laksmana - 2008 - corporate board governance and voluntary disclosure of executive compensation practices

The absence of clear regulatory guidance has led to cross-sectionaldifferences in the amount of information reported by the boards, providing aresearch avenue to examine whether quality

Corporate Board Governance and Voluntary

1 Introduction

The rise of recent corporate scandals has focused considerable public attention onthe role of boards of directors in corporate governance McKinsey & Company’sglobal investor opinion surveys (2000, 2002), for example, report that institutionalinvestors perceive board practices to be at least as important as financial perfor-mance when evaluating companies for investment The surveys also highlight thatinvestors demand that boards take greater responsibility for communicating materialmatters to shareholders and be responsive to investor requests for information ongovernance issues

In the present study, I examine whether certain board and compensation mittee characteristics, as proxies for board governance quality, are associated withthe extent of board disclosure of compensation practices Executive compensationdisclosure provides a natural setting to examine disclosure of board practices Toaddress the criticisms about board failure to use the “pay for performance” stan-dard, the U.S Securities and Exchange Commission (SEC) made boards moreaccountable for their decisions by requiring a report justifying their compensationpolicies (SEC 1992) The SEC, however, does not specify the items to be disclosed

com-or the repcom-orting fcom-ormat to be followed Thus, boards have considerable latitude inwhat details to report, making the reported items, for all practical purposes, voluntarydisclosures The absence of clear regulatory guidance has led to cross-sectionaldifferences in the amount of information reported by the boards, providing aresearch avenue to examine whether quality of board governance is associated withvariations in board disclosure of compensation practices

The present study differs from prior disclosure studies (e.g., Chen and Jaggi2000; Eng and Mak 2003; Ajinkya, Bhojraj, and Sengupta 2005; Karamanou andVafeas 2005) by examining disclosures as the outcomes of board decisions ratherthan the (assumed) effect of boards on management decisions The present studyfocuses on compensation-related disclosures because boards of directors (especiallythe compensation committees) have a responsibility to report the basis of theiractions for determining executive compensation in the companies’ proxy statements

* Accepted by Laureen Maines This study is based on my dissertation at Georgia State University.

I would like to thank my dissertation committee members, Lawrence D Brown, Jennifer R Joe, Omesh Kini, and especially Ram S Sriram (chair), for their guidance and support I am grateful to Laureen Maines (associate editor) and two anonymous referees for their constructive comments and suggestions An earlier version of this paper benefited from comments made by participants at the 2005 Annual Meeting of the American Accounting Association and the 2005 Ohio Region Meeting of the American Accounting Association.

1148 Contemporary Accounting Research

To make effective disclosure decisions, boards and compensation committees need

to devote a significant amount of time and resources (i.e., personnel and their edge base) to set compensation disclosure policies, examine potential disclosureitems, consider the consequences of several disclosure options, and make the finaldecisions I posit that the time and resource commitment of directors to performthese tasks is positively associated with the extent of compensation practice dis-closure I use three proxies to measure the time and resource commitment of boards:the proportion of busy outside directors (measured by number of directorships),meeting frequency, and board (compensation committee) size The first two are meas-ures for time commitment of directors and board diligence Board (committee) size

knowl-is a proxy for knowledge base and ability to dknowl-istribute workload and assignments.Boards of directors also need the power to act independently from manage-ment to serve the best interests of shareholders The present study examines boarddisclosure decisions where shareholders and managers may have conflicts ofinterests Although shareholders demanding greater disclosure on executive compen-sation practices show enthusiasm for the compensation committee report, corporateexecutives expressed concerns that the report was “an undue intrusion into theinternal affairs of the company” (SEC 1992, 2992) Management resistance togreater disclosure of compensation practices suggests potential disagreements withboards on the extent of disclosure I posit that more independent boards and com-pensation committees are more likely to make objective decisions by supportinggreater disclosure Because chief executive officers (CEOs) have some influence

on the director nomination process, I also investigate whether the presence of inant CEOs is associated with less compensation transparency

dom-I develop a disclosure index consisting of 23 compensation-related items dom-Inote that firms reported, on average, 32 percent of the total items in the first year(i.e., 1993) that boards were required to issue a compensation report I also notethat boards reported more compensation-related disclosures in 2002 than they did

in 1993 The development of this disclosure index is a unique feature of this study.However, like other disclosure indexes, its construction is subjective I, therefore,validate the index by showing that the disclosure scores are inversely related to two

This provides evidence that greater compensation disclosure reduces informationasymmetry

On the basis of the 1993 sample, I find that boards and compensation tees with the authority to exercise independent oversight of management providemore disclosure of executive compensation practices The results show that CEOinfluence in the director selection process and, more weakly, board (compensationcommittee) independence status play some role in explaining board disclosurepractices From both sample periods, I find that board disclosure increases with theamount of time and resources dedicated to board duties More specifically, board(compensation committee) meeting frequency and board (compensation commit-tee) size are positively associated with compensation practice transparency.The results of this study contribute to a greater understanding of why someboards are more likely to voluntarily disclose their executive compensation practices

commit-Corporate Governance and Executive Compensation Disclosure 1149than others First, the study provides further support that boards need the power toact independently from top management for them to make effective decisions Theresults suggest that less CEO influence in the director nomination process leads togreater transparency because directors have greater ability to exercise independentjudgement when there are potential disagreements with management These resultsare consistent with prior studies documenting that independent-dominated boardsare more likely to influence management disclosure decisions than management-dominated boards (Chen and Jaggi 2000; Ajinkya et al 2005; Karamanou andVafeas 2005) Second, the results highlight the importance of the board governanceprocess for effective decision making The study documents that having less time

to meet as a group and having fewer directors serving on the board lead to lesstransparency Frequent board meetings would facilitate greater information sharingamong directors and having more directors serving on boards would allow betterworkload distribution and committee assignments

The remainder of this paper is organized into five sections Section 2 providessome background information and discusses the hypotheses Section 3 addressesthe sample selection process, data sources, development of the disclosure index,and its validity assessment Section 4 discusses the main analyses, including theboard governance measures, the ordinary least squares (OLS) regression model,and the control variables, and presents the OLS regression results Section 5 presentsthe additional analysis using the two-stage regression model Section 6 concludesand discusses the implications of the study and future research

2 Background and hypothesis development

Board report on executive compensation practices

On October 15, 1992, the SEC enacted new disclosure rules governing executive

explain the basis of their decisions in setting executive compensation The reportaddresses a frequent criticism about the lack of association between senior man-agers’ performance and their pay The SEC believes that the report will “enhanceshareholders’ ability to assess how well directors are representing their interests”and “to inform shareholders of the (Compensation) Committee’s good-faith rationale

The board’s report on executive compensation practices should provide holders with information on how management performance is measured andwhether pay is aligned to performance The report requires directors to disclose thebasis for their actions in setting the compensation of executive officers, includinghow corporate performance and executive compensation are aligned In the case ofthe CEO, the report must also address the extent to which CEO compensation isperformance-related and the quantitative and qualitative performance measuresused to determine the compensation The board, however, does not have to reportspecific performance targets or other compensation disclosures that would jeopard-ize the firm’s competitive advantages Other than requiring boards to report therationale for their decisions, the SEC does not provide specific guidance on the spe-cific items that must be reported and the reporting format that must be followed

share-1150 Contemporary Accounting Research

Board governance and compensation practice disclosures

To make effective decisions, boards need the power to act independently frommanagement Unlike insider and affiliated directors whose career or personal inter-ests are tied to the firm management, outside directors should be free from personalconflicts of interest and be able to exercise independent judgement when there aredisagreements with management Boards also need to devote sufficient time andresources to perform their monitoring duties Effective board decisions necessitatethat directors dedicate adequate time and resources to learn the issues at handbefore board meetings, engage in substantive discussion and decision making dur-ing meetings, and perform follow-up reviews after meetings (Conger, Finegold,and Lawler 1998; Beasley, Carcello, Hermanson, and Neal 2007)

Prior research has shown that the appointment of independent outside

independent outside directors value their reputation as directors because good utation signals the value of their services to the director labor market Reputationwill be more important to them than to inside directors whose careers are tied tothe CEOs or affiliated outside directors who have family or business ties to the

the board and its committees is associated with the quality of the financial ing process Board independence and / or audit committee independence areinversely related to the likelihood of earnings manipulation (Dechow, Sloan, andSweeney 1996) and financial statement fraud (Beasley 1996; Wright 1997), theabsolute values of abnormal accruals (Klein 2002a), and the likelihood that firms

associ-ated with a decrease in the likelihood of auditor dismissal following going-concernopinions (Carcello and Neal 2003) and an increase in the likelihood of auditors’issuing going-concern opinions (Carcello and Neal 2000) Finally, board inde-pendence is positively associated with the issuance, frequency, and accuracy ofmanagement earnings forecasts (Ajinkya et al 2005; Karamanou and Vafeas 2005)

A number of studies find that corporate executives are sensitive to stakeholdercriticism on compensation-related issues and distort proxy disclosures to managestakeholder impressions (e.g., Lewellen, Park, and Ro 1996; Murphy 1996; Yer-mack 1998; Baker 1999) Greater compensation practice disclosure increases theability of investors to monitor and punish managers for underperformance In addi-tion, greater disclosure could diminish management’s ability to negotiate favorablechanges in future contracts with the board Unjustified deviations from publiclydisclosed policies, such as awarding a “special” bonus when top management fails

to achieve previously reported performance targets and lowering performance gets to make them easier to reach (Byrne, Lavelle, Byrnes, and Vickers 2002),could attract adverse publicity and shareholder criticisms Although shareholdersdemand greater disclosure of executive compensation practices and show enthusi-asm for the compensation committee report, corporate executives have expressedconcerns that the report is “an undue intrusion into the internal affairs of the com-pany” (SEC 1992, 2992)

tar-Corporate Governance and Executive Compensation Disclosure 1151Requiring boards and compensation committees to report details of executivecompensation imposes greater responsibility for them to justify their compensationpolicies and become more effective at performing their monitoring duties Inde-pendent, outside directors whose reputations are at stake are more likely to reportmore details of compensation practices than are inside or affiliated directors whenthere are disagreements with management on the extent of disclosure Becauseimproved compensation transparency is in the best interests of shareholders, inde-pendent directors who value their reputation are more likely to provide greatercompensation disclosure than inside or affiliated-outside directors Therefore:

com-pensation committee report is positively associated with board and compensation committee independence.

Although directors have reporting responsibility to disclose compensationpractices, top management, especially the CEO, could affect board decisions onthe extent of information to be reported CEOs have incentives to protect their joband increase their other benefits by using their influence in the director nominationprocess They often dominate this process in an attempt to acquire “weak” boardswhose directors are less likely to challenge the CEOs and are more inclined to sup-port their decisions (Mace 1986; Lorsch and MacIver 1989; Monks and Minow1996) In the context of the present study, CEOs with significant influence overthe director nomination process could use their power to limit compensationdisclosure Thus:

com-pensation committee report is negatively associated with CEO power over the director nomination process.

Prior studies show that directors who have reputations as effective tive) monitors are rewarded (punished) with increases (decreases) in the number ofdirectorships held (Gilson 1990; Shivdasani 1993; Harford 2003; Farrell andWhidbee 2000) Furthermore, Ferris, Jagannathan, and Pritchard (2003) find a sig-nificantly positive market reaction to the initial appointment of outside directorswith multiple directorships to a board without an incumbent “busy” director, sug-gesting that the experience and reputation capital of such directors are valuable tothe board

(ineffec-While the number of multiple directorships seems to be associated with tor expertise, directors serving on too many boards could spread themselves toothinly As the number of directorships increases, directors could devote less time toperforming their duties at a single board Busy boards are associated with anincreased likelihood of financial fraud (Beasley 1996), excessive CEO compensa-tion (Core, Holthausen, and Larcker 1999), substandard financial performance,and negative abnormal returns (Fich and Shivdasani 2006), implying that the pres-ence of overcommitted directors reduces the oversight of management

direc-1152 Contemporary Accounting Research

Executive compensation is a vital mechanism to discipline management.Boards and their compensation committees have the primary responsibilities to setappropriate incentive contracts, to ensure that the contracts are optimally struc-tured, and to adjust the contracts to better align performance measures to pay.Directors holding multiple directorship positions are likely to have more experi-ence to perform these tasks and have greater incentive to maintain good oversightand decision-making performance To perform these duties effectively, however,boards and their compensation committees must spend adequate time and effort.Mace (1986) and Lipton and Lorsch (1992), on the basis of their experience serv-ing on corporate boards, conclude that typical directors do not devote enough time

to carry out their duties This problem is more severe for outside directors holdingtoo many directorships These competing explanations, therefore, lead to the follow-ing nondirectional hypothesis:

HYPOTHESIS 3 The extent of executive compensation disclosure in the sation committee report is associated with “busy” boards and compensation committees.

effective-ness (Conger et al 1998) A few studies suggest that the frequency of board andcommittee meetings is a relatively good proxy for board diligence Vafeas (1999)finds that boards meet more frequently after stock price declines and that firm per-formance increases following years of higher number of board meetings Fre-quency of meetings is also associated with the quality of financial reporting Boardand audit committee meeting frequency is positively associated with external auditfees (i.e., more audit work for greater assurance) (Carcello, Hermanson, Neal, andRiley 2002) and is negatively associated with discretionary accruals (Xie, David-son, and DaDalt 2003) I posit that board and compensation committee meetingfrequency is positively associated with the extent of compensation practice disclo-sure when directors have more time together as a group to discuss various aspects

of compensation disclosure Thus:

com-pensation committee report is positively associated with board and compensation committee meeting frequency.

Two views exist regarding the relationship between board (committee) sizeand its monitoring effectiveness Some believe that smaller boards facilitate morefrequent and intense information sharing and processing than do larger boards(Lipton and Lorsch 1992; Jensen 1993) This view is consistent with organiza-tional behavior research, such as Hackman 1990, highlighting that productivitylosses could arise when working groups grow larger Yermack (1996), for example,provides empirical evidence on the effectiveness of small boards He finds thatsmaller boards are more likely to provide CEOs with stronger compensation incen-tives and dismiss them for poor performance than are larger boards

Corporate Governance and Executive Compensation Disclosure 1153

In contrast, more recent studies suggest that larger boards have a greaterknowledge base to fulfill their advisory role (Coles, Daniel, and Naveen 2005) andgreater ability to distribute workload and committee assignments to perform theirmonitoring duties (Klein 2002b; Anderson, Mansi, and Reeb 2004) than do smallerboards Recent studies show that board (committee) size is inversely related to thecost of debt (Anderson et al 2004) and positively associated with the issuance ofmanagement earnings forecast updates (Karamanou and Vafeas 2005) In the con-text of the present study, larger boards (compensation committees) would havemore resources (i.e., directors with various backgrounds and skills) to performtheir tasks than smaller boards However, larger boards could become less effectivethan smaller boards because of coordination and free rider issues These competingexplanations lead to the following nondirectional hypothesis:

com-pensation committee report is associated with board and comcom-pensation committee size.

3 Sample and disclosure scores

Sample and data sources

The initial sample for this study consists of firms in nonregulated industries listed

on the Standard & Poor’s (S&P) 500 as of December 31, 1992 Two examinationperiods, 1993 and 2002, are used The 1993 proxy season is selected because theboard’s report on compensation practices was required for firms filing any proxystatements after January 1, 1993 By selecting the 1993 proxy year, I could examineboards’ first response to the new disclosure requirement The 2002 proxy season isselected because, at the time of data collection, it was the most recent period forwhich proxy statements were available

The data on board disclosures, board and compensation committee istics, director information, stock ownership by blockholders and institutions, andshareholder votes were hand-collected from proxy statements Searches of theDow Jones Interactive / Factiva database were performed to identify firms’compensation-related news Executive compensation and other accounting/marketdata were obtained from COMPUSTAT, ExecuComp, and the Center for Research

character-in Security Prices (CRSP)

In selecting the sample, I excluded firms if they are in regulated industries (110

reports, director data, and compensation, as well as accounting and / or other dataare not available for the examination periods The final sample for the 1993 and

2002 proxy seasons consists of 218 and 232 firms, respectively The sample sents six major industries Manufacturing is the largest group (65 percent) Theremaining firms are from wholesale and retail (14 percent), service (9 percent),transportation (5 percent), mining (6 percent), and construction (1 percent)

repre-1154 Contemporary Accounting Research

Compensation disclosure scores

To measure the extent of board disclosure, I constructed a database of disclosureindex based on information reported in compensation committee reports Priorresearch used disclosure indexes to measure voluntary disclosure, but they focusedmainly on financial reporting practices (e.g., Botosan 1997; the Association forInvestment Management and Research (AIMR) disclosure rankings) To myknowledge, no published study to date has used an index to specifically measuredisclosures related to management performance and reward The selection of itemsincluded in my compensation disclosure checklist was guided by the SEC rules forthe Board Compensation Committee Report (SEC 1992); discussion of the ele-ments of incentive compensation plans by Butler and Maher 1986 and Hilton,Maher, and Selto 2003; the Conference Board’s Best Practices for EstablishingExecutive Compensation (Brancato, Peck, and Hervig 2001); and leading companyproxy statements

Table 1 presents the disclosure checklist and percentage of firms reportingeach item during the 1993 and 2002 proxy years Firms receive one point for thepresence of each item in their compensation committee report The extent of boarddisclosure on compensation practices is measured as the sum of scores obtained.Each item is equally weighted because user preferences are not known In both

1993 and 2002, the basis for determining salary (item 4), type of general and cific measures for determining annual rewards (items 8 and 9), and discussion ofwhether annual rewards are granted on achievement of performance targets (item15) are the items reported most often; more than 84 percent of corporate boardsdisclose them In contrast, weights assigned to performance measures (items 11, 18,and 19), specific performance targets (items 13 and 21), and award formulas (items

spe-14 and 22) are among the least reported items A number of firms disclosed that theykept these items confidential to avoid compromising their competitive positions.Chi-square tests were performed to compare the percentage of firms disclosingeach item in 1993 with that in 2002 Compared with 1993, there have been signifi-cant increases in the percentage of firms disclosing the employment of compensationconsultants (item 2), the basis for determining salary (item 4), types of perfor-mance evaluation (items 5 and 6), weights on performance measures (items 10, 11,

18, and 19), and range or absolute value of rewards (items 12 and 20) in 2002 (all

Validity assessment of board disclosure scores

Because the development of the disclosure index requires subjective assessments, Iexamine its validity by correlating the disclosure scores with bid – ask spread andstock return volatility, two common proxies for information asymmetry (Glostenand Milgrom 1985; French and Roll 1986) High bid–ask spread and return vola-tility are associated with high information asymmetry Because greater disclosurewill reduce investors’ uncertainty about how incentive systems are set to promotemanagement accountability, the disclosure scores should be inversely related toboth bid–ask spread and return volatility

Corporate Governance and Executive Compensation Disclosure 1155TABLE 1

Compensation practice disclosure items

Compensation process

Employment of a compensation consultant 2 45.87 60.34 * Top management’s involvement in compensation

Types of measures disclosed (e.g., financial,

nonfinancial, corporate, business-unit,

Specific measures disclosed (e.g., ROA,

operating income, customer satisfaction) 9 86.70 85.78 Weight assigned on performance measures

Weight assigned on each type of measures 10 14.68 31.90 * Weight assigned on each measure 11 4.13 20.26 * Range or absolute value of rewards 12 34.86 52.16 *

Detail formula or steps disclosed 14 6.88 10.78 Discussion whether awards are granted on

achievement of performance targets 15 84.86 88.79

Long-term incentive

Performance measures

Types of measures disclosed (e.g., financial,

nonfinancial, corporate, business-unit,

Specific measures disclosed (e.g., ROA,

operating income, customer satisfaction) 17 37.61 36.21

(The table is continued on the next page.)

% firms disclosing

1993 proxy (n 218)

2002 proxy (n 232)

1156 Contemporary Accounting Research

have been documented in prior research to be correlated with bid – ask spreads

(e.g., Leuz and Verrecchia 2000) All data are from the 1993 and 2002 fiscal years

SPREAD is the annual average of the daily absolute difference between the closing

percentage of daily traded shares (i.e., daily traded shares divided by total shares

stock returns over the fiscal year

SPREAD and SCORE for the 1993 and 2002 data are 0.23 and 0.20,

SPREAD on SCORE and the control variables using both untransformed and

one-tailed), providing evidence that a greater number of disclosure items is associated

with lower information asymmetry

TABLE 1 (Continued)

Weight assigned on performance measures

Weight assigned on each type of measure 18 1.83 10.78 *

Weight assigned on each measure 19 1.83 9.05 *

Range or absolute value of awards 20 11.47 22.41 *

Discussion whether awards are granted on

achievement of performance targets 23 42.20 44.40

Notes:

Chi-square tests are performed to test for differences in percentage of firms reporting

disclosure items in 1993 and 2002.

* Significant at the 0.01 level.

† Significant at the 0.05 level.

% firms disclosing

1993 proxy (n 218)

2002 proxy (n 232)

Corporate Governance and Executive Compensation Disclosure 1157

significant in either period, but the correlations between the natural logs of

VOLATILE and SCORE are negative and significant in both periods (p-value

SCORE is likely nonlinear Panel B of Table 2 reports the results of the regressions

of VOLATILE on SCORE, MKVAL, and VOLUME using both untransformed and

consistent with those reported in panel A of Table 2; a higher disclosure score is

related to lower information asymmetry Overall, the results suggest that the

dis-closure scores are valid

TABLE 2

Regression of bid–ask spreads and stock return volatility

Panel A: Regression of bid–ask spreads on board disclosure scores

1158 Contemporary Accounting Research

4 Main analyses

Board governance measures

I classify the board governance variables into five categories: (a) board and pensation committee independence; (b) CEO power over the director nominationprocess; (c) time commitment of directors (“board/compensation committee busystatus”); (d) board and compensation committee diligence; (e) board and compen-

com-sation committee size The variables in the first category include B_INDDIR,

C_INDDIR, and IND_NOM B_INDDIR is defined as the ratio of the number of

independent outside directors to the total number of directors The definition of pendent, outside directors excludes those who are current employees of the firm,and those with family or business interests tied to top management or the enter-prise, including relatives of management, former employees, professional advisers

inde-to the company (e.g., consultants, atinde-torneys), officers of significant suppliers orcustomers of the company, and interlocking directors (i.e., two CEOs or their sub-

ordinates sit on each other’s boards) C_INDDIR measures compensation committee independence, defined in the same manner as B_INDDIR IND_NOM equals one if

the board has a nominating committee that is composed solely of independentdirectors, and zero otherwise Shivdasani and Yermack (1999) provide evidencethat boards without a nominating committee or whose CEOs serve on this commit-tee have a larger proportion of insiders and affiliated directors than those with afully independent nominating committee

TABLE 2 (Continued)

Notes:

The t-values are in parentheses.

* Significant at the 0.01 level One-tailed test is used whenever the directional

SPREAD is the annual average of the daily absolute difference between the closing bid and

ask prices divided by the mean of the bid and ask prices for the 1993 (2002) fiscal year.

SCORE is the sum of scores received from all compensation practice disclosure items (items

1–23) in the 1993 (2002) proxy year.

MKVAL is the total market value of the firm’s equity at the end of the 1993 (2002) fiscal

year.

VOLUME is the mean percentage of daily traded shares (i.e., daily traded shares divided by

total shares outstanding) over the 1993 (2002) fiscal year.

VOLATILE is the standard deviation of daily stock returns over the 1993 (2002) fiscal year.

Corporate Governance and Executive Compensation Disclosure 1159The second category consists of three variables measuring the degree of CEO

power over the director nomination process: B_%APPOINT, C_%APPOINT, and

CEOTENURE B_%APPOINT and C_%APPOINT are measured as the percentage

of outside directors on the board and the compensation committee, respectively,that were appointed after the current CEO took office Outside directors joining theboard after the current CEO took office are assumed to be appointed by the CEObecause most CEOs have influence in the director nomination process (Mace 1986;

Monks and Minow 1996) CEOTENURE is defined as the number of years the rent CEO has served as the firm’s CEO Baker and Gompers (2003), for example,

cur-documented that the proportion of independent outsiders on boards decreases withthe length of CEO tenure

Four variables are used to measure the time commitment of directors (“board/

compensation committee busy status”): LnB_OTHDIR, Ln C_OTHDIR, B_BUSY, and C_BUSY LnB_OTHDIR and LnC_OTHDIR are the natural logs of the average

number of other directorships held by independent directors serving on the board

and the compensation committee, respectively B_BUSY and C_BUSY are dummy

variables, equal to one if the majority of outside directors on the board and thecompensation committee, respectively, serve on three or more other boards, andequal to zero otherwise Although several studies use the average number of direc-torships to identify busy boards (e.g., Shivdasani 1993; Beasley 1996; Ferris et al.2003), Fich and Shivdasani (2006) suggest that it is a noisy measure when there is

a wide dispersion in the number of directorships held by outside directors on theboards

The next category captures board diligence and consists of two variables,

Ln B_MEETINGS and Ln C_MEETINGS, defined as the natural logs of board

meeting frequency and compensation committee meeting frequency, respectively.Finally, the last category measures board and compensation committee size

Ln B_SIZE and Ln COMP_SIZE are the natural logs of the number of directors

serving on the board and the compensation committee, respectively

Because some variables are highly correlated, I use principal component sis (PCA) to classify the original variables into multiple aspects of board governancequality and obtain factor scores that would be used in the regression analyses Iretain all factors with an eigenvalue greater than one I use an oblique rotation tohave a factor structure in which each original variable loads highly on only onefactor, but allow the retained factors to be correlated

analy-For the 1993 sample, the analysis results in five factors that retain 70.82 percent

of the total variance in the original governance variables The Kaiser-Meyer-Olkin(KMO) measure of the overall sampling adequacy is 0.68 A KMO measuregreater than 0.60 suggests that factoring is appropriate (Sharma 1996, 116) Vari-

ables B_INDDIR, C_INDDIR, and IND_NOM have high loading (i.e., greater than 0.60) on factor 1 (hereafter called INDEPENDENCE) Variables B_%APPOINT,

C_%APPOINT, and CEOTENURE have high loadings on factor 2 (hereafter called CEOPOWER) Factor 3 is labeled BUSYDIR and has four variables with high load-

ings, Ln B_OTHDIR, B_BUSY, Ln C_OTHDIR, and C_BUSY Factor 4 is called

DILIGENCE and has two variables, LnB_MEETINGS and LnC_MEETINGS, with

1160 Contemporary Accounting Research

high loadings Finally, factor 5 (hereafter called BOARDSIZE) has two variables with high loadings, LnB_SIZE and LnC-SIZE The retained factors are consistent

with my initial groupings of the governance variables and those of Larcker,

Richard-son, and Tuna 2007 In the regression analyses, INDEPENDENCE, CEOPOWER,

BUSYDIR, DILIGENCE, and BOARDSIZE are the computed scores of the board

in the 1993 analysis Two variables measuring the frequency of board and

com-pensation committee meetings, LnB_MEETINGS and LnC_MEETINGS, and two variables measuring board and compensation committee size, Ln B_SIZE and LnC_SIZE, load highly on the fourth factor (hereafter called DILIGENCE/SIZE) Four variables based on the computed factor scores, INDEPENDENCE,

CEOPOWER, BUSYDIR, and DILIGENCE / SIZE, are be used in the regression

analyses

Regression model

The relationship between the disclosure score and board governance factors isexamined using the following regression:

SCORE  0 Board Governance Factors i Controls j  (1)

SCORE is the sum of scores received from all disclosure items shown in

Table 1 The scores are for disclosures made during the 1993 (2002) proxy seasonfor executive compensation in the fiscal year (FY) 1992 (2001) All independent

variables, except VOTE (defined below), are from FY 1992 (2001).

I include several control variables that might be correlated with SCORE.

OP_COM is defined as the proportion of stock options to total CEO compensation.

The disclosure items examined in this study (e.g., performance measures, weights,and ranges) may be more relevant to cash and other performance-based compensa-tion than option grants Therefore, firms relying heavily on stock options for CEOcompensation would be likely to have a fewer number of details to disclose

SUBJECT equals one if the board used subjective performance evaluation in

determining CEO compensation, and zero otherwise Information about the use ofsubjective performance evaluation was obtained from the compensation committeereport Subjective performance evaluation refers to the use of complete discretionfor determining compensation if boards do not rely on any objective standards Inthis case, subjective performance evaluation would result in fewer details aboutcompensation practices In contrast, boards could employ objective standards butuse discretion or subjective judgement to adjust the compensation for factors out-side management’s control In this case, the use of subjective evaluation could lead

Corporate Governance and Executive Compensation Disclosure 1161

to greater disclosure because more details on both the quantitative standards and

OVERPAID is measured as CEO actual cash and stock-based compensation

minus the standard (benchmark) pay for FY 1992 (2001) When CEOs are compensated, boards may provide greater disclosure to justify the payment andavoid criticism about their compensation practices This argument is consistentwith the social psychology research (e.g., Schlenker 1980; Schlenker and Weigold1989) that people feel more pressure to justify their decisions when the decisionshave significant consequences for them (e.g., reputation damage) Following Core

over-et al 1999 and Baker 1999, overpayment (underpayment) is a pay component inexcess of (below) a benchmark pay The benchmark pay is determined by regress-ing the total CEO cash and stock-based compensation in FY 1992 (2001) on a set

of economic determinants of compensation: sales, market-to-book, return on assets(ROA), stock return, standard deviation of ROA, and standard deviation of stockreturn The benchmark pay is the predicted value of CEO compensation computedusing the estimated coefficients of this regression

NEWS is a nominal variable equal to one if an article criticizing the firm’s

compensation policies appeared in the financial press (i.e., Barron’s, BusinessWeek,

Dow Jones Business News, Dow Jones News Services, Forbes, Fortune, and Wall Street Journal) during FY 1992 (2001), and zero otherwise (Byrd, Johnson, and

Porter 1998) When compensation plans are being criticized, boards are morelikely to increase disclosure to explain in detail their compensation policies

VOTE is equal to one if the executive compensation plan is put to a shareholder

vote, and zero otherwise Information about the shareholder vote was obtainedfrom the 1993 (2002) proxy statement Because major changes to executive compen-sation plans are usually subject to shareholder approval, the extent of disclosurecould be associated with the upcoming shareholder vote on compensation-relatedissues

%BLOCKOWN and %INST_OWN are the total percentage of outstanding

shares owned by blockholders and institutional investors, respectively, reported inproxy statements Blockholders (institutional holders) are defined as individual(institutional) investors owning a significant percentage (e.g., 5 percent) of out-standing shares Significant blockholder and institutional ownership could putpressure on boards to disclose the details of executive compensation practices, sug-

gesting a positive association between SCORE and both %BLOCKOWN and

%INST_OWN In contrast, the presence of blockholders and institutional holders

could reduce the concerns of outside shareholders about unjustified executive

compensation, suggesting a negative association between SCORE and the two

variables The latter is consistent with Tosi and Gomez-Mejia 1994 documentingthat individual and institutional investors with significant ownership exercise directcompensation monitoring that replaces the need for greater disclosure

PERFORM, a proxy for firm performance, is the ROA in FY 1992 (2001),

defined as income before extraordinary items divided by total assets Empiricalresults on the association between disclosure and firm performance are mixed(e.g., Lang and Lundholm 1993; Skinner 1994) Firm performance could have a

1162 Contemporary Accounting Research

positive effect on disclosures because more profitable firms may differentiatethemselves from others by disclosing more information In contrast, firms withpoor performance may also increase disclosure to reduce negative publicity andadverse price reaction

LnF_SIZE is a measure of firm size, defined as the natural logarithm of total

assets at the end of FY 1992 (2001) Larger firms may have more details to reportbecause they have more complex compensation plans than do smaller firms

PROP is the proportion of sales revenue of the four largest firms in the

indus-try (two-digit Standard Industrial Classification [SIC] code) to total indusindus-try sales(i.e., four-firm concentration ratio) Prior research has used this metric to measure

competitive or proprietary costs (e.g., Harris 1998) High (low) value of PROP

disclo-sure decision provide mixed predictions for the association between disclodisclo-sure andproprietary costs Firms in highly competitive environments (i.e., with higher pro-prietary costs) provide less information to protect their competitive advantages(Verrecchia 1990) In contrast, firms with low proprietary costs may have fewerincentives to provide more information because they want to protect their abnor-

mal profits and market shares (Hayes and Lundholm 1996) Finally, EXCHANGE

is a dummy variable for stock exchange, equal to one if the company’s commonstock is traded on the New York Stock Exchange, and zero otherwise

Descriptive statistics and correlation among variables

Table 3 summarizes the descriptive statistics of the board disclosure scores, theoriginal board governance variables, and the control variables for the 1993 sample(panel A) and the 2002 sample (panel B) Compared with 1993, boards report more

disclosure items in 2002 The differences in the mean and median of SCORE for the two periods are statistically significant (p-value< 0.01) In addition to the over- all disclosure scores (SCORE ), Table 3 reports the scores for each subcategory

(i.e., pay-for-performance, annual, and long-term compensation practice disclosures).The average firm in 2001 has a greater proportion of independent directorsserving on its board and compensation committee, has a smaller proportion of out-side directors appointed after the current CEO took office, and is more likely tohave a fully independent nominating committee than that in 1992 (all significant atthe 0.01 level) The outside directors serving on boards and compensation commit-tees in 2001 have a greater number of other directorship positions than those serving

compensa-tion committees (with more than 50 percent of outside members serving on three

or more other boards) are not significantly different between the two time periods.The frequency of board and compensation committee meetings remains thesame between the two periods A typical board and a typical compensation com-mittee held about seven and four meetings, respectively Despite increased boardand compensation committee independence, the average number of directors serving

however, the average number of compensation committee members remains stant A typical compensation committee has about four members

con-Corporate Governance and Executive Compensation Disclosure 1163TABLE 3

1164 Contemporary Accounting Research