 n the recent wave of corporate scandals,
from Enron to Tyco, poor corporate governance structures have
clearly been a contributing
factor. The tales of excess compensation, poor capital allocation,
and, occasionally, outright theft, have shone a harsh spotlight
on the relationships between chief executive officers and the
boards of directors. Too frequently, directors – who are
supposed to represent the shareholders – have acted in
ways that enrich CEOs and other favored executives while impoverishing
common shareholders. On many boards, two (or more) directors
serve together on a different company’s board. For example, General Motors Corp.’s
April 2002 proxy revealed that the GM board had two mutual interlocks:
CEO John F. Smith, Jr., and Director George M.C. Fisher were
also directors on the board of Delta Air Lines, Inc.; and Smith
and Director Alan G. Lafley were also on the board of the Proctor & Gamble
Co. (where Lafley is the CEO). We dub these associations “mutual
interlocks.” And in a sample of 366 large companies, 87
percent had at least one mutual interlock in 1991.
Director
interlocks have clear consequences for shareholders. Our
empirical analyses show that CEO compensation
tends to increase
and CEO turnover tends to decrease when the CEO’s board
has one or more pairs of board members who are mutually interlocked
with another company’s board. Why? On the one hand, mutual
interlocks could be an indication of and a contributor to CEO
entrenchment, from which higher compensation and lower turnover
naturally follow. On the other hand, mutual interlocks may indicate
the strengthening of important and valuable strategic alliances.
And the higher CEO compensation and lower turnover may be a just
reward for orchestrating such alliances. We believe that the
first interpretation is more accurate.
Director Interlocks
Researchers from several disciplines have been looking into interlocks
for several decades. And at first it appeared that interlocks
were a sign of weakness. In one of the earliest such U.S. studies,
economist Peter Dooley in 1969 found that less-solvent firms
were likely to be director-interlocked with banks. Later studies
also reported that firms with high debt-to-equity ratios, or
that had an increased demand for capital were likely to have
interlocks. The reason: Financially stressed firms may seek to
add bank officers to their boards to receive more favorable consideration.
Or banks may demand board seats so they can monitor firms more
closely.
Organizational behavior experts have
examined the extent to which a board is an instrument of
management interests. Some have argued
that companies use board interlocks as a mechanism to improve
contracting relationships, or to reduce the information uncertainties
created by resource dependencies between firms. This stream of
research suggests that the composition of boards, including interlocks,
is largely determined by the efforts of CEOs to influence the
selection of new directors so that they are responsive to that
particular CEO’s interests.
Financial economists have examined interlocks
as well. Kevin Hallock of the University of Illinois found
that CEOs serving
in employee-interlocked firms earn higher salaries than they
otherwise would. Nevertheless, existing research has not documented
a connection between director interlocks and total CEO compensation.
And in our survey of previous studies, we did not find any associations
between interlocks of various kinds and firm performance. That
leads us to believe that interlocks aren’t designed to
serve a firm’s strategic goals, and don’t serve them
in practice.
Compensation and Turnover
Several recent studies have examined the relation between top
executive compensation and board composition. And they report
mixed results. For example, some authors have found a positive
association between CEO compensation and the percentage of outside
directors on the board. Other studies have found no relation
between a board majority of outside directors and top management
compensation. The level of incentive-based executive compensation
appears to be positively connected with firm performance, and
incentive-based compensation appears to be used more extensively
by outsider-dominated boards.
“After
poor firm performance, CEOs are more likely to be dismissed
if the board of directors has a majority of outsiders.”
|
Other
scholars have found an inverse relation between the probability
of a top management change
and prior stock price performance.
After poor firm performance, CEOs are more likely to be dismissed
if the board of directors has a majority of outsiders. Empirical
analyses indicate that the probability of top management turnover
is reduced if the top executives are members of the founding
family or if they own higher levels of stock. Executive turnover
is also negatively related to the ownership stake of officers
and directors in the firm and positively related to the presence
of an outside blockholder. Other studies have found that the
likelihood of CEO departure is inversely associated with both
the dollar value of stock option compensation in relation to
cash pay and the amount by which a CEO’s compensation is
higher than would be expected from comparisons with the compensation
of other CEOs. But thus far, no study has considered the possible
effects that boards with mutual director interlocks have on CEOs’ total
compensation and turnover.
The Data
We looked at CEO compensation and CEO
turnover for 452 industrial firms, first compiled by NYU
Stern professor David Yermack. These
firms were drawn from Forbes magazine’s lists of the largest
500 U.S. companies in categories such as total assets, market
capitalization, sales, or net income. The data set includes all
companies meeting this criterion at least four times during the
1984-1991 period. Compensation data were collected from the corporation’s
SEC filings. Directors who were full-time company employees were
designated as “inside” directors. Individuals closely
associated with the firm – for example, relatives of corporate
officers, or former employees, lawyers, or consultants, or people
with substantial business relationships with the company – were
designated as “gray” directors. All the rest were
designated “outside” directors. We also drew on the
data assembled by Kevin Hallock, who analyzed 9,804 director
seats held by 7,519 individuals in 1992. We took as our final
data set the 366 industrial firms for the 1991 proxy season that
appeared in both the Yermack and the Hallock data sets. (Utility
and financial firms were excluded from the study because government
regulation may lead to a different role for directors.)
In order to examine how director interlocks
may affect CEO compensation, we used a measure of total remuneration
that included salary
and bonus, other compensation, and the value of option awards
when granted. We believe that this sum is a more accurate measure
of what boards intended to pay, which could be different from
what CEOs earn, since CEOs often exercise options early, thereby
sacrificing a significant portion of the award’s value.
As an estimate for CEO turnover, we
used a dependent variable that was set equal to one if a
CEO leaves office during the last
six months of the current fiscal year or the first six months
of the subsequent period. In order to control for retirement-related
voluntary departures, we included in the analysis the CEO’s
age. Turnover events occurred in 9.0 percent of the sample (thirty-three
firms).
Considering Interlocks
 The key explanatory variable of this
study was the number of mutual interlocks on the firm’s
board. While two boards can be interlocked if they share
one director, they are mutually
interlocked if they share at least two directors. For any given
board, a director could be part of more than one pair of mutual
interlocks, so it is quite possible that a board may have a greater
number of mutual interlocks than directors. In our sample of
industrial firms, board sizes ranged from four to 26, with an
average of 12.18. The number of mutual interlocks ranged from
zero to 42, with an average of 12.15.
Other independent variables used in
the study were based on their likely relevance and effects
on CEO compensation and CEO turnover,
as established by other authors. As in numerous other studies,
Tobin’s Q (the market value of assets divided by the replacement
cost of assets) was used as a proxy for the growth opportunities
of the firm.
Table 1 presents descriptive statistics of the key variables
in this study and their correlation with the number of board
director interlocks. As seen, the mean number of directors who
are CEOs of other firms is 1.94. This result is similar to that
reported by James Booth and Daniel Deli, who found the mean to
be 1.87 for 1989-90 data. The mean number of outside directors
serving on the board was 6.94, which is also consistent with
the previous literature.
Two Hypotheses
If the CEO dominates the selection process of directors to the
board, and if the CEO is in fact filling the director positions
with sympathetic members, then we would expect a positive association
between the fraction of these favorable board members and the
compensation of the CEO. In other words, our first hypothesis
stipulates that boards with a larger number of mutual director
interlocks will pay a higher compensation package to the CEO.
Our second hypothesis states that there is an inverse association
between the presence of mutual interlocks and the likelihood
of CEO departure.
What do the results show? The correlations reported in Table
1 suggest the existence of a relationship between the number
of mutual director interlocks and the compensation of the CEO.
It is not surprising that larger boards have more interlocks
and that a preponderance of interlocks appears to be positively
connected with outside directors and with directors who are CEOs
of other organizations. Mutual director interlocks appear to
be curtailed by close ownership and governance structures. Our
results show a negative and significant correlation between this
variable and the indicators for CEO-as-founder and for non-CEO
chairman.
Since director interlocks could just be indicators of strategic
power relationships between firms at the highest level, it cannot
be automatically concluded that CEOs and interlocked directors
exploit networks of board memberships for their personal gain
simply because these multiple board affiliations exist. In fact,
CEOs could be rewarded with additional compensation and long
job durations for successfully establishing mutual interlocks
that serve the strategic goals of the firm.
But the data show a significant negative
relationship between the number of mutual interlocks and
the number of “gray” directors,
many of whom could represent companies that have supplier or
customer relationships with the company. This negative relationship
reinforces our skepticism as to the likelihood that the mutual
interlocks serve the strategic goals of the firm.
Extra Compensation
 To test our first hypothesis, we ran
an ordinary least square regression to estimate the effect
that mutually interlocking
boards have on the total compensation of the CEO. These calculations
took into account factors such as interlocks, firm size, tenure
of the CEO, firm performance, and stock ownership of the CEO.
The results of this estimation are presented in Table
2. As expected,
the number of mutual director interlocks is found to be significant
and positively associated with total compensation. This finding
suggests that the links created by the mutual interlocking relations
between boards actively benefit the CEO. In other words, with
the aid of mutual interlocks, CEOs are able to extract significantly
larger compensation packages from their firms. The robustness
of this result is upheld through further investigations of the
components of the CEO’s pay package.
When we repeated the analysis using
the natural log of only the sum of the CEO’s salary and bonus as the dependent variable,
the coefficient for mutual directors was positive and significant.
That suggests that even just the sum of the CEO’s basic
salary and bonus tends to increase as a consequence of the mutual
director interlocks. In fact, we found that a mutual interlock
adds an average of $143,000 (approximately 13 percent) to the
average CEO salary and bonus.
The evidence presented in Table
2 is
in line with the view that mutual interlocks may indeed assist
the CEO in extracting lucrative
remuneration packages from the firm. The networks and traffic
of influences created by mutual interlocking directorships have
probably been utilized by CEOs in exerting control over the majority
of board members. This finding suggests that directors may not
be making decisions that benefit the firm’s shareholders
the most. Mutually interlocking directorships could be weakening
the control mechanisms put in place to ensure that directors
fulfill their fiduciary duty and act in the best interest of
the shareholders.
When we ran other regressions with these
data we found that stock option compensation appears not
to be judiciously used by boards
in compensating their CEOs in the presence of mutual interlocks.
We believe this reflects cronyism and weakens the board’s
monitoring function. This interpretation is consistent with the
view of academics and corporate governance activists who perceive
interlocks generally as corrupt. Thus, although other studies
find that markets react favorably to the adoption of stock option
plans to compensate top executives, we find that stock options
can be misused if the board’s monitoring activities are
weakened by interlocks.
Other results in these regressions are
consistent with the previous literature. We found that CEO
pay is inversely related to the
fraction of equity held by the CEO. And as economists Sherwin
Rosen, Clifford W. Smith, Jr., and Ross L. Watts have found in
other studies, we find that large companies and firms with greater
growth opportunities pay more to their CEOs. A company’s
net-of-market stock return was found to have a positive and significant
association with total CEO compensation, consistent with previous
studies.
CEO Turnover
To test our second hypothesis, we investigated
whether the presence of mutual interlocking directorships
decreases the board’s
ability to monitor the CEO, thereby decreasing the likelihood
that the CEO will depart. We analyzed the data, including CEO
and company characteristics that should be associated with the
probability of turnover. Michael Jensen and Kevin Murphy have
suggested that one obvious CEO feature likely to affect the turnover
process is age. To control for this influence we included the
CEO’s age in the estimation. And to control for firm performance,
we included the firm’s current and previous year stock
returns net-of-market as well as the current period return on
assets. Further control variables included proxies for growth
opportunities (the ratio of research and development {R&D}
over sales), the ratio of long-term debt to total assets, company
size, and the fraction of common stock held by the CEO or his
immediate family.
Table 3 presents coefficient estimates for the CEO turnover model.
And the results are as hypothesized: The coefficient on the mutual
interlock variable is negative and significant as predicted,
implying that the presence of mutual board interlocks is inversely
associated with the probability of CEO turnover. We interpret
this result to indicate that mutually interlocking directorships
weaken the monitoring power that the board has over the chief
executive. Further, mutual interlocks contribute to the possible
entrenchment objectives of the CEO. This result is in agreement
with the notion that boards are ineffective in controlling the
CEO, who is likely to control the nomination and selection process
of the directors.
These results are consistent with other theories and research
on CEO turnover. As previous studies have noted, we found that
CEOs are less likely to leave office if they own a large fraction
of equity in the firm or if company performance is strong. And
we found that age is positively associated with the probability
of CEO turnover. Firm size, as proxied by the natural log of
the firm assets, does not appear to play a role in the likelihood
that the CEO leaves office.
Conclusion
Academics and the popular press have suggested that corporate
boards are ineffective in monitoring CEOs, since CEOs frequently
dominate the director selection process. Boards filled with CEO-sympathetic
director appointees are likely to overcompensate and undermonitor
the chief executive. Our view is that the mutually interlocking
directorships that are prevalent among firms are responsible
for the production of sympathetic directors. These directors
have the opportunity to pay and re-pay each other favors because
of their multiple board memberships and may well be doing so
in league with the CEOs who nominated them.
“Interlocking
directorships weaken the monitoring power that the board
has over the chief executive.”
|
Our results
indicate that the power alliances created by directors with
multiple memberships
are used by self-serving CEOs to extract
handsome remuneration packages from firms and to strengthen their
entrenchment. Boards that overcompensate and undermonitor the
CEO are not fulfilling their fiduciary duties to the shareholders.
As a result, board mutual interlocks weaken the firm’s
governance structure, promote cronyism, and exacerbate the firm’s
agency problems.
The results reported here indicate that it is at least plausible
that mutual director interlocking relationships between different
corporate boards might affect the voting patterns and decisions
that these boards make on other matters besides CEO compensation
and turnover.
Overall, our research suggests that
inter-board relationships should be more closely scrutinized
to determine whether these
relationships encourage decisions that enhance shareholder wealth
or instead facilitate empire building by self-serving CEOs. If,
as we suspect, the latter is the case, then closer monitoring – private
and/or public – of boards is needed.
Eliezer M. Fich, Stern Ph.D. 2000, is visiting assistant professor
of finance at the Kenan-Flager Business School at the University
of North Carolina. Lawrence J. White is Arthur E. Imperatore
professor of economics at NYU Stern.
This article is adapted from an article that appeared in the
Fall 2003 Wake Forest Law Review. |
