founded within three years, several hundreds of them failed during the following few
years. This was perceived as a deep crisis of modern corporate capitalism. And it was a new
phenomenon for contemporaries: before 1870, the foundation of a joint-stock company was
in most German states only possible with a royal concession. Such a concession was often
di¢ cult to obtain and the number of joint-stock companies in Germany was very small.
This changed in 1870, when free incorporation of joint-stock companies was implemented.6
The boom-and-bust cycle of the German IPO market of the early 1870s encouraged
emotional debates in German parliaments, the …nancial press, and between academics. For
example, both, Wilhelm Oechelhaeuser (1876), a German parliamentarian, law professor,
and expert on joint-stock companies, and Ernst Engel (1875), president of the Prussian
statistical o¢ ce, point to the importance of the 1870 legal reform, and the low moral
standards of company founders.
One major component of the criticised law was the switch from a concession to a
normative system of company foundation. Before the 1870 reform act a royal concession,
which was di¢ cult to obtain, was necessary in most German states to found a jointstock company, whereas after the liberalisation nearly every citizen could found such a
company. Furthermore, the new law abolished the extensive audit of foundation projects
by public authorities and the state supervision of joint-stock companies. Instead, company
supervision was handed-over to a shareholder body, the supervisory board (Aufsichtsrat),
see e.g. Renaud 1875 or Gareis 1880 for textbook treatments of Germany’s corporate law.
Thus, the 1870 reform established the German system of corporate control of today;
which di¤ers from the UK and US system (see Fohlin, 1998, and Fear and Kobrak, 2006,
for details). The reform introduced a two-tier board system, a system of an executive and
6
This means also that from 1871 on corporate law was uniform for the German Empire and a federal law.
This compares to the US where free incorporation was introduced …rst in Conneticut in 1837. Since then
regulation of corporations is state wise in the US, which historically lead to a more lessez-faire treatment
in standards (Grandy, 1989).
7
a supervisory board. The executive board (Vorstand ) is responsible for the daily management of the …rm, whereas the supervisory board (Aufsichtsrat) monitors the executive
board and takes part in strategic decision-making. It is thus in a way comparable to the
function of non-executive directors.
According to the 1870 law, the supervisory board was a permanent board of shareholders that should monitor the executive board. On this activity, it had to report at
the annual meeting of shareholders (§ 225 ADHGB7 ; Renaud 1875, pp 626; Gareis 1880,
pp 214). However, shareholder’s in‡uence on the composition of the supervisory board
could be restricted. Only the …rst supervisory board had to be elected by a shareholder
meeting. Thereafter, the supervisory board could coopt further members at will without
the shareholders taking part in this decision (Renaud 1875, p 628). The …rst shareholder
meeting was regularly held before the …rm was quoted on the stock exchange and thus
the founders e¤ectively appointed the …rst supervisory board. Therefore, the actual shareholders were not necessarily represented at the supervisory board, a momentous restriction
of monitoring rights.
Responsible for the daily management of the …rm was the executive board, the Vorstand
(§§ 209, 227 ADHGB; Renaud 1875, pp 524; Gareis 1880, pp 210, Makower 1877, p 224).
However, a transfer of nearly all decision rights to the supervisory board was possible. If
this was done, nobody could be held legally accountable for corporate activities. Even
intentional misbehavior had no legal consequences then, because at most members of the
executive board could be punished for intent.
The only way for shareholders to exercise their rights was to vote at the annual general
meeting, the Generalversammlung (§§ 209, 224 ADHGB; Renaud 1875, pp 458; Gareis
1880, pp 207). The legal default was that every share had one vote at this meeting.
However, voting and participation rights could be restricted by the company’s charter.
For example, the number of votes of each shareholder could be limited, or the number of
shares giving one vote could be larger than one. The …rst regulation restricted the voting
power of block-holders, whereas the latter one annulled the voting rights of small-holders.
Additionally, information rights of shareholders were weak under the 1870 legislation.
Only publication of a balance sheet was stipulated, but it was neither necessary to release
a pro…t-loss statement nor to release a management report. Furthermore, while valuation
of assets was regulated, non-realized pro…ts could be balanced leaving discretionary power
with the executive and supervisory boards.
Altogether, corporate governance under the 1870 law was insu¢ cient: voting and participation rights of shareholders at the annual meeting could be restricted, the supervisory
board was not necessarily elected by shareholders, the power sharing was not clear-cut be7
ADHGB = Allgemeines deutsches Handelsgesetzbuch (General German commercial law).
8
tween the weakly accountable executive board and the not accountable supervisory board,
share-holders were not necessarily well informed.
Nonetheless, the contemporary debate about the de…cits of the 1870 act centered
on problems of IPOs. There was substantial asymmetric information between company
founders and stock market investors, which led to information rents for founders, giving
them an incentive to found unsound companies. Already in 1872-73, contemporaries were
strongly alarmed with the insu¢ cient regulation of company foundation.8
As a reaction, a new company law was enacted in 1884. The focus of the reform of law
was on company foundation (see Deutscher Reichstag, 1884), nonetheless corporate governance standards were raised. The 1884 joint-stock companies act signi…cantly improved
monitoring and information rights of shareholders and it broadened accountability of the
two boards.
Since 1884, every shareholder has voting and participation rights at the annual meeting,
but still a restriction of the maximum number of votes of a shareholder was possible
under the 1884 law (§§ 190, 221 ADHGB). With the new legislation, every member of the
supervisory board had to be elected by the shareholders (§§ 191, 224 ADHGB) and could no
longer be coopted. Finally, a clear and strict legal separation of supervisory and executive
board was established (§§ 192, 225a ADHGB). The executive board was responsible for
the daily management and strategic decisions. The supervisory board’s main duty became
monitoring, but it maintained a right to participate in strategic-decision making.
Turning to information rights, the publication of a balance sheet, of a pro…t-loss statement, and of a management report became compulsory with the new 1884 legislation.
Moreover, asset valuation was more strongly restricted. Basically, only realized pro…ts
could be balanced. It was thus not possible to generate book pro…ts and consequently
distribute these unrealized pro…ts to shareholders and managers. Altogether, scope and
quality of information rights improved after 1884.
Also the threat of punishment in case of executive’s misbehavior was altered by the
reform (Deutscher Reichstag 1884, pp 343). First of all, the new law clari…ed that intentional misbehavior of executives or supervisory board members was a criminal o¤ence,
which was to be punished by imprisonment and a …ne (§ 249 ADHGB). In particular, the
maximum penalty for intentional misreporting of accounting …gures was increased from
three months imprisonment and a 3,000 Mark …ne to one year imprisonment and a 20,000
8
See e.g. Glagau (1876) for a polemic account or Deutscher Reichstag (1873) for an early parliamentary
discussion. The discussion of the 1870s is summarised in the motivations and reasons for the 1884 reform
act (Deutscher Reichstag, 1884). The motivation starts with a discussion of the general need for reform (pp
236-246), followed by discussions about the character of shares (pp 246-260), the foundation of a company
(pp 261-281), issuance of equity (pp 282-286), the relationship between boards and shareholders after the
foundation (pp 286-301), and accounting standards (pp 301-306).
9
Mark …ne (§ 249a ADHGB).9 In addition, share price manipulations and the selling of
votes for the annual meeting was interdicted (§§ 249d, 249e ADHGB).
Although the e¤ectiveness of the 1884 act was questioned by some contemporaries, this
discussion focused rather on possible extensions of the reform, than on the general direction
of the reform itself (Warschauer 1902; Philippovich 1909). In general, the reform was
perceived as an improvement of corporate control and corporate governance (Hessberger
1889, p 57). In particular, the discretionary power of managers declined together with
higher costs of misbehavior after 1884 and this can be expected to have a substantial
in‡uence on the economic relationship between principals and agents.
What is important for our research question and strategy is to keep in mind that while
the reform in 1884 was mainly intended to solve problems of corporate foundation, it had
large and important implications also for corporate governance. It is unlikely that the
reform was motivated by an excessive management compensation. In fact, the legislators
did not consider existing compensation schemes or problems with those to be a reason
for the 1884 reform at any point.10 Therefore, the reform should qualify as a natural
experiment for compensation schemes. The improvement of corporate control can be
expected to in‡uence compensation schemes o¤ered to managers. At the same time, it
is not a result of management compensation that paid out a too large share of pro…ts to
managers.
4
Data Description
4.1
Sources
Our data base is composed from a number of sources for accounting, stock market, and
executive data on German corporations between 1871 and 1911.
The main data source are ”Saling’s Börsenpapiere”, a stock market annual for Germany which was occasionally published before 1880 and published annually from that
year onwards until World War I. This widely circulated handbook was targeted towards
investors and bankers. It contains accounting and stock market data for all companies
quoted at the Berlin stock exchange. In addition, it provides information about the company’s charters that is relevant to potential investors, e.g. information about voting rights
and distribution of pro…ts. Finally, it reports the names of the executive board members.
Today, this rich source is an important data base for quantitative economic history (see
9
These …gures compare to a per capita NNP of roughly 400 Mark annually or a typical compensation
for executive board members in manufacturing of roughly 15 000 Mark. Legislators defeated a higher
maximum penalty since this would immoderately weaken monitoring incentives for shareholders.
10
The only remark in the whole motivations and reasons for the 1884 reform that comes close to a debate
on compensation schemes is a debate on compulsory shareholding of executives. The act …nally decided
against such a compulsory requirement, so as to not interfere in the optimal allocation of managers to
companies (Deutscher Reichstag, 1884, p 332).
10
e.g. Tilly, 1989; Edwards and Ogilivie 1996; Fohlin 1998; Baten and Schulz 2005).
To make this data usable, we partly rely on data sets collected from this source by
Rettig (1978) for manufacturing, mining, and non-…nancial services and Bittner (2005)
for mining and manufacturing. Rettig (1978) structured his sample as to represent the
size and geographical variation of …rms reported in ”Saling’s Börsenpapiere”in 1900. Bittner (2005) adds some manufacturing …rms to Rettig’s sample using the same criterion.11
Rettig (1978) and Bittner (2005) focus on accounting and stock market data. We supplemented their data sets with information about bonus payments (Tantieme) to executive
and supervisory board members. In addition, we collected the names and tenure of each
executive. Moreover, we collected information from the company charters for all …rms for
the years 1871 until 1911. In particular, we investigated the rules of the company charters
when it came to voting rights and pro…t distribution.
In particular, rules of pro…t distribution are important to construct our …nal data. All
…rms announced a minimum dividend of four to six percent of the paid-up share capital as
a guaranteed dividend. The pro…ts above this threshold level are the determination base
for bonus payments to executives. Therefore, our pro…t measure is the accounting pro…t
less the announced minimum performance.
The data sets provided by Rettig (1978) and Bittner (2005) start in 1880. Given that
we want to analyze the impact of a reform in 1884, it is necessary to prolong Rettig’s
and Bittner’s sample backwards from 1880 until 1871. If the company had been founded
after 1871, we tried to obtain all data back until foundation. The relevant accounting,
stock market, and executive data for the 1870s were collected from the few available issues
of ”Saling’s Börsenpapiere” and from the ”Berliner Börsenzeitung”, Germany’s leading
…nancial daily at the time.12
Furthermore, we add some banks to the sample, which were neglected by Rettig (1978)
and Bittner (2005). The relevant data were collected from two secondary sources. Bosenick
(1912) presents accounting and stock market data for the nine largest German banks –the
so-called ”great banks”for the years 1871-1910 (see Fohlin 1998). We augment Bosenick’s
data by information on board membership reported by Reitmayer (1999), who gives these
data for the years 1871-1913.
While most data items could be used as reported in the sources, our main variable of
interest, bonus payments to executives, had to be reconstructed in few cases. Some …rms
report only the bonus payment to both boards, the executive and supervisory board, as
a whole. Since the bonus payments to the supervisory board were governed by a rule
11
Moreover, he adds a substantial number of mining companies which we do not use further.
“Saling’s Börsenpapiere” are available for 1870,1873, 1875, and 1878. Many corporations published
their accounts in the “Berliner Börsenzeitung”. However, collecting data from this source is rather challenging and the resulting data set therefore less comprehensive.
12
11
that was stated in the company’s charter, we could reconstruct the bonus payment to
the supervisory board on this basis. The bonus paid to the executive board was then
calculated as the residual of the bonus to the two boards.
From our base sample of …rms, we remove all companies that are public utilities,
transport, or mining, because these industries were either regulated by the government
(see Milward 2004) or cartelized (see Peters 1989).13 Moreover, we remove all observations
where pro…ts or bonus payments fall outside a
2:5 standard deviations intervall around
their …rm-speci…c mean values. This reduced the sample to 48 corporations, 39 from
manufacturing and nine banks with a total of 855+324 observations.
4.2
Data Quality
For the US and Britain, historical accounting data is known to be of problematic quality,
since accounting principles and coded standards developed mostly during the 20th century. By contrast, most of todays’s German system of accounting principles and its legal
codi…cation was developed already about 50 years earlier in the late 19th century.
British companies limited by shares had to publish annual balance sheets only from
1900 onwards. Accounting standards were introduced in 1906. However, already before
that date, accounts were generally inspected by auditors (Matthews et al. 1997).
In the US, disclosure requirements were not introduced before World War I in almost
all states. The …rst state introducing compulsory disclosure was Massachusetts in 1903.
This re‡ects that also, incorporation standards were comparatively high in Massachusetts.
However it also meant that most companies choosed to incorporate in other states, e.g.
Delaware and New Jersey. To counteract the resulting problem of information quality,
the New York stock exchange attempted introducing publicity of listed companies in 1900
since the board of the New York stock exchange noted that most of the largest listed companies did not publish accounts. Nevertheless, the stock exchange was not very successful
enforcing the new regulation and some …rms preferred de-listing over disclosure. Moreover
while the New York stock exchange introduced disclosure, it did not enforce accounting
standards (Baskin 1988).
This di¤ers substantially for Germany. German shareholders and the general public
were much better informed about the …nancial situation of companies than their British
or American counterparts in the 19th century. Compulsory publication of an annual
balance sheet and minimum accounting standards were introduced in Germany already
in 1861. The publication of an annual pro…t and loss statement was prescribed from
1884 onwards, but many corporations published pro…t-and-loss statements before that
date. Moreover, accounting standards were clear and undisputed and nearly at the same
13
Joskow et al. (1996) show that public regulation signi…cantly alters compensation schemes.
12
standard than accounting standards in modern Germany (Fear and Kobrak 2006). The
general acceptance of accounting standards may for example be judged by the number of
court trials regarding accounts. Between 1871 and 1913 only twelve lawsuits were tried
at the Imperial court (Reichsoberhandelsgericht, from 1880 Reichsgericht), the highest
German civil court.14
4.3
Descriptive Statistics
Table 1 presents descriptive statistics of our data set. The sample period runs from 1871
until 1913 and contains 1,179 observations. Out of these, 855 are for manufacturing …rms
and 324 for banks. However our analysis focusses on the manufacturing sample. The …rm
size of the manufacturing …rms is measured by the value of total assets and the stockmarket value.15 On average, manufacturing …rms had total assets of about 11.3 million
Mark and a stock market value of about 8.2 million Mark. Banks were in both dimensions
much larger and had average total assets of about 298 million Mark and a stock market
capitalization of circa 95.6 million Mark.
Performance relevance for bonus payments is measured by the accounting pro…t less
the guaranteed minimum dividend payments. On average, this variable was slightly less
than 500,000 Mark for manufacturing …rms and about 4.2 million Mark for banks. Out of
these excess pro…ts, each executive of a manufacturing …rm received an average bonus of
about 11,400 Mark, whereas an executive of bank received a bonus of on average nearly
144,000 Mark.
For the manufacturing sample, we include dept over total assets as to proxy for the
dependence of manufacturing …rms on external …nance. Firm relying more on external
…nance can be expected to be monitored more closely by the debt holders. Hence, they
may o¤er di¤erent types of contracts to their managers.
Of particular interest is a sub-sample of our manufacturing …rms. Typically, the board
of executives consists of a number of members and we only have information about the
average payment made to all memebers of the board. However, there are a number of
…rms that have a board of executives consisting only of a single CEO. In these cases, we
observe the actual bonus payment of this CEO, so that we can check for the robustness of
our …ndings to compositional considerations on the basis of this sub-sample. Compared
to the whole sample–of course–single manager …rms are smaller measured by the value of
assets and stock market capitalization on average.
14
15
Eight lawsuits refer to the pre-1884 legislation and only four the post-1884 legislation.
All data are de‡ated to 1913 prices using Ho¤mann’s (1965) NNP de‡ator.
13
Table 1: Descriptive Statistics
Total asset (in million Mark)
Shareholder value (in million Mark)
Share of debts in total assets (in percent)
Pro…ts above treshold level (in Mark)
Bonus per executive (in Mark)
Number of observation
5
Manufacturing
Mean Std.dev.
11.3
17.2
Banks
Mean Std.dev.
298.0
291.0
8.2
14.4
95.6
77.5
34.1
14.8
n.a
n.a.
497 102
955 898
4 243 720
4 803 932
11 444
17 860
134 804
148310
855
324
Estimation
5.1
Main results
Our basic speci…cation of the estimation is a random e¤ects Tobit model where we use
the whole universe of manufacturing …rms that is in our sample. We use all data on these
…rms from 1870 until 1910. The Random-E¤ects-Tobit speci…cation captures the crosssectional heterogeneity of the sample as well as the fact that a bonus paid to executives
is non negative by de…nition. The regression model thus is
wit = max 0;
0
+
0
it
+
1 it It 1885
where wit is the average bonus paid to an executive,
it
+ xit +
i
+ uit ;
is pro…ts, and xit are other controls.
However, we check for the robustness of our …ndings from this speci…cation and complement it by four additional model estimations in the section following this one. The
…rst robustness check is to see whether the results found are speci…c to the manufacturing
sector by using the data on banks that we have for comparison. Secondly, we check if
our results are driven by compositional e¤ects as the size of boards varies over time. For
this reason we look at those …rms with only a CEO and no other executives. Thirdly, we
check whether our results are sensitive to the assumption of homoscedasticity among the
two sub-period. For this purpose, we split the sample in the two sub-periods 1870-1884
and 1885-1910 and perform the estimation for both sections separately. Finally, we check
if it matters for our results that the contracts may have a dynamic character. Therefore
in this …nal speci…cation, we consider a dynamic setup using a two-step Arrelano-Bond
(1991) type GMM estimator. This estimator, however, cannot account for the censoring
14
Table 2: Basic regression results: Random e¤ects Tobit-estimator
Manufacturing
Coe¢ cient p-value
0.0310
0.000
Manufacturing, reduced
Coe¢ cient
p-value
0.0312
0.000
-0.0197
0.008
-0.0189
0.007
9040
0.000
10 108
0.000
-0.0010
0.000
-0.0012
0.000
Shareholder value
0.0000
0.9150
Debt ratio*pro…ts
-0.0016
0.8670
Debt ratio
12 343
0.0970
3 383
0.466
-33.2216
0.9910
6752
0.006
Pro…ts
Pro…ts*post-1884
Post 1884
Assets
Constant
No. of obs.
No. of groups
Wald
2
Log-likelihood
850
883
38
39
90.47
144.5
-5 970.237
-6 163.87
of the left-hand-side variable. Taking into account both, dynamics and censoring is not
feasible with the number of observations available. To all four alternative speci…cations,
our qualitative …ndings are robust.
The fundamental regression results are displayed in Table 2. For manufacturing …rms,
an increase of excess pro…ts of 1,000 Mark leads to an additional bonus payment of about
31 Mark to each executive. The 1884 reform act dummy variable interacted with the excess
pro…t variable has, as expected, a negative sign and is highly signi…cant for both sectors.
For the manufacturing sector, the coe¢ cient is between 18.89 Mark and 19.70 Mark per
1,000 Mark additional pro…ts. This means that the 1884 joint-stock companies reform act
reduced the pay-performance correlation by about two thirds. These coe¢ cients form the
minimum level of incentive related compensation in manufacturing and banking respectively. The reform improved corporate governance mechanisms. Monetary motivation of
managers was substituted by better institutions, inducing a fall in the pay-performance
correlation. Only this shift of the correlation coe¢ cient can be interpreted as an incentive
15
pay during the pre-reform period.
From our set of control varianles, …rm size measured by shareholder value has no
signi…cant in‡uence on the pay-performance relationship. Therefore, we do not consider
this variable further. The second …rm size measure, total asset, has a statistically as well
as economically signi…cant in‡uence on the pay-performance relationship. An increase of
total assets by 1000 Mark decreases bonus pay by about 1.19 Mark. The coe¢ cient is
negative, implying the larger …rms reward performance less and ”Empire building”is not
rewarded.
The share of debts in total liabilities has no in‡uence on the steepness of the relationship between bonus payments and excess pro…ts, but in one speci…cation a weakly
signi…cant in‡uence on the level of bonus payments. Higher level of debts should re‡ect
monitoring incentives of banks and other debt holders. The close monitoring by bankers
and the higher liquidation risk of highly leveraged …rms should reduce the need for monetary motivation of managers. However, the e¤ect is instable and we do not consider it
further.
Finally, bonus payments are signi…cantly higher in the post-1884 period than before.
In particular, manufacturing …rms pay on average 9-10 000 Mark more in the later period
than in the earlier period. This higher level of bonuses re‡ects the overall growth of …rms
and the economy.
To sum up, manufacturing …rms pay bonuses related to excess pro…ts of the …rms.
The correlation coe¢ cient of the bonus-excess pro…t correlation signi…cantly falls after
the implementation of better corporate governance institutions in 1884. This reform did
not in‡uence characteristics of mangers and the e¤ect can therefore be interpreted as an
incentive pay paid before the implementation of the new law.
5.2
Robustness tests
As a …rst robustness check we look at the sample of data from the banking sector that
we have collected for comparison. The results are displayed in Table 3. For banks, the
the pre-1884 bonus share in pro…ts is with 32.90 Mark for each executive per 1000$ pro…t
nearly identical to the pre-1884 pro…t share in manufacturing. The reform in 1884 has a
qualitatively similar e¤ect in the banking sector as in manufacturing. However, quantitatively it is only roughly a third, and the post1884 pro…t share that is distributed to the
manegers is only 6.88 Mark per 1,000 Mark lower than before 1884.
Our second robustness check investigates, if the reform e¤ect found may be due to
compositional in‡uences. We investigate manufacturing corporations managed by one
executive separately to account for the fact that we do not know the exact distribution
of pro…t shares within the board. From our data, we only know the total bonus payment
16
Table 3: Robustness check I: Evidence from Banking and Single Manager Firms, Random e¤ects
Tobit-estimator
Pro…ts
Coe¢ cient
0.0329
Banks
p-value
0.000
One-manager-…rms
Coe¢ cient p-value
0.0412
0.000
Pro…ts*post-1884
-0.0069
0.032
-0.0261
0.016
Post 1884
46511
0.000
10,039.90
0.002
Assets
-0.0001
0.006
-0.0001
0.649
2,205.89
0.809
2,282.38
0.506
Debt ratio
Constant
41 360
No. of obs.
324
365
No. of groups
9
26
2
566.7
66.62
Log-likelihood
-3 716.3868
-2,682.21
Wald
0.000
to the whole executive board and the number of board members. We than assume an
equal distribution of pro…ts shares between executives. Yet, it might be possible that one
executive, the CEO, received a higher share in pro…ts than other executives. In this case,
changes in the composition of the board may be driving our results. To account for this,
we construct a sub-sample of manufacturing …rms managed by only on executive. About
40 percent of our observations belong to this category. Our main result, a signi…cant
in‡uence of the 1884 legal reform on executive compensation, is also visible in this subsample. The pro…t share payed to these CEOs before and after 1884 is larger than the
average pro…t share of executives for all manufacturing …rms (0.041 and 0.015 vs. 0.031
and 0.011). However, also in this sub-sample the reform has a drives down the pro…t share
tremendously. In fact, the absolute reform e¤ect is even larger for the single-manager-…rms
(0.026 vs. 0.020). However, the relative e¤ect of the reform is about the full sample as
well as for the single manager sample. The legal reform reduced the correlation between
excess pro…ts and bonus payments by about two thirds.
Our third robustness check is to compare the two sub-samples, using all pre-1884 and
all post-1884 observations for manufacturing …rms but splitting the sample at the time
of the reform, performing the regression analysis separately for both periods. Results are
17
Table 4: Robustness check II: Sample Split for both Sub-Periods, Random e¤ects Tobit-estimator
Pro…ts
Pre-1884 observations
Coe¢ cient
p-value
0.0301
0.000
Post-1884 observations
Coe¢ cient p-value
0.0107
0.000
Assets
0.0002
0.716
-0.001
0.000
Debt ratio
-2,608.18
0.874
25,073.62
0.000
Constant
-5,522.33
0.345
2,478.57
0.367
No. of obs.
150
733
No. of groups
31
39
2
19.84
67.82
Log-likelihood
-890.59
-5,262.06
Wald
displayed in Table 4. This sample split approach tells us whether the homoscedasticity
assumption for both time periods may be crucial. The estimated relationship between
excess pro…ts and bonus payments is in accordance with the previous results much larger
in the pre-1884 period than after the corporate governance reform. The di¤erence between
the two point estimates is about 0.02, which almost exactly equals the e¤ect measured by
the dummy-variable approach used in the pooled regressions. However, the sample split
approach allows to observe other di¤erences between the pre- and post-1884 sub-samples.
Company size and the debt ratio do not in‡uence bonus payments before the reform, but
thereafter. After 1884, we …nd a signi…cant positive relationship between the level of assets
and the level of the debt ratio and the level bonus payments.
Finally, we use the two-step Arrelano-Bond estimator to account for possible dynamic
structures of our data. The results, reported in Table 5, supports our hypothesis that the
1884 legal reform reduced the need for monetary motivation of managers. The bonus payments positively depend on the excess pro…ts, the legal reform substantially reduced the
need for monetary motivation. The interaction of excess pro…ts and the 1884 dummy variable is signi…cantly negative. This coe¢ cient displays the minimum incentive component
of the pay-and-performance correlation.
18
Table 5: Regression results:Two-step Arellano-Bond estimator
Bonus, t
1
Manufacturing
Coe¢ cient p-value
0.18152
0.000
Pro…ts
0.02667
0.000
Pro…ts*1884 dummy
-0.02168
0.000
1884 dummy
-635.45
0.803
Assets
0.00001
0.587
Debt
4,925.84
0.103
Constant
223.08
0.004
No.of obs.
645
No. of groups
37
2
31.95
Sargan
6
Conclusion
In this paper, we have used a fundamental reform of the corporate governance code –
the 1884 joint-stock companies act in Germany –as an identi…cation scheme for incentive
contracts using di¤erence-in-di¤erence estimation technique. We …nd an economically
as well as statistically signi…cant correlation between pro…ts of manufacturing …rms and
banks and the bonus payments to executives of these corporations, even if we control for
other variables, e.g. stock market value, total assets, or liabilities of the …rms.
The correlation coe¢ cient indicates that a 1,000 Mark increase in pro…ts leads to an
increase of bonus payments of about 30-32 Marks for a large sample of German manufacturing …rms and banks before 1884. This share of pro…ts distributed to each manager,
however, does not neccessarily measure incentives for executives. The estimated coe¢ cient
also absorbs the correlation between pay and performance induced by other variables, e.g.
business cycle e¤ects. We applied a di¤erence-in-di¤erence estimation technique to separate the incentive component from other components of the correlation. Our identi…cation
scheme is a legal reform in Germany in 1884. This reform alters the corporate governance
institutions, but it did not change the characteristics of manages or the business cycle
in‡uence on the pay-performance correlation. Thus, any shift in the correlation between
19
managerial compensation and …rm performance can be interpreted as an incentive component under the old regime of corporate governance. This incentive component was
substituted by better institutions under the new system of corporate governance in Germany.
In fact, the estimated in‡uence of pro…ts on managerial compensation declined by
about two thirds after 1884 for manufacturing …rms. This result proves robust both
qualiatively and quantitatively to alternative estimation approaches and sample splits.
For a sample of banks, which was collected for comparison, the e¤ect is quantitatively
somewhat smaller but also signi…cant. As for the post-reform period, naturally we can
not exactly di¤erentiate between general-equilibrium, business cycle, and incentive e¤ects
in the empirical pay-performance correlation. Yet, the large impact the reform itself has
points towards a substantial incentive proportion in the sensitivity found, as the reform can
not be expected to have wiped out all problems of corporate control but only a substantial
fraction. The shift in the estimated e¤ect of pro…ts on managerial compensation can
be attributed to the incentive component of managerial compensation before the reform.
Better institutions in 1884 allowed the monetary component to be substituted.
In the light of our results, one may be alarmed of an overoptimistic interpretation of
an increase in the pay-performance relation observed in the recent decade. It may be that
contracts were written suboptimal before. Our results, however, suggest that managerial contracts were formed in optimal reaction to changes in the contractual environment
already in the 19th century. Most importantly, they highlight the interconnection of all
elements of corporate governance, incentive pay being but one.
20
References
Arellano, Manuel and Stephen Bond (1991): ”Some Tests of Speci…cation for Panel Data:
Monte Carlo Evidence and an Application to Employment Equations”, Review of
Economic Studies, 58 (2), 277-97.
Aggraval, Rajesh K. and Andrew A. Samwick (1999): ”The other Side of the Trade-o¤:
The Impact of Risk on Executive Compensation”Journal of Political Economy, 107
(1), 65-105.
Baltagi, Badi H. and Ping X. Wu: "Unequally Spaced Panel Data Regressions with AR(1)
Disturbances", Econometric Theory, 15(6), 814-23.
Bittner, Thomas (2005): "An event study of the Rhenish-Westphalian Coal Syndicate",
European Review of Economic History, 9, 2005, 337-364
van der Borgth, Richard (1883): "Statistische Studien über die Bewährung der Actiengesellschaften", Verlag von Gustav Fischer, Jena.
Bosenick, Alfred (1912): ”Neudeutsche gemischte Bankwirtschaft - Ein Versuch zur
Grundlegung des Bankwesens”, ”Volume 1”, Schweizer Verlag, Munich.
Burhop, Carsten (2004a): "Die Kreditbanken in der Gründerzeit",
Stuttgart.
Steiner Verlag,
Burhop, Carsten (2004b): "Die Vergütung des Führungspersonal deutscher Groß
banken,
1871-1913", mimeo Uni Münster.
Burhop, Carsten (2004c): ”Executive Remuneration and Firm Performance: The Case of
Large German Banks, 1854 -1910”, Business History, 46(4), 525-43.
Conyon, Martin J. (1997): ”Corporate Governance and Executive Compensation”, International Journal of Industrial Organizatrion, 15, 493-509.
Core, John E., Robert W. Holthausen, and David F. Larcker (1999): ”Corporate Governance, Chief Executive O¢ cer Compensation, and Firm Performance”, Journal of
Financial Economics, 51, 371-406.
Core, John, Wayne R. Guay, and David F. Larcker (2003): ”Executive Compensation and
Incentives: A Survey”, Federal Reserve Bank New York Policy Review April 2003,
27-50.
Garen, John E. (1994): ”Executive Compensation and Principal Agent Theory”, Journal
of Political Economy, 102 (6), 1175-99.
Gareis, Karl (1900): "Handelsgesetzbuch vom 10. Mai 1897 nebst dem Einführungsgesetz
vom 10. Mai 1897 unter Ausschluss des Seerechts", 2nd ed. Beck, Munich.
Glagau, Otto (1877): ”Der Börsen- und Gründungsschwindel in Deutschland”, Verlag von
Paul Frohberg, Leipzig.
21
Grossmann, Volker (2003): ”Managerial Job Assignment and Imperfect Competition in
Asymmetric Equilibrium”, mimeo, University of Zurich.
Hall, Brian J. and Je¤rey B. Liebmann (1998) ”Are CEOs really paid like bureaucrats?”,
Quarterly Journal of Economics, 113 (3), 653-91.
Hessberger, Karl (1889): "Die volkswirtschaftlichen Wirkungen der Aktiengesetznovelle
vom 18. Juli 1884 auf die Gründungen von Aktiengesellschaften", Fulda.
Himmelberg, Charles P. and Glenn Hubbard (2000): ”Incentive Pay and the market for
CEOs: An Analysis of Pay-for-Performance Sensitivity”, Columbia Business School
Working Paper.
Holmstrom, Bengt (1999): "Managerial Incentive Problems: A Dynamic Perspective", Review of Economic Studies, 66(1), 169-82.
Hopt, Klaus J. (1980): ”Ideelle und wirtschaftliche Grundlagen der Aktien-, Bank- und
Börsenrechtsentwicklung im 19. Jahrhundert”, In: Coing, Helmut / Wilhelm, Walter
(eds.), ”Wissenschaft und Kodi…kation des Privatrechts im 19. Jahrhundert”, Bd. 5,
”Geld und Banken”, Fritz Knapp Verlag, Frankfurt a. M..
Hsiao, Cheng (2003): "Analysis of Panel Data", Cambridge University Press, Cambridge.
Jensen, Michael C. and Kevin J. Murphy (1990): ”Performance Pay and Top-Management
Incentives”, Journal of Political Economy, 98 (2), 225-64.
John, Kose and Yiming Qian (2003): ”Incentive Features in the Banking Industry”, Federal Reserve Bank New York Policy Review April 2003, 109-121.
Joskow, Paul L., Nancy L. Rose, and Catherine D. Wolfram (1996): "Political constraints
on executive compensation: evidence from the electric utility industry", 27(1), 16582.
Kato, Takao (1997): ”Chief Executive Compensation and Corporate Groups in Japan: New
Evidence from Microdata”, International Journal of Industrial Organization, 15, 45567.
Kraft, Kornelius and Antonia Niederprüm (1999a): ”Ist die Vergütung von Managern im
Zeitablauf ‡exibler geworden?”, Zeitschrift für betriebswirtschaftliche Forschung, 51,
1999, 787 - 804.
Kraft, Kornelius and Antonia Niederprüm (1999b): ”Determinants of management compensation with risk-averse agents and dispersed ownership …rm”, Journal of Economic Behavior & Organization, 1999, 17-27.
Lambert, Richard A. and David F. Larcker (1987): ”An Analysis of the Use of Accounting and Market Measures of Performance in Executive Compensation Contracts”,
Journal of Accounting Research, 25 , 85-125.
Lazear, Edward P. (1986): "Salaries and Piece Rates", Journal of Business, 59(3), 405-31.
22
Lazear, Edward P. (2003): ”Output-Based Pay: Incentives, Retention or Sorting?”, IZA
DP. No. 761.
Maddala, G.S. and Shaowen Wu ”A Comparative of Unit-Root Tests with Panel Data and
a New Simple Test”, Oxford Bulletin of Economics and Statistics, Special Issue, 61
(0), 631-52.
Milward, Robert (2004): "European governments and the infrastructure industries, c.
1840-1914", European Review of Economic History, 8, 2004, 1-28
Murphy, Kevin J. (1999): ”Executive Compensation”, In Orley Ashenfelter and David
Card (eds.): ”Handbook of Labor Economics”Vol. 3B, North-Holland, Amsterdam,
2485-2563.
Peters, Lon LeRoy (1989): "Managing competition in German coal, 1893-1913", Journal
of Economic History, 49, 419-434
Philippovich, Eugen von (1909): ”Grundrißder politischen Ökonomie”, Volume 2, Part 1,
4th Edition, J.C.B. Mohr, Tübingen.
Prendergast, Canice (1999): "The Provision of Incentives in Firms", Journal of Economic
Literature, 37(1), 7-63.
Reitmayer, Morten (1999): ”Bankiers im Kaiserreich”, Vandenhoeck&Ruprecht, Göttingen.
Ring, (1890): ”Das Aktienrecht in Deutschland”, In: Conrad, J. et al. (eds.), ”Handwörterbuch der Staatswissenschaften”, Vol. 1, 85 - 101, Verlag von Gustav Fischer, Jena.
Schmid, Frank A. (1997): " Vorstandsbezüge, Aufsichtsratsvergütung und Aktionärsstruktur", Zeitschrift für Betriebswirtschaft, 67, 67 –83.
Schwalbach, Joachim and Ulrike Graß
ho¤ (1997): ”Managervergütung
und
Unternehmenserfolg”, Zeitschrift für Betriebswirtschaftslehre, 67, 1997, 203 217.
Shleifer, Andrei and Robert W. Vishny (1997): ”A Survey of Corporate Governance”,
Journal of Finance, 52 (2), 737-783.
Simon, Carol J. (1989): "The e¤ect of the 1993 securities act on investor information and
the performance of new issues", American Economic Review, 79, 1989, 295-318
Tellkampf, J. L. (1876): "Vorschläge zur Verbesserung des Actien-Gesellschaftswesens",
Berlin.
Talley, Eric L. and Johnson Gudrun (2004): "Corporate Governance, Executive Compensation, and Securities Litigation", USC CLEO Research Paper No. C04-4.
Warschauer, Otto (1902): ”Die Reorganisation des Aufsichtsratswesens in Deutschland”,
Freier Verlag, Berlin.
23
Zhou, Xianming (2000): ”CEO pay, …rm size, and corporate performance: evidence from
Canada”, Canadian Journal of Economics, 33(1), 213-251.
24