Printer Friendly

Executive compensation and corporate performance in electric and gas utilities.

* Recent studies on public utilities do not find a positive relationship between managerial compensation and corporate profitability (see Hirschey and Pappas [16], and Carroll and Ciscel [9]). This literature interprets these findings as suggesting that incentives for profitability are not needed in public utilities, since regulation provides assured profits. However, there are at least three reasons to cast doubt on these findings.

First, these studies examine the pay-performance relationships for managers across different firms. Economic theories of incentive contracting suggest than an efficient contract should reward managers based on factors such as managerial ability, managerial responsibility, firm size and past performance, in addition to current performance. Murphy [27] argues that in the absence of a theory specifying the relevant observable variables, cross-sectional models of pay-performance relationships are subject to a serious omitted-variables problem. However, if these omitted factors do not change over time for individual executives, we can correctly assess the pay-performance relationship by analyzing time-series regressions for individual executives. Recent studies on unregulated firms find a strong positive relationship between managerial compensation and stockholder returns (see, e.g., Benston [5], Coughlan and Schmidt [12], Murphy [27], and Antle and Smith [2]). These findings contrast with earlier studies on unregulated firms that use a cross-sectional approach and find, at best, a weak relationship between executive compensation and firm profitability (see, e.g., Ciscel and Carroll [11] for a survey). These studies support similar claims in the popular press (see, e.g., Loomis [23]).

Second, most of the studies on public utilities confine their attention to only part of the compensation package, usually salary and bonus, rather than examining the total pay package. This approach ignores potentially performance-sensitive pay components, such as contingent or deferred remuneration, and stock-related incentive plans. Third, most of these studies use accounting-based performance measures, which tend to be historical rather than forward-looking.

This paper reexamines the relationship between top executive compensation and corporate performance in public utilities by applying new methodologies that directly address these problems. Our findings are mostly consistent with the hypothesis that senior managers are rewarded for pursuing stockholder wealth maximization.

Section I reviews the pertinent literature and develops the testable hypotheses. In Section II, we present our methodology. The data is described in Section III. The results of our analysis are in Section IV. Some additional tests are described in Section V. We conclude in Section VI.

I. Previous Research on the

Pay-Performance Relationship

for Top Managers

The issue of managerial behavior in the unregulated firm has been extensively debated. In one view, the separation of ownership from control allows managers to pursue self-serving goals that may not benefit stockholders (Berle and Means [6], Jensen and Meckling [18], and Jensen [17]). Managers are seen to seek job security, perks, prestige, and control rather than stockholders' welfare. Baumol [3] and [4], and Marris [25] suggest sales maximization as the primary managerial objective, since growth in sales is likely to result in a larger firm and increase prestige and possible perk consumption. (1) They argue that compensation plans, proposed and adopted through managerial influence, reflect these goals. According to this view, there is a positive relation between compensation and growth in sales. Furthermore, according to some, (e.g., Loomis [23]), there is no link between compensation and profitability. We will refer to this view as the sales maximization hypothesis.

An alternative view holds that competitive market forces, as well as managerial compensation contracts, tend to align the interests of managers with those of stockholders. Smith and Watts [28] survey executive compensation plans and find that the plans are explicitly designed to motivate managers to maximize stockholders' wealth. Such plans include stock options, phantom stocks, stock appreciation rights and other performance plans based on stock performance. According to this view, firms that do not use such compensation schemes are likely to be less efficient, and would tend to fail more often. In addition, competitive forces, such as the discipline induced by labor markets (Fama [14]), and the threat of corporate takeovers, (Manne [24]), induce managers to pursue stockholders' interests. We refer to this view as the stockholder wealth maximization hypothesis.

In an early cross-sectional study of unregulated firms, McGuire, Chiu and Elbing [26] find support for the sales maximization hypothesis with evidence that compensation is correlated more with sales than with profitability. However, their study, and the empirical tests that followed, share three drawbacks. First, profitability is measured with accounting data and not stock market returns. Second, compensation in these studies usually includes salary and bonus, but omits other important components such as contingent and deferred remuneration. And, finally, they employ cross-sectional tests that ignore differences in individual-specific attributes (positions and responsibilities of managers). Lewellen and Huntsman [21], studying industrial firms, were two of the early researchers who dealt with some of these problems. They concluded that "profits appear to have a strong and persistent influence on executive rewards, whereas sales seem to have little, if any, such impact" [21, p. 718]. More recently, Benston [5], Coughlan and Schmidt [12], and Murphy [27] reexamine this issue for industrial firms using time-series data on the total compensation packages of individual managers and find evidence that executive compensation is positively related to stock performance, as well as to sales.

The above issues have not been comprehensively addressed in the studies on public utilities. Hirschey and Pappas [16] estimate cross-sectional regressions with data on 69 utilities for 1977. The dependent variable, compensation, is measured as salary plus bonus plus deferred compensation. Common stock-based rewards are not considered. The independent variables are accounting profits (net income after taxes) and sales (total revenues). They find that managerial compensation is positively related to sales and is negatively related to profitability. In a modified test design, Carroll and Ciscel [9] obtain similar results. Both studies attribute their findings of a lack of incentives to maximize profits to the claim that utilities are assured recovery of costs and a "fair" return on their capital. (2) Thus, this implies that there is no need to motivate managers to maximize profits. Another explanation that has been offered is that utilities have limited investment opportunities, so that it is not so important to base pay on firm performance (Smith and Watts [28]). However, others, such as Baumol [4] and Bonbright [7], argue that there is a need to motivate managers of public utilities because of the potential for savings, and thus higher profitability, during regulatory lags. (3) Furthermore, it can be argued that stockholders need to motivate managers so that they obtain higher allowed rates of return from public utility commissions.

We reexamine the role of compensation schemes in motivating managers in public utilities, using market-based data and new methodologies. We investigate whether managers are rewarded for maximizing stockholder's wealth. Specifically, we test the hypothesis that executive compensation in electric and gas utilities is positively related to returns on common equity. We also examine the alternative hypothesis that managers are compensated to maximize sales (growth in sales).

II. Methodology

In this section we develop a methodology to test the relationship between managerial compensation and returns to stockholders.

According to the sales and stockholder wealth maximization hypotheses, increases in managerial compensation depend on growth in sales and stock returns, respectively. To reduce the role of manager-specific characteristics, such as age, education, experience, and geographical location of the utility, all of which affect the level of compensation, we focus on changes in compensation over time for an individual executive. Assuming that compensation for a given year is the reward for performance during the year, we can describe the pay-performance relation with a first-differences time-series equation:

[]=[a.sub.i]+[b.sub.i][]+[c.sub.i][]+ [[epsilon]], (1)

where [] is the rate of growth in total annual compensation from year t- 1 to year t for manager i, [] is the rate of return on common stock in year t for manager i's firm, and [] is the rate of growth in sales for his firm from year t - 1 to t. The total annual compensation is the sum of salary, bonus, securities, properties, benefits, contingent remuneration, and deferred compensation for a manager.

Following Gibbons and Murphy [15], one would expect that the incentive scheme rewards a manager based on the firm's relative rather than absolute stock performance. So that we may test for the reward for firm-specific performance rather than industry-wide (or general) economic conditions, we replace [] with the industry-adjusted rate of return, ([] - [R.sub.It]), where [R.sub.It] is the rate of return on an industry index of common stocks. To test the sales maximization hypothesis, We use [], the rate of growth in sales (Baumol [4] and Marris [25]).

Next, following Murphy [27], we assume that the pay-performance relation is the same for all managers. (4) We can then use the pooled time-series cross-sectional regression:

[]=a+b([]-[R.sub.It]) + c [] + [[]. (2)

The choice of a first-difference model, Equation (2), over a model based on the level of compensation offers some econometric advantages. Level regressions with pooled time-series cross-sectional data suffer from heteroscedasticity and autocorrelation of residuals. These problems are reduced by using a first-difference model (see, e.g., Christie [10] and Dechow and Sloan [13]).

There are, however, two potential problems with pooling data across managers. First, changes in compensation of managers from the same firm may be correlated. For example, their bonuses may be drawn from a common bonus pool. While this does not bias the slope coefficient, it may overstate its t-statistic. Second, the compensation packages of managers in different positions or ranks may be related to different performance measures. For example, it may be the case that CEOs are compensated primarily on the basis of stock price performance, while the compensation of vice-presidents is tied to growth in sales. (5) Such information may be lost in the aggregation. Consequently, we estimate Equation (2) separately for managers occupying different executive positions. A positive b, the coefficient of the adjusted rate of return, implies that the compensation of the manager is positively related to changes in stockholders' wealth. A positive c coefficient implies that changes in compensation are positively related to growth in sales.

Finally, the specification in Equation (2) is a constrained regression, since it sets the coefficients of [] and [R.sub.It] to be b and -b, respectively. In order to capture the unconstrained impacts of [] and [R.sub.It] on [], we also use the following regression:

[]=a+[b.sub.1][]+[b.sub.2][R.sub.It]+[]+ [[epsilon]]. (3)

Given the return on firm i's stock, we expect a manager's pay raise to be smaller if the industry's return is higher. This implies that the coefficient of [R.sub.It] in Equation (3), [b.sub.2], is negative. As before, we expect that pay raises are positively related to returns on the firm's stock, and so [b.sub.1] should be positive.

III. Data

The sample for this study was selected from all the utility firms with SIC codes 4911 and 4931 on the COMPUSTAT tapes which were listed on the New York Stock Exchange or the American Stock Exchange in 1985. One hundred and eighteen such firms were identified. Proxy statements were requested directly from these firms for the ten-year period 1975-1984. The requirement that the firm exist for ten years raises the possibility of survivorship bias. However, in this industry, it is unlikely to be serious because of the safety net provided by regulation. During this period, diversification into nonutility businesses was not common, and, consequently, the result is a relatively homogeneous sample of regulated firms. While all except ten firms responded to our request, a fair number of them did not provide complete records for the ten years. Our final sample consists of 69 utilities for which we have compensation data for the entire ten-year period.

In Exhibit 1, we provide descriptive statistics for the year 1980 for the firms in our sample and compare them with those that were excluded because of unavailability of compensation data. Since size is a characteristic that can significantly affect management practices, including the


pay-performance relation, we provide in the exhibit a number of measures for size. We find that statistically the two groups are similar in size, as measured by sales, number of employees, shares outstanding, market value of equity, or total book value of assets. (6) We find similar results for other years, as well. (7) This comparison suggests that our sample is not atypical of the industry as a whole. However, our sample differs from the samples in other studies which draw firms from manufacturing industries. For example, Murphy's [27] sample drawn from Fortune 500 industrial firms has mean sales of $4,966 million in 1983 dollars, whereas the mean sales of firms in our sample is $767 million, which is $927 million in 1983 dollars.

Compensation Data. For each of the 69 firms in our sample, we obtained compensation data for the top executives whose compensation is disclosed in corporate proxy statements for each of the years 1975-1984. Proxy statements for each of the years 1975-1984. Proxy statements disclose this information for up to the five highest-paid executives. We were able to obtain compensation data for a total of 2,864 executive-years. Exhibit 2 provides details on the selection of our sample. Since two successive years data are needed in order to compute the growth rate in compensation of an executive, we were able to compute it for 2,147 executive-years.

Exhibit 3 provides descriptive statistics in 1990 dollars on the annual compensation of managers (using the Consumer Price Index to convert dollar values to 1990 equivalents). The following three components of compensation (drawn from data in proxy statements) are shown in the exhibit:

(i) Salary plus bonus, SALB;

(ii) Long-term compensation, LTC, which we define as the sum of three items reported in the proxy statements: (a) securities, properties and benefits, (b) contingent remuneration, and (c) deferred compensation; and,

(iii) Total annual compensation, TCOMP, defined as the sum of salary plus bonus and long-term compensation, (i) plus (ii).

Whereas SALB is primarily the compensation for current performance, LTC provides a reward that is more contingent on long-run performance, an objective that is consistent with the outlook of stockholders. Importantly, LTC contains items such as stock awards, that provide incentives for pursuing stockholders' wealth maximization.

Our choice of these particular components is guided by the following reasons. The SEC did not require separate reporting of salary and bonus over most of our sample period. As a result, only the sum of the two is reported in the proxy statements for all years. Consequently, we restrict our analysis to the sum of salary and bonus. We find the usage of stock options so rare among utilities that we ignore it as a component of total compensation. Finally, we ignore pension benefits because of the difficulty in estimating their contribution toward the compensation for a particular year from the limited information disclosed in proxy statements. This is a potential limitation of the study.

Exhibit 3 also provides information on compensation by executive position. Proxy statements contain compensation data for up to the five highest-paid executives in the firm, whose compensation exceeds a threshold defined by the SEC. Those usually consist of the chief executive officer, chairman, president, and one or more vice-presidents. Any analysis of the top executive positions is complicated by the fact that a given executive at times occupies multiple positions. For example, the CEO almost always holds the position of either chairman or president. (8) In order to conduct the analysis by executive positions, we define the following groups:

* All: All managers;

* CEO-T: CEOs who may also hold the position of chairman and/or president;

* CHM-T: Chairmen who also hold the position of CEO and/or president;

* CHM-ONLY: Chairmen who do not hold any other position;

* PRES-ONLY: Presidents who do not hold any other position;

* VP-ONLY: Vice-presidents who do not hold any of the above positions. (9)

We form these groups in an attempt to classify managers according to their relative importance in the firm, after taking into consideration the number of managers holding different positions. For example, we expect that the CHM-T group, chairmen who simultaneously hold the CEO and/or president position, are likely to be the topmost managers in their firms. (10) CEO-T, a group that overlaps


most important group of managers. Chairmen who do not hold any other position, CHM-ONLY, may be approaching retirement, having relinquished the CEO position earlier (see Vancil [29]).

In Exhibit 3, we find that compensation practices in utilities differ from those found among industrial firms by Murphy [27]. The total compensation as well as the proportion of long-term compensation in the total packages are much lower in utilities than in industrial firms. We now discuss the pay packages of the top management group and each of the top management positions.

All Managers. The average total annual compensation, TCOMP, of public utility executives amounts to $202,600, which represents only 34.3% of the corresponding compensation of executives employed by industrials, based on Murphy's [27] sample after adjusting for inflation. (11) Salary and bonus, SALB, in public utilities amounts to about 39% of that in industrials in Murphy [27]. It is usually the largest component in the compensation packages of public utility managers. SALB represents 93.5% of TCOMP in utilities versus 80.1% for industrials.

From the pay-performance perspective, the difference between TCOMP and SALB, i.e., LTC, is particularly useful since it contains the reward for long-term performance, such as stock grants and other contingent compensation. There is some evidence that the stock market reaction to the introduction of long-range managerial schemes is positive, as would be expected if these plans help in aligning the interest of managers with stockholders (Brickley, Bhagat, and Lease [8]). The average public utility manager received only $18,500 as LTC, versus $117,450 in Murphy's [27] sample of industrials. LTC, as a fraction of total pay, was 6.5% in utilities in contrast to tenure in our q equations. (18) We also control for the median tenure of the directors. A board which does not turn over rapidly is potentially a symptom of an entrenched management. Conversely, longer director tenures might simply signal that directors in the firm invest in firm-specific human capital to make them more proficient directors. Finally, we include a dummy variable which takes the value one if two or more of the directors are related in any of the years of our sample. We include this variable because family companies are sometimes believed to be fundamentally different from other companies (Mace [27]), and we wished to control for this difference.

The results of these equations are similar to those discussed above for both shareholdings and composition. Shareholdings still seem to affect q. The effects once again are larger when we treat the shareholdings variables as endogenous. There does not appear to be a relation between board composition and q, even treating board composition as endogenous using lagged composition as instruments.

CEO tenure does not seem to affect profitablity at low levels of tenure. However, for CEOs who have been on the job more than 15 years, each additional year reduces profitability. When we treat the other variables in the equation as endogenous (the second column), this effect is significant at the five percent level. This suggests that CEOs who remain on the job for too long become entrenched and reduce corporate performance.

Boards with longer median tenures tend to have higher q's. However, when we treat these variables as endogenous, the effect goes away for the inside directors, but not for outside directors. This result is consistent with the notion that the acquisition of firm-specific knowledge over time by outside directors improves firm performance.

One potential objection to these results is that they are not generated by relations between the control structure and corporate performance. Rather, they may be generated by some spurious correlation between our variables and our measure of q. We test this objection by rerunning our regressions using an earnings-based measure of profitability. While accounting numbers are imperfect measures of economic profits, (Fisher and McGowan [13]), we nonetheless feel that accounting measures of profitability can provide an independent check of the results using q.

The accounting measure of profitability we use is earnings before interest and taxes. (19) This measure is a particularly good one for our purposes because it is not sensitive to changes in capital structure or special tax treatments. We normalize this number by our estimate of the replacement cost of the firms assets.

Piecewise linear equations predicting earnings levels are shown in Exhibit 3. The pattern of the signs is similar to thos in Exhibit 2. Profitability still increases at very small ownership levels (less than one percent) and declines at moderate ownership levels (one percent to five percent). Once again these effects are larger and more significant when instrumental variables are used to control for the endogeneity of the shareholdings.

The one variable which is significant in the pooled ordinary least-squares earnings equation which is not significant in the analogous q equation is the fraction of outsiders between 40% and 60%. However, the effect goes away completely when we treat it as endogenous (the second column). Therefore, this correlation is most likely due to firms who perform poorly adding outsiders (Hermalin


and Weisbach [16]) rather than firms with outsiders on their board performing poorly.


The analysis of the previous sections reveals some interesting results. The most striking is that there appears to be no relation between board composition and performance. Admittedly, this could simply be due to insufficiently powerful tests. On the other hand, our results do suggest that even if such a relation does exist, it is small, with little economic significance. Although a "negative" result, it is important to consider the reasons behind it and their implications four our understanding of corporate governance and incentive theory.

One obvious explanation for this result is that board composition simply does not matter. Inside and outside directors are equally bad (or, possibly, good) at representing the shareholders' interests. This explanation is certainly consistent with top management's control of the board-selection process.

Yet, it is less consistent with the growing literature suggesting that outside directors play an important role in monitoring management (Weisbach [37], Brickley, et al [4], and Shivdasani [34]). A way to reconcile these other papers with ours is to recognize that there are good reasons to have inside directors. Mace [27] and Vancil [35] have argued that inside directors on the board facilitate the succession process. In addition, Mace [27] has suggested that inside directors help CEOs maximize value by providing both advice and knowledge about the day-to-day operations of the company. If each board is optimally weighted between insiders and outsiders, we would not expect to find a cross-sectional relation between board composition and performance in equilibrium. Moreover, the outside-director-as-monitor literature has focused on extraordinary events such as unusually bad performance or adoption of poison pills, from which it is difficult to ascertain the value of the day-to-day monitoring by outside directors.

A third explanation for our negative result is that firms, despite any variation in the severity of their underlying agency problems, reduce theiragency problems to approximately the same level of residual agency. Since their residual agency problems -- which are all that matter for performance -- are the same, variation in performance will be uncorrelated with actions taken to reduce underlying agency problems such as board composition. This explanation would suggest that although a strong outsider-dominated board reduces agency costs, it is impossible to find evidence for this by regressing performance on board composition.

The results concerning stock ownership by top management and CEO tenure are consistent with the arguments presented above. At low levels of ownership (less than one percent), corporate performance improves with increases in ownership. At these levels, management's interests are increasingly aligned with the shareholders'; but with less than one percent of the stock, management does not own enough stock to insulate it from other disciplinary devices such as the takeover market. Beyond one percent, corporate performance declines with ownership, possibly because the increasng insulation from disciplinary devices more than offsets the increased alingment of interests. These results, which were obtained using an econometric technique controls of the endogeneity of managerial shareholdings, serve to confirm the findings of other authors (Morck, et al [29] and McConnel and Servaes [28]) and to suggest that the results of this prior research were not driven by its assumption of exogenous shareholdings.

Much work remains to be done in this area. First, stronger tests should be developed to discern whether board composition has any effect on firm performance. Second, the relation between board composition and incentive contracts should be investigated -- in particular, are these controls complements or substitutes. Finally, more work is needed on the extent to which firm performance is determined by the incentives provided to managers.

Benjamin E. Hermalin is an Assistant Professor of Economics at the University of California at Berkeley, Berkeley, California, and Michael s. Weisbach is an Assistant Professor of Economics at the University of Rochester, Rochester, New York.

(1) In fact, Hermalin [15] has shown that different governance structures can be optimal even for identical firms in the same industry. The reason is that, in his model, the competition is such that not all firms would benefit by providing their managers with strong incentives; but if no firms were to provide incentives, some firms would, then, find it in their interests to provide strong incentives.

(2) The literature suggests that these two variables do measure what we intend them to measure. Jensen and Murphy [21] show that the vast majority of direct incentives faced by top managers come through stock ownership. Weisbach [37], Brickley, et al [4], Brickley and James [5], Shivdasani [34], Byrd and Hickman [6], and Rosenstein and Wyatt [30] all provide evidence suggesting that board structure is an important determinant of the level of direct monitoring of top management.

(3) On the other hand, directors could seek to establish a reputation for not rocking-the-boat, which might make them more attractive to other firms, whose managements are looking to avoid scrutiny and interference. Moreover, as Holmstrom [18] has shown, simply because an agent (e.g., a director) is concerned about his reputation does not imply that he will take actions that are in the principal's (the shareholders') interests.

(4) Kaplan and Reishus [23] have examined this reputation idea by studying whether CEOs of poorly performing firms are asked to serve as directors of other companies with the same frequency as CEO of good-performing firms. They find that poor performance -- measured as a reduction in dividends -- leads to fewer outside directorships for the CEO.

(5) Along these lines, firms vary in many dimensions, and consequently, the underlying degree of divergence between the shareholders' interest and management's interest will also vary. We would expect that the firms in which the underlying degree of divergence is greatest to be the ones in which the strongest measures for closing this divergence would be most used -- their managers would have the stronger pecuniary incentives and the most outside directors on their boards. Even the strongest measures will not fully eliminate this divergence. Indeed, those firms with the greatest degree of divergence could still be the firms with the greatest degree even after all firms adopt measures to reduce this divergence. Optimizing firms will adopt these measures up to the point where their marginal benefit equals their marginal cost. So, if there are decreasing returns to these measures, as seems plausible, then firms that start with the greatest divergence of interest will still have the greatest residual divergence of interest and the worst performance even with the strongest control measures (e.g., the greatest proportion of outside directors). This explanation works against our finding a relation between performance and the proportion of outside directors.

(6) An alternative approach to this problem would be to analyze the change in the stock price around the additions of outside directors. Rosenstein and Wyatt [30] use this approach and find that stock prices rise around the announcement of new outside directors, which is consistent with the monitoring hypothesis.

(7) Examples of this include reserving top management positions or board seats for family members, creating foundations to honor family members, and the use of company resources for family members.

(8) See "CEO Disease," Business Week, April 1, 1991, pp. 52-60.

(9) Tobin's q has been used as a measure of profitability by a number of studies (see, for example, Morck, Shleifer, and Vishny [29], McConnell and Servaes [28], or Lindenberg and Ross [25]).

(10) For more detail on the estimation of the market value of long-term debt, as well as the rest of the construction of q, see Salinger and Summers [31].

(11) One might object that this assumption could biasthe results, because true depreciation is not exponential and deviations of true depreciations from estimated depreciations could be correlated with the independent variables. To test the importance of this objection, we recomputed q under the assumption that the firm's reported depreciation was the actual depreciation. The q's were very similar to the q's used below; the correlation coefficient between the two sets of q's was 0.995. When the equations were reestimated using the q's with the firm's depreciation, the results were almost identical to those reported.

(12) For a detailed discussion of how the sample was constructed and how the directors were classified, see Hermalin and Weisbach [16].

(13) For most directors, this number was identical to the number of shares beneficially owned. However, for a number of directors, the two were considerably different. The most common reason was family holdings. Often, a director will deny beneficial ownership of shares held by another member of his family (typically a wife or minor children). Even if the director does not exercise direct voting power of such shares, it would seem likely that they would be voted as he suggested, and so we attributed them to the director.

Another frequent reason for differences between beneficial and voting ownership was trusts and foundations. Often a director would be a trustee of a trust or foundation in which he had no beneficial interest, but in which he controlled that trust or foundation's voting rights. In such cases we tended to include the holdings of the trust or foundation with the holdings of the director. In cases where the director's control of the voting rights was felt to be minimal or unclear (for example, when director A was a trustee of director B's family trust), we did not add in the trust's shares, but rather noted the association and the size of the trust's holdings.

(14) To standardize our attribution of shareholdings, we established a set of guidelines which covered most of the cases we encountered. The goals of the guidelines were: (1) to attribute shares to the director who was likely to have the greatest say in how they were voted; (2) to avoid double counting; and (3) to establish a way to note that directors were associated with a set of stock when it was either unlikely they had absolute control over the set, or when disentangling control was impossible given the information on the proxy. As part of meeting goals (1) and (2), when two (or more) directors clearly controlled a set of stock, we attributed the stock to the employee director, when one director was a full-time employee of the firm, while the other was not; or to the director who was a member of top management, when one director was top management, while the other was not; or to the senior family member, when the directors were related.

Goal (3) was met by recording the set of stock in question and noting which directors were connected with it. For example, in 1980, Bendix owned roughly 20% of Asarco, and consequently two members of Bendix' board served as directors of Asarco. Bendix' holdings were recorded separately and it was noted that the two dirctors in question were connected with these holdings. Other examples are firms dominated by a single family where the family's shareholdings are in complex trusts and holding companies. For three companies, Carnation, Winn-Dixie Stores, and Zayre, we chose to record only the sum of the family holdings and note the family directors associated with those holdings rather than attempt to disentangle complicated family ownership.

(15) Occasionally, precise starting and finishing dates were reported in the proxies as part of the director's biography. For those CEO's beginning prior to 1971, and for whom we only knew the starting year, we assumed they began July 1st of that year to minimize bias when carrying out certain calculations.

(16) We understand that the proportion of outside directors is likely to be a crude proxy for the "effectiveness" of the board. We use it because of two reasons: first, it is the measure of independence most commonly discussed in both the academic and institutional presses; second, it is the most relevant for proposals about board regulation. An alternative measure used successfully by Shivdasani [34] is the number of additional directorships held by each outside director, which is a measure of the value of the reputation of each director which should affect his incentives to uphold his reputation. See Weisbach [38] for more discussion.

(17) Morck, et al [29] use a slightly different division. Since we have data on very low levels of ownership by top management, while Morck, et al do not, we chose our divisions to make use of this data. The results using their divisions are similar, though less pronounced.

(18) From Warner, et al [36] and Weisbach [37], we know that CEO tenure is jointly endogenous, as these papers find a relation between poor performance and CEO turnover. That effect, however, is sufficiently small so that we do not expect it to create any serious biases. We have estimated our equations treating CEO tenure as both exogenous and endogenous, and have obtained nearly identical results. For the sake of space, only the results treating CEO tenure as exogenous are shown here.

(19) This variable is constructed by adding data items 15, 16, 18, and 49 on the annual COMPUSTAT industrial tape.


[1] American Law Institute, "Principles of Corporate Governance and Structure: Restatement and Recommendations, Tentative Draft No. 1," Philadelphia, The American Law Institute, 1982.

[2] B.D. Baysinger and H.N. Butler, "Corporate Governance and the Board of Directors: Performance Effects of Changes in Board Composition," Journal of Law, Economics, and Organization (Spring 1985) pp. 101-124.

[3] W. Brainard, J. Shoven, and L. Weiss, "The Financial Valuation of the Return to Capital," Brookings Papers on Economic Activity (Issue 2, 1980), pp. 453-502.

[4] J.A. Brickley, J.L. Coles, and R.L. Terry, "The Board of Directors and the Enactment of Poison Pills," Working Paper, University of Rochester, 1991.

[5] J.A. Brickley and C. James, "The Takeover Market, Corporate Board Composition, and Ownership Structure: The Case of Banking," Journal of Law and Economics (April 1987), pp. 161-181.

[6] J. Byrd and K. Hickman, "Do Outside Directors Monitor Managers? Evidence from Tender Offer Bids," Working Paper, Washington State University, 1991.

[7] G. Crystal, In Search of Excess, New York, Norton, 1991.

[8] H. Demsetz, "The Structure of Ownership and the Theory of the Firm," Journal of Law and Economics (June 1983), pp. 375-393.

[9] H. Demsetz and K. Lehn, "The Structure of Corporate Ownership: Causes and Consequences," Journal of Political Economy (December 1985), pp. 1155-1157.

[10] D. Dunn, "Directors Aren't Doing Their Jobs," Fortune (March 16, 1987), pp. 117-119.

[11] E. Fama, "Agency Problems and the Theory of Firm," Journal of Political Economy (April 1980), pp. 288-307.

[12] E. Fama and M. Jensen, "Separation of Ownership and Control," Journal of Law and Economics (June 1983), pp. 301-326.

[13] F.M. Fisher and J.J. McGowan, "On the Misuse of Accounting Rates of Return to Infer Monopoly Profits," American Economic Review (March 1983), pp. 82-97.

[14] J. Hausman, "Specification Tests in Econometrics," Econometrica (November 1978), pp. 1251-1271.

[15] B. Hermalin, "Heterogeneity and Organzational Form," Working Paper, University of California at Berkeley, 1991.

[16] B. Hermalin and M. Weisbach, "The Determinants of Board Composition," Rand Journal of Economics (Winter 1988), pp. 589-606.

[17] B. Holmstrom, "Moral Hazard and Observability," Bell Journal of Economics (Spring 1979), pp. 74-91.

[18] B. Holmstrom, "Managerial Incentive Problems: A Dynamic Perspective," in Essays in Economics and Management in Honor of Lars Wahlbeck, Helsinki, Swedish School of Economics, 1982.

[19] R. Holthausen and D. Larcker, "Financial Performance and Organizational Structure," Working Paper, University of Pennsylvania, 1991.

[20] M.C. Jensen and W.H. Meckling, "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure," Journal of Financial Economics (October 1976), pp. 305-360.

[21] M.C. Jensen and K. Murphy, "Performance Pay and Top-Management Incentives," Journal of Political Economy (April 1990), pp. 224-264.

[22] M.C. Jensen and K. Murphy, "CEO Incentives -- It's Not How Much You Pay, But How," Harvard Business Review, (May-June 1990), pp. 138-153.

[23] S. Kaplan and D. Reishus, "Outside Directorships and Corporate Performance," Journal of Financial Economics (October 1990), pp. 389-410.

[24] S. Kole, "Managerial Ownership and Firm Performance: Incentives or Rewards," Working Paper, University of Rochester, 1991.

[25] E. Lindenberg and S. Ross, "Tobin's q Ratio and Industrial Organization," Journal of Business (January 1981), pp. 1-32.

[26] P.W. MacAvoy, S. Cantor, J. Dana, and S. Peck, "ALI Proposals for Increased Control of the Corporation by the Board of Directors: An Economic Analysis," in Statement of The Business Roundtable on the American Law Institute's Proposed "Principles of Corporate Governance and Structure: Restatement and Recommendations," New York, Business Roundtable, 1983.

[27] M.L. Mace, Directors, Myth and Reality, Boston, Harvard Business School Press, 1971.

[28] J. McConnell and H. Servaes, "Additional Evidence of Equity Ownership and Corporate Value," Journal of Financial Economics (October 1990), pp. 595-613.

[29] R. Morck, A. Shleifer, and R. Vishny, "Management Ownership and Market Valuation: An Emprical Analysis," Journal of Financial Economics (January/March 1988), pp. 293-316.

[30] S. Rosenstein and J.G. Wyatt, "Outside Directors, Board Independence, and Shareholder Wealth," Journal of Financial Economics (August 1990), pp. 175-192.

[31] M. Salinger and L. Summers, "Tax Reform and Corporate Investment: A Microeconometric Simulation Study," in Behavioral Simulation Methods in Tax Policy Analysis, M.S. Feldstein (ed.), Chicago, University of Chicago Press, 1983.

[32] H. Schaller, "A Reexamination of the Q Theory of Investment Using U.S. Firm Data," Journal of Applied Econometrics (December 1990), pp. 309-323.

[33] S. Shavell, "Risk Sharing and Incentives in the Principal and Agent Relationship, "Bell Journal of Economics (Spring 1979), pp. 55-73.

[32] A. Shivdasani, "Board Composition, Ownership Structure, and Hostile Takeovers," forthcoming, Journal of Accounting and Economics.

[35] R. Vancil, Passing the Baton: Managing the Process of CEO Sucession, Boston, Harvard Business School Press, 1987.

[36] J.B. Warner, R.L. Watts, and K.H. Wruck, "Stock Prices and Top Management Changes," Journal of Financial Economics (January/March 1988), pp. 461-492.

[37] M. Weisbach, "Outside Directors and CEO Turnover," Journal of Financial Economics (January/March 1988), pp. 431-460.

[38] M. Weisbach, "Corporate Governance and Hostile Takeovers," forthcoming, Journal of Accounting and Economics.
COPYRIGHT 1991 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Corporate Compensation Policy Special Issue
Author:Agrawal, Anup; Makhija, Anil K.; Mandelker, Gershon N.
Publication:Financial Management
Date:Dec 22, 1991
Previous Article:The effects of board composition and direct incentives on firm performance.
Next Article:On the capital structure of leveraged buyouts.

Related Articles
De-fanging CEO compensation.
Tracking pay for performance.
The compensation committee's destiny.
Monitoring versus bonding: shareholder rights and management compensation.
Compensation Obscenity.
Entrepreneurship and executive compensation: turning managers into owners.
The IRS tackles excessive executive compensation.
CalSTRS calls for guidelines on executive pay.
The case for an executive compensation standards board: if private industry does not act quickly to self regulate executive pay and set high...

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters