CEO Compensation and Firm Performance: A Research Proposal

Introduction

CEO compensation has long attracted attention from a wide range of groups in both general society and corporate circles. Investors who fund various organizations typically want to ensure that a hired CEO's remuneration and its associated incentives are having a direct and positive influence on the general performance of the firm in question.

When a CEO receives generous compensation yet delivers poor performance, the response from the media, investors, and the general public is often fierce. Global reactions to the misappropriation of funds tied to certain leadership and governance styles have therefore prompted considerable academic debate regarding the actual determinants of CEO pay and what effect it should have on corporate performance. Most of the literature, however, points out clearly that the most important factor influencing the size of a CEO's compensation is the size of the corporation itself. The compensation–performance relationship is therefore a fundamental issue requiring study in order to ensure the profitability and continuity of global corporations. Previous studies do suggest, however, that only a weak relationship exists between CEO pay and the level of performance of any given corporation.

It has been observed in recent years that a high level of corporate governance is necessary to ensure that CEO compensation truly matches corporate performance. This concern has been fuelled largely by cases of leadership failure coupled with very high levels of CEO compensation. The best examples are the various scandals that affected corporations in both the United States and the United Kingdom. Specific examples of corporate failure marked by very high CEO pay include the Maxwell Corporation, Enron, and Tyco (Girma et al., 2007, p. 65). The failures of these major corporations were all linked to elevated executive compensation (Matsumura and Shin, 2005).

The purpose of this research is to determine the extent to which CEO remuneration at a large corporation — such as Coca-Cola — influences the general performance of the organization. The study examines CEO salary, bonus, and long-term compensation in relation to various organizational, financial, and CEO-specific factors, including corporate reputation, employee growth, company size, return on equity, CEO tenure, and CEO stock ownership.

The research technique to be adopted is based on a longitudinal panel analysis of a sample of 100 firms covering the period from 2000 to 2004. This technique is, however, rarely utilized in the context of socially centered research, as noted by Ryman and Bell (2003). The primary method of data collection for this study will be archival research. It is worth noting that there is a considerable distance between the researcher and the actual research reality; the researcher will apply real CEO salary statistics against actual company performance datasets, both of which are removed from the researcher's direct observation. The archival technique is favored due to its permanency, and it also has the merit of allowing future researchers to apply the same datasets in the analysis of other phenomena or to build on existing research. An additional source of data is the financial databases available through hand collection from various annual remuneration reports.

The datasets for the two major parameters are then analyzed using first-differenced and subsequently system Generalized Method of Moments (GMM) regression techniques, as postulated by Arellano and Bond (1991). The problem of unobserved corporation-specific effects is addressed by using corporation-differenced data, while the use of lagged performance variables and careful instrument selection addresses the issues of persistence and endogeneity.

Data Collection and Methodology

Several equations are then formulated to correspond with the various hypotheses drawn, and dummy variables are introduced at the necessary points in these equations to account for even minor determinants and errors that may be present in the analysis.

CEO compensation in this research proposal is treated as a means of evaluating a CEO's salary. Company owners strive to compensate their CEOs fairly in order for them to continue running the company without financial strain while still acting in the company's best interest. This is owed to the fact that management is "not always constrained to act with the owner's welfare in mind" (Lewellen and Huntsman, 1970). Because the owner is not directly involved in running the company, the manager — who acts on the owner's behalf — should be well compensated. CEO compensation becomes a complex issue because both owners and CEOs work hard to achieve the best outcomes for their company. Owners should therefore look for ways — such as providing a base salary supplemented by performance-based bonuses or commissions through incentive-based contracts — to align these interests. As Murphy (1986) notes, the level of managerial effort will depend on an executive's incentive contract.

CEO compensation has grown gradually over the last 50 years and now spans disciplines such as economics, law, accounting, and organizational structure (Murphy, 1999). Financial economists such as Jensen and Murphy (1990) have studied the relationship between CEO compensation and company performance extensively. Financial economists have also investigated the effects of variables such as director decisions, capital base, dividend policies, mergers, and diversification on CEO compensation (Murphy, 1999, p. 2). CEO compensation is thought to be the most effective way of addressing the classic corporate problem of separating ownership from control (Jensen and Meckling, 1976).

Hall and Liebman (1997) revisited Jensen and Murphy's work and found a stronger relationship between CEO pay and company performance than Jensen and Murphy (1990) had identified. This difference arose because Jensen and Murphy (1990) used data from 1969 to 1983, which predates the explosion in the use of stock option grants that began in the 1980s. Hall and Liebman used expanded data covering the 1980s and into the 1990s. Between 1980 and 1994, the median value of stock option grants ranged from zero to $325,000. In 1980, 30 percent of CEOs received options; by 1994, that figure had risen to 70 percent (Hall and Liebman, 1997). This study therefore supports the view that stock options are a very important component of incentive-based contracts.

This proposal evaluates the correlation between CEO compensation and its effects on company performance, guided by two underlying assumptions. The first holds that CEO compensation is determined by past performance, while the second holds that incentives are not the primary driver and that executive productivity instead depends on "managerial ability which is initially unknown, but revealed over time" (Murphy, 1989).

Literature Review

CEO compensation affects company performance and should reflect a commitment on the part of both owners and managers. When a good compensation package is in place, a manager is incentivized to apply his or her full capabilities, enabling the company to perform better. Simultaneously, the owner should strive to maintain a sound capital base and sound dividend policies, among other measures, to enable the smooth running of the company.

Earlier researchers have attempted to evaluate this phenomenon using various theoretical frameworks, including neoclassical theory, managerial theory, relative performance evaluation theory, principal-agent theory, human capital theory, social comparison theory, managerial power theory, information processing theory, and tournament theory.

A large number of researchers who evaluate CEO pay and performance use the principal-agent framework as the basis of their research. Agency theory provides an effective way of aligning CEO interests more closely with those of owners by enabling a clear reward system that acknowledges and compensates performance accordingly.

The objective of this study is to examine this relationship by controlling for persistence and simultaneity. When conducting empirical work with panel data, two factors must be taken into consideration. First, it is likely that CEO compensation and firm performance are correlated with the current realization of unobserved firm-specific effects (Marschak and Andrews, 1944). Additional factors — such as the effectiveness of research and development efforts — differ from firm to firm and could influence either CEO compensation or firm performance, or both. Second, CEO compensation tends to be highly persistent over time, and remuneration and firm performance are typically jointly determined.

Hypothesis 1: As a CEO gains more years of experience, his pay will increase.

It is also assumed that CEOs hired from outside the firm, based on the company's prior performance, will tend to be more highly compensated than those sourced internally. This will depend on the company's performance in previous years (Joskow and Rose). This assumption is also consistent with the view that a firm's market value greatly depends on its performance in the following year.

Hypothesis 2: A CEO who is externally hired will be compensated more than a CEO who was internally promoted.

There is also a presumption that a CEO may have been promoted from within or hired externally from a firm at which he or she accumulated many years of experience. This study will assign a value of 1 if the CEO was hired internally. According to Rose and Shepard (1994), a manager hired externally earns a salary approximately 15 percent higher than one who was promoted internally.