پاداشهای مدیریتی و رفتار برای سود، سازمان های دولتی و غیر انتفاعی: شواهد از صنعت بیمارستان

Studies of mixed industries frequently focus on differential behavior between for-profit and either nonprofit or governmental producers. Substantially less is known about differences among governmental, religious nonprofit, and secular nonprofit organizations. We examine the compensation of hospital CEOs to assess the extent to which these three organizational forms pursue similar objectives. Compensation levels, the use of salaries versus bonuses as proxies for weak versus strong incentives, and the criteria organizations use to determine bonuses are analyzed. We conclude that the CEO incentive contracts at religious nonprofit, secular nonprofit, and governmental hospitals imply substantive differences in the behavior of these organizations.

The role of different forms of institutions has been fundamental to economics since at least the days of Adam Smith. The choice between relying on private markets and on government continues to concern both economics and public policy debates. The debate has increasingly been expanded to encompass the option of relying on a third form of institution, private nonprofit organizations. In recent decades, as the nonprofit sector’s share of gross domestic product has increased while the public sector share declined (US Census Bureau, 1999), it has become increasingly important to understand both the forces that affect the choice among organizational forms and the efficiency consequences of these choices. Research on mixed industries—examples of which include hospitals, nursing homes, colleges, day care centers, and museums—has focused heavily on the distinction between for-profit firms and organizations that are subject to a ‘nondistribution constraint,’ which prohibits organizations from distributing profits to their managers, trustees, or owners. It is important to note that such a constraint does not prohibit organizations from actually earning positive profits. Two common ownership types—religious private nonprofit and secular private nonprofit—are both legally bound by an explicitly stated nondistribution constraint. Governmental organizations, while not legally subject to the same nondistribution constraint as are private nonprofits, are subject to political and legal constraints that also effectively constrain governmental organizations from distributing their profits. Thus, the religious nonprofit, secular nonprofit, and governmental institutional forms are all constrained with respect to the distribution of their profits, and, as a consequence, studies that rely on nondistribution constraints to explain differences in organization behavior cannot account for any systematic differences that may exist among these forms. While there are many industries in which one or more of the three types of these distribution-constrained organizations are present, typically along with for-profit firms, relatively little is known about whether and how these various forms of organization differ in their behavior and outputs and, hence, the degree to which they are substitutes. Such information would be useful in assessing how the presence of multiple organizational forms influences consumer welfare. Additionally, better knowledge of differences (or similarities) in behavior across religious nonprofit, secular nonprofit, and governmental organizational forms would contribute to the development of a more general model of the mixed industry that is capable of explaining the presence and distribution of multiple organizational forms.2 By examining the structure of executive compensation in one large mixed industry (hospitals), we seek to answer the question of whether different types of distribution-constrained organizations—religious nonprofit, secular nonprofit, and governmental—behave similarly. We postulate that the incentives provided by any organization through its executives’ compensation are likely to reflect the organization’s objectives—which we define as those established by the board of directors (at nonprofits) or higher-level governmental authority—whenever organizations are similarly constrained with respect to the set of contracts that they can write. Incentives, in turn, influence managerial behavior, which can ultimately be expected to influence the nature of outputs. Thus, if different organizational forms pursue identical objectives and are similarly constrained, one would expect them to offer their executives identical incentives. Equivalently, if we observe systematic differences in compensation contracts, this would suggest that substantive differences exist across organizational forms that likely affect organization behavior and the nature of outputs. Our analysis reveals both strong similarities across organizational forms in certain dimensions of compensation and sizable differences in others. Understanding executives’ incentive structures across institutional forms is of interest in its own right, as well as for other reasons. Such knowledge would contribute to a better understanding of managerial labor markets, the relationships between incentives and organization behavior in input and output markets, the use of strong relative to weak incentives to reward workers, the modeling of organizational behavior in mixed industries, the wisdom of applying anti-trust law to nonprofit organizations, the efficiency of public subsidies to both religious and secular nonprofits, the effects of privatizing governmental activities to nonprofit relative to for-profit producers, and the effects of conversions of organizations between nonprofit and for-profit status. Recent research has examined managerial reward structures in the hospital industry, comparing for-profit and nonprofit organizations (Arnould et al., 2000, Brickley and Van Horn, 2000 and Roomkin and Weisbrod, 1999). In its empirical work, this study is closest in spirit to the work of Roomkin and Weisbrod, and yet it asks an entirely separate question. Unlike Roomkin and Weisbrod, we focus on the extent to which the three different forms of distribution-constrained organizations differ in the incentives that they provide to their executives as a means of addressing the larger question of whether they are likely to pursue different goals. By contrast, the Roomkin and Weisbrod paper focuses only on distinctions between for-profit and nonprofit organizations. If religious nonprofit, secular nonprofit, and governmental organizations all behave similarly, nothing is lost in such an approach. If these three types of distribution-constrained organizations do differ in their behavior, however—and the results of our study suggest that they do—then any approach that emphasizes solely the distinction between for-profits and nonprofits will fail to capture the true extent of the heterogeneity in the mixed industry. Our analysis is distinct from the work of Roomkin and Weisbrod in two other important ways. First, we focus in greater depth on the performance measures that different hospitals report using in the determination of executive bonuses. Due to limitations of the data on for-profits, the prior study was not able to do this effectively. Second, our analysis of differences in compensation among the three forms of distribution-constrained organizations—relative to differences between each of those types of organizations and for-profit firms—provides some evidence on the importance of nondistribution constraints relative to other influences in determining differences in managerial reward structures across organizational forms. The next section outlines the theoretical framework that motivates our empirical analysis. Section 3 discusses the data and methodology, setting forth the hypotheses to be tested and describing the empirical models. Results and discussion follow in the 4 and 5 sections, respectively.

We have presented evidence consistent with the hypothesis that organizational behavior—as defined by the interaction of institutional objectives and constraints, and as reflected by a number of financial compensation variables—differs systematically not only between for-profits and other types of organization, but also among three types of organization that are subject to nondistribution constraints: religious nonprofits, secular nonprofits, and governmental organizations. While a variety of models may explain the differences that we find, a model in which organizational objectives vary systematically across institutional form is particularly attractive when it comes to explaining all of our findings, taken together. Our results have implications for several different areas of economics research. Most immediately, our results highlight the necessity of incorporating measures of organizational form when modeling executive compensation or studying compensation contracts. Whether due to differences in objective functions or institutional constraints, these differences are meaningful and influence the nature of the contract. For similar reasons, our results have implications for the literature on comparative institutional behavior. While researchers in this field have traditionally relied on the presence or absence of a nondistribution constraint to generate differences in behavior across organizational forms, our findings suggest that a more general model of mixed industries is needed—one that distinguishes between religious and secular nonprofits, and between each of these and governmental suppliers. Since the differences that we find among religious nonprofit, secular nonprofit, and governmental organizations cannot be attributed to the presence or absence of nondistribution constraints—as all three types are subject to them—other systematic differences must exist, in objective functions or constraints other than the nondistribution constraint. Further research into the sources of these differences would be useful. Finally, our results have implications for government policy. If different institutional forms pursue different objectives, then the process of privatizing government operations might include consideration of nonprofit sector alternatives to private for-profit organizations. With respect to the application of antitrust law, our results suggest that treating all nonprofits or, more broadly, all private organizations identically under antitrust law may not be optimal, as differential objectives will in many circumstances imply differential behavior. For example, a merger between two religious nonprofits may have very different behavioral implications than a merger between a for-profit and a secular nonprofit. We have reason to believe that the work presented here and in a companion paper by Roomkin and Weisbrod (1999) have broad applicability outside the hospital industry. Our informal inquiries suggest, for example, that in the post-secondary education industry, managers of for-profit enterprises are more likely to be compensated with performance-based bonuses than are their counterparts in public or private nonprofit universities.28 Similarly, in the fitness-center industry the use of performance-based bonuses is common at for-profit centers but rare at nonprofit YMCAs. Systematic studies of the use of performance-based bonuses and of the criteria on which they are based would be useful in these and other mixed industries such as nursing homes, museums and galleries, theater and the arts, and day care.