Organizational commitment, human resource practices, and organizational characteristics.
Numerous studies have shown that organizational commitment predicts important variables, including absenteeism, organizational citizenship, performance, and turnover (e.g., Mathieu and Zajac, 1990; Morrow, 1993; Wright and Bonett, 2002). Colbert and Kwon (2000) noted that organizational commitment has been related inversely to both intent to search for job alternatives and intent to leave one's job (also see Quarles, 1994). Also, it reduces absence frequency (Farrell and Stamm, 1988). In addition, organizational commitment has been related to more positive organizational outcomes, including job satisfaction (Williams and Hazer, 1986) and attendance motivation (Burton et al., 2002). These studies underscore organizational commitment's importance and thus the need to understand better its antecedents.
Many firms have moved rapidly from Tayloristic work systems, in which employees exercise limited discretion within narrow job descriptions, toward High Commitment Work Systems (HCWSs) that require considerable discretion, initiative, and judgment under reduced supervision (Osterman, 1994, 1995). Lincoln and Kalleberg (1985) described a similar earlier trend inspired by the apparent success of Japanese-style management. These trends have spurred increased interest in organizational commitment, and will likely enhance its importance in coming years (Osterman, 1995). Still, many issues concerning organizational commitment remain poorly understood (Price, 1990). For example, scholars contend that many organizations adopt Human Resources (HR) practices intended to maximize employee commitment (e.g., Guest, 1992), yet, with few exceptions (e.g., Ogilvie, 1986), little systematic research examines the influence of HR practices, including practices intended to promote commitment. Further, although prior work suggests that organizational structure influences commitment (Berger and Cummings, 1990), its effects have received only limited attention (see Lincoln and Kalleberg (1985) for a notable exception).
We examine effects of HR practices and organizational characteristics on commitment. Nearly all past studies rely on samples drawn from one (usually) or very few worksites (cf. Agarwala, 2003; Becker et al., 1996; Caldwell et al., 1990; Eaton, 2003). This severely restricts inferences on factors that tend to be constant within organizations, but vary across organizations, notably HR practices and organizational structures. The National Organizations Survey (NOS) used a hyper-sampling strategy to develop a nationally representative sample of employees and workplaces (Kalleberg et al., 1996). It provides matched employee-employer surveys, including employer responses on HR practices and organizational characteristics. In addition to providing meaningful variance on HR practices, this helps overcome common method variance and causal inference problems that plague single-source studies.
Our study provides two major contributions. First, we test hypothesized relationships that have received little attention in past studies. Specifically, we examine the influence of HR practices and organizational structure on organizational commitment. Second, our study uses a diverse national probability sample of U.S. workers and their employers, providing greater generality than prior studies.
In the following section, we discuss theoretical foundations of organizational commitment, and present hypotheses for its relationships with various HR practices and organizational characteristics. Next, data, measures, and methods are discussed. Then, results are presented. Finally, we conclude with a discussion section addressing limitations, contributions of our study, and future research.
THEORY, LITERATURE, AND HYPOTHESES
Organizational commitment is an individual attitude that reflects one's identification with and involvement in a particular organization (e.g., Mowday et al., 1979). It can be characterized by three related factors: 1) strong belief in the organization's goals and values, 2) willingness to exert extra effort on its behalf, and 3) strong desire to maintain membership. The extant literature has included important debates, or alternative perspectives, on several important issues, including the focus and conceptualization of organizational commitment. Various foci of employee commitment are possible. These include commitments to one's employer, profession, immediate supervisor, and union (Morrow, 1993). Our concern is with commitment to the employer. This aligns with our emphasis on organizational-level variables (i.e., HR practices, organizational characteristics) rather than supervisor or coworker attributes. Also, we emphasize a "global" or unidimensional conception of organizational commitment rather than a faceted approach. Commitment facets are sufficiently correlated such that one may speak meaningfully of overall (unidimensional) commitment (Morrow, 1993), and thus focus on it as a useful concept, even though it may be useful for some purposes to examine its facets in greater detail.
Critics have argued that research on organizational commitment lacks strong theoretical grounding (cf. Mathieu and Zajac, 1990; Morris and Sherman, 1981). At least three compatible theoretical streams apply. First, common to all social exchange theories is the notion that individuals seek favorable outcomes relative to their inputs (Cook et al., 1993). Consistent with social exchange theory, perceived organizational support (POS) addresses the organization's commitment to its employees. Specifically, POS refers to the extent employees perceive that employers value employee contributions and care about their wellbeing (Eisenberger et al., 1986). Finally, psychological contract theory refers to the implicit, reciprocal rights and obligations that individuals perceive within exchange relations (Rousseau, 1995), and can also be used to understand commitment (Rousseau and Wade-Benzoni, 1995). Each perspective suggests that commitment is contingent on perceived exchanges. More favorable exchanges should strengthen employee attraction to the employment relationship and increase commitment. In addition, all three theories allow for exchange of either instrumental or affective outcomes. Finally, implicit or explicit within each theory is the notion that employees hold expectations of what employers should provide (e.g., comparison level (CL) standards or expectations, expectations of support, psychological contract expectations).
Meta-analyses (e.g., Matheiu and Zajac, 1990) and more qualitative reviews (e.g., Morrow, 1993) show that prior study of antecedents varies greatly. Mathieu and Zaiac (1990) were able to incorporate 41 prior estimates relating commitment and employee age, but only three on "organizational characteristics" (size and centralization). None of the antecedents examined could be termed a "human resource practice." This is both interesting and problematic, given that such practices are a principal means organizations use to influence commitment (Guest, 1992).
Variables such as age, gender, and tenure are often studied because they vary within a single site, are readily available, and are easy to measure. Prior studies typically focus on a single site where HR practice variation is limited or non-existent; thus, HR practices are not examined (Caldwell et al., 1990). Although a few studies have examined effects of specific HR practices (e.g., Agarwala, 2003; Bills, 1987; Bemmels, 1995; Eaton, 2003; Florkowski and Schuster, 1992; Lincoln and Kalleberg, 1985), HR measures tend to be unique and not comparable across studies. Tiffs makes the accumulation of findings difficult at best. Hence, little is known about the influence of HR practices and most organizational characteristics. Despite this, prior research provides a basis for predictions on effects of HR practices and organizational characteristics. Testing such predictions is needed to establish the empirical relationship between HR practices and commitment (Meyer and Allen, 1997).
Human Resource Practices
Prior work (especially Delaney and Huselid, 1996) suggests distinct HR concepts that should influence commitment. Using that research as a foundation, the present study posits that internal labor markets, hiring selectivity, training, grievance resolution mechanisms, benefits, employee involvement, incentive pay, union pressure, compensation cuts, and downsizing all affect organizational commitment.
Internal Labor Markets. An internal labor market (ILM) exists when external hiring is limited to entrylevel positions, job "ladders" exist to advance internally, and promotion from within is favored (e.g., Althauser, 1989). Bills (1987) suggested that a primary motivation to establish and maintain ILMs is to secure employee commitment. Since ILMs favor current employees over external rivals, employees should perceive ILMs positively and reciprocate this support with commitment (Eisenberger et al., 1986). Also, individuals may see ILMs fulfilling an employer's duty to provide security and opportunity for development and advancement. We expect employees appreciate their favored status in an ILM and, consequently, feel higher commitment.
H1: Internal labor market practices are positively related to organizational commitment.
Hiring Selectivity. Hiring selectivity refers to the rigor of hiring as indicated by applications per vacancy. Selective hiring may indicate numerous phenomena, including extreme care in matching applicants with job requirements or firm culture, an internal development strategy (i.e., selection for careers), and conscious effort to "skim the cream" of the labor market and offer higher compensation to dissuade shirking. Hiring selectivity and associated screening also may form a "rite of passage," an initiation rite that elicits increased attachment. Employees' awareness that the organization has incurred substantial costs in selection also may enhance their sense of valuation by the organization and their perceptions that it is committed to them (Eisenberger et al., 1990). The effect may be a felt obligation or voluntary desire (or both) by employees to repay" the organization for its investment through commitment. Further, based on attraction-selection-attrition (ASA) theory (Schneider et al., 1995), rigorous selection systems should produce high person-organization fit, enhancing attachment. Finally, employee awareness of rigorous hiring may enhance personal competence perceptions, and meta-analytic evidence shows that perceived personal competence is strongly related to commitment (Matheiu and Zajac, 1990). This suggests that selectivity enhances employee commitment. Prior research supports this prediction (e.g., Caldwell et al., 1990).
H2: Hiring selectivity is positively related to organizational commitment.
Training. A strong training emphasis implies an internal labor market and career opportunities (Althauser, 1989). Such emphasis is indicated by formal training, numbers participating, training effectiveness, and firm attempts to match current training practices. Based on human capital theory, employees should perceive employer training investments as symbols of continued future employment and higher wages (e.g., Strober, 1990). Employees should value training for the security, advancement, and higher earnings implied, and thus be more committed. Prior research shows organizational climates that emphasize continuous training and updating elicit high commitment (e.g., Maurer and Tarulli, 1994).
H3: An emphasis on employee training is positively related to organizational commitment.
Grievance Resolution Mechanisms. Many firms have formal procedures for resolving worker disputes. These provide a mechanism to redress grievances, and promote "voice" over "exit." Such procedures encourage the view that the organization seeks fairness and is open to calls for change from below. Because this concern for due process and fairness signals the organization's support and its commitment to employee concerns, grievance mechanisms should increase employee commitment (Eisenberger et al., 1986; Koys, 1991). Further, grievance procedures fulfill due process and fairness expectations (Rousseau, 1995). Consistent with these arguments, Bemmels (1995) found grievance procedures positively related to commitment.
H4: The provision of formal dispute resolution procedures is positively related to organizational commitment.
Benefits. As noted, certain HR practices provide both instrumental and affective outcomes. Although benefits certainly have instrumental worth to employees, they also may signal the organization's affective valuation of employees (i.e., organizational support). Further, since many benefits are discretionary, employees should reciprocate with commitment (Eisenberger et al., 1986). Koys (1991) found stronger commitment when benefits are seen as discretionary. finally, some prior results suggest a positive relationship (e.g., Grover and Crooker, 1995).
H5: The level of benefits provided to employees is positively related to commitment.
Employee Involvement. Employee involvement is often viewed as an integral part of High Commitment Work Systems. Involvement in decisions provides a sense of ownership of and commitment to both those decisions and the organization (Guest, 1992). Clearly, employee involvement processes should engender the perception that the organization values employee contributions. Thus, we predict that involvement increases commitment. Additionally, employee involvement might be perceived by workers as a discretionary positive benefit. In the sense that involvement opportunities are voluntarily provided, employee commitment to the employer-employee relationship should increase. Further, Lawler and Youn (1996) found that when parties to an exchange relationship work together toward a superordinate goal, relational cohesion increases. In that employee involvement serves the dual purposes of enriching jobs and improving organizational processes and outcomes, employee involvement will be positively related to employee commitment to the organization. Previous research has supported a moderately strong link between related measures (e.g., autonomy) and commitment (Kalleberg et al., 1994; Wallace, 1995). Employee involvement is often realized in the form of increased worker responsibility and autonomy (e.g., Stone, 2005). Therefore, we predict:
H6a: Employee responsibility is positively related to organizational commitment.
H6b: Employee autonomy is positively related to organizational commitment.
Incentive Pay. One argument for incentive pay systems is that they harmonize employee and employer interests by aligning incentives. Further, incentive pay systems should promote equity feelings because workers are paid for performance contributions. Accordingly, employees should view incentive pay as a form of support and show increased commitment in return. Despite well-known problems with incentive systems, this basic idea suggests that workers will be more committed in firms where performance is an important earnings influence. Prior research by Florkowski and Schuster (1992) provides some support for the positive effects of an incentive pay system on commitment.
H7: Incentive pay systems are positively related to organizational commitment.
Union Pressure. Union pressure refers to the extent that unions influence a firm's current practices, specifically practices on training and wages. Workers realize how they benefit from union pressure or activity outside their own workplaces (e.g., when wage increases are granted following a union-management settlement elsewhere). Union pressure may promote perceptions that the organization begrudgingly addresses employee concerns. Workers may thus view organizations as lacking commitment to employees, and reduce their commitment. Finally, workers in firms with high union pressure may credit gains received to unions rather than employers, and thus we expect:
H8: Union pressure is negatively related to organizational commitment.
Compensation Cuts. As Wheeler and McClendon (1991) noted, a pay cut is a clear threat to well-being, and encourages hostility. Compensation cuts constitute a clear violation of the psychological contract. Social exchange theory suggests unilateral employer-initiated reductions in the exchange rate will decrease employee attachment. Further, reduced compensation levels also likely fall below employees' comparison levels (CLs) or expected levels of pay. Compensation cuts also signal a lack of employer support, in that employee contributions are devalued and concerns for employees are seemingly abandoned. Generally, various compensation reductions (e.g., direct cuts in pay and benefits; increased deductibles, co-payments, or employee contributions) should diminish employee commitment.
H9: Compensation cuts are negatively related to organizational commitment.
Downsizing. Downsizing indicates reduced commitment of firms to workers and this causes workers to respond with reduced commitment to employers (Rousseau, 1995). The effect within a given firm is not clearcut, however. Some research suggests that layoff survivors who had the greatest commitment prior to the layoff had greater negative reactions than those who were less committed (Brockner, 1992). Also, "survivors" may be grateful, and new hires (some victims of downsizing elsewhere) may be grateful to their new employers. Thus, we leave this as an empirical question.
H10a: Downsizing is negatively related to organizational commitment among the retained employees.
H10b: Downsizing is positively related to organizational commitment among the retained employees.
Bureaucratic structuring of activities refers to a key dimension of bureaucracy (the other being centralization; Pugh et al., 1968). Structuring is generally thought to indicate greater efficiency (Price and Mueller, 1986), but its impact on worker attitudes is unclear (cf. Adler and Borys, 1996). On the one hand, greater efficiency may mean worker concerns are addressed more effectively, while, on the other, bureaucratic structuring might imply that workers are removed from decisions assigned to specialists and experience the frustration of dealing with layers of specialized bureaucrats. According to Lawler and Youn (1993), the intensity of interaction between parties to exchange influences relational cohesion and each party's commitment. If structuring depersonalizes the exchange, commitment should fall. Although there is evidence that bureaucratic structuring increases commitment in voluntary associations (Knoke, 1990), on balance the benefits of structuring accrue to non-employee stakeholders. Consistent with this reasoning and previous results (e.g., Lincoln and Kalleberg, 1985), we expect:
H11: Bureaucratic structuring is negatively related to organizational commitment.
Decentralization. Previous discussion references issues associated with decentralization, specifically, employee involvement. Although sharing of decision making may operate through such variables, decentralization is more pervasive. For example, employees may value decision making that is "local" and feel committed to such decisions, even if not directly involved. As such, shared decision-making power should engender perceptions of organizational support, beyond effects stipulated earlier. Further, because organizations typically have no formal obligation to distribute decision making, those that do should be perceived positively (e.g., a goodwill symbol). Evidence mildly suggests that workers are more committed to decentralized organizations (Mathieu and Zajac, 1990).
H12: Decentralization is positively related to organizational commitment.
Not-for-profit Status. Not for profit (NFP) organizations often exist to serve goals regarded as nobler than the goals of for-profit organizations. Employees are often drawn to NFPs to serve a "cause" (e.g., to help others). Moreover, one may seek employment to serve a cause and subsequently generalize that commitment to the organization. According to ASA theory, employees in organizations whose values and goals are congruent with their own should form strong attachments (Schneider, 1987). Because a major aspect of commitment is one's identification with organizational goals and values, commitment should be higher in NFPs.
H13: Organizational commitment is higher among employees in not-for-profit organizations than among employees in for-profit organizations.
Figure I provides a graphic summary of the research relationships examined here. As shown, HR practice and organizational characteristics are predicted to have direct effects on organizational commitment. Figure I also presents control variables. Those will be discussed briefly below.
[FIGURE 1 OMITTED]
The data used in this study are from the 1991 National Organizations Survey (NOS) database. The NOS consists of two surveys, linked via a hyper-sampling design. Respondents to the 1991 General Social Survey (GSS; a national probability sample of adults) were asked to identify their employers. The "NOS proper" was then administered to GSS respondents' employers (also in 1991), yielding a nationally representative sample of employing establishments. (For details on sampling and procedures see Kalleberg et al., 1993, 1996). The GSS component provides data on various employee characteristics and attitudes. Demographic variables such as age, education, and gender are provided, along with a six-item Likert-style scale for organizational commitment based closely on the Organizational Commitment Questionnaire or "OCQ" (Mowday et al., 1979).
The real strength of the NOS arises, however, in the richness of the NOS-proper component, the employer survey, and matched employee-employer responses. The quality of available measures varies across issues, but the NOS-proper provides multi-item scales for many HR practices and other organizational characteristics. Just under 400 cases are available with matching surveys and complete responses to items in the organizational commitment measure.
Our measures for most independent variables closely follow previous work by Delaney and Huselid (1996). When possible, we used employer responses to measure HR practices and organizational characteristics. Employee responses (from the GSS) were used to measure organizational commitment and control variables (e.g., job satisfaction). To obtain greater confidence in the scales, we submitted multi-item measures to exploratory factor analysis. In the majority of cases, the constructed scales were, in fact, uni-dimensional. Other scales were examined carefully (based on content) to ensure that multiple dimensions within the single measure were part of a higher-order construct captured by the composite scale. (A detailed description of all measures is provided in an appendix available from the authors.)
Human Resource Practices. Two items were used to measure internal labor markets by assessing the extent of promotion-from-within. A single question was posed for up to three different job types, yielding up to three items used to form a scale measuring hiring selectivity ("In the past two years, about how many applications have you considered for an opening?"). Four items were used to measure training emphasis (e.g., "Within the last two years, how many employees participated in formal job training?"). A single item addressed grievance resolution procedures ("Are there formal procedures for resolving disputes between employees and their supervisors or co-workers?"). Twelve items measured benefits (e.g., medical benefits, sick leave with full pay, pension programs). Two scales were constructed to measure employee involvement. First, a seven-item scale was constructed for worker responsibility (e.g., training of others, evaluation of others' performance). Second, a four-item scale was constructed for worker autonomy (e.g., having substantial decision input, working independently). A single question was posed for up to three different job types, yielding up to three items used to form a scale measuring incentive pay ("How important is job performance in determining individual earnings?"). Four items addressed whether HR practices are influenced by union pressure (e.g., "Is any of the formal training offered because of union contracts?"). Five items were used to measure compensation cuts (e.g., "Over the past two years, has the organization lowered salaries or wages, or given smaller raises to pay for medical or fringe benefits?"). Finally, one item assessed downsizing ("Compared to one year ago, has the number of full-time employees in the organization increased, decreased, or remained the same?").
Organizational Characteristics. A single item assessed organizations' not-for-profit status ("Not-for-profit status of the organization"--yes or no). A composite index based on items used to assess organizational specialization (eight items), departmentalization (eight items), vertical hierarchy (one item), and formalization (seven items) constituted a measure of bureaucratic structuring. Finally, eight items were used to measure decentralization of decision making.
Organizational Commitment. Mowday et al. (1979) developed the Organizational Commitment Questionnaire (OCQ), a self-report instrument. Three facets were originally proposed, but Mowday et al. (1982) maintain that the OCQ measures a single global construct. Other related measures consistently demonstrate strong consistency with the OCQ (Morrow, 1993). Six items based on the OCQ (Mowday et al., 1979) were used to measure organizational commitment (Davis and Smith, 1991). Except for very minor wording changes, the GSS commitment items were identical to OCQ items. Factor analysis indicated that the six-item OCQ based measure was, in fact, unidimensional.
Other possible antecedents of organizational commitment, which may correlate with the variables of interest, are controlled to obtain unbiased estimates. We control for variables that have shown moderate or strong influences on commitment (Mathieu and Zajac, 1990). Single items were used to measure education, gender, minority status, job complexity, and organizational tenure. Multi-item scales were used for job satisfaction, perceived personal competence, and the perceived fairness of rewards. Although organizational level was formed by consulting multiple items, they were not combined in an additive fashion. Consequently, the organizational level variable does not have an alpha.
Ordinary least squares (OLS) was used to estimate relationships for two samples. One sample consisted of cases remaining after case-wise deletion for missing data (N = 199). The second used the replacement of missing data for independent variables via mean substitution (N = 381). Although no particular variable presented extraordinary missing data problems--no single predictor had less than 334 valid values--the cumulative effect of missing data using case-wise deletion was quite large. Given the sample size gains from missing data replacement (N = 199 versus N = 381), we present results for both samples. For directional hypotheses, we used one-tailed tests. For H10a and H10b and the control variables we used two-tailed tests.
This section will present both descriptive and analytic results. Descriptive statistics, including means, standard deviations, correlations, and coefficient alpha (if applicable) are provided in Table 1. (Descriptive statistics following missing data replacement are available from the authors on request.)
Regression results are presented in Table 2. As shown, the overall model was significant (p < .01) for both samples. Explained variance in organizational commitment was 40.8 percent in both, and results for individual predictors were similar across samples. Results for the control variables showed that job satisfaction had a strong impact ([beta] = .50 to .52, p < .01) on commitment. Perceived personal competence was marginally significant and positive in Sample 1 (p < .10) but not in Sample 2 (marginal significance level, MSL = .11). Perceived fairness of rewards also was related directly to commitment (p < .05) in Sample 1, but not in Sample 2 (MSL = .14). None of the remaining controls was significant.
Human Resource Practices and Organizational Commitment
Results from the regression analyses indicated that internal labor markets were not significant in either sample. Thus, HI was not supported. Results also failed to support a significant effect for hiring selectivity. Thus, H2 was not supported. In fact, the coefficient for the selectivity variable was large relative to its standard error in both samples with the "wrong" sign; that is, the results tend to "disconfirm" H2. Hypothesis 3, which stipulated that employee training was positively related to commitment, was not supported in either sample. Hypothesis 4, which posited that grievance resolution procedures are positively related to organizational commitment, was supported in Sample 2 (p < .05), but its coefficient was not significant in Sample 1 (MSL =. 17). Hypothesis 5, which predicted that employee benefits increase organizational commitment, was not supported in either sample. Hypothesis 6a, which argued that responsibility is positively related to organizational commitment, was supported in Sample 1 (p < .01) and marginally supported in Sample 2 (p < .10). Hypothesis 6b, which stipulated that individual autonomy is positively related to organizational commitment, was supported in Sample 2 (p < .01) and marginally supported in Sample 1 (p < .10). None of the remaining hypotheses for HR practices received support, with the exception of H9 (concerning compensation cuts' adverse effects on organizational commitment), which received marginal support in Sample 1 (p < .10) but not in Sample 2 (MSL = .11).
Organizational Characteristics and Organizational Commitment
Results from the regression analyses did not support H11, which stated that bureaucratic structure is negatively related to organizational commitment. However, H12, which predicted that decentralization of decision making is positively related to commitment, was supported in Sample 1 (p < .05) and Sample 2 (p < .01). Finally, H13, positing higher commitment in not-for-profits, was not supported. Overall, the [R.sup.2] change due to organizational characteristics was approximately .02 in both samples, a significant change (p < .05) for Sample 2, but not in Sample 1 (MSL = .16).
Removing Job Satisfaction from the Analysis
Including job satisfaction as a predictor of commitment may "stack the deck" against finding HR practice effects; that is, these practices affect job satisfaction and, thus, by including job satisfaction as a control, only direct effects of HR practices on commitment would show in our findings. Direct effects are in fact what we posited, but the large effect for job satisfaction led us to examine equations without the job satisfaction predictor to assess the sensitivity of our results for hypothesized effects. Without job satisfaction, the [R.sup.2] values dropped dramatically (e.g., to .21 for Sample 1), but the results for HR practices and organizational characteristics changed little. In Sample 1, for example, without job satisfaction the grievance mechanisms variable became significant at the .10 level, whereas it was not significant (MSL = .17) with job satisfaction included. In either case, the responsibility, autonomy, compensation cuts, and decentralization variables were significant with the signs shown in Table 2, and no other HR practice or organizational characteristic variables were significant.
Synergy Among HR Practices
One possible explanation for the modest evidence for HR practice effects involves synergy among HR practices--individual practices may be less important than their configuration. Following Delaney and Huselid (1996), we created an index where each firm "scored" one point for each instance where its practice on one of seven "progressive" HR practices (e.g., internal labor markets, selective hiring) exceeded the sample mean. The results for this variable were not significant in Sample 1 (MSL = .12) or in Sample 2 (MSL = .29). Results for other variables were similar to those in Table 2. Thus, as in Delaney and Huselid's study of firm performance, there was no evidence of a synergistic effect for HR practices.
Employee Reports versus Employer Reports and Measurement Quality
Most of the multi-item scales used demonstrated acceptable reliability, but some were constructed from individual items that were not part of recognized, validated scales. Although we examined the dimensionality and reliability of multi-item scales, there are grounds to question the validity of some measures. Perhaps the case of the single-item measure for grievance resolution practices best illustrates this point. Our sample's employers responded to a yes/no question: "Are there formal procedures for resolving disputes between employees and their supervisors or co-workers?" An affirmative response can mean many things, from an anonymous suggestion/complaint box to a formal procedure with codified employee rights and provisions for impartial resolution by an arbitrator (i.e., from "window dressing" to a genuine "system of industrial jurisprudence"). Given this, the positive but modest evidence merits some respect, as it seems likely that our results understate the organizational commitment effects of formal voice mechanisms. Although other measures may be less problematic in such regards, this may be a general problem in measuring HR practices from informant reports. In effect, what employers report and what employees experience may diverge. Nominally similar HR practices may diverge as a function of top management commitment to these practices, among other reasons.
To explore this issue further, we experimented in the area of employee benefits, the only area of HR practices where our data provided comparable reports from both employees and employers. The sensitivity of our results to reliance on imperfect reports from employers may be seen by examining the comparative predictive power of employee reports. Overall, the results in specifications substituting the employee-based measure for the employer-based measure were very similar to those reported in Table 2, yet the coefficient for the benefits scale based on employee reports was positive and at least marginally significant (p < .07 or better) in both samples. This lends some credence to the possibility that a reason for the disappointing results may be shortcomings in the employer-based measures. Unfortunately, it is not possible to separate this possibility from the confounding effects of common method variance that may inflate results for employee-based measures (see below).
Much of prior literature suggests that HR practices and other organizational characteristics influence organizational commitment. However, little research has tested these predictions. Further, virtually all prior research used single-site samples. Thus, the ability to generalize has been limited. To address these shortcomings, we examined the effects of HR practices and organizational characteristics using a diverse national probability sample of U.S. workers and their employing organizations.
Overall, the [R.sup.2] change due to HR practice measures in both samples is on the order of .05, a significant change in a joint F-test (p < .01) in Sample 2, but not in Sample 1. The results for these joint tests are notable in summarizing the collective effects of HR practices, and particularly in terms of overcoming co-linearity problems that can make tests for individual variables unreliable. That is, to the extent that HR practices co-vary strongly, their individual effects are difficult to estimate with precision, but the joint F-test includes their common as well as unique effects. Thus, along with the modest correlations among HR practices and other predictor variables indicated in Table 1, the joint F-test results lead us to discount co-linearity as an explanation for the HR practices results. Note that bivariate correlations between HR practice measures and commitment in Table 1 are modest to weak. This is not an instance where predictor-criterion relations are obscured by correlation among predictors.
Taken as a whole, the results provide limited support for the proposed effects of HR practices and organizational characteristics. Specifically, for HR practices, results showed that grievance resolution procedures, compensation cuts (inversely), and employee involvement predicted commitment. Further, the only organizational characteristic that predicted commitment was decentralization.
Many hypotheses were not supported. These findings may call into question the attitudinal commitment effects of some HR practices (e.g., generous benefits, internal labor markets) and organizational characteristics (e.g., bureaucratic structure, not-for-profit status). At the same time, we acknowledge that hindsight suggests other possible explanations. Nonsignificant or marginal results for many HR practices may lie in the nature of the variables. Specifically, given the relatively fixed, stable quality of most HR practices, it may be that the influence of these organizational-level factors is a bit far removed from the attitudinal variable of organizational commitment (cf. Mathieu and Zajac, 1990). That is, certain human resource practices may constitute more distant rather than more immediate influences. Accordingly, observed effects of HR practices may not be very strong. Becker et al.'s (1996) findings on supervisory commitment as a predictor of performance indicates that more immediate influences may have stronger effects.
Aside from the quality of measures, there is the question of omitted variables. Although the NOS provided at least a reasonable proxy for many commitment antecedents identified in previous research, some gaps are evident. We already noted the absence of direct employee involvement measures. Another notable gap is initial employee socialization, and we were unable to identify any reasonable proxy for this concept. Future research needs to further examine these variables.
Another issue is statistical power. Based on our sample sizes and number of predictors, tables for a power level of .80 (e.g., Green, 1991) suggests that this study was capable of detecting moderate or strong effects, but perhaps not small effects. Limitations in measures noted above may combine with limited power to account for some disappointing findings. Despite some positive and interesting findings, questions remain about HR practice or organizational characteristic effects on commitment.
A final possible limitation is the age of the data. HR practices and employee expectations have certainly changed since the surveys were administered in 1991. Although the dataset is older, the results are still applicable to current business practices. Most or all of the HR practice variables tested in this study are still widely used in organizations and are subjects of continuing research debate. Also, considering the breadth of organizations sampled, it is unlikely that either employee expectations or HR practices have significantly changed across the range of workplaces surveyed. However, future research could attempt to replicate these findings using a more recently sampled population.
Despite limitations, this study makes important contributions. A major contribution is that it examines the influences of numerous HR practices and organization-level factors on organizational commitment rarely examined in prior research or that have been examined only in a single site. As such, the study provides some insight on the "commitment maximizing" effects of HR practices and organizational characteristics suggested by some scholars (e.g., Guest, 1992). A second contribution lies in the relatively unique quality of the data (despite limitations noted). Specifically, the fact that data cross levels from multiple, independent sources and were collected at different times contributes to the validity of the findings, the confidence that can be placed in them, and the conclusions that can be drawn. Further, because the study results are based upon data obtained from a nationally diverse probability sample, the generality is higher than in most prior studies. Finally, the nature of the relationships examined herein also allows stronger causal inferences to be drawn than is possible in most commitment research. Specifically, the probable direction of influence is much stronger for the effects of HR practices and organizational characteristics on organizational commitment than for the reverse effects. This is consistent with Mathieu and Zajac's (1990) call for improved causal analysis and inference.
For practitioners, these results indicate the importance of employee "voice" through grievance resolution procedures, employee involvement, and decentralization. In particular, if one views employee involvement programs to be generally consonant with the decentralization of decision making (a structural method for increasing worker autonomy) and programs designed to increase employee "voice," the overall pattern of results converges to provide support for an "empowerment" effect. Thus, it appears that organizational practices and characteristics that provide for the expression of worker interests are the strongest organizational determinants of employee commitment.
The cumulative effect of the preceding research design qualities is a level of methodological rigor that is often absent in organizational commitment research, which frequently relies on self-report data collected at a single point in time, leaving results open to common method variance explanations. Mathieu and Zajac (1990) noted in their meta-analytic review that the average adjusted correlation of organizational commitment with other variables, when both were measured with self-report instruments, was 2.66 times larger than the same average adjusted correlation when different measurement methods or data sources were used. It is notable that Mathieu and Zajac's (1990) meta-analytic findings showed that the average adjusted correlation across commitment and related variables, when using different measurement sources and methods, was a mere. 132. Therefore, despite the potential limitations noted earlier, our findings may, in fact, be more meaningful than first appears.
Finally, both established theoretical bases and prior organizational commitment research were used to ground the hypotheses in the present study. Although the notion of exchange between employer and employee has been recognized in the area of organizational commitment (cf. Mowday et al., 1982), none of the prior work that we reviewed provided a clear and explicit theoretical integration of organizational commitment with social exchange theory. Given that social exchange theory addresses both instrumental and affective exchanges, it could serve as a basis for integrating conceptions of calculative or continuance commitment (which are based on instrumental or economic concerns) and the attitudinal conception of commitment examined here. For example, social exchange theory examines how relationships initially based on instrumental concerns may later become more affectively oriented and could prove useful in understanding the dynamics of how commitment develops over time (cf. Mathieu and Zajac, 1990; Mowday et al., 1982). An explicit treatment of commitment based on social exchange theory also might be useful in generating new, insightful hypotheses. For example, theory predicts that interaction frequency or intensity leads to a stronger relational affect and cohesion, and the objectification of the relationship as a unique, valued entity. Future research, therefore, might examine whether the frequency or intensity of organizational-individual interactions predicts commitment (e.g., the relative effects of social activities or communications on member commitment).
* We thank Dan Gallagher for helpful comments on an earlier draft of this paper. Cynthia Fukami provided useful comments on the literature. We are grateful to John Delaney, Mark Huselid, and David Knoke for assistance in using the National Organizations Survey. Marc Street provided research assistance. We thank Kevin Halligan for special assistance.
Adler, P. and B. Borys. 1996. "Two Types of Bureaucracy: Enabling and Coercive." Administrative Science Quarterly 41: 61-89.
Agarwala, T. 2003. "Innovative Human Resource Practices and Organizational Commitment: An Empirical Investigation." International Journal of Human Resource Management 14: 175-197.
Althauser, R. 1989. "Internal Labor Markets." In Annual Review of Sociology 15. Eds. W. Scott and J. Blake. Palo Alto, CA: Annual Reviews, Inc. pp. 143-161.
Becker, T., R. Billings, D. Eveleth and N. Gilbert. 1996. "Foci and Bases of Commitment: Implications for Job Performance." Academy of Management Journal 39: 464-482.
Bemmels, B. 1995. "Dual Commitment: Unique Construct or Epiphenomenon?" Journal of Labor Research 16: 401-422.
Berger, C. and L. Cummings. 1990. "Organizational Structure, Attitudes, and Behavior." In Personality and Organizational Influence. Eds. B. Staw and L. Cummings. Greenwich, CT: JAI Press. pp. 287-326.
Bills, D. 1987. "Costs, Commitment, and Rewards: Factors Influencing the Design and Implementation of Internal Labor Markets." Administrative Science Quarterly 32: 202-221.
Brockner, J. 1992. "Managing the Effects of Layoffs on Survivors." California Management Review 34: 9-28.
Burton, J, T. Lee and B. Holtom. 2002. "The Influence of Motivation to Attend, Ability to Attend, and Organizational Commitment on Different Types of Absence Behaviors." Journal of Managerial Issues 14 (2): 181-197.
Caldwell, D., J. Chatman and C. O'Reilly. 1990. "Building Organizational Commitment: A Multifirm Study." Journal of Occupational Psychology 63: 245-261.
Colbert, A. and I-W. Kwon. 2000. "Factors Related to the Organizational Commitment of College and University Auditors." Journal of Managerial Issues 12 (4): 484-501.
Cook, K., L. Molm and T. Yamagishi. 1993. "Exchange Relations and Exchange Networks: Recent Developments in Social Exchange Theory." In Theoretical Research Programs: Studies in the Growth of Theory. Eds. J. Berger and M. Zelditch, Jr. Stanford, CA: Stanford University Press. pp. 296-322.
Davis, J. and T. Smith. 1991. General Social Surveys, 1972-1991: Cumulative Codebook. Storrs, CT: Roper Center for Public Opinion Research.
Delaney, J. and M. Huselid. 1996. "The Impact of Human Resource Management Practices on Perceptions of Organizational Performance." Academy of Management Journal 39: 949-969.
Eaton, S. 2003. "If You Can Use Them: Flexibility Policies, Organizational Commitment, and Perceived Performance." Industrial Relations 42: 145-167.
Eisenberger, R., P. Fasolo and V. Davis-Lamastro. 1990. "Perceived Organizational Support and Employee Diligence, Commitment, and Innovation." Journal of Applied Psychology 75: 51-60.
--, R. Huntington, S. Hutchison and D. Sowa. 1986. "Perceived Organizational Support." Journal of Applied Psychology 71: 500-507.
Farrell, D. and C. Stamm. 1988. "Meta-analysis of the Correlates of Employee Absence." Human Relations 41:211-227.
Florkowski, G. and M. Schuster. 1992. "Support for Profit Sharing and Organizational Commitment: A Path Analysis." Human Relations 45: 507523.
Green, S. 1991. "How Many Subjects Does It Take to Do a Regression Analysis?" Multivariate Behavioral Research 26:499-510.
Grover, S. and K. Crooker. 1995. "Who Appreciates Family-responsive Human Resource Practices: The Impact of Family-friendly Practices on the Organizational Attachment of Parents and Nonparents." Personnel Psychology 48: 271-288.
Guest, D. 1992. "Employee Commitment and Control." In Employment Relations: The Psychology of Influence and Control at Work. Eds. J. F. Hartley and G. M. Stephenson. Oxford, England: Blackwell Publishers, Inc.
Hammer, T. H. and A. Avgar. 2005. "The Impact of Unions on Job Satisfaction, Organizational Commitment, and Turnover." Journal of Labor Research 26: 241-266.
Jandeska, K. E. and M. L. Kraimer. 2005. "Women's Perceptions of Organizational Culture, Work Attitudes, and Role-modeling Behaviors." Journal of Managerial Issues 17 (4): 461-478.
Kalleberg, A., D. Knoke, P. Marsden and J. Spaeth. 1996. Organizations in America: Analyzing Their Structures and Human Resource Practices. Newbury Park, CA: Sage. --, --, -- and --. 1994. "The National Organizations Study: An Introduction and Overview." American Behavioral Scientist 37: 860-871.
--, --, -- and --. 1993. National Organizations Survey (NOS), 1991 [computer file]. Champaign, IL: University of Illinois Survey Research Laboratory [producer]. Ann Arbor, MI: Interuniversity Consortium for Political and Social Research [distributor], 1994.
Knoke, D. 1990. Organizing for Collective Action. New York, NY: Aldine de Gruyter.
Koys, D. 1991. "Fairness, Legal Compliance, and Organizational Commitment." Employee Responsibilities and Rights Journal 4: 283-291.
Lawler, E. and J. Youn. 1996. "Commitment in Exchange Relations: Test of a Theory of Relational Cohesion." American Sociological Review 61: 89-108.
-- and --. 1993. "Power and the Emergence of Commitment Behavior in Negotiated Exchange." American Sociological Review 58: 465-481.
Lincoln, J. and A. Kalleberg. 1985. "Work Organization and Workforce Commitment: A Study of Plants and Employees in the U.S. and Japan." American Sociological Review 50: 738-760.
Mathieu, J. and D. Zajac. 1990. "A Review and Metaanalysis of the Antecedents, Correlates, and Consequences of Organizational Commitment." Psychological Bulletin 108: 171-194.
Maurer, T. and B. Tarulli. 1994. "Investigation of Perceived Environment, Perceived Outcome, and Person Variables in Relationship to Voluntary Development Activity by Employees." Journal of Applied Psychology 79: 3-14.
Meyer, J. and N. Mlen. 1997. Commitment in the Workplace: Theory, Research, and Application. Thousand Oaks, CA: Sage Publications.
Morris, J. and J. Sherman. 1981. "Generalizability of an Organizational Commitment Model." Academy of Management Journal 24: 512-526.
Morrow, P. 1993. The Theory and Measurement of Work Commitment. Greenwich, CT: JM Press.
Mowday, R., L. Porter and R. Steers. 1982. Employee-organization Linkages. New York, NY: Academic Press.
--, R. Steers and L. Porter. 1979. "The Measurement of Organizational Commitment." Journal of Vocational Behavior 14: 224-247.
Ogilvie, J. 1986. "The Role of Human Resource Management Practices in Predicting Organizational Commitment." Group and Organizational Studies 11: 335-359.
Osterman, P. 1995. "Work/Family Programs and the Employment Relationship." Administrative Science Quarterly 40: 681-700.
--. 1994. "How Common is Workplace Transformation and How Can We Explain Who Does It?" Industrial and Labor Relations Review 47: 173-188.
Payne, S. C. and A. H. Huffman. 2005. "A Longitudinal Examination of the Influence of Mentoring on Organizational Commitment and Turnover." Academy of Management Journal 48: 158-168.
Price, J. and C. Mueller. 1986. Handbook of Organizational Measurement. Marshfield, MA: Pitman Publishing, Inc.
Price, K. 1990. "Declining Employee Commitment: What It Means, Why It Is Happening, and What We Can Do About It." In Human Resource Strategies for Organizations in Transition. Eds. R.J. Niehaus and K. F. Price. New York, NY: Plenum Press. pp. 183-197.
Pugh, D., D. Hickson, C. Hinings and C. Turner. 1968. "Dimensions of Organization Structure." Administrative Science Quarterly 13: 65-105.
Quarles, R. 1994. "An Examination of Promotion Opportunities and Evaluation Criteria as Mechanisms for Affecting Internal Auditor Commitment, Job Satisfaction and Turnover Intentions." Journal of Managerial Issues 6(2): 176-194.
Rousseau, D. 1995. Psychological Contracts in Organizations: Understanding Written and Unwritten Agreements. Thousand Oaks, CA: Sage.
--and K. WadeBenzoni. 1995. "Changing Individual-Organizational Attachments." In The Changing Nature of Work. Ed. A. Howard. San Francisco, CA: Jossey-Bass. pp. 290-322.
Schneider, B. 1987. "The People Make the Place." Personnel Psychology 40: 437-453.
--, H. Goldstein and D. Smith. 1995. "The ASA Framework: An Update." Personnel Psychology 48: 747-773.
Stone, D. 2005. "Employee Involvement." In The Blackwell Encyclopedia of Management: Human Resource Management. Ed. S. Cartwright. Malden, MA: Blackwell Publishing. pp. 110-111.
Strober, M. 1990. "Human Capital Theory: Implications for HR Managers." Industrial Relations 29: 214-239.
Wallace, J. 1995. "Organizational and Professional Commitment in Professional and Nonprofessional Organizations." Administrative Science Quarterly 40: 228-255.
Wheeler, H. and J. McClendon. 1991. "The Individual Decision to Unionize." In The State of the Unions. Eds. G. Strauss, D. G. Gallagher and J. Fiorito. Madison, WI: Industrial Relations Research Association. pp. 4783.
Williams, L. and J. Hazer. 1986. "Antecedents and Consequences of Satisfaction and Commitment in Turnover Models: A Re-analysis Using Latent Variable Structural Equation Methods." Journal of Applied Psychology 71: 219-231.
Wright, T. and D. Bonett. 2002. "The Moderating Effects of Employee Tenure on the Relations Between Organizational Commitment and Job Performance: A Meta-analysis." Journal of Applied Psychology 87:1183-1190.
J. Frank Dame Professor of Management
Florida State University
Dennis P. Bozeman
Associate Professor of Management
University of Houston
Associate Professor of Management
California State University Los Angeles
James A. Meurs
Florida State University
Table 1 Means, Standard Deviations, Reliabilities, and Correlations (N = 199) Variable Mean SD 1 2 3 4 1 Org. Commitment 2.774 0.480 (77) 2 Internal Market 0.033 0.640 -07 (71) 3 Selectivity 0.110 0.764 03 23 (67) 4 Training 0.092 0.802 -06 33 27 (89) 5 Grievance Mechanisms 0.759 0.429 03 33 12 39 6 Benefits 0.692 0.177 -01 44 21 31 7 Responsibility 2.070 1.988 17 -07 05 -04 8 Autonomy 11.447 2.775 30 00 -01 -13 9 Incentive Pay 2.645 0.564 09 -01 05 -08 10 Union Pressure 1.408 0.589 04 19 07 25 11 Compensation Cuts 0.176 0.184 -22 18 -06 23 12 Downsizing 0.236 0.426 00 11 -02 10 13 Not-for-Profit 0.397 0.491 -02 08 08 19 14 Structuring 0.107 0.726 -01 46 32 49 15 Decentralization 3.521 0.978 06 16 34 21 16 Education 13.724 2.689 07 03 21 04 17 Gender 0.528 0.500 -03 -02 -02 -05 18 Minority 0.106 0.308 -08 16 07 07 19 Job Complexity 44.995 14.289 08 -04 10 06 20 Org. Tenure 7.819 8.732 -02 05 -15 03 21 Org. Level 0.452 0.715 20 -20 -09 -18 22 Job Satisfaction 0.029 0.679 55 -06 08 -13 23 Perceived Competence 3.889 0.659 10 -05 09 -04 24 Fairness of Rewards 0.595 0.357 13 00 -02 -03 Variable 5 6 7 8 9 10 1 Org. Commitment 2 Internal Market 3 Selectivity 4 Training 5 Grievance Mechanisms (--) 6 Benefits 37 (79) 7 Responsibility -06 00 (78) 8 Autonomy -03 -02 15 (78) 9 Incentive Pay -06 01 -01 13 (84) 10 Union Pressure 28 16 03 03 -28 (79) 11 Compensation Cuts 10 27 -03 -15 -04 13 12 Downsizing 09 11 -07 -06 -06 21 13 Not-for-Profit 36 00 -05 04 -30 16 14 Structuring 47 41 05 -03 -03 19 15 Decentralization -01 13 -01 -06 14 -06 16 Education -01 08 09 20 02 02 17 Gender -06 -16 -10 -10 02 -07 18 Minority 19 09 -14 -19 00 -03 19 Job Complexity 07 -01 07 19 04 00 20 Org. Tenure 17 12 06 06 -01 22 21 Org. Level 00 -08 29 29 09 -08 22 Job Satisfaction -04 -06 02 31 09 -06 23 Perceived Competence -01 02 14 06 01 -01 24 Fairness of Rewards -08 -19 -11 05 -04 11 Variable 11 12 13 14 15 16 1 Org. Commitment 2 Internal Market 3 Selectivity 4 Training 5 Grievance Mechanisms 6 Benefits 7 Responsibility 8 Autonomy 9 Incentive Pay 10 Union Pressure 11 Compensation Cuts (42) 12 Downsizing 09 (--) 13 Not-for-Profit -02 -06 (--) 14 Structuring 13 14 19 (79) 15 Decentralization 02 06 -09 45 (91) 16 Education 06 -01 25 -01 07 (--) 17 Gender -05 -02 15 -13 -06 02 18 Minority 04 19 -01 14 14 -08 19 Job Complexity -01 -04 36 01 01 60 20 Org. Tenure -01 08 -10 02 02 -22 21 Org. Level -16 -07 -05 -07 -03 23 22 Job Satisfaction -27 -04 -08 -11 -05 -02 23 Perceived Competence -07 -17 -03 -03 00 -02 24 Fairness of Rewards -03 -13 08 -08 -10 -08 Variable 17 18 19 20 1 Org. Commitment 2 Internal Market 3 Selectivity 4 Training 5 Grievance Mechanisms 6 Benefits 7 Responsibility 8 Autonomy 9 Incentive Pay 10 Union Pressure 11 Compensation Cuts 12 Downsizing 13 Not-for-Profit 14 Structuring 15 Decentralization 16 Education 17 Gender (--) 18 Minority 10 (--) 19 Job Complexity 14 -14 (--) 20 Org. Tenure -09 00 -10 (--) 21 Org. Level -11 -06 23 12 22 Job Satisfaction -03 -13 -04 00 23 Perceived Competence 00 10 -04 07 24 Fairness of Rewards 19 -22 02 -04 Variable 21 22 23 24 1 Org. Commitment 2 Internal Market 3 Selectivity 4 Training 5 Grievance Mechanisms 6 Benefits 7 Responsibility 8 Autonomy 9 Incentive Pay 10 Union Pressure 11 Compensation Cuts 12 Downsizing 13 Not-for-Profit 14 Structuring 15 Decentralization 16 Education 17 Gender 18 Minority 19 Job Complexity 20 Org. Tenure 21 Org. Level (--) 22 Job Satisfaction 15 (--) 23 Perceived Competence 23 -02 (68) 24 Fairness of Rewards -14 13 -22 (56) Notes: Decimals omitted for correlations and reliabilities (shown on diagonals for multi-item scales). Correlations of approximately .14 or larger in absolute value are significant at the .05 level or better (two-tailed test). Correlations of approximately .18 or larger in absolute value are significant at the .01 level or better. Table 2 Standardized Regression Results for Organizational Commitment Sample 1 (N = 199) Independent Variables [beta] [delta] F([delta] [R.sup.2] [R.sup.2]) Human Resource Practices .058 1.55 Internal Labor Markets -0.068 Selectivity in Hiring -0.128 Training .020 Grievance Mechanisms .078 Benefits .073 Responsibility .150 ** Autonomy .093 (+) Incentive Pay .044 Union Pressure .094 Compensation Cuts -.095 (+) Downsizing .016 Organizational Characteristics .018 1.76 Bureaucratic Structure -0.021 Decentralization .156 * Not-for-Profit Status -.011 Control Variables Education .024 Gender (Female) -.014 Minority Status .032 Job Complexity .057 Organizational Tenure -0.096 Organizational Level .010 Job Satisfaction .504 ** Perceived Personal Competence .118 * Fairness of Rewards .136 * Overall F 5.245** [R.sup.2] .408 Adjusted [R.sup.2] .330 Sample 2 (N = 381) Independent Variables [beta] [delta] F([delta] [R.sup.2] [R.sup.2]) Human Resource Practices .049 2.68** Internal Labor Markets .042 Selectivity in Hiring -.120 Training .025 Grievance Mechanisms .117 * Benefits -0.038 Responsibility .063 (+) Autonomy .151 ** Incentive Pay -0.022 Union Pressure .022 Compensation Cuts -0.055 Downsizing .020 Organizational Characteristics .016 3.25* Bureaucratic Structure -0.067 Decentralization .147 ** Not-for-Profit Status -.002 Control Variables Education -.011 Gender (Female) -.027 Minority Status -.021 Job Complexity .030 Organizational Tenure -.055 Organizational Level .061 Job Satisfaction .516 ** Perceived Personal Competence .071 Fairness of Rewards .066 Overall F 10.673** [R.sup.2] .408 Adjusted [R.sup.2] .370 Notes: (+) p < .10, *p < .05, **p < .01. The [delta][R.sup.2] (change in [R.sup.2]) and associated F-statistics are for removal of the specified sets of variables, that is when all other variables are retained in the equations.