The Importance of Context
Our reviews of performance appraisal and merit plan research and practice indicate that plan success or failure are substantially influenced by the context within which they are embedded. Research on performance appraisal now encompasses a broader set of organizational factors, along with the individual and task factors that it has traditionally studied (Murphy and Cleveland, 1991). Research on pay now stresses the importance of viewing pay and pay for performance plans in the context of an organization's personnel system, its structure and managerial styles, and its strategic goals (Balkin and Gomez-Mejia, 1987a, 1990; Carroll, 1987). Managers of performance appraisal, merit pay, and variable pay plans stress that these plans must fit or be consistent with the organization's personnel practices, culture, and strategic mission or goals if they are to work as the organization intends. Both researchers and managers acknowledge the influence of environmental conditions on organization decisions about adopting and implementing these plans.
The rationale underlying this concern with context is a simple one. In Chapter 5, we noted that theory and research on individual motivation show that individuals are motivated by pay to the extent that they value pay, understand performance goals, and believe that pay is contingent on that performance. Variations in an organization's context attributable to its strategy, structure, job design, culture, management systems, personnel systems, and work force culture and characteristics can strengthen or attenuate the links between pay and individual motivation. Design and implementation of performance appraisal and merit plans that fit or are consistent with context factors tend to strengthen these links.
Unfortunately, on the basis of the existing research evidence, it is difficult to be specific about what fit really is. However, it may be useful to briefly summarize several points about the concept of fit that are relevant to the federal government's interest in performance appraisal and pay for performance and that appear to receive support in the existing quantitative, clinical, and/or practice literatures.
This chapter is organized around the three categories of related contextual factors identified in reviews of research and practice on performance appraisal and pay for performance: (1) the nature of the organization's work—primarily its technology and job designs; (2) the broader features of organizational context such as size, management systems, personnel systems, and work force culture and characteristics; and (3) features of the organization's environment, such as its economic growth, the presence of unions and associations, and the pressures exerted by multiple public regulators and interest groups. We do not offer any comprehensive review of the very diverse research literatures that might be brought to bear on the influence of context on performance appraisal and merit pay. Instead, our discussion focuses on the factors that may be particularly relevant to the federal government (Perry and Porter, 1982) and provides some general research findings suggesting how these factors may influence performance appraisal and merit pay. We end with a description of the federal bureaucratic context and its implications for performance appraisal and merit pay.
TECHNOLOGICAL FIT: THE NATURE OF THE ORGANIZATION'S WORK
An organization's technologies and the pace of change characteristic of those technologies will influence the way an organization defines its jobs and work methods (Scott, 1981). The performance evaluation measures most suitable to an organization will depend in part on the effects of technology on job complexity, interdependence, and stability; on job goal specificity; and on the ease of measuring or supervising job performance (Dornbusch and Scott, 1975; Murphy and Cleveland, 1991). For example, stable jobs characterized by low complexity, in which performance goals can be easily specified for each employee and in which employee performance is easily observed, are compatible with more quantitative, carefully scaled individual evaluations based on specific work output or behavior.
However, using highly specific, individual performance appraisals and incentives with jobs that are complex and involve multiple and ambiguous goals can result in employees ignoring important aspects of their jobs or distorting job information to make their performance look good. Blau (1955) provided a classic description of how explicit performance measures in a public employment agency induced interviewers to behave in ways that were consistent with their
performance goals but destructively competitive and nonproductive for the organization. Pfeffer and Baron (1988) suggested that one factor promoting the increased reliance on "contingent" labor (such as part-time, temporary, leased, or subcontracted labor) was managerial performance appraisals that assessed outputs per capita, in which the denominator of that ratio was based only on full-time equivalents. In both these cases the use of quantitative performance measures created incentives for employees to behave in ways that were rational but organizationally detrimental.
Many scholars have pointed out that managers and professionals in public-sector organizations face conflicting, diffuse goals that make it difficult to develop meaningful performance criteria (e.g., Buchanan, 1975). It has also been suggested that it may be easier to establish concrete and appropriately challenging goals in jobs in which the bottom line is measurable (staff sales, units produced) than in more typically bureaucratic jobs, such as managing strategic planning or policy development. Yet the examples of the possible unintended consequences of explicit performance measures cited above are equally pertinent in the private and public sectors. Indeed, one could argue that the potential problems are greater for private-sector employers because they have moved away from developing standards and elements based on job analysis for their managerial and professional employees, and instead rely on management-by-objective kinds of appraisal. The premium that such appraisal systems place on specific, narrowly defined goals and the likelihood that negotiated goals will tend to be lenient would seem also to increase the likelihood that important aspects of the job will be ignored or the appraisal otherwise distorted.
BROADER ORGANIZATIONAL FACTORS
While the relationships among an organization's technology, job designs, performance evaluation plans, and pay for performance plans have been the contextual factors most directly examined, many others are thought to influence an organization's success in adopting and implementing these plans. For the purposes of this chapter, we classify them broadly as: (1) factors related to organization strategy, goal clarity, and cohesiveness; (2) factors related to organization size, structure, and management systems—including personnel systems; and (3) factors related to work force climate and employee-labor-management relations.
Organization Strategy, Goal Clarity, and Cohesiveness
The business policy and strategy literature suggests that organizations vary in their perceptions of their environments and in their definitions of the strategic goals meant to help them compete in those environments. Many different strategic orientations have been identified in this literature (see, for example,
Schendel and Hofer, 1979; Lamb, 1984; Porter, 1985; and Harrigan, 1988). Two primary strategic postures have been applied in studies of the association between strategy and performance evaluation and pay systems in private-sector firms: a dynamic, growth-oriented model and a steady-state model. Most of these studies are cast at the headquarters level and examine executive compensation systems. For our purposes they are interesting because they suggest that different strategic goal orientations are associated with different emphases on performance evaluation and pay for performance plans. We emphasize the word suggest here, since these studies are, at best, descriptive and cannot be viewed as generalizable.
A 1985 study by Kerr, for example, used a multiple case study methodology to classify firms according to their corporate strategies and to distinguish patterns of performance evaluation and pay plans within strategic classes. The 20 firms that Kerr classified were pursuing either an "evolutionary/dynamic growth" or an "steady-state/maintenance" strategic approach to their market environments. Evolutionary strategies were defined as emphasizing increasing market growth through active pursuit of new markets via acquisitions, joint ventures or mergers, and innovative products or services. Steady-state strategies were defined as emphasizing holding onto current market positions through internal development of technology, improvements in products or services, increasing work force productivity, internal coordination, and economies of scale.
Kerr found that executives in firms successfully pursuing evolutionary strategies were more likely to be evaluated strictly on quantitative, organization-level measures of strategic performance tied to bonus plans that offered high returns (40 percent or more of base salary). Executives in firms successfully pursuing steady-state strategies were more likely to be evaluated against a mix of subjective and quantitative performance measures cast at both the individual and the organization levels. Their bonuses paid out at a lower rate (20 percent of base salary). (Kerr's results were consistent with earlier work on corporate strategy and executive pay such as Berg , and Pitts .)
In a 1984 study of electronics manufacturing firms, Balkin and Gomez-Mejia found that firms pursuing strategic innovation and growth goals through new research and development were more likely to offer their engineers and scientists a higher proportion of their pay in the form of incentives (bonuses, profit-sharing, and stock) than firms with less investment in innovation and new development.
In both these studies, organizations pursuing riskier (i.e., evolutionary, innovative) strategies were evaluating their managers or professionals on quantitative, specific, organization-level performance goals. They offered them pay incentives that would be paid out only if the organization was successful, but would then pay out very well. We can speculate that by tying performance evaluations strictly to strategic goal attainment and by offering high payouts, organizations are sending a signal to current and prospective employees about
the importance of more entrepreneurial, innovative behavior to the organization. Organizations pursuing maintenance-oriented strategies evaluated their managers on a mix of more qualitative individual behaviors and quantitative organization-level goals; bonus payouts were typically a lower proportion of base salaries than in firms pursuing riskier strategies. We can likewise speculate here that by using this combination of performance evaluation and bonuses, these organizations are sending a signal to employees about the importance of professional management skills in meeting specific organization performance goals, and of getting along as individuals within the performance norms shaped by the work force culture. These remain, at this point, speculations, although many researchers have voiced them (Salter, 1973; Galbraith, 1977; Lawler, 1981; Balkin and Gomez-Mejia, 1987a; Carroll, 1987).
There are also theoretical perspectives suggesting that organizations vary in their ability to define strategic goals so that they are likely to be understood and seen as legitimate by their employees, their other stakeholders, and the public. For example, organizations in highly institutionalized sectors or those relying significantly on public trust may be more likely to adopt very formal, precise performance evaluations in response to external pressures or regulations (Meyer and Rowan, 1977). In such cases, organizations may use performance appraisals and pay for performance plans (like merit plans) to make their management decisions appear more legitimate to both employees and other organization stakeholders. (See, for example, Tolbert and Zucker's 1983 study of the adoption of civil service reforms.)
The reason for this, some scholars would argue, is that when organization goals are most difficult to define and job performance is thus difficult to evaluate against some agreed-upon criteria, organizations feel compelled to adopt more formal, precise evaluations in order to assure their constituents that they are operating rationally and efficiently. For instance, a government agency with a fairly straightforward mission and relatively easily defined performance criteria, such as the Internal Revenue Service, might exhibit less formal and precise performance evaluation than one with a less clearly defined mission and performance criteria, such as the Environmental Protection Agency. It is precisely the difficulty in identifying effective job performance in the latter case that induces decision makers to emphasize formal evaluation. Moreover, it is symbolically important for employees and other organization stakeholders to perceive that meaningful evaluation criteria are used and that differential outcomes are not capricious (Salancik, 1977; March and March, 1978). As March (1981:232) writes, "decision making is, in part, a performance designed to reassure decision makers and others that things are being done appropriately." In reality, desired performance may be difficult or impossible to specify or identify a priori, especially in higher-skilled and information-intensive lines of work. Ironically, this may make it all the more likely for an organization to try
and do so in order to leave the impression among members that things are not done arbitrarily.
Organizational Structure, Management Systems, and Size
Structure and Management Systems
The perspectives on organizational choice of strategic goals and organizational ability to define strategic goals assume that such choices are influenced by internal structure, management systems, and personnel systems. By virtue of their history, growth patterns, strategic goals, and the environmental challenges they face, organizations make decisions about their physical and geographic structures, their job designs and hierarchies, their management systems, and so forth. While the conceptual writing and the research undertaken to examine the relationships between organization strategy, structure, management, and environment are extensive, we focus here only on selected work used in normative proposals about the relationship between organization structure, management systems, and performance appraisal and pay systems. This includes work by Burns and Stalker (1961), Miles and Snow (1978, 1983), Balkin and Gomez-Mejia (1987a, 1990), and Carroll (1987).
Burns and Stalker's classic study (1961) proposed two ideal types: (1) the organic organization—young, innovative, aggressively pursuing growth in highly uncertain environments—and (2) the mechanistic organization—less risk-oriented, more stable, more focused on internal efficiencies. They described the organic organization as one in which jobs or tasks are undifferentiated; performing them requires employees with general problem-solving or professional skills. Job definition is flexible and changes with the organization's goals and technology. Decision-making responsibility is decentralized, and employee input is not only valued but expected. Management hierarchies are flat; evaluations are tied to external professional standards and broadly defined organization goals. Policy and work rule standardization and formalization are low; communication is open, offering information rather than supervisory instruction.
Burns and Stalker's mechanistic organization, by contrast, is described as one in which jobs or tasks are highly defined, requiring employees with specialized, functional skills and specific organizational experience. Job definition is stable but difficult to change, as change would require new skills. Decision making tends to be centralized, with each supervisor having a distinct span of control and set of responsibilities, and employee input is low. In the mechanistic genre, hierarchies are steep, and control systems tend to be behavioral and tied to employee loyalty and diligence in carrying out assignments. Policy and work rule standardization and formalization are high; communication tends to be restricted, vertical (and one-way: top to bottom), and focused on instruction rather than an information exchange. Burns and Stalker proposed that the mechanistic
type provided the context an organization needed to capitalize on the managerial and technical efficiencies possible in more stable, certain environments.
Miles and Snow (1978, 1983) proposed a similar pattern in the context variables of an organization's structure and management system in their prospector and defender types, adding more detail on the personnel management systems appropriate to each. (Miles and Snow's work echoes that of Doeringer and Piore, 1971, on variations in a firm's internal labor market development.) In their large-scale, systematic case studies and writings, Miles and Snow distinguished two distinct strategic goal orientations—one that emphasizes innovation and market growth (prospector firms) and one that emphasizes holding current market position by pursuing cost efficiencies, quality, and productivity improvements (defender firms). The personnel management systems of the more entrepreneurial organizations emphasize general skills, hiring at all levels of the organization, higher investments in recruiting than in training and development, and performance measures tied to innovation and competitive market outcomes; retention is not considered a primary personnel management goal. Defender firms are described as emphasizing job-specific skills; promotion from within; retention; higher investments in selection, training, and development; and performance measures tied to cost efficiencies, social norms, and historical standards. In short, in the Miles and Snow typology, prospectors spend personnel dollars to buy a work force; defenders, to make or build one.
Drawing on this and other work, Carroll (1987) and Balkin and Gomez-Mejia (1987a, 1990) have proposed that organizations pursuing growth-oriented, innovative strategies, which have organic structures and management systems and personnel practices that emphasize buying an entrepreneurial work force will be best served by performance evaluations that emphasize competitive, organization-level performance and by pay systems that emphasize group incentives and bonuses. Organizations pursuing cost efficiencies and maintenance strategies, with mechanistic structures and management systems and personnel practices that emphasize internal skill development and the importance of work force norms, would be better served by more traditional performance appraisal and merit plans or other policies that recognize an employee's long-term contributions to the organization.
There is considerable anecdotal literature that supports these prescribed patterns of association or fit between performance evaluation and pay for performance on one hand and organizational strategy, structure, management, and personnel systems on the other (Cook, 1981; Salschieder, 1981; Ellig, 1982; Smith, 1982). However, there is little research specifying the exact dimensions of fit among organizational systems, nor are there generally accepted theories concerning how such fit contributes to organizational performance.
Case study research on high-performance organizations and on organizational innovation also suggests that effective performance appraisal and pay allocation practices must be closely aligned with an organization's culture,
structure (e.g., number of layers of management, job structure), management style (e.g., centralization versus decentralization), and work force (Beer et al., 1990). A number of commentators have argued that the success of so-called high-commitment organizations illustrates the power of well-integrated personnel systems to increase motivation and organizational effectiveness (Walton, 1979, 1980). In some cases, new manufacturing facilities have been built from the ground up, with all of the elements of the organization planned and designed to be congruent from the outset to increase motivation, teamwork, and effectiveness. Although there have been instances of failure and regression over time, the record of these high-commitment work systems suggests that motivation, attachment, quality, and productivity are positively affected when the human resources policies and practices of the organization are highly congruent.
Motivation in high-commitment organizations seems to be governed not by one dimension such as pay, or a relationship with the boss, or the nature of the work, but by a multiplicity of organizational practices such as organization design, pay practices, management style, information and feedback, employee involvement, and the types of employees recruited and socialized into the organization. The internal consistency of these practices is thought to reinforce employee perceptions of the organization's fairness and concern for equity (Greenberg, 1986b).
The work discussed so far does not capture the size or scale, the scope of operations, the complexity of joint working arrangements, and the diversity of work forces typical of many large, modern organizations. In particular, in large organizations with diverse operating units and work forces, there is always the question of where in the organization's structure decentralization of performance appraisal and pay systems is most likely to facilitate the achievement of strategic objectives. We know, for example, that even within a discrete business unit, personnel systems, including performance evaluation and pay for performance plans, may vary by employee group (Hewitt Associates, 1989).
The business policy studies of the 1960s and 1970s illustrated two basic approaches to corporate structuring and control of large, diverse businesses: one in which corporate management took a hands-off or holding company approach to managing business divisions; the other in which the corporate management tried to set basic policy guidelines and used both performance evaluation and pay systems to tie division managers to corporate as well as divisional goals (Chandler, 1962; Berg, 1965; Pitts, 1976). Recent case studies of globalizing or transnational firms have noted that, while some firms try to manage and coordinate diverse businesses and work forces by developing more elaborate bureaucratic and centralized structures and controls, most have moved to global statements of corporate values that are intended to guide, but not dictate, business
unit actions at a decentralized level (Doz and Pralahad, 1981; Galbraith and Kazanjian, 1986; Bartlett and Ghoshal, 1988; Evans, 1989). The work of Vancil and Buddrus (1979) also supports decentralized control of performance appraisal and pay for performance plans based on the nature of the work being performed (e.g., team-based, task interdependence, task concreteness, stability of technology, etc.).
Work Force Climate and Employee-Management-Labor Relations
The research we reviewed earlier on pay for performance plans indicated that such factors as employees' confidence and trust in management, their opportunities to participate in setting performance goals, and the availability of channels for appeals of performance appraisal ratings and merit allocations can influence both their motivation to perform and their assessments of the fairness of performance appraisals and pay for performance plans. There is considerable case study and anecdotal literature documenting problems that can occur when individual and group incentive plans, for example, are implemented in an organization unit in which employee-management-labor relations have been traditionally hostile (Whyte, 1955; Lawler, 1973; Schuster, 1984b; Mitchell et al., 1990). Problems include the development of work force norms restricting performance, and gaming or providing false performance information in order to get plan payoffs without changing actual performance. This literature provides some warning to organizations attempting to implement new pay for performance plans in hostile work climates that they must understand the risks involved. Lawler (1981) suggests that, in such situations, organizations should improve the work climate before implementing pay changes.
While there are a host of environmental factors that may influence organizational arrangements, we focus here on three sets of institutional forces of particular interest to performance evaluation and pay for performance systems: economic pressures and growth; the presence of unions and professional associations; and the pressure of laws and regulations governing personnel systems.
Economic Pressures and Growth
Our review of research on pay for performance plans suggested that the economic environment the organization faces and its projected employment and financial growth can influence employees' acceptance of pay for performance plans. Lawler's (1973) review of case studies on individual incentive plans suggested that employees were less likely to accept the plans (and thus be
motivated by them) when they believed the plans might eventually result in reducing the organization's demand for people in their jobs. Likewise, the case studies of gainsharing plans suggest that employees are more likely to accept these plans when there is some form of job guarantee attached or the organization's future economic success and growth look promising (Schuster, 1984b). Conventional merit plans also offer more incentive potential for employees when the organization is growing. As we noted in our review of practice, the opportunity to promote high-performing employees who are also high in their salary range makes it more likely that merit plans will, over time, provide higher-performing employees with higher pay levels. Some organizations, faced with limited employment growth, are now considering avoiding restrictions on merit allocations for employees already high in their salary ranges by offering some portion of merit increases as lump sums (i.e., not added into base salaries). In short, some assurances that pay for performance plan payouts are feasible and that job security is not jeopardized by the plan appear to be important to employee acceptance and motivation under pay for performance plans. Both may be influenced by the organization's economic and growth prospects.
Unions and Professional Associations
Unions in the United States have resisted performance appraisal systems and pay for performance arrangements because they view them as cloaking managerial exploitation (hence worker distrust of performance appraisal ratings) and reducing worker solidarity by substituting wage competition (merit or incentive plans) for a community of interest among laborers (Stone, 1974). Unions aim to raise the wage levels of the whole collective, rather than the wages of individual members. The practical effect of union resistance and aims is well documented. Most surveys of incentive systems in use (including merit plans) indicate that unionized employees are far less likely than nonunionized employees to be covered by such pay arrangements (Bureau of National Affairs, 1981, 1984). Freeman and Medoff (1984), in a comprehensive study of unions, noted that unionization tended to reduce wage differentials among union members, while raising their overall wage level relative to that of first line management (also see Kalleberg and Lincoln, 1988). To the extent that pay for performance plans might increase any disparities between the rewards of managers and those of the employees they supervise, the relatively high degree of unionization in the federal government might make employees more resistant to pay for performance plans (Advisory Committee on Federal Pay, 1990).
The 1980s saw some slight reduction in union resistance to alternative pay arrangements in the private sector. The particular pay arrangements conceded, however, typically were profit-sharing and lump sum plans, which do not differentiate among individual employees (Mitchell, 1985).
Industrial unions typically emphasize the power of numbers in reaching wage bargains; craft unions and professional associations typically emphasize skills. Professional associations might be expected to resist centralized or standardized performance measurement systems and related pay for performance plans on the grounds that only members of the profession can appropriately judge performance. To the extent that such systems might reduce the power or wages of their groups, they would be resisted. The exception here might be personnel professionals who have a particular stake in the institution of these systems (Baron et al., 1986).
Overall, then, the extent of unionization and professionalization in an organization's labor markets will tend to reduce support for the adoption of performance evaluation and pay for performance plans. In the federal government, there are four associations that represent managers and professionals and at least four employee unions. Survey responses of government managers suggest that, although there is agreement in concept with merit pay, there is dissatisfaction with its administration to date.
Laws and Regulations Governing Personnel
External laws and regulations impose additional goals for organizations (for example, equal employment opportunity) and often prescribe internal structures and controls for achieving them. These imposed goals and internal systems may be incompatible with the organization's other goals and internal systems. In the federal government, details of many personnel programs are dictated by law and audited by the General Accounting Office and the Office of Personnel Management. The extent to which external laws and regulations have pressured organizations to adopt internal structures and programs that are at cross-purposes with mainline organizational goals is debatable. The laws and regulations prohibiting employment discrimination, for example, have brought dramatic changes in the way companies conduct and document their personnel management procedures. Although these laws were designed to implement important constitutional and policy goals, they have also had an impact—many would argue a positive impact—on such things as the quality of employment tests and the resources that companies devote to human resource management. Other legal protections may not be compatible with effective performance appraisal systems. For example, the legal protections available to federal employees have put significant pressure on the design and administration of the performance appraisal system.
In the case in which employees have explicit procedures for grievance and due process, there is generally more emphasis placed on the development of clear and concrete performance standards and dimensions that have at least the appearance of validity to both the supervisor and the employee. There is always the danger that, in an environment with heavy legalistic protections
for employees, the performance appraisal system will be asked to provide an unrealistic level of measurement rigor. The results of our analysis of the technology of performance appraisal (see Chapter 4) suggest that moderate levels of validity can be achieved under highly controlled conditions, but that we have probably reached the point of diminishing returns in the search for measurement precision. Moreover, although research analyzing court decisions on performance appraisal systems indicates that appraisal focusing on specific behaviors or results are more likely to find judicial approval, recent measurement and cognitive research fails to support a preference for these approaches over the appraisal of broad traits.
There are some specific concerns with regard to the protections afforded federal employees. While due process requirements giving employees the right to appeal their evaluations are common in the public sector and are related to the concept of fairness in the public and private sectors, some of the bases for appealing performance appraisals under the Civil Service Reform Act may hamper effective managerial discretion. For example, the Civil Service Reform Act requires that performance standards be objective to the maximum extent feasible. Lack of objectivity of the performance standard can be the basis for appealing an unsatisfactory performance evaluation. We have discussed in Chapter 4 the inappropriateness of the terms objective and subjective, particularly with reference to managerial appraisal. We have also established that managerial performance does not lend itself to job-specific measurement. Providing employees a right to appeal their performance appraisals if the standards are not objective enough is likely to be a time-consuming exercise with no valid or beneficial outcome.
Using very precise individual performance measures and incentives systems for managerial and professional jobs can have potentially negative consequences for the organization; many organizations use more global appraisals combined with merit plans for such jobs.
Organizations differ in their ability to articulate strategic goals that provide direction throughout the management hierarchy in setting meaningful performance appraisal goals. Some organizations—especially public-sector organizations—find it difficult to articulate overall mission or strategic goals.
Public-sector organizations may use more formal, precise performance appraisals in an effort to make management decisions appear legitimate both to employees and to other constituents. While this may be useful in satisfying some constituents (for example, Congress) it may make employees skeptical of their performance appraisals and any pay system based on them, and it may reduce management incentives to administer the systems as the organization intends.
The literature related to fit suggests that there is a general match between certain patterns of organization strategy, structure, management on one hand and performance evaluation and pay systems on the other. For example, traditional performance appraisal and merit pay plans appear to be most suited to steady-state organizations, which emphasize skill development and work force norms. Group incentive systems appear better suited to innovative entrepreneurial organizations.
Many large firms with diverse goals and work forces have moved towards decentralized management strategies, with the home office providing policy and audit functions and the local units designing and implementing performance evaluation and pay systems.
This general discussion of contextual factors shaping performance appraisal and pay practices suggests not only that performance appraisal and pay practices must be aligned with the rest of an organization and its environment but also, presumably, that the reverse is true. In other words, to the extent that the federal government is seriously devoted to pay for performance, success in implementing it is unlikely unless the broader context supports it.