he nature and extent of differences between employees and organizations of the public and private sectors has long been a topic of academic debate. Although much of the debate has centered on structural differences between public and private organizations, there has also been a longstanding interest in the differences between public and private sector employees. During the past decade, the New Public Management (NPM) movement (Barzelay 2001; Osborne and Gaebler 1992) has downplayed sectoral distinctions through its implicit suggestion that “management is management,” regardless of sector. The NPM paradigm has its roots in the business concepts of customer service and total quality management and in public-choice theory, a branch of economics that views individuals as self-interested and rational decision makers who primarily seek to maximize their personal utility (Argyriades 2003; Borins 2002). Consequently, career choices are viewed as the outcome of rational and self-interested job seekers’ choice between the competing options of public and private sector jobs in order to obtain the maximum set of personal rewards (Argyriades 2003; Perry and Wise 1990). This perspective implies that sectoral distinctions are becoming increasingly irrelevant to job seekers, that is, a job is a job, regardless of the sector in which the organization operates. The contrasting view, espoused by proponents of public-service motivation (e.g., Frederickson and Hart 1985; Perry and Porter 1982; Perry and Wise 1990), is that individuals are drawn to careers in public service primarily by a unique set of altruistic motives such as wanting to serve the public interest, effect social change, and shape the policies that affect society. This perspective views public service as a distinct profession or calling to which certain types of people are morally compelled. This implies that job seekers do not necessarily view private sector and public sector jobs as competing options; an individual who is drawn to a career in public service would choose a public sector job even if the economic rewards were not competitive with comparable jobs in the private sector. The opposition of these two perspectives begs the question of whether there is a unique set of values that differentiates individuals who choose public service careers from those who elect to work in the private sector. Our knowledge of the existence and nature of such sectoral value differences has important implications for public sector organizations. With a workforce that is aging faster than the labor force in general (Lowe 2001), public organizations face the challenge of attracting and retaining young Knowing what types of people are most drawn to careers in public service is an important step in recruiting and retaining new public servants., – people in public-service careers to replace the aging baby boom cohort as it approaches retirement (Lewis and Frank 2002). Knowing what types of people are most drawn to careers in public service is step in recruiting and retaining new public servants. If NPM theorists are correct, then public organizations will need to find ways to compete with the extrinsic rewards offered by private sector organizations in order to recruit talent. If, however, individuals are drawn to the public sector by a motivation to serve the public interest, public organizations will need to focus their recruitment efforts on the challenges and rewards of service for the public good. If this is the case, public organizations will be able to attract highquality employees despite their inability to compete with the economic rewards offered by private sector organizations (Alonso and Lewis 2001). The goal of this study is to investigate whether there are identifiable sectoral differences in the general values, work values, and organizational commitment of employees. This article makes three unique contributions to the literature on value differences across sectors. First, by importing established constructs from the fields of organizational behavior and social psychology, we hope to bring a new perspective to this important topic. To our knowledge, no previous study has investigated sectoral differences in both general values and work values. Second, unlike past studies, which have included widely varied occupations within their definition of public service, we focus exclusively on knowledge workers.’ Limiting the sample to knowledge work occupations helps to control for a number of variables that are related to the type of work one does (e.g., education, income, and general socioeconomic status), which may confound observed value differences. Third, past research has varied widely in its definition of the public sector. Some researchers have limited public sector samples to employees of government agencies (e.g., Alonso and Lewis 2001; Blank 1985; Rainey 1979; Young, Worchel, and Woehr 1998), whereas others have included such professions as education and health care within their definitions of public service (e.g., Karl and Sutton 1998). “his disparity in the definitions of the public sector makes comparison between samples difficult. The present study is unique in that it considers public administrators (i.e., people employed directly by government agencies) separately from employees in theparapublic sector3 (i.e., those employed in the extended public service, including publicly funded education and health care) and investigates differences between these groups. The article begins with an overview of the concepts of interest-general values, work values, and organizational commitment-and a review of past findings concerning sectoral differences on these variables. We then describe the sample and methodology for the present study and discuss our findings. We conclude with a discussion of the implications of our findings and directions for future research. Theoretical Framework Our objective in this article is threefold: First, we seek to determine whether there are significant differences in the general values held by public servants, parapublic-sector employees and private sector employees. This will give us an indication of whether individuals working in the public service hold a value set that differs from their private sector counterparts. Second, we seek to determine whether public sector, parapublic-sector, and private sector employees differ in what they seek from their work. This will provide some indication of whether people are drawn to one sector or the other on the basis of their work values. Third, we seek to determine whether employees in the three sectors differ with respect to their level of commitment to their organizations. Including this attitudinal outcome in the study will allow us to determine whether public sector, parapublic-sector, and private sector organizations vary in their ability to engender feelings of dedication and involvement in their employees. Each of these concepts is explored in greater detail in the following sections. General Values A value can be defined as an “enduring belief that a specific mode of conduct or end state of existence is personally or socially preferable to an opposite or converse mode of conduct or end state of existence” (Rokeach 1973, 5). Values are goals or criteria that we use to determine the desirability of certain actions or motives in our lives. Values may be considered cognitive expressions of human needs, but they also have an affective component, evoking strong emotion when they are contravened (Bilsky and Schwartz 1994; Locke 1976; Rokeach 1973). Unlike attitudes, which are specific to some target object, such as a person, activity, or thing, general values transcend specific contexts (Rokeach 1973). The general nature of values places them at the core of one’s psychological identity (Ball-Rokeach, Rokeach and Grube 1984; Bilsky and Schwartz 1994). Thus, values are an integral component of our understanding of an individual’s motivations (Locke 1991). If there are fundamental psychological differences between public sector, parapublic sector, and private sector employees, we would expect to find these differences in their general value sets. We use the term general values here to distinguish them from the concept of work values, which is discussed later. Although there are many conceptions and typologies of general values, two models and their correspondent measures have gained prominence in academe. The first was developed by Milton Rokeach, whom many consider to be the seminal contributor to the modern study of values. Rokeach (1973) identified two broad categories of values: instrumental values, which are beliefs about the desirability of different modes of conduct, and terminal values, which are beliefs about 606 Public Administration Review 9 JulyIAugust 2006 the desirability of different end goals of existence. He developed the Rokeach Value Survey, a popular value measure that requires respondents to rank order lists of 16 instrumental and 16 terminal values. More recently, Schwartz (1992) developed a model of general values that classifies values on the basis of their motivational content. Building on past research, notably that of Rokeach (1973), Schwartz and colleagues (Schwartz 1992; Schwartz and Bilsky 1987; Schwartz and Sagiv 1995) identify 10 types of values (see table 1), each motivated by a unique combination of human needs. Schwartz’s 10 value types and the instrument that measures them (the Schwartz Value Survey) have gained widespread support among values researchers over the past decade. Sectoral Differences in General Values Since the mid-1980s, increasing attention has been paid to the role of values in the disciplines of business and public administration. However, the roles of values in these two fields have been quite different. Business theorists view values as a means of attaining organizational performance and excellence (e.g., Peters and Waterman 1982). ‘This interest is evidenced by the proliferation of management literature concerning “corporate culture” and “values-based management,” which suggests that professional employees are best governed by a strong set of overarching values rather than by policies, procedures, and structures (Anderson 1997; Peters and Waterman 1982; Schein 1992). Public administration theorists, on the other hand, see values as a means of strengthening the decision making and ethics of public servants and enabling them in their role as stewards of the public interest (e.g., Kernaghan 2000, 2003; Tait 1997; Van Wart 1998). Relatively few studies have examined general value differences between employees in different sectors. One notable study was conducted by Rawls, Ullrich, and Nelson (1975), who assessed psychological differences between business school graduate students electing to Table I Definitions of Value Types Value Type enter the private and nonprofit (including the public) sectors. They found the two groups differed significantly on 8 of 16 values taken from the Rokeach Value Survey. The nonprofit group ranked three values higher than the private sector group: cheerful (light-hearted), forgiving (willingness to pardon others), and helpful (working for the welfare of others). The private sector group ranked five values higher than the public sector group: comfortable life (prosperous life), ambitious (aspiring), clean (neat, tidy), obedient (dutiful), and responsible (dependable). It is somewhat surprising, given these findings, that little subsequent investigation of value differences among sectors has occurred. Given the lack of extant theory and research on general value differences between public, parapublic, and private sector employees, it is difficult to surmise a priori what value differences, if any, might be found. Therefore, we propose the following hypothesis: HI: There will be no significant difference in the general values (achievement, benevolence, conformity, hedonism, power, self-direction, security, stimulation, tradition, and universalism) of employees in the private, public, and parapublic sectors. Work Values Work values can be defined as generalized beliefs about the desirability of certain attributes of work (e.g., pay, autonomy, working conditions), and work-related outcomes (e.g., accomplishment, fulfillment, prestige). Like general values, work values act as the criteria that an individual uses in selecting appropriate work-related behaviors and goals. Past research has suggested that general values and work values are separate but related constructs (Elizur and Sagie 1999; Sagie and Elizur 1996); work values are derived from broader general values (George and Jones 1997; Roe and Ester 1999). Although a wide variety of work value typologies have been proffered, there appears to be a consensus on at Definition Achievement Personal success through demonstrating competence according to social standards Benevolence Preservation and enhancement of the welfare of people with whom one is in frequent personal contact Conformity Restraint of actions, inclinations and impulses likely to upset or harm others and violate social expectations or norms Hedonism Pleasure and sensuous gratification for oneself Power Social status and prestige, control or dominance over people and resources Self-direction Independent thought and action-choosing, creating, and exploring Security Safety, harmony, and stability of society, of relationships and of self Stimulation Excitement, novelty, and challenge in life Tradition Respect, commitment, and acceptance of the customs and ideas that traditional culture or religion provide Universalism Understanding, appreciation, tolerance, and protection for the welfare of all people and for nature Source: Schwartz (1992). A Comparison of Values and Commitment 607 least two fundamental types of work values: intrinsic work values, which pertain to the inherent psychological satisfactions of working, such as interesting work, challenge, variety, and intellectual stimulation, and extrinsic work values, which relate to the material aspects ofwork, such as pay, benefits, and job security (Dagenais 1998; Elizur 1984; Herzberg, Mausner, and Snyderman 1959; Ros, Schwartz, and Surkiss 1999; Zytowski 1970). There is less consensus on other work values, but a number of researchers have suggested the following: social work values, which pertain to relations with coworkers, supervisors, and other people (Elizur 1984; Pryor 1979; Ros, Schwartz, and Surkiss 1999; Super 1970); altruistic work values, which involve the desire to help others and to make a contribution to society (Dawis and Lofquist 1984; Super 1970); and prestige values, which relate to status, influence, and power (Dawis and Lofquist 1984; Pryor 1979; Ros, Schwartz, and Surkiss 1999; Super 1970). Sectoral Differences in Work Values The basic theory concerning values and occupational choice dictates that individuals seek occupations that fit with their individual work values (Dawis and Lofquist 1984; Judge and Bretz 1992; Schwarzweller 1960; Simpson and Simpson 1960; Super 1953). This implies that individuals whose values match those of public-service organizations will elect to work in the public service, whereas those who value what the private sector has to offer will seek employment in that sector. Thus, we would expect to find distinct sets of work values evident in the respective sectors. In this section, we review the theoretical and empirical evidence concerning the relationship between work values and sector of employment. Each of the five types of work values mentioned previously (extrinsic, intrinsic, altruistic, prestige, and social) are discussed in turn. Extrinsic Work Values Two common stereotypes of public sector employees suggest that they are more concerned with job security and less concerned with pay than their private sector counterparts. It is assumed that people who value pay will seek employment in the private sector, which is generally perceived to pay more than the public or parapublic sectors for comparable jobs (Lewis and Frank 2002). Similarly, because public sector employment is considered relatively secure by comparison to the private sector, it is assumed that people who value security and are relatively risk averse will seek work in the public sector (Baldwin 1991; Bellante and Link 1981). Previous research has consistently found that private sector employees do indeed value high salaries significandy more than do public sector employees (Frank and Lewis 2004; Karl and Sutton 1998; Khojasteh 1993; Lewis and Frank 2002; Newstrom, Reif, and Monczka 1976; Rainey 1982; Rawls, Ullrich, and Nelson 1975). However, contrary to the stereotype, previous empirical studies have repeatedly found no sectoral difference in the importance of job security (Frank and Lewis 2004; Karl and Sutton 1998; Lewis and Frank 2002; Rawls, Ullrich, and Nelson 1975). In fact, Newsrrom, Reif, and Monczka (1976) and Khojasteh (1993) found that public servants value security less than private sector employees. Karl and Sutton (1998) argue that the inconsistency of these findings is a function of the labor market conditions in the period during which the research was conducted. They argue that public servants, having faced massive layoffs over the past decade, now place as great an emphasis on job security as do their private sector counterparts. These findings regarding extrinsic work values lead us to the following hypotheses: H12: Private sector employees will place greater importance on pay and benefits than will parapublic or public sector employees. H3: There will be no difference among private sector, public sector, and parapublic-sector employees in the importance placed on job security. Int7insic Work Values Intrinsic work values include aspects related to the nature of one’s work, such as intellectual stimulation, creativity, and challenge. The theory of occupational choice suggests that people who value intrinsic rewards highly will seek employment in the sector that best fulfills that value. Because we have no theoretical reason to presume that one sector is perceived to offer greater intrinsic rewards than the others, we would expect to find limited differences in the intrinsic work values of employees from the various sectors. Unfortunately, relatively few studies have considered sectoral differences in intrinsic work values, and those studies have considered a small range of intrinsic variables. For example, Khojasteh (1993) found that public sector managers place greater value on achievement than do their private sector counterparts. Research has also found that public servants value interesting work more than private sector employees (Frank and Lewis 2004; Karl and Sutton 1998; Lewis and Frank 2002). With the exception of Khojasteh (1993), whose sample included only managers, previous studies have included a wide range of professions from blue-collar to managerial positions. The fact that none of these studies made an attempt to control for variations in work type begs the question of whether the observed differences are attributable to variations in work type among sectors. This point is especially salient with 608 Public Administration Review e July I August 2006 respect to the intrinsic aspects of work, which vary significantly between blue-collar and white-collar jobs. Given our focus on knowledge workers in the present study, we would expect to find relatively little difference between the sectors on the importance placed on intrinsic work values, inviting the following hypothesis: H4: There will be no significant difference in the intrinsic work values (e.g., interesting work, intellectually stimulating work, creativity, challenge, etc.) of employees in the private, public, and parapublic sectors. Altruistic Work Values A key tenet of the notion of public-service motivation (Perry and Wise 1990) is that people who seek work in the public sector are fundamentally motivated to serve the public interest. Given the altruistic focus of parapublic-sector organizations, this tenet can be extended to parapublic employees as well. Past research that has examined altruistic work values has consistently found that public servants value helping others and making a contribution to society more than private sector employees (Frank and Lewis 2004; Karl and Sutton 1998; Lewis and Frank 2002). Therefore, we offer the following hypothesis: H5: Employees in the public and parapublic sectors will place greater importance on making a contribution to society than will private sector employees. Prestige Work Values The only prestige work value to be included in prior research is advancement opportunities. Unfortunately, the results have been contradictory. Frank and Lewis (2004) found that private sector employees value advancement opportunities more than public sector employees, whereas Khojasteh (1993) found that public sector managers value advancement opportunities significantly more than private sector managers. Karl and Sutton (1998) found no significant difference between employees of the two sectors on this value. Because findings with Social Work Values Social work values were examined in only two studies reviewed for this article. Rawls, Ullrich, and Nelson (1975) found that nonprofitsector employees (which include both public sector and parapublic-sector employees) place greater importance on personal relations than did private sector employees. More recently, Khojasteh (1993) found no significant difference between public sector employees and private sector employees with respect to the importance they place on interpersonal relations. Given the lack of theoretical evidence and the contradictory results on social values, we propose the following hypothesis: H7: There will be no significant difference among private sector, public sector, and parapublic-sector employees with respect to the importance placed on social work values (e.g., coworker relations, a fun workplace). Organizational Commitment Organizational commitment relates to the strength of one’s identification with and involvement in a particular organization. It is manifest in (1) a strong belief in and acceptance of the organization’s goals and values; (2) a willingness to exert a considerable amount of effort on behalf of the organization; and (3) a desire to remain within the organization (Porter et al. 1974, 603). Organizational commitment has long been a topic of interest for management researchers, and it has been linked to performance (Mowday, Porter, and Steers 1974; Porter, Crampon, and Smith 1976), lower employee absenteeism (Blau and Boal 1987; Mowday, Porter, and Steers 1974), and lower turnover (Hom, Katerberg, and Hulin 1979; Porter et al. 1974; Steers 1977; Tett and Meyer 1993). In the context of the present study, organizational commitment provides an important link between the values of the individual employee and those of his or her organization. In the case of public and parapublic organizations, which promote values related to public service and the greater good of society, it is particularly imperative that employees have high levels of respect to this value item have been contradictory and there is no theoretical evidence to argue for a directional hypothesis, we propose that no significant differences will be found. H6: There will be no significant difference among private sector, public sector, and parapublic-sector employees with respect to the importance placed on prestige values (e.g., In the case of public and parapublic organizations, which promote values related to public service and the greater good of society, it is particularly imperative that employees have high levels of commitment to these values if organizations are to fulfill their purposes. commitment to these values if organizations are to fulfill their purposes. However, the few studies that have directly compared levels of organizational commitment between public sector employees and private sector employees have consistently shown private sector employees to demonstrate higher levels of organizational commitment than public servants (e.g., Buchanan 1974; Liou and prestigious work, influence over organizational outcomes, opportunities for advancement, etc.). Nyhan 1994; Perry and Rainey 1988; Rainey 1979; Zeffine 1994). It has been argued that this relative A Comb arison of Values and Commitment 609 lack of organizational commitment in the public service presents a crisis and therefore merits further attention (Liou and Nyhan 1994; Staats 1988). Zeffane (1994) argues that organizational commitment is largely affected by management style. He found significant differences between the public and private sectors in the degree to which managers were perceived as emphasizing two management styles: (1) flexibility and adaptation (greater in the private sector) and (2) rules and regulations (greater in the public sector). Both of these management styles had significant positive associations with organizational commitment. The net effect of these relationships was greater organizational commitment in the private sector. Buchanan (1974) argues that the relative lack of organizational commitment among public sector managers is the result of the weak performance-outcome link in public management. Because the goals of public agencies are broad and diffuse rather than specific and tangible, managers in the public sector are less able to gauge the value of their efforts in meeting overall organizational requirements. Because public managers must often focus on process rather than outcomes, they identify less with the broader goals of their organization and more with their immediate work. This leads them to be less committed to the organization’s higher-level goals. The extant theoretical and empirical evidence concerning sectoral differences in organizational commitment leads us to the following hypothesis: H8: Organizational commitment will be higher for private sector employees than for public sector and parapublic-sector employees. Methodology Participants The research participants included 549 knowledge workers employed in large Canadian organizations (500 or more employees). The sample was restricted to knowledge workers to control for the effects of the type of work performed by the participants (i.e., blue-collar jobs versus white-collar jobs). Participants were classified as knowledge workers and included in the sample if they identified themselves as employed in managerial, professional, technical, or administrative positions. Clerical, manual, retail, production, and other workers were not included in the sample. Participants were categorized into sectors based on the nature of the organizations for which they worked. Private sector organizations included firms from a number of industries, including retail, financial 610 Public Administration Review * JulyIAugust 2006 services, telecommunications, information technology, hospitality, transportation, and insurance. Previous research comparing the private and public sectors has varied widely in its definition of public sector employees. To avoid confusion related to the breadth of our definition of the public service, we split the broader public service into two categories: public administration (referred to here as public servants) and the parapublic sector. Public servants were directly employed by departments and agencies of federal, provincial, and municipal governments. Parapublic organizations included such employers as universities, schools, hospitals, and long-term care facilities, all nonprofit organizations that receive government funding but are not directly operated by any level of government. Public servants represented 42 percent of the total sample, private sector employees composed 22 percent, and parapublic-sector employees made up 36 percent. Past research has shown that one’s preference for public-service work is affected by age, educational level, and gender (e.g., Blank 1985; Lewis and Frank 2002). To control for the confounding effects of these variables, all three were included as covariates in the analyses performed in this study. Educational level was coded as a dichotomous variable. Participants who had at least a university or college degree were placed in one category, and those with less education were placed in another. Gender was coded as a categorical variable, and age was included as a continuous variable. The sample was 35 percent male and 65 percent female. Measures General Values The general values of respondents were measured using the Schwartz Value Survey, a popular instrument that coincides with Schwartz’s (1992) model of 10 general values. Although the original version of the survey (Schwartz 1992) contained 56 items, a more parsimonious 44-item version, suggested by Schwartz and Sagiv (1995), was used here. Respondents were asked to rate the importance of 44 value items “as a guiding principle in my life” on a nine-point scale as follows: -1 = opposed to my values; 0 = not important; 1, 2 = unlabeled; 3 = important; 4, 5 = unlabeled; 6 = very important; 7 = of supreme importance. In accordance with Schwartz’s (1992) instructions, responses for the 44 individual items were used to generate 10 value-type scores corresponding to 10 ten value types of the Schwartz model. The value-type scores were calculated as the summed average of the scores for items corresponding to each value type as given by Schwartz. Work Values Work values were measured using items from the comprehensive work value survey developed by Lyons (2003). The items in this measure represent a wide variety of work aspects selected through a review of established work value measures. Lyons factor analyzed the measure’s 31 items and found that 22 items loaded on the following five factors: intrinsic work values (nine items), extrinsic work values (three items), prestige work values (five items), altruistic work values (three items), and social work values (two items). These 22 items, grouped according to the factor structure obtained by Lyons, were utilized in the present study. In order to compare our results to previously reported findings, we analyzed the individual items rather than creating indexes for each of the five work value types. A factor analysis of these 22 items produced minor changes from the factor structure obtained previously. Five dusters of variables were found. In particular, there were six intrinsic work value items (cluster 1), four items related to prestige aspects of the job (cluster 2), three items related to altruistic work values (cluster 3), three items related to the extrinsic aspects of work (cluster 4), and two items pertaining to the social environment of work (cluster 5). Four items (work variety, accomplishment, fulfilling work, and travel opportunities) did not load as expected and were not included in the five clusters of work values that were analyzed. Respondents were asked to indicate the value they placed on each item by rating its importance “with respect to work” according to the same rating scale employed in the Schwartz Value Survey. Organizational Commitment Respondents’ levels of organizational commitment were measured using the nine-item version of Porter et al.’s (1974) Organizational Commitment Questionnaire. This instrument contains a number of statements related to individuals’ attitudinal commitment to their organization. Respondents were asked to rate their agreement with each statement on a five-point Likert-type scale with the following anchors: I = strongly disagree, 3 = neutral, 5 = strongly agree. ‘The nine items were then averaged to calculate an overall organizational commitment score. The reliability coefficient (Cronbach’s alpha) was .90, which is consistent with previous published findings (Cook et al. 1981). Analyses Sectoral differences in general values and work values were investigated using multivariate analysis of covariance (MANCOVA). For general -values, a one-way (sector) MANCOVA was computed with the 10 value types measured by the Schwartz Value Survey as the dependent variables and age, gender, and level of education as covariates. For work values, separate between-subject MANCOVAs were computed for each of the five work value variable clusters, with the relevant work value items as the dependent variables and age, gender, and educational level as covariates. Type III sum of squares was used to accommodate the imbalance in cell counts across sectors (Tabachnick and Fidell 2001). Where significant sectoral differences were observed on the omnibus tests, protected F-tests were employed to examine sectoral differences on the individual general value and work value variables. For individual variables that showed significant sectoral differences, post hoc Bonferroni tests were conducted to examine pair-wise differences between the private, public, and parapublic-sector groups. We tested for sectoral differences in organizational commitment using a one-way (sector) analysis of covariance (ANCOVA), with organizational commitment as the dependent variable and gender, age, and educational level as the covariates. Results General Values The mean general value scores and standard errors for each of the three sectoral groups are given in table 2. The omnibus F-test for the general values MANCOVA revealed that general values did not differ significantly by sector, F(20, 992) = 1.56, = .054 for Wilks’s lambda, controlling for the effects of age, educational level, and gender. Thus, as hypothesis 1 suggested, there were no significant sectoral differences in general values. Each of the covariates had a significant effect on values (p < .001 for education and age andp < .05 for gender). Work Values The results of the five work value MANCOVAs and the follow-up F-tests are presented in table3, along with the mean work value importance scores and the standard errors for the three sectors. Extrinsic Work Values MANCOVA 1 involved three extrinsic work value items: salary, job security, and benefits. The multivariate test showed no significant sectoral differences in this set of items, F(6, 974)= 2.05, p = .057 for Wilks’s lambda). Thus, contrary to hypotheses 2 and 3, no significant differences were observed among sectors with respect to the importance of job security, pay, or benefits. Intrinsic Work Values MANCOVA 2 incorporated six intrinsic work value items: intellectually stimulating work, challenging work, interesting work, continuous learning at work, creativity in work, and using one’s abilities at work. The omnibus F-test indicated that this set of items varied significantly between sectors, F(12, 960) = 2.53, p < .003 for Wilks’s lambda. The control variables of gender and age were significant while educational level was not. Thus, contrary to hypothesis 4, significant differences were A Comparison of Values and Commitment 611 Table 2 Mean General Value Scores and Standard Errors Public sector Private sector Parapublic sector Mean Std. Error Mean Std. Error Mean Std. Error Achievement 4.60 .08 4.52 .11 4.52 .09 Benevolence 5.15 .07 5.14 .09 5.22 .07 Conformity 4.51 .08 4.57 .11 4.81 .09 Hedonism 4.53 .09 4.41 .12 4.38 .10 Power 2.08 .08 2.19 .11 1.98 .09 Security 4.64 .07 4.72 .10 4.85 .08 Self-direction 5.11 .07 4.89 .09 4.97 .07 Stimulation 3.87 .09 3.61 .13 3.73 .10 Tradition 3.20 .09 3.48 .12 3.51 .10 Universalism 4.91 .07 4.79 .10 5.04 .08 observed among the three sectors with respect to intrinsic work values. (mean = 5.10). Challenging work was also more important to public servants (mean = 5.36) than to parapublic-sector employees (mean = 5.00) Follow-up F-tests for the six individual intrinsic work values revealed significant sectoral differences (p < .05) Altruistic Work Values The dependent variables in on two of the six intrinsic work values: intellectually MANCOVA 3 were the altruistic work value items: stimulating work and challenging work. Post hoc work that makes a contribution to one’s society, work tests revealed that public sector employees placed that is consistent with one’s moral values, and fairness more importance on intellectually stimulating work in the application of policies and programs. The (mean = 5.44) than did parapublic-sector employees F-test for this analysis revealed significant sectoral Table 3 Mean Work Value Scores, Standard Errors, and MANCOVA and ANCOVA Results Public Sector Private Sector Parapublic Sector Sector Item Mean Std. Error Mean Std. Error Mean Std. Error Sig.(F) MANCOVA 1: Extrinsic work values .06 Having benefits (e.g., vacation pay, health/dental insurance, 5.29 .10 5.60 .14 5.40 .11 – pension plan, etc.) that meet your personal needs Having the assurance of job security 5.13 .11 5.02 .16 5.28 .12 – Doing work that affords you a good salary 5.17 .09 5.29 .13 5.04 .10 – MANCOVA 2: Intrinsic work values .00″* Doing work that is intellectually stimulating 5.44 .09 5.12 .13 5.10 .10 .02* Working on tasks and projects that challenge your abilities 5.36 .09 5.29 .13 5.00 .10 .02* Doing work that you find interesting, exciting, and engaging 5.52 .08 5.28 .12 5.31 .09 .12 Having the opportunity to continuously learn and develop 5.21 .09 5.30 .13 5.23 .11 .86 new knowledge and skills Doing work that involves creativity and original thought 4.62 .11 4.80 .16 4.44 .13 .20 Doing work that allows you to use the abilities you have 4.96 .10 4.57 .14 4.5 .11 .07 developed through your education and experience MANCOVA 3: Altruistic work values .00** Doing work that makes a helpful contribution to society; 5.00 .10 4.50 .14 5.46 .11 .00″** makes a difference Doing work that is consistent with your moral values 5.48 .10 5.46 .14 5.52 .11 .92 Working in a setting where policies and programs are 5.45 .09 5.37 .13 5.47 .10 .83 administered with fairness and impartiality MANCOVA 4: Prestige work values .00** Having the authority to organize and direct the 2.67 .13 2.98 .18 2.70 .15 .35 work of others Doing work that is prestigious and regarded highly by others 2.86 .13 3.46 .18 2.96 .15 .02* Having the ability to influence organizational outcomes 4.15 .12 4.10 .17 3.77 .13 .10 Having the opportunity for advancement in your career 4.41 .12 4,82 .17 3.86 .17 .00*** MANCOVA 5: Social work values .35 Working in an environment that is lively and fun 4.31 .11 4.60 .16 4.19 .13 – Working with agreeable and friendly coworkers with 4.52 .11 4.80 .16 4.53 .12 – whom you could form friendships *p <.05; **p <.01; ***p < .001. 612 Public Administration Review a JulyIAugust 2006 differences, F(6, 9 7 2 ) = 5.07,,p < .001 for Wilks’s lambda. Gender and age were significantly associated with these two items (p < .0 1), but level of education was not. Follow-up F-tests revealed significant sectoral differences in the importance of work that makes a contribution to society (p < .001). Post hoc tests showed that, as hypothesis 5 suggested, work that makes a contribution to society was valued more by both parapublic-sector employees (mean = 5.46) and by public servants (mean = 5.00) than by private sector employees (mean = 4.50). Interestingly, parapublicsector employees valued making a contribution to society significantly more than public servants, a finding that was not proposed a priori. Prestige Work Values MANCOVA 4 included the four prestige work value items: authority, prestigious work, influence, and opportunities for advancement. The omnibus F-test revealed significant sectoral differences on this set of work value items, F(8, 966) =4.08, p < .001 for Wilks’s lambda. The effect of age on these items was significant (p < .05), but the effects of gender and education were not. Thus, contrary to hypothesis 6, there were significant differences among the sectors with respect to prestige values. Follow-up -F-tests revealed significant sectoral differences in the importance of two of the four prestige work values: prestigious work (p < .05) and opportunities for advancement (p < .001). Post hoc tests revealed that parapublic employees placed significantly less importance on advancement (mean = 3.86) than did both public servants (mean = 4.41) and private sector employees (mean = 4.82). Prestigious work was significantly more important to private sector employees (mean = 3.46) than to public servants (mean = 2.86). Social Work Values MANCOVA 5 involved two social work value items: friendly coworkers and a fun work environment. The multivariate F-test showed that employees in the various sectors did not vary significantly on these items, F(4, 978) = 1.12, p =.345 for Wilks’s lambda. Thus, as hypothesis 7 suggested, there were no significant differences among the sectors with respect to social work values. Organizational Commitment The results of the organizational commitment ANCOVA indicated a significant sectoral difference (p < .000). Gender was the only significant covariate (p < .05). Post hoc Bonferroni tests revealed that private sector employees reported significantly higher levels of organizational commitment (mean = 3.79) than did either public servants (mean = 3.41,p < .01) or parapublic employees (mean = 3.49, p < .01). The difference between public servants and parapublic employees was not significant. Thus, as hypothesis 8 suggested, both parapublic employees and public servants were less committed to their organizations than were private sector employees. Discussion This study asked whether there is a unique value set that differentiates public sector, parapublic-sector, and private sector employees. On the whole, the results of this study suggest that the answer is a qualified no, as there were limited overall differences in the values of employees from the various sectors. No significant sector differences were observed in any of the 10 value types proposed by Schwartz’s (1992) value model when the effects of age, gender, and educational level were controlled for. Because all three of these demographic variables were significantly related to general values, it appears that each better differentiated people on the basis of their values than did sector. This finding suggests there is no fundamental value predisposition that guides people into careers in one sector or another. Perhaps the most notable finding is that employees in the public and parapublic sectors are no more altruistic and no less self-interested than employees in the private sector. This calls into question the argument for public sector motivation, which suggests that public and parapublic employees are following a unique calling that is not evident in private sector employees. The findings with respect to work values are somewhat more revealing. Even after the impacts of age, gender, and educational level were controlled for, employees from the various sectors differed significantly in the importance they placed on 5 of the 18 work value items that were analyzed. As expected, there were sectoral differences in the importance of making a contribution to society. Employees of the broadly defined public service (i.e., both the public and parapublic sectors) valued contributing to society more than employees in the private sector. The importance of making a contribution was significantly higher for parapublic employees than for public servants. Given the types of “nurturing” jobs found in the parapublic sector (i.e., education and health care), this finding is not surprising. The findings regarding contribution to society lend some support to the notion of a public-service motivation, as they suggest that people motivated by altruistic goals are more likely to be found in public sector or parapublic sector jobs. It is impossible in a cross-sectional study such as this one to attribute causality to these relationships. However, it seems most plausible that individuals who place importance on contributing to society are drawn to public and parapublic service. The reverse implies that individuals who opt to work in the public and parapublic sectors are somehow socialized within their organizations A Comparison of Values and Commitment 613 to place importance on these values. The latter explanation is possible, but it is less consistent with the theoretical link between work values and stable general values. Private sector employees were found to value prestigious work more than employees in either of the other sectors. Our finding suggests that prestige-seeking individuals are drawn more to the private sector than to the parapublic and public sectors. ‘This may reflect a decline in citizen’s esteem for public institutions over recent Private sector decades (Volcker 1989; Zussman found to value 1997), which, in turn, may be drawing prestige seekers to the more than emj private sector. Alternatively, in of the ott light of the poor reputation of public institutions, public servants may rationalize their employment in the public sector by downplaying the importance of prestige in reporting their work values. Clearly, further research is needed to explore the nature of this value difference. Previous research by other authors did not look specifically at the work values of parapublic employees. The results of the present study indicate some interesting differences that may have been averaged out in past research that combined these sectors as a single group. Specifically, we found that employees in the parapublic sector placed significantly less importance on advancement than did private or public employees. This coincides with Frank and Lewis’s (2004) finding that private sector employees value advancement opportunities more than public servants. One explanation for this finding lies in the nature of parapublic professions: For many parapublic employees, such as nurses and teachers, the opportunities for upward career advancement are relatively limited. It is possible that people who greatly value advancement opt out of parapublic professions. It is also possible that people in parapublic professions rationalize their lack of advancement opportunities by downplaying their importance. Qualitative data are required to gain a fuller understanding of this phenomenon. We also found that parapublic employees placed significantly less importance on intellectually stimulating work and challenging work than did employees in the public sector. The overall pattern suggested by these findings is that parapublic employees appear to be less interested in at least some of the intrinsic aspects of work than are their public-service colleagues. Given the findings regarding the importance of making a contribution to society, it appears that the true reward for employees in the parapublic sector is the altruistic nature of their work rather than the intrinsic benefits of the work itself. Again, qualitative research is needed to better explain the nature of these differences. e pl pl Despite the interesting sectoral differences discussed previously, there were no significant sectoral differences in 13 of the 18 work value items we analyzed. A number of these nonsignificant findings are notable in light of past research. There was no significant difference in the importance of job security among employees from the three sectors. This corroborates past empirical findings and lends further opposition to the stereotype of public servants as obsessed with job security. -nployees were Contrary to past findings, there restigious work were no significant sectoral differences in the importance of oyees in either. salary. It is possible that this •r sectors. contradiction of past findings is the result of limiting our sample to knowledge workers, who include the highest-paid employees of each sector. For such workers, it is conceivable, if not likely, that pay is of relatively little immediate concern, as they are relatively well paid. Had we included clerical and manual laborers, we may have found greater differences. Finally, past research has consistently shown that private sector employees demonstrate greater commitment to their organizations than do public sector employees. Our results support this finding: Both public sector and parapublic-sector employees reported significantly lower levels of organizational commitment than did private sector employees. This suggests that relative to private sector employees, fewer public sector and parapublic-sector employees feel that their personal goals and values are compatible with those of their organizations. Zeffane (1994) noted a lack of clear theoretical explanation for the lower relative organizational commitment among public sector employees. One explanation lies in potential variations in what constitutes “the organization.” Although organizational commitment in the private sector is easily construed as commitment to one’s “firm,” the target of one’s commitment is less obvious in the public sector. It is possible that public servants direct their commitment at the institution of government as a whole rather than the specific department or agency in which they work. It is reasonable to assume that a public servant may be dedicated to serving the public without being particularly dedicated to any specific public organization. Future research on commitment in the public service should broaden the concept to include commitment to public service in general. “Three limitations of the present study merit consideration in future research. First, although this study is unique in looking at the parapublic sector as an independent group, we did not consider the values of other types of nonprofit employees, such as those 614 Public Administration Review a JulylAugust 2006 working in voluntary organizations, charities, and other nongovernmental organizations. Our intent was to give separate consideration to employees working in the extended public sector, apart from those employed directly by government. It has been argued that the nonprofit sector (the so-called third sector) will become increasingly important in the modern economy as governments continue to focus on fiscal restraint and educated employees continue to seek meaningful work that affects society (Rifldn 1995). Therefore, it is important that this sector be given consideration in future research. Perhaps the inclusion of this sector in research of this type will reveal a fuller picture of value patterns that correspond to the degree of publicness of various organizations. Second, our study included only Canadian employees. Though there are undoubtedly similarities between Canadians and citizens of other developed countries, the nature of public and private organizations and the relationships between the sectors is unique to a given national context. Although Canadians may be expected to share a wide variety of values with their American neighbors, The fact that the nature of Canada’s Westminster-style parliamentary system were no more and the ubiquitous public fund- less self-intere ing of parapublic organizations in sector emplc Canada may affect the sectoral come as a dis differences observed here. This research merits repetition in other proponents national contexts to better estab- public-servi lish the nature of sectoral value differences. p a St a] 0 ,e Third, the sample for this study was restricted to knowledge workers. This allowed us to control for a variety of confounding influences, such as socioeconomic status and occupational choice (apart from the choice of sector). However, this restriction diminishes the generalizability of the results to other types of occupations. Because both the public and private sectors include employees from a wide variety of occupations, ranging from unskilled labor to knowledge work, it would be useful to examine the relationship between work type and individual values and commitment in the context of sectoral differences. It might be asked, for example, whether the work value differences observed in this study are evident between blue-collar workers in other sectors as well. Although this study identified differences in the work values of employees in different sectors, qualitative research is needed to fully interpret and explain the nature and causes of these differences. Of particular interest is the issue of causality. It is important to know whether work value differences among employees in different sectors are the result of occupational choices based on existing work values or the product of socialization and rationalization once the occupational choice is made. Furthermore, Lewis and Frank (2002) note that holding a public-service job does not directly constitute an endorsement of public-service values and rewards, as one’s preference for a job in a particular sector does not ensure that one is available. It is likely that there are individuals in both the public and private sectors who would prefer to be employed in the opposite sector if the opportunity were available to them. We do not know to what extent this mismatch is present in the workforce. The measurement of participants’ preferences for work in one sector or another may be a better determinant of sectoral differences than the actual sector of employment. Conclusion On the whole, the results of this study suggest there are limited differences between private sector employees and public sector employees with respect to the three constructs measured. The finding of no systematic differences in the general ublic servants values of public sector employees and private sector employees Itruistic and no suggests that, at the most general ed than private psychological level, employees in ,ees will likely both sectors are highly similar ppointment to when demographic differences f the “unique are considered. The fact that motivation,” public servants were no more altruistic and no less selfinterested than private sector employees will likely come as a disappointment to proponents of the “unique public-service motivation.” In their recruitment efforts, public sector organizations cannot rest on the assumption that public service attracts a different breed of person. The work value findings suggest that employees in the parapublic sector place less emphasis on advancement opportunities and intrinsic work vales than do public servants and private sector employees. Instead, they value the contribution that their work makes to society and the personal fulfillment they derive from their work. Health care and educational institutions would do well to emphasize the altruistic motives of this type of work as a means of attracting and retaining people in these professions. This is an important finding, as public administration research often lumps parapublic employees within the spectrum of public service. Our results provide a strong indication that this sector merits its own attention, particularly where work values are concerned. It must not be presumed that public-service motivation applies equally to employees across the parapublic and public sectors. This study added further support to the common finding that private sector employees are more A Comparison of Values and Commitment 615 committed to their organizations than public sector employees. Our results further indicate that parapublic-sector employees are less committed to their organizations than private sector employees. The implication of this finding is that employees in the public and parapublic sectors do not identify as strongly with the values and goals of their organizations as do private sector employees. This is a serious issue that public and parapublic organizations must address. As noted previously, organizational commitment is antecedent of a number of other important work outcomes, such as performance, reduced absenteeism, and employee turnover. Furthermore, committed employees are critical to the fulfillment of organizational mandates, particularly in the realm of knowledge work, where employees are not easily interchangeable. The trend toward values-based management necessitates, more than ever, a workforce that is motivated and committed to organizational goals. Romzek (1990) argues that government organizations are at risk of dropping below a critical threshold of dedicated employees. The results of this study punctuate her warning.