Explaining Public Service Motivation

Abstract

Both leadership and public service motivation (PSM) are important issues in contemporary public administration. By connecting these issues, the aim of this article is to assess the impact of one particular type of transformational leadership behavior—promoting public values—on PSM development. Two hypotheses were tested with a dataset of 3,506 state civil servants in Belgium. The analysis shows that a positive relation exists and that it is moderated by a set of basic psychological needs (autonomy and competence), whereby it demonstrates the validity of an institutional theory of PSM.

Keywords

public service motivation supervisor basic psychological needs satisfaction self-determination theory public values

Introduction

Public service motivation (PSM) or the motivation to “do good for others and shape the well-being of society” (Perry & Hondeghem, 2008, p. 3) has quickly grown into one of the core concepts of public administration and public management research, as it has been related to many different outcomes that are considered to beneficial for public-sector organizations (see Perry & Hondeghem, 2008 for an overview). However, given its important contemporary status, little empirical research has been devoted to exploring the institutional or organizational origins of PSM, despite empirical and theoretical evidence pointing toward such a relationship (Moynihan & Pandey, 2007; Perry, 1997; Perry & Vandenabeele, 2008; Vandenabeele, 2011). According to Pandey and Stazyk (2008), “research has yet to elaborate upon [. . .] pathways for antecedent [. . .] of public service motivation.” Besides, the lack of an institutional focus, another issue that has remained underdeveloped, is relating PSM to other motivational perspectives. Although Perry and Vandenabeele (2008) relate PSM to particular theories of self-regulation, many of the proposed pathways have not been explored so far. As most research has focused on Person–Environment fit research (Bright, 2007; Bright, 2008; Christensen & Wright, 2011) or goal-setting theory (Wright, 2007), other theories of self-regulation have not been studied yet.

This article would like to further this research by investigating the potential institutional origins of the concept and at the same time linking it to the self-determination theory (SDT) of self-regulation (Deci & Ryan, 2004). It would like to provide an answer to these questions by studying the interesting question of the relationship between PSM and leadership behavior, and the role basic psychological needs satisfaction plays in relationship. Although there has been substantially less attention for leadership in the public sector compared with the private sector (Terry, 2003), it is nevertheless an important issue, both with academics and practitioners (Van Wart, 2003). In particular, the focus will be on the relationship between the values of direct supervisors and the PSM levels of their subordinates, and to what extent this relationship is moderated by basic psychological needs satisfaction. This perspective is an illustration of the (often informal) process of institutional socialization, being not only an interesting academic question but also a highly relevant practical matter, as leadership behavior and motivation are assumed to be some of the major variables in determining organizational effectiveness (Rainey & Steinbauer, 1999).

First, this article will address some theoretical issues concerning both independent variables—leadership and basic needs satisfaction—and the dependent variable, PSM. Second, the data and methodology are further discussed, and particular attention is devoted to the statistical analysis. Third, after presenting the results of the analyses, these are discussed into considerable depth and some general conclusions are formulated.

Theoretical Overview

This theoretical framework provides an overview of the literature concerning the important concepts in this article. First, various types of leadership are discussed and assessed on their motivational merits, in particular regarding SDT. Second, the relationship with PSM is discussed, and a set of hypotheses based on this review will be presented.

Transformational Leadership in the Public Sector

In current management research, leadership has many guises. However, two main types can be distinguished. Until the late 1970s, leadership was primarily focused on transactional leadership (Rainey, 1997). This type of research was “transactional” in the sense that “leaders approach followers with an eye to exchanging” (Burns, 1978, p. 4). Afterward, the focus shifted toward transformational leadership (Conger & Hunt, 1999). This type of leadership research aimed at a different kind of leader. Transformational leadership results in “shifts in the beliefs, the needs, and the values of followers” (Kuhnert & Lewis, 1987, p. 648).

In this article, the analysis will be limited to transformational leadership as the aim is to investigate the impact of leadership on a value-laden concept as PSM. In an organizational context, this is one of the aims of transformational leadership. By influencing the self-concept of organizational members through leader behavior, the goal is to obtain social identification and to internalize certain values within the organizational members (Shamir, House, & Arthur, 1993), which in turn can motivate followers. Bass (1999) also distinguishes four elements in transformational leadership—individualized consideration, idealized influence, intellectual stimulation, and inspirational motivation—and particular inspirational motivation is a connection to PSM. Next to this, transformational leadership also strengthens moral behavior of institutional members (Dvir, Eden, Avolio, & Shamir, 2002), which demonstrates its particular aptitude for PSM-related research.

However, within the domain of public administration, skepticism prevailed (Van Wart, 2003). Some voiced their doubts on whether transformational leadership was at all suited for the public sector. After all, certain conditions influence both the emergence and the effectiveness of transformational leadership, and these conditions not always fit a public environment, dramatically decreasing the emergence of charismatic leadership in a bureaucracy (Javidan & Waldman, 2003). Others also state that, as the organizational features or a public-sector organization substitute for leadership (Kerr & Jermier, 1978), leadership is less influential (or needed) in public-sector environment (for an overview, see Hooijberg & Choi, 2001). More recently, transformational leadership has nevertheless made its way into the field of public administration as a means to establish a value-based management in the public sector (Denhardt & Campbell, 2006; Paarlberg & Perry, 2007; Trottier, Van Wart, & Wang, 2008).

A last issue to which particular attention needs to be devoted is that of transformational leadership, in the sense of transforming an employee’s identity, is not limited to the highest levels of the organization. Where transactional leaders are mainly presented as small group, “direct supervisor”–style leaders, transformational leaders are often conceived as higher up the hierarchy, “top executive”–style leaders (Van Wart, 2003). However, numerous examples have demonstrated that also direct supervisors, located much lower in the organizational hierarchy, have a substantial impact on someone’s work-related identity. Supervisors are used as an intra-organizational referent, from which essential value-laden information can be obtained (Filstad, 2004; Settoon & Adkins, 1997). Therefore, direct supervisors make as much sense as socializing agents as executive leaders do in representing a set of institutionalized values.

Transformational Leadership and PSM: SDT Enters the Equation

The idea of public servants who have a drive to contribute to the general interest has been around for ages. It can be traced back to Aristotle and Plato and other historic writers who have dealt with it in their works (Horton, 2008). In the 20th century, authors have also found this concept appealing to describe (at least some of) the motivations of present-day civil servants (Chapman, 1988; Downs, 1967; Mosher, 1968). However, it was not until Perry and Wise (1990) defined PSM as “an individual’s predisposition to respond to motives grounded primarily or uniquely in public institutions” (p. 368), that it became a more formally established concept in its own right.

Although little empirical work has been done on the institutional antecedents of PSM (Pandey & Stazyk, 2008), the available empirical (Moynihan & Pandey, 2007; Vandenabeele, 2011) and theoretical work (Perry, 2000; Perry & Vandenabeele, 2008; Vandenabeele, 2007) suggests that institutions play an important role in its development. Institutions—in its singular form defined as “a formal or informal, structural, societal or political phenomenon that transcends the individual level, that is based on more or less common values, has a certain degree of stability and influences behavior” (Peters, 2000, p. 18)—are held responsible for the transmission of values from the institution, a structured and value-based interaction pattern, in which they are embedded to the individual’s identity as a member of that institution.

Elaborating on this idea, it can be combined with the concepts of the SDT by Deci and Ryan (2004). SDT distinguished between various types of controlled motivation in which an external source forces an individual to behave in a certain way, and autonomous motivation in which one voluntarily—based on extrinsic or intrinsic sources—behaves in a certain fashion. PSM can be considered as an autonomous type of identity regulation that is based on internalized values that are the foundation of an institution and its related identity. This type of self-regulation corresponds to the institutional logic of appropriateness (March & Olsen, 1989; March & Olsen, 1995), where individuals act in accordance to institutional values because they have internalized these (Berger & Luckmann, 1966; Scott, 2001). In terms of PSM, people will be public service–motivated because they have internalized public service values. According to Ryan and Deci (2005), this process of internalization (or the socialization of an autonomous regulated identity) is enhanced by the satisfaction of the individual basic psychological needs of autonomy, competence, and relatedness. These are assumed to be present within each individual, and they are considered to be the foundation for individuals’ growth-oriented movement and the process of internalization. To the extent that these needs are satisfied, an identity (and the value(s) it is based on) will be internalized better, resulting in a more autonomous regulation of the identity (Ryan & Deci, 2005).

The inspirational motivation part of transformational leadership (or socialization by leaders) can help to illustrate this institutional theory of PSM. Moreover, it has, on a theoretical basis, been linked to transformational leadership that explicitly relates leadership, PSM, and SDT (Paarlberg & Lavigna, 2010) The values promoted by the supervisor can be considered to be an instantiation of the institutional values. Scott (2001) refers to the influence of symbolic or relational systems (in this case, the social beliefs and the norms promoted by the supervisor) as “carriers of institutions” to explain agency within institutions and provide meaning to the actions.

This results in a set of hypotheses concerning the impact of leadership promotion of public service values on the level of individual PSM (see Figure 1).

Figure 1.

Graphical representation of the hypotheses.

First, transformational leadership can be considered as a way to promote and socialize institutional values. Therefore, Hypothesis 1 (H1) is formulated thus:

H1: Promotion of institutional public service values by supervisors influences positively the individual level of PSM.

This hypothesis is further refined in Hypothesis 2 (H2) that adds another institutional element as a moderator and is thus formulated as follows:

H2: The relationship between the leadership promotion of public service values and PSM is positively moderated by the basic psychological needs satisfaction within the organization.

Method

The data on which the analysis in this article is based are of a quantitative nature. Therefore, a short description of the data collection and the sample, as well as a discussion of the measurement instruments and the statistical techniques is provided.

Sample and Data

The data on which this article is based were collected in a survey among Flemish state civil servants of the central ministries and some associated agencies in spring 2006.¹

The survey itself was an Internet-based e-mail survey.² As all employees of state government (even the lowest levels) have access to the Internet, the risk of coverage error was reduced. A drawback is that web surveys return in general a lower response rate compared with mail surveys (Crawford, Couper, & Lamias, 2001). Couper (2000) attributes this to the lack of tried and tested techniques for increasing response, the discouraging impact they have on respondents who are less familiar with the Internet or confidentiality issues raised by electronic communication.

To increase absolute responses and response rates, a number of strategies were applied. First, the population was sampled based on census (entire population), increasing the sample to the maximum available. Second, both the technology and the design of the survey were chosen to secure maximum technological and psychological accessibility, reducing the perception of burden (Crawford et al., 2001). The consent of the chief executive officer, a voucher lottery, anonymity preservation, maximum visibility of the researchers, and a low technological threshold were applied to achieve a maximal sample size. Third, one week after the first e-mail, a reminder was sent out.

When reviewing the data quality, suspicious entries (doubles, inconsistent or monotonous response patterns) were deleted and 3,506 usable responses were retained. Response rates for the organizations varied from 21% to as much as 90%.³ The bulk of the data (over 90%) was collected with civil servants of the core ministries (which were at the time the largest group of civil servants in our surveyed population). The average respondent was 42 years old at the time of the survey and had been in service for just over 11 years. Both sexes are more or less equally represented in the survey (52% being male). When looking at the grades (Grade A requires a master’s degree or equivalent, Grade B requires a bachelor’s degree, Grade C requires high school education, and Grade D requires no education), most of the respondents belonged to Grade A, while the smallest group belonged to Grade D. This could be evidence of a bias toward Grade A, especially because higher level employees tend to participate more in web surveys (Couper, Blair, & Triplett, 1999). However, no accurate population data are available, and therefore no assessment of representativeness can be made. Nevertheless, this possible bias would only result in biased conclusions if grade itself would be a moderator in any of the proposed relationships. As this is not the case (analysis demonstrated no significant moderator effects), the possible lack of representativeness is no threat to this study’s conclusions. This matches with Brehm’s (1993) findings that correcting for demographic biases demonstrates very little impact in the case of correlational analyses.

Dependent and Independent Variables in the Analysis

The concept of PSM was first described by Perry (1996), who identified four dimensions of PSM, “attraction to politics and policy making,” “public interest,” “compassion,” and “self-sacrifice.” Later research (Bright, 2007; Camilleri, 2006, 2007; Coursey & Pandey, 2007; Coursey, Perry, Brudney, & Littlepage, 2008; Vandenabeele, 2008) more or less confirmed this factorial structure, although in some cases, subtle or more marked differences could be noted. This article will use the typology and instrument by Vandenabeele (2008) consisting of the dimensions “attraction to politics,” “public interest,” “compassion,” “self-sacrifice,” and “democratic governance.” (Table 1)⁴

Table 1.

Public Service Motivation Measurement Scale (α = .79).

I do not care about politicians

“Politics” is a dirty word

To me, before anything, good civilians should think of society

Serving the public interest is an important drive in my daily life (at work or outside work)

I voluntarily and unselfishly contribute to my community

To me, serving the public interest is more important than helping individual persons

Much of what I do is for a cause bigger than myself

I am prepared to make enormous sacrifices for the good of society

I feel people should give back to society more than they get from it

Making a difference in society means more to me than personal achievements

To me, helping people who are in trouble is very important

I seldom think about the welfare of other people whom I don’t know personally

Without solidarity, our society is doomed to fall apart

Fighting poverty is an important duty of government

To me, patriotism includes seeing to the welfare of the others

Everybody is entitled to a good service, even if it costs a lot of money

Even in the case of major disasters, public service should be maintained

It is important that public servants account for all the costs they make

Although PSM consists of a number of dimensions, this article only considers an aggregate instrument of PSM, not the individual dimensions. This instrument involves averaging the score on the set of PSM items, scored from 1 to 5 (1 = I totally disagree and 5 = I totally agree, or the inverted scores, if applicable). Although this measure provides less information on the individual dimensions of PSM, an aggregate approach has been frequently applied in PSM research (Brewer & Selden, 2000; Kim, 2005; Lewis & Frank, 2002; Naff & Crum, 1999). Also, correlations between factors are high, ranging from .29 to .93, resulting in substantial redundancy. In total, 18 items, relating to five dimensions of PSM, are averaged, with an acceptable Cronbach’s alpha of .79. Finally, this approach will also limit the size of the analysis, not only making it clearer but also limiting the possible amount of Type I errors.

To measure leadership promotion of public service values (limited to the direct supervisor)—as one element, inspirational motivation, of transformational leadership—respondents were asked to what extent their direct supervisors discussed a set of public service values with them (Table 2). These values correspond to the five dimensions distinguished in the PSM instrument and were also measured on a 5-point scale (from 1 = I totally disagree to 5 = I totally agree). However, they are not the same items that were used in the PSM instrument, as many respondents would find it impossible to comment on such detailed items as used in the PSM instrument on behalf of their supervisor. The aim was rather to give an overview of the values promoted by the supervisor, as a reflection (of the perception) of the institutionalized public service values within this organization. As is the case with PSM, these were also aggregated into a single seven-item instrument, with a Cronbach’s alpha of .84.

Table 2.

Leadership Promotion of Public Service Values (α = .84).

My direct supervisor discusses the influence we exert on public policy with me.

My direct supervisor discusses our contribution to the public interest with me.

My direct supervisor discusses the importance of compassion to fellow citizens who have a hard time with me.

My direct supervisor discusses the necessity to forgo my own interests with me.

My direct supervisor discusses the importance of equal treatment with me.

My direct supervisor expects me to account for my actions.

My direct supervisor discusses the importance of honesty and trustworthiness with me.

The final construct that is measured in this study is basic psychological needs satisfaction. The basis for the measurement instrument is the Basic Psychological Needs at Work Scale (Deci et al., 2001). This is a 21-item scale, measuring the three basic psychological needs (competence, autonomy, and relatedness). To validate the instrument, the dataset was randomly split, resulting in one calibration dataset and one validation dataset, preventing capitalization on chance (MacCallum, Roznowski, Mar, & Reith, 1994). The model itself was analyzed by means of a diagonally weighted least squares estimation, to account for the ordinal character of the data (Jöreskog, 2005). The analysis resulted in a 14-item instrument, measuring the three hypothesized dimensions (Table 3). Goodness-of-fit values were well within acceptable levels (see appendix) and both discriminant and convergent validities were assessed using confidence intervals of inter-factor correlation and significant lambda values, respectively (Anderson & Gerbing, 1988).

Table 3.

Factor Model for Basic Needs Satisfaction at Work.

	λ	Composite α
Autonomy		.74
I feel like I can make a lot of inputs to deciding how my job gets done.	.49*
I am free to express my ideas and opinions on the job.	.64*
My feelings are taken into consideration at work.	.64*
I feel like I can pretty much be myself at work.	.80*
Relatedness		.84
I really like the people I work with.	.84*
I get along with people at work.	.81*
I pretty much keep to myself when I am at work (I).	.43*
I consider the people I work with to be my friends.	.68*
People at work care about me.	.77*
People at work are pretty friendly toward me.	.55*
The people with whom I work seem to like me	.49*
Competence		.69
People at work tell me I am good at what I do.	.52*
I have been able to learn interesting new skills on my job.	.69*
Most days, I feel a sense of accomplishment from working.	.74*

Note. (I) = inverted coded.

p < .05.

Next to these variables, a number of control variables were also included in the analysis (gender, year of birth, and grade). While “gender” is a dichotomous dummy variable, with males being the reference group (scoring “0”), “grade” has four categories (Grade A requiring master’s degree, Grade B requiring bachelor’s degree, Grade C requiring high school education, and Grade D requiring no education). Grade D is excluded from the analysis, and therefore the reference category (Hardy, 1993). “Year of birth” is a metric variable.

Statistical Method

Regression analysis is the most appropriate method for examining the relation between a set of independent variables and a single dependent variable. Multiple ordinary least squares (OLS) regression is probably the most widely used variant of regression analysis.

To test the hypotheses stated earlier in this study, the analysis should be able to incorporate interactive effects. More specifically, when testing H2, the analysis should be capable of assessing the effect of promotion of public service values on PSM for different levels of the moderator variables, in this case, the basic psychological needs satisfaction. An obvious methodology would be interactive or moderated regression (moderated multiple regression [MMR]) using product terms of the independent and the moderator variables (Baron & Kenny, 1986; Jaccard, Turisi, & Wan, 1990).

However, in comparison to experimental settings, MMR is often underachieving in detecting moderator variables in survey data, although compelling theoretical reasons for such a relationship exists. Despite the fact that the general model statistics (Model F, R², t-values) are in these cases relatively robust, the estimates of the regression parameters (β’s and standard errors) are especially prone to unreliability, due to the joint distribution of the data (also reducing statistical power; for an overview, see McClelland & Judd, 1993 or O’Connor, 2006). Also, as cross-products are highly correlated with both independent variables, the risk of multicollinearity is substantially increased. To avoid multicollinearity, Jaccard et al. (1990) propose standardization of the independent and dependent variables. However, this does not fully alleviate problems of MMR. Therefore, and despite the compelling evidence that MMR is the single best method for detecting moderator effects (Stone-Romero & Anderson, 1994), sometimes it is not always capable of identifying—and specifying the direction of—hypothesized interactions and its interpretation becomes cumbersome.

Therefore, other ways of analyzing interactions can be added to supplement—not replace—MMR and provide additional information. One of the possible approaches is the comparison of subgroups (subgroup correlation comparison [SCC]). Although the statistical power of SCC is substantially lower than MMR (Stone-Romero & Anderson, 1994), it does not suffer from multicollinearity as no product terms are involved. Also, because of its bivariate nature, it suffers less from correlated and therefore often confounding independent variables (in cases where multicollinearity is less of a threat). SCC enables to investigate and readily interpret the direction of the alleged interaction effects, by statistically comparing sizes of correlation coefficients of a higher and a lower scoring—split at the mean of their respective moderator—subgroup (tested by means of a Q-statistic, approximately distributed as a chi-square with k − 1 degrees of freedom) (Jaccard et al., 1990). Due to the relatively large sample size, the issues regarding statistical power play less of a role, which enables SCC to provide additional information on moderator effects.

Analysis of the Data and Presentation of the Results

When analyzing the correlation table (Table 4), one can observe that a positive correlation exists between all main variables. This warrants further investigation of the main effect and possible moderators.

Table 4.

Correlation and Descriptive.

		N	M	SD	1	2	3	4	5	6	7	8	9	10	11
1	Gender	3,437	1.48	0.50	1.00
2	Year of birth	3,437	1964	10.80	.23	1.00
3	Grade A	3,437	0.41	0.49	−.07	.11	1.00
4	Grade B	3,437	0.20	0.40	.04	.15	−.42	1.00
5	Grade C	3,437	0.29	0.45	.00	−.16	−.53	−.33	1.00
6	Grade D	3,437	0.09	0.29	.07	−.15	−.27	−.16	−.21	1.00
7	Leadership values promotion	3,437	3.28	0.82	.00	−.06	−.03	.03	−.02	.04	1.00
8	Autonomy satisfaction	3,437	5.03	1.11	.00	.05	.06	.04	−.09	−.01	.31	1.00
9	Competence satisfaction	3,437	5.33	1.06	−.04	.00	.08	.02	−.08	−.05	.34	.64	1.00
10	Relatedness satisfaction	3,437	4.89	0.74	.03	.03	−.01	.02	−.04	.05	.26	.56	.45	1.00
11	PSM	3,437	3.68	0.42	−.15	−.15	.15	−.04	−.06	−.10	.22	.16	.27	.16	1.00

Note. PSM = public service motivation. Values of over |.04| are significant at p > .05.

The regression analysis (Table 5) consists of four hierarchically nested models.⁵ Model 1 assesses the influence of gender, age (as measured in year of birth), and grade as controls. For gender and grade, male and Grade D—no education required—are the respective referents. All control variables display a significant influence on the level of PSM, explaining a substantial part of the variance (R² = .064). Women display lower levels of PSM, as younger respondents do, whereas people in higher grade have higher levels of PSM. These effects remain consistent throughout the analysis.

Table 5.

Multiple Moderated Regression Model for Need Satisfaction and Leadership Value Promotion.

	Model 1		Model 2		Model 3		Model 4
	β	SE	β	SE	β	SE	β	SE
Gender	−.08***	0.01	−.09***	0.01	−.08***	0.01	−.08***	0.01
Year of birth	−.01***	0.00	−.01***	0.00	−.01***	0.00	−.01***	0.00
Grade A	.21***	0.02	.23***	0.02	.21***	0.02	.21***	0.02
Grade B	.12***	0.03	.12***	0.03	.11***	0.03	.11***	0.03
Grade C	.08***	0.03	.09***	0.02	.09***	0.02	.09***	0.02
Leadership values promotion			.11***	0.01	.07***	0.01	−.10*	0.05
Autonomy satisfaction					−.03**	0.01	−.10***	0.03
Competence satisfaction					.09***	0.01	.07*	0.03
Relatedness satisfaction					.04***	0.01	.03	0.04
Autonomy × Leadership							.02*	0.01
Competence × Leadership							.01	0.01
Relatedness × Leadership							.00	0.01
F model	48.00***		71.04***		68.18***		53.29***
F ratio	—		176.36***		56.01***		7.41***
n	3,504		3,476		3,474		3,474
R ²	.064		.109		.151		.156
Adj. R²	.063		.108		.148		.153

p < .05. **p < .01. ***p < .001.

Models 2 and 3 add the main effects of the main and moderating variables to the regression equation. Again, both leadership promotion of public values add a significant and substantial portion to the explained variance (additional R² is .045 and .042 respectively). Leadership values promotion has a positive relationship with PSM, as do competence and relatedness need satisfaction, whereas autonomy satisfaction displays a negative relationship (due to the high correlation with the satisfaction of the other needs, a confounding effect is demonstrated; the correlation table shows that there is in fact a positive correlation).

Model 4 adds the moderator, operationalized as a product term, which results in an increase of the R² of .005. Despite the low effect size, this is a significant increase—as can be read from the F-ratio statistic—of the total R² and can therefore be evaluated as an indication of a moderator effect. The changing coefficients for the main effects should not be considered, due to multicollinearity of the product terms and the main variables. The total moderating model—control variables included—accounts for 15.6% of the total variance.

As stated earlier, it may be useful to supplement this analysis with other techniques for assessing moderator effects. One such method is the SCC, which compares the correlation coefficient r between two variables in a lower and a higher scoring group of the hypothesized moderator variable. The dataset was split for the score on the moderator variables (based on the mean score), rendering six datasets, three lower scoring groups, and three higher scoring groups. The analysis results in three comparisons of r-values between a lower and a higher scoring group (Table 6).

Table 6.

Subgroup Correlation Comparison of Leadership on PSM (Basic Needs Satisfaction Subgroups).

	Lower scoring group			Higher scoring group
	n	r ^a	Z-score	n	r ^a	Z-score	Z′	Q
Autonomy	1,668	.15	0.15	1,810	.23	0.23	0.19	5.53*
Competence	1,344	.12	0.12	2,134	.19	0.20	0.17	4.20*
Relatedness	1,761	.17	0.17	1,717	.21	0.22	0.19	1.57

Note. PSM = public service motivation.

Correlation is significant at p < .001.

p < .05.

For autonomy satisfaction, the lower scoring group exhibits a lower r (.15) than the higher scoring group (.23). As can be derived from Table 6, this difference is statistically significant. The same goes for competence satisfaction, with respective r’s of .12 and .19. However, for relatedness, the difference between the correlations is not statistically different. These results supplement the image provided by the MMR.

Discussion

Both the regression analysis and the SCC demonstrate that promotion of public service values by the direct supervisor is related to individual-level PSM. This observation fits the framework developed earlier, which states that the institutional values, being a part of the supervisor’s organizational and therefore institutional role, are transmitted to the employee’s identity. The findings thus corroborate to a large extent H1 as, according to the data, leadership at a supervisory level is indeed able to influence employees’ PSM levels, which are a part of the identity, substantially. This is in keeping with the findings of Bruce (1994), who found that (attitudes toward) exemplary behavior by supervisor is positively correlated to (in that particular case ethical) employee behavior.

The hypotheses concerning an interaction effect of institutional socialization and basic needs satisfaction is also corroborated up to a certain extent. Both the MMR and SCC found evidence of such a moderator effect. Where the MMR was not able to distinguish between the various possible moderators, the SCC can make some distinction and find an effect of both autonomy and competence satisfaction on the relationship. Relatedness satisfaction did not display such a finding (despite the difference between the two correlations being in the hypothesized direction, it is not sufficiently large to be significant), although there is a direct effect of relatedness on satisfaction. This does not fully corroborate SDT, which indicates that this is necessary for internalization (Gagné & Deci, 2005; Ryan & Deci, 2005).

However, evidence may not be what it seems as the likelihood of the hypotheses does not only depend on the results of the MMR or the SCC. To be convincing, the study as a whole needs to satisfy to a sufficient extent the conditions of causality (Wright, 2008). A first condition is the condition plausibility, which states that there should be a reasonable theoretical justification. Second, the condition of association requires that if the criterion variable varies, the outcome variable should also vary. A third condition is that of temporal sequence, requiring that the criterion should precede the outcome. A final condition is that of non-spuriousness, which requires that alternative explanations should be excluded, in particular underlying third variables. Regression models in a cross-sectional dataset are therefore not fully suited to satisfy these conditions. The requirement of plausibility and association present little problems, the other two bring more difficulties about. However, both the design of the study and theoretical considerations can help to overcome these difficulties. First, by creatively designing survey items, depending on recall data, one can create a temporal sequence in a cross-sectional survey (as is demonstrated by Perry, Brudney, Coursey, & Littlepage, 2008, when asking about previous volunteering experience). In a similar way, the items for leadership assume previous experience as opposed to the PSM items, which assume a present attitude toward public service. Second, theory on previous research may help to alleviate problems, when researching a previously assessed relationship and its direction. The relationship between transformational leadership and motivation has been well documented (Judge & Piccolo, 2004; Kark, Chen, & Shamir, 2003), and it is, therefore, reasonably to assume the relationship will have the same direction in the present study. Finally, theoretical consideration as well as design might also help to overcome problems with non-spuriousness. Theory might aid in selecting the right antecedents by looking at the mechanisms in previous studies (as has been done in the theoretical framework). Also, by fine-tuning the models by means of moderator and mediator variables, spuriousness can be avoided (Wright, 2008). In this particular study, not only moderating effects are controlled for developing the model hierarchically but the study also controls for the main effects of some of the aspects of the SDT, which could constitute a rival explanation (the main effect of leadership is not removed when adding basic needs satisfaction to the regression model). However, evidently, it is not possible to control for all rival theories in a single study, which only stresses the need for a sound theoretical framework. In conclusion, despite the fact that cross-sectional datasets are not the optimal strategy, the findings do corroborate the hypotheses stated earlier as the conditions of causality are sufficiently satisfied.

The corroboration of the hypotheses stated earlier is important in several respects. First, it is important to know that transformational leadership behavior—in particular, the elements that refer to inspirational motivation—has an influence on the individual PSM levels. This not only provides evidence for an institutional theory of PSM but it also provides evidence that PSM can be developed within an organizational environment and that leaders are in a key position to influence this development. These findings also provide support for a wider approach toward nurturing and developing PSM within the environment of public-sector organizations, as advocated by Paarlberg, Perry, and Hondeghem (2008). Second, and related to the first point, is that leadership in the public sector can act in a transformational way, even if it is only by promoting certain public values. This should draw attention to other transformational elements of public-sector leadership—as has been done by some of those who have stressed the importance of transformational leadership styles in the public sector (Paarlberg & Perry, 2007; Trottier et al., 2008) and perhaps abandoning earlier strategies of “uniqueness” of public-sector leadership in favor of more general established concepts (without losing sight of the particularities of the public sector). Moreover, this kind of leadership and its associated behaviors is not limited to senior leaders or chief executive officers. Based on the data used in this study, the area of transformational leadership can be opened up to lower-level direct supervisor type of leaders. It is those who run day-to-day business in government offices who are able to influence the identity and motivation of their subordinates.

Next to this, it is also important that PSM theory is further related to other motivational frameworks. PSM theory has been developed on an interdisciplinary basis (Perry & Vandenabeele, 2008), “borrowing” and “integrating” from various fields in social sciences. It is therefore important to see that a motivational perspective as SDT empirically contributes to further development and testing of PSM theory, but it is equally important to see that PSM and its development cannot be reduced to any one of its building blocks. The findings from this present study indicate that satisfaction of basic psychological needs contribute to the development of PSM, but at the same time, PSM cannot be fully explained as basic needs satisfaction, as it has a direct but not fully mediating effect as well as a moderating effect.

Finally, related to the previous point, this provides support for a comprehensive and interdisciplinary institutional theory of PSM. Multiple structural value-based relationships appear to contribute to PSM development, and the leader–follower or supervisor–subordinate relations is only one example of such institutions, whereas education, work organization, politics, and family are other examples of such antecedents (Pandey & Stazyk, 2008; Perry, 1997; Vandenabeele, 2011). It is therefore necessary to further our knowledge of PSM and, in particular, of its antecedents, to identify additional institutions that influence PSM.

Conclusion

The results of this study should appeal to both scholars and practitioners. In particular, the finding that institutional elements seem to have a substantial and significant influence on the individual PSM level is important. In the case of this research, the results in this article demonstrate that the institutional values of a public organization, crystallized in the behavior of supervisors, have a socializing effect on the individual, and therefore may generate PSM. These results therefore provide evidence of “transformational leadership” at the direct supervisor level, showing that this concept, as middle managers are able to transform the employee’s identity, has unjustly been reserved for executive leaders only. This finding should urge practitioners at any level to actively pursue the development of PSM in their own organization, as suggested by Paarlberg et al. (2008), giving them more confidence in the potential outcome of their efforts.

However, other institutional elements moderate this relationship. The extent to which basic psychological needs are satisfied within this institutional context determine how strong the socialization effect is. When the environment of the employee is experienced as granting a certain degree of autonomy to the individual and enabling people to achieve real results, the socializing effect of public service values from within the institution is enhanced. Apparently, not only the “what” (which values) but also the “how” (perception of the environment in which is socialized) is important in developing an institutional, in this case a public service, identity. Again, this should signal the importance of people management not only at street-level but also higher up in the organizational structure. After all, addressing the basic psychological needs of the individual in its institutional setting provides the most optimal conditions for transferring organizational values.

Despite these interesting conclusions, some limitations rest on the results of the research in this chapter. First, the conclusions are based on a cross-sectional database. This is a possible threat to the internal validity of the conclusions. Although this was somewhat remedied by embedding the conclusion in a sound theoretical framework and by controlling for other variables that point to other possible explanations, the danger of a lower internal validity is still looming at the back. Therefore, additional research should be carried out to strengthen these findings, in different environments and with both cross-sectional and longitudinal ways of collecting data. Nevertheless, the findings presented in this article provide interesting insights that can help to further develop a public administration theory and practice of PSM and how to harness it.

Footnotes

Appendix

Table A1.

Fit Statistics for Basic Needs Satisfaction (BNS) Model.

		χ²(SB)	df	GFI	RMSEA	RMSEA CI	RMSEA p < .05	CFI	NFI
Null model BNS	Calibration	1706.79	186	.971	.068	[.065 − .071]	.000	.940	.934
BNS model	Calibration	514.88	74	.990	.058	[.054 −.063]	.002	.979	.976
BNS model (loose validation)	Validation	511.72	74	.990	.058	[.053 −.063]	.003	.978	.978
BNS model (partial validation)	Validation	547.49	85	.988	.056	[.051 −.060]	.017	.980	.977

Note. SB = Satorra-Bentler; GFI = Goodness-of-Fit Index; RMSEA = root mean square error approximation; CI = confidence interval; CFI = Comparative Fit Index; NFI = Normed Fit Index.

Author’s Note

Previous versions of this article have been presented at the 3rd Trans-Atlantic Dialogue 2007, the International Research Society for Public Management XII 2008 conference, the European Group of Public Administration annual conference 2008 and the Public Management Research Conference 2009.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research has been funded by Research Foundation Flanders (FWO).

Notes

Author Biography

Wouter Vandenabeele is an assistant professor at the Utrecht University School of Governance, the Netherlands, and a visiting professor at the Public Management Institute at KU Leuven University, Belgium. His research interests concern motivation of public servants—in particular public service motivation—and strategic human resources management in the public sector.

References

Anderson

J. C.

Gerbing

D. W.

(1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103, 411-423.

Baron

R. M.

Kenny

D. A.

(1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Bass

B. M.

(1999). Two decades of research and development in transformational leadership. European Journal of Work & Organizational Psychology, 8, 9-32.

Berger

P. L.

Luckmann

(1966). The social construction of reality. New York, NY: Anchor Books.

Brehm

J. O.

(1993). The phantom respondents: Opinion surveys and political representation. Ann Arbor: Michigan University Press.

Brewer

G. A.

Selden

S. C.

(2000). Why elephants gallop: Assessing and predicting organizational performance in federal agencies. Journal of Public administration Research and Theory, 10, 685-711.

Bright

L. A.

(2007). Does person–organization fit mediate the relationship between public service motivation and the job performance of public employees? Review of Public Personnel Administration, 27, 361-379.

Bright

L. A.

(2008). Does public service motivation really make a difference on job satisfaction and turnover intentions of public employees. American Review of Public Administration, 38, 149-166.

Bruce

(1994). Ethical people are productive people. Public Productivity & Management Review, 17, 241-252.

10.

Burns

J. M.

(1978). Leadership. New York, NY: Harper & Row.

11.

Camilleri

(2006). Towards developing an organizational commitment-public service motivation model for the Maltese public service employees. Public Policy and Administration, 21(1), 63-83.

12.

Camilleri

(2007). Antecedents affecting public service motivation. Personnel Review, 36, 356-377.

13.

Chapman

(1988). Ethics in the British civil service. London, England: Routledge.

14.

Christensen

R. K.

Wright

B. E.

(2011). The effects of public service motivation on job choice decisions: Disentangling the contributions of person-organization fit and person-job fit. Journal of Public Administration Research and Theory. Advance online publication. doi:10.1093/jopart/muq085

15.

Conger

J. A.

Hunt

J. A.

(1999). Overview charismatic and transformational leadership: Taking stock of the present and future (Part 1). Leadership Quarterly, 10, 121-127.

16.

Couper

M. P.

(2000). Web surveys: A review of issues and approaches. Public Opinion Quarterly, 64, 464-494.

17.

Couper

M. P.

Blair

Triplett

(1999). A comparison of mail and e-mail for a survey of employees in US statistical agencies. Journal of Official Statistics, 15, 39-56.

18.

Coursey

Pandey

S. K.

(2007). Public service motivation measurement: Testing an abridged version of Perry’s proposed measurement scale. Administration & Society, 39, 547-568.

19.

Coursey

Perry

J. L.

Brudney

J. L.

Littlepage

(2008). Psychometric verification of Perry’s public service motivation instrument: Results for volunteer exemplars. Review of Public Personnel Administration, 28, 79-90.

20.

Crawford

S. D.

Couper

M. P.

Lamias

M. J.

(2001). Web surveys: Perceptions of burden. Social Science Computer Review, 19, 146-162.

21.

Deci

E. L.

Ryan

R. M.

(2004). Handbook of self-determination research. Rochester, NY: University of Rochester Press.

22.

Deci

E. L.

Ryan

R. M.

Gagné

Leone

D. R.

Usunov

Kornazheva

B. P.

(2001). Need satisfaction, motivation, and well-being in the work organizations of a former Eastern Bloc country. Personality and Social Psychology Bulletin, 27, 930-942.

23.

Denhardt

J. V.

Campbell

K. B.

(2006). The role of democratic values in transformational leadership. Administration & Society, 38, 556-572.

24.

Downs

(1967). Inside bureaucracy. Boston, MA: Little, Brown.

25.

Dvir

Eden

Avolio

B. J.

Shamir

(2002). Impact of transformational leadership on follower development and performance: A field experiment. Academy of Management Journal, 45, 735-744.

26.

Filstad

(2004). How newcomers use role models in organizational socialization. Journal of Workplace Learning, 16, 396-409.

27.

Gagné

Deci

E. L.

(2005). Self-determination theory and work motivation. Journal of Organizational Behavior, 26, 331-362.

28.

Garnett

J. L.

Marlow

Pandey

S. K.

(2008). Penetrating the performance predicament: Communication as a mediator or moderator of organizational culture’s impact on public organizational performance. Public Administration Review, 68, 266-281.

29.

Hair

J. F.

Black

W. C.

Babin

B. J.

Anderson

R. E.

(2009). Multivariate data analysis. Upper Saddle River, NJ: Prentice Hall.

30.

Hardy

M. A.

(1993). Regression with dummy variables. Newbury Park, CA: SAGE.

31.

Hooijberg

Choi

(2001). The impact of organizational characteristics on leadership effectiveness models: An examination of leadership in a private and a public sector organization. Administration & Society, 33, 403-430.

32.

Horton

(2008). History and persistence of an idea and an ideal. In Perry

J. L.

Hondeghem

(Eds.), Motivation in public management: The call of public service (pp. 17-32). Oxford, UK: Oxford University Press.

33.

Jaccard

Turisi

Wan

C. K.

(1990). Interaction effects in multiple regression. Newbury Park, CA: SAGE.

34.

Javidan

Waldman

D. A.

(2003). Exploring charismatic leadership in the public sector: Measurement and consequences. Public Administration Review, 63, 229-242.

35.

Jöreskog

(2005). Structural equation modeling with ordinal variables using LISREL. Retrieved from http://www.ssicentral.com/lisrel/techdocs/ordinal.pdf

36.

Judge

T. A.

Piccolo

R. F.

(2004). Transformational and transactional leadership: A meta-analytic test of their relative validity. Journal of Applied Psychology, 89, 755-768.

37.

Kark

Chen

Shamir

(2003). The two faces of transformational leadership: Empowerment and dependency. Journal of Applied Psychology, 88, 246-255.

38.

Kerr

Jermier

J. M.

(1978). Substitutes for leadership: Their meaning and measurement. Organizational Behavior and Human Performance, 22, 375-403.

39.

Kim

S. M.

(2005). Individual-level factors and organizational performance in government organizations. Journal of Public Administration Research and Theory, 15, 245-261.

40.

Kuhnert

K. W.

Lewis

(1987). Transactional and transformational leadership: A constructive/developmental analysis. Academy of Management Review, 12, 648-657.

41.

Lewis

G. B.

Frank

S. A.

(2002). Who wants to work for government? Public Administration Review, 62, 395-404.

42.

MacCallum

R. C.

Roznowski

Mar

C. M.

Reith

(1994). Alternative strategies for cross-validation of covariance structure models. Multivariate Behavioral Research, 29, 1-32.

43.

March

J. G.

Olsen

J. P.

(1989). Rediscovering institutions. New York, NY: The Free Press.

44.

March

J. G.

Olsen

J. P.

(1995). Democratic governance. New York, NY: The Free Press.

45.

McClelland

G. H.

Judd

C. M.

(1993). Statistical difficulties of detecting interactions and moderator effects. Psychological Bulletin, 114, 376-390.

46.

Mosher

F. C.

(1968). Democracy and the public service. New York, NY: Oxford University Press.

47.

Moynihan

D. P.

Pandey

S. K.

(2005). Testing how management matters in an era of government by performance. Journal of Public Administration Research and Theory, 15, 421-439.

48.

Moynihan

D. P.

Pandey

S. K.

(2007). The role of organizations in fostering public service motivation. Public Administration Review, 67, 40-53.

49.

Naff

K. C.

Crum

(1999). Working for America—Does public service motivation make a difference. Review of Public Personnel Administration, 19(4), 5-16.

50.

O’Connor

B. P.

(2006). Programs for problems created by continuous variable distributions in moderated multiple regression. Organizational Research Methods, 9, 554-567.

51.

Paarlberg

L. E.

Lavigna

(2010). Transformational leadership and public service motivation: Driving individual and organizational performance. Public Administration Review, 70, 710-718.

52.

Paarlberg

L. E.

Perry

J. L.

(2007). Values management: Aligning employee values and organizational goals. American Review of Public Administration, 37, 387-408.

53.

Paarlberg

L. E.

Perry

J. L.

Hondeghem

(2008). From theory to practice: Strategies for applying public service motivation. In Perry

J. L.

Hondeghem

(Eds.), Motivation in public management: The call of public service (pp. 268-293). Oxford, UK: Oxford University Press.

54.

Pandey

S. K.

Stazyk

E. C.

(2008). Antecedents and correlates of public service motivation. In Perry

J. L.

Hondeghem

(Eds.), Motivation in public management: The call of public service (pp. 101-117). Oxford, UK: Oxford University Press.

55.

Perry

J. L.

(1996). Measuring public service motivation: An assessment of construct reliability and validity. Journal of Public Administration Research and Theory, 6, 5-23.

56.

Perry

J. L.

(1997). Antecedents of public service motivation. Journal of Public Administration Research and Theory, 7, 181-197.

57.

Perry

J. L.

(2000). Bringing society in: Toward a theory of public service motivation. Journal of Public Administration Research and Theory, 10, 471-488.

58.

Perry

J. L.

Brudney

Coursey

Littlepage

(2008). What drives morally committed citizens? A study of antecedents of public service motivation. Public Administration Review, 68, 445-458.

59.

Perry

J. L.

Hondeghem

(Eds.). (2008). Motivation in public management: The call of public service. Oxford, UK: Oxford University Press.

60.

Perry

J. L.

Vandenabeele

(2008). The behavioral dynamics of public service motivation. In Perry

J. L.

Hondeghem

(Eds.), Motivation in public management: The call of public service (pp. 56-79). Oxford, UK: Oxford University Press.

61.

Perry

J. L.

Wise

L. R.

(1990). The motivational bases of public service. Public Administration Review, 50, 367-373.

62.

Peters

B. G.

(2000). Institutional theory in political science: The new institutionalism. London, England: Continuum.

63.

Rainey

H. G.

(1997). Understanding and managing public organizations. San Francisco, CA: Jossey-Bass.

64.

Rainey

H. G.

Steinbauer

(1999). Galloping elephants: Developing elements of a theory of effective government organizations. Journal of Public Administration Research and Theory, 9, 1-32.

65.

Ryan

R. M.

Deci

E. L.

(2005). On assimilating identities to the self: A self-determination perspective on internalization and integrity within cultures. In Leary

M. R.

Tagney

J. P.

(Eds.), Handbook of self and identity (pp. 253-272). New York, NY: Guilford.

66.

Scott

W. R.

(2001). Institutions and organizations. Thousand Oaks, CA: SAGE.

67.

Settoon

R. P.

Adkins

C. L.

(1997). Newcomer socialization: The role of supervisors, co-workers, friends and family members. Journal of Business and Psychology, 11, 507-516.

68.

Shamir

House

R. J.

Arthur

R. B.

(1993). The motivational effects of charismatic leadership: A self-concept based theory. Organization science, 4, 577-594.

69.

Spector

P. E.

(2006). Method variance in organizational research: Truth or urban legend? Organizational Research Methods, 9, 221-232.

70.

Stone-Romero

E. F.

Anderson

L. E.

(1994). Relative power of moderated multiple regression and the comparison of subgroup correlation coefficients for detecting moderating effects. Journal of Applied Psychology, 79, 354-359.

71.

Terry

L. D.

(2003). Leadership of public bureaucracies—The administrator as conservator. Armonk, NY: M.E. Sharpe.

72.

Trottier

Van Wart

Wang

(2008). Examining the nature and significance of leadership in government organizations. Public Administration Review, 68, 319-333.

73.

Vandenabeele

(2007). Towards a theory of public service motivation: An institutional approach. Public Management Review, 10, 545-556.

74.

Vandenabeele

(2008). Development of a public service motivation measurement scale: Corroborating and extending Perry’s measurement instrument. International Public Management Journal, 11, 143-167.

75.

Vandenabeele

(2011). Who wants to deliver public service? Toward an institutional theory of public service motivation. Review of Public Personnel Administration, 31, 87-107.

76.

Van Wart

(2003). Public sector leadership theory: An assessment. Public Administration Review, 63, 214-228.

77.

Wright

B. E.

(2007). Public service and motivation: Does mission matter? Public Administration Review, 67, 54-64.

78.

Wright

B. E.

(2008). Methodological challenges associated with public service motivation research. In Perry

J. L.

Hondeghem

(Eds.), Motivation in public management: The call of public service (pp. 80-98). Oxford, UK: Oxford University Press.