The Limits of Public Service Motivation: Confidence in Government Institutions Among Public Servants

Abstract

Do government employees have more confidence in government than their counterparts in other sectors? The public service motivation (PSM) literature suggests that they should, but this assertion has never been tested. In this paper, we examine how public servants feel about the federal government and compare these feelings with those of employees in other sectors. Although we find that public servants hold some parts of government in higher regard than their counterparts, this finding does not apply to every branch of government. The patterns of confidence in institutions we find among America’s public servants suggest that there is a limit to PSM.

Keywords

confidence in government public service motivation public/private differences

Do government employees have more confidence in government institutions than their counterparts in other sectors? The voluminous public service motivation (PSM) literature suggests that they should. After all, public servants engage in a variety of prosocial behaviors. They are more likely to be motivated by altruistic concerns (Rainey & Steinbauer, 1999) more apt to self-sacrifice (Perry & Wise, 1990), more in touch with political activities (Ertas, 2014a), and more likely to volunteer (Ertas, 2014b) and generally have higher levels of “social trust” (G. A. Brewer, 2003) than their private-sector counterparts. It is not much of a conceptual leap to suggest that they should also be more confident in the very institutions for which they work (G. A. Brewer, Selden, & Facer, 2000). Surprisingly, throughout the PSM research agenda, there are few mentions of confidence in government among public servants and no extant empirical tests of this fundamental hypothesis.

This oversight is important for two reasons. Perhaps most importantly for our purposes, discerning whether public citizens are more confident in government than employees in other sectors will help address long-standing questions about the extent and potential limits of PSM. We expect that government employees engage in a host of prosocial behaviors¹—particularly those related to the very institutions that employ them. If they do not, then our understanding of the “call of public service” might need to be revisited.

In addition, answering this question will help inform the vast literature on confidence and trust in government. Indeed, scores of studies find that trust in government is in decline (e.g., Hetherington, 2005), leading some prominent political professionals to conclude that we are facing a “crisis of trust” in the United States (Schoen & Caddell, 2013). Assessing whether public servants have experienced a similar decline in confidence will help us understand the causes and consequences of this decline in trust.

In this article, we examine how public servants feel about the federal government and compare these feelings with those of employees in other sectors. Although we find that public servants do hold some parts of government in higher regard than their non-public-sector counterparts, this finding does not apply to every branch of government. In the end, we conclude that the patterns of confidence we find among America’s public servants suggest that there is a limit to PSM. Although public employees are no doubt different in many ways, when it comes to confidence in government institutions, they appear to be rational actors who make measured decisions about who receives their confidence.

Confidence and Trust in Government Institutions

Political pundits and mainstream news organizations have documented historically low levels of confidence and trust in government. For example, the Pew Center for the People and the Press finds that trust in government peaked in the mid-1960s when almost 80% of the people responded that they trusted government most of the time or just about always. Today, these numbers are almost a perfect inverse of what they were in the 1960s; in October 2013, just 19% of the population expressed high levels of trust in government (Pew Center for the People and the Press, 2013). This decline is important for a number of reasons, but culturalists like Almond and Verba (1963) would view this as a sort of canary in the coal mine, alerting us to a larger decline in civic-republican culture.²

Over 40 years ago, Arthur Miller (1974) argued that trust in government was low and would continue to fall as long as the average American remained dissatisfied with the policy alternatives proffered by the two major political parties. Although there have been some dissenting voices, most conclude that trust in government varies systematically with government performance (Hetherington, 1998). Some argue that economic interests reign supreme (Citrin & Green, 1986), and more recent work focuses on the primacy of international affairs (Hetherington & Rudolph, 2008).

Not surprisingly,

when people are satisfied with the policy direction that the country is taking, when the economy is growing, when the president has attractive personal characteristics, when the country is threatened and when media criticism is relatively low, political trust increases. (Hetherington, 2005, p. 16)

Variables meant to tap into an individual’s socioeconomic status, such as gender, race, age, education, and income, have little, if any, influence on trust in government (Hetherington, 2005). One of the few consistent predictors of trust in government is knowledge of the policy in question. F. L. Cook, Jacobs, and Kim (2010) find that it is possible to increase both knowledge and trust in government institutions by providing citizens with more information. Given that public-sector employees may become familiar with different types of information than private- or non-profit-sector employees, however, it is unclear whether these findings regarding information might work differently for employees of different sectors.

The implications of declining trust and trust in government have been cataloged in the public administration literature as well. According to Robert Ruscio (1996), “lack of trust leads to, among other things, excessive micromanagement . . . flexibility and discretion become clearly constrained . . . it hinders informal relationships and leads to an excessive dependence on rules, formal procedures, regulations, and legalisms (p. 463).” Others find that individual levels of trust are associated with opinions of specific policies. For example, Cooper, Knotts, and Brennan (2008) argue that people are more likely to cede power to government over zoning when they trust local government. Similarly, Herian, Shank, and Abdel-Monem (2014) find that people express more support for electronic health records when they trust government. Clearly, confidence and trust in government are important for political as well as policy reasons.

Traditionally, research on confidence and trust in government has been reliant on a relationship between trust and government performance. Because this relationship has historically been found to be minimal/non-existent (Christensen & Lægreid, 2005; Glaser & Hildreth, 1999; Van de Walle & Bouckaert, 2003; Yang & Holzer, 2006), the basis for analyzing trust in government must be broadened.

In his study, Van Ryzin (2011) analyzes the relationships between “process” and performance/outcomes in the public sector, as well as their respective relationships with the broader notion of trust of government. Performance/outcome indicators include respondents’ perceived success in health care provision, crime control, and combating unemployment, among others; process indicators employed include Likert-type scale measures of fairness, preferential treatment, and perceived corruption. These procedural interests were also reflected in P. R. Brewer and Sigelman’s (2002) study comparing trust in federal government among five groups in the Washington, D.C., area—federal employees, family members, friends, state/local government employees, and other American citizens. Appreciation for governmental processes was reflected in their findings, as federal and state/local government employees “were more likely to believe that the federal government is efficient, uncorrupted, and well intentioned” in their actions (P. R. Brewer & Sigelman, 2002).

A noteworthy finding stemming from the P. R. Brewer and Sigelman (2002) study is that higher levels of trust are seen in both federal and state/local government employees. This suggests a sense of commonality/camaraderie in appreciation of government processes, even though outcomes and end-objectives vary greatly.

Van Ryzin juxtaposes American findings against those of the aggregate of 32 other nations. Focused specifically on the United States, findings indicate that the process indicators have a greater effect on respondents’ trust in government than the outcomes of government action, which lacked statistical significance. This disconnect between outcomes and trust corroborates Van de Walle and Bouckaert’s (2003) finding that the link “between performance and trust can only be made when very specific conditions are present” (p. 891). Given that they are on the “inside” of governmental processes, it is reasonable to believe that government employees in our study would not only follow suit with the populous but also have an even stronger relationship (Van Ryzin, 2011).

Applied specifically to the judicial branch, Van Ryzin’s (2011) findings on the importance of governmental processes in building higher levels of trust in government are corroborated by Tyler (1990). In Tyler’s study, perceptions of the “legitimacy of legal authorities” (Van Ryzin, 2011) were more determined by perceptions of “procedural justice” than outcomes of the judicial process. Tyler defines this “procedural justice” as “neutrality, lack of bias, honesty, efforts to be fair, politeness, and respect for citizens’ rights” (p. 7).

Public-sector employees are managing governmental processes in all three branches of government, positive perceptions of processes have a demonstrated importance in positively affecting levels of trust in government, and thus these public employees’ perceptions of their own work could influence their own trust in government. This establishes a pre-existing level of trust among public servants (Van de Walle & Bouckaert, 2003) that could reasonably be considered a mutually reinforcing cycle. This is somewhat reflective of P. R. Brewer and Sigelman’s (2002) “self-image enhancement” that contributes to the government (all levels) employee/non-government employee disparity in trust.

The separation of process from outcomes, along with the potential insular cycle of process and trust, warrants discussion of Putnam’s (1971) concept of “political elite” within society. The individuals Putnam describes as the elite in American political culture are those “much more interested, much more involved, and much more influential in public affairs than their fellows” (Putnam, 1971, p. 651). Putnam explicitly identifies “professional politicians” as being among the most prominent members of this elite group, but his proposed definition is equally (if not more) applicable to public-sector employees. Being their full-time employment and livelihood, interest, involvement, and influence are inherent to their positions.

The proposition of the “political elite” was an attempt to make a practical connection between political behavior and ideological stances, thereby addressing a “troubled literature” (Putnam, 1971). This became a foundational piece of literature for future research, but relied on a definition that aggregated elected officials and appointed civil servants. It has been well established that there are inherent differences between Putnam’s “political elite” and others. In the same spirit, Putnam (1973) later described a subset within the political elite that is critical in any discussion of perceptions. Gauging trust among those elected officials with ultimate authority over the government has inherent differences from gauging trust among those appointed officials entrusted to execute the directives of elected officials.

Thus far, we have discussed confidence and trust in government as a generalized concept, but other work examines the determinants of trust and confidence in specific branches of government. Not surprisingly, the U.S. Congress has probably received the lion’s share of this attention—probably because it is Congress that the public holds in the lowest regard. Hibbing and Theiss-Morse (1995) argue that it is the very nature of participatory democracy that drives down support for the people’s branch, later arguing that Americans would have more confidence in Congress (Hibbing, 2002) and other institutions (Hibbing & Theiss-Morse, 2002) if these branches of government practiced “stealth democracy.” This is the idea that Americans “want democratic procedures to exist but not to be visible on a routine basis” (Hibbing & Theiss-Morse, 2002, p. 2). Richardson, Houston, and Hadjiharalambous (2001) examine confidence in all three branches of government and find that there are different predictors of confidence, depending on the institution in question. Americans, it appears, hold discreet opinions on discreet institutions.

PSM: A Brief Review

These dichotomies of performance/outcomes as indicators of trust and elected/appointed as subsets of the “political elite” (Putnam, 1971) segue nicely into discussion of the scholarly literature on PSM. Long debated among researchers, PSM is a critical part of this discussion of trust in government among government employees.

According to Perry and Wise (1990, p. 368), PSM encompasses “an individual’s predisposition to respond to motives grounded primarily or uniquely in public institutions and organizations.” The central assumption of the PSM literature is that public servants are fundamentally different than their counterparts in other sectors. This assumption has been successfully tested in a variety of arenas, including the value of intellectual stimulation (Lyons, Duxbury, & Higgins, 2006), aversion to risk (Bellante & Link, 1981), and greater emphasis on commitment to work than commitment to organization (Boyne, 2002). There is even some evidence that the PSM may not emanate from nascent levels of motivation but rather from being placed in a role working for the government (G. A. Brewer & Brewer, 2011).

According to this literature, confidence and trust in government institutions should be a key piece of this PSM. Indeed, a belief that public servants are motivated by altruistic concerns should increase confidence and trust in government. Regrettably, although many have drawn linkages between the importance of trust in government and basic concepts in public management (e.g., Brewer & Selden, 1998; Terry, 1998; Wamsley et al., 1990; cited in G. A. Brewer et al., 2000), we can find no empirical tests of this connection between public service and trust in government. It stands to reason that public servants would feel more positively toward government than their non-public-sector counterparts, because they understand the ambiguous goals, the inherent conflicts between due process and efficiency, and other tensions unique to public service, but we simply do not know whether this is the case. G. A. Brewer (2003) cites this very problem by stating that “overall, we know little about public servants’ civic attitudes and behavior beyond conventional wisdom and anecdotal accounts . . .” (p. 12).

Bozeman and Su (2014) identify four potential “drivers” of PSM—interests “inherent to the individual, a matter of socialization, related to occupation, or working in a context in which the focus is providing service” (p. 7)

These four models align with the performance/outcomes dichotomy, with the first two being process oriented (personal interest in the public good and “socialization” to the processes and objectives of the sector) and the later two being outcome oriented (occupational performance expectations and service delivery).

Houston’s (2011) recent distinction between types of motives in PSM is also relevant to our discussion. Houston argues “public service motivation translates into highly valuing obligation-based intrinsic motives but not enjoyment-based ones” (p. 769). This has a clear connection to the performance/outcomes disparity discussed earlier. More specifically, the “obligation” felt by public-sector employees to provide a more effective, efficient, and equitable process trumps “enjoyment” (however it may be defined). Confidence in institutions is clearly an obligation-based intrinsic motive rather than one that is “enjoyment-based,” providing more credence to our hypothesis that government employees will display higher levels of confidence in government institutions.

Data

Although scholars have used a variety of measures of trust and confidence in government, the most popular, by far, are the American National Election Studies (ANES) trust battery and the General Social Survey (GSS) questions about confidence in the leaders of various government institutions. Although trust and confidence in government are not the same concepts, they are closely aligned—both empirically and theoretically. They are also closely related to other concepts frequently measured in the literature such as generalized trust (Uslaner, 2002) and social capital (Putnam, 2000).

The ANES battery is probably the most frequently used measure in political science (e.g., Hetherington, 1998) and public administration (G. A. Brewer, 2003) but it is more sharply affected by “short-term evaluations of political events and leaders” (Cook & Gronke, 2005, p. 784) than other competing measures. The GSS measure,³ which asks about confidence in the leaders of institutions, taps into “generalized confidence” (Cook & Gronke, 2005, p. 784) and provides a stable indicator of the sort of prosocial attitude toward government we wish to measure. In addition, the GSS collects a variable that indicates whether a person works for a government, for-profit, or non-profit organization. The GSS has both the confidence and the public service variables available for data sets from 2000 to 2012.

The GSS, therefore, provides a readily available and reliable survey instrument to test our hypotheses. For the analysis that follows, we focus on GSS questions about the three major branches of the federal government: the executive, judicial, and legislative branches. Although we include a host of independent variables, we are most interested in the variable indicating whether the respondent is employed by a public or non-public organization. We expect that public service respondents will have more confidence in all leaders of government institutions than their counterparts in other fields, and we expect that the difference will be largest for the executive branch as it is likely closest to the nature of public service they experience in their jobs. Descriptive statistics for all variables used in our analysis are presented in Table 1.

Table 1.

Description of Variables.

	Observations	M	SD	Minimum	Maximum
Sex	18,945	1.55	0.50	1	2	1 = male; 2 = female (The SEX variable, excluding Don’t Know and Refused)
Age	18,945	47.12	17.5	18	99	self-reported age (The AGE variable, excluding DK and Refused)
Education	18,945	13.62	5.19	0	99	coded 0-20 based on years of education (The EDUC variable, excluding and D/K and refused)
Black	18,945	0.15	0.15	0	1	self-identified Black respondents are coded 1, all others are coded 0 (recode of the RACE variable)
White	18,945	0.77	0.42	0	1	self-identified White respondents are coded 1, all others are coded 0 (recode of the RACE variable)
Income	18,945	14.86	18.20	1	99	ordinal scale of income with higher numbers representing more income (the INCOME variable, excluding DK and refused)
Religious attendance	18,945	3.60	2.77	0	8	ordinal scale from never to several times a week (The ATTEND variable, excluding DK and refused)
Government worker	17,929	0.18	0.39	0	1	1 = works for government; 0 = all else (The WRKGO variable, excluding DK, No answer, and N/A)

Results

To gain a better understanding of the state of confidence in government over this time period, we begin our analysis by examining the data without considering the influence of public-sector service. Figure 1 presents the percentage of the general population who have a “great deal” of confidence in the leaders of each of the three institutions. The most obvious story is the anomalous case of 2002. Mostly likely in response to the shift of concern from domestic to international issues after 9/11 (Chanley, 2002), citizens expressed the highest levels of confidence in all three branches of government in the 2002 ANES. Although the size of the increase for the executive and legislative branches is striking, respondents expressed only slightly more confidence in the judicial branch in 2002 than they did in 2000—perhaps because their base level of confidence was already relatively high.

Figure 1.

Confidence in federal institutions 2000 to 2012.

By examining each institution separately, you can see that respondents have the highest level of confidence in the judicial branch, regardless of the year. Interestingly, respondents evidently drew little distinction between the executive and legislative branches in 2000, 2002, and 2008, whereas they seemed to have more confidence in the executive branch in 2004, 2006, 2010, and 2012. Based on this GSS measure, the legislative branch has experienced the largest decline is confidence. Since 2008, fewer than 10% of Americans express a “great deal of confidence in the legislative branch,” and the numbers have continued to decline in subsequent years.

Next, we examine each of these cross-sections by separating government and non-government employees. Figure 2 displays these data for the legislative branch, whereas Figure 3 displays them for the executive branch and Figure 4 for the judicial branch. Examining them one at a time, it is clear that status as a government employee had no obvious influence on confidence in the legislative branch from 2000 to 2008. In each of these years, the percentage of respondents who indicated a great deal of confidence in the legislative branch differed by no more than a percentage point. The data begin to diverge in 2010 and remain distinct in 2012 (although the size of the difference is clearly smaller than in 2010). Of perhaps the most interest is the direction of this relationship—whereas we might expect that public employees would hold higher confidence in the legislative branch, it is actually the non-public-sector employees who appear to have more confidence in the first branch of government. Although we do not know the causal mechanism here, it is possible that the divided congress that occurred in 2010 and 2012 led to some of this decline. It is important to note, however, that none of these differences are statistically significant.

Figure 2.

Confidence in the legislative branch by employee type 2000 to 2012.

Figure 3.

Confidence in the executive branch by employee type 2000 to 2012.

Figure 4.

Confidence in the judicial branch by employee type 2000 to 2012.

Figure 3 presents the descriptive data for the executive branch. Similar to the legislative data, public-sector and non-public-sector employees tend to hold similar attitudes from 2000 to 2008. There is also a familiar rise in both lines in 2002, again likely related to the “rally round the flag” effect following 9/11. It is also possible that some of this rise is a function of the change in administration that occurred in late 2000 to early 2001. It is well established that the public opinion toward the president tends to be largest immediately following his election, and it makes sense that this could translate to the executive branch in general. The change in administrations in 2008-2009 could also explain some of the patterns we see then. Similar to the legislative data, we begin to see the two lines separate in 2010 and 2012, but that is where the similarities end. Whereas public employees appear to display lower levels of trust in the legislative branch beginning in 2008, they display higher levels of trust in the executive branch beginning the same year. These bivariate differences are statistically significant (p < .05).

Figure 4 displays data for confidence in the judiciary spanning 2000 to 2012. There are distinct patterns of support for the judicial branch in all years but one (2008), although the size of the difference was relatively small and not statistically significant in 2002 and 2006. Although confidence in the judiciary among government employees took a steady decline from 2000 to 2010, non-public-sector employee confidence was more stable across this same time series. In 2012, government employees appear to have more confidence in the judiciary than non-public employees, but the pattern was reversed as recently as 2008. In sum, the judicial story is clearly distinct from the other two branches of government. Some of this could be a result of the distinct environment of the judiciary—a workforce that, according to the White House Office of Management and Budget, includes roughly 17,500 employees in FY 2011. Given the number of vacancies that have become the norm in lower courts, the “third branch” of government has clearly experienced its unique set of challenges.

Even though the judicial branch includes district and appellate courts as well, the “face” of the branch is the nine-member Supreme Court. The decline in trust in the judicial branch among public employees coincided with the rather politicized appointments and confirmations of Justices Roberts, Alito, Sotomayor, and Kagan. It is certainly possible that public employees were disconcerted by the increasingly politicized face of the Court.

Thus far, the data suggest that the relationship between public service and confidence in institutions varies across time and by institution. From 2000 to 2008, there were intermittent differences between how confident public and non-public employees were in the judicial branch and no differences in their confidence in the executive and legislative branches. In 2010 and 2012, public employees expressed less confidence in the legislative branch, but more confidence in the executive branch—at least compared with their non-public-sector counterparts. The results for the judicial branch are more mixed.

Although these results are potentially important, they do not control for any confounding factors. In particular, we know from the literature on confidence and trust in government that partisanship provides a lens that colors people’s opinions of government institutions. A simple analysis of GSS data also reveals that government employees are more likely to identify with the Republican Party than their non-public-sector counterparts (χ² = 113.89, p < .001). From this, it could be that some of the trends we see are really just partisanship in disguise.

To address this concern, we next present the results of a series of ordinal logistic regression models for each year in our study.⁴ The dependent variable for all of the models in Table 2 represents confidence in the executive branch. Table 3 presents results for the legislative branch and Table 4 for the judicial branch. Our independent variable of theoretical interest is the familiar coding for whether or not the respondent is a government worker (1 = yes; 0 = no). We also include many of the other variables that are often associated with confidence in government institutions, including partisanship, sex, age, education, race, income, and religious attendance.⁵

Table 2.

Explaining Confidence in the Executive Branch.

	2000	2002	2004	2006	2008	2010	2012
Republican Party ID	−.192***	.290***	.365***	.392***	.237***	−.340***	−.316***
Republican Party ID	(.026)	(.037)	(.038)	(.027)	(.030)	(.032)	(.033)
Government worker	−.025	.095	.130	.148	−.048	.514***	.311**
Government worker	(.136)	(.174)	(.178)	(.122)	(.146)	(.140)	(.149)
Sex	.075	−.151	−.099	−.009***	.142	−.111	−.006
Sex	(.098)	(.136)	(.140)	(.003)	(.114)	(.115)	(.116)
Age	−.008***	−.002	−.011**	−.009***	−.016***	−.007**	−.019***
Age	(.003)	(.004)	(.004)	(.003)	(.003)	(.003)	(.003)
Education	.008	.005	−.047*	−.022*	−.019	.002	.047**
Education	(.011)	(.018)	(.025)	(.013)	(.015)	(.011)	(.019)
Black	−.026	.044	−.154	−.458**	−.136	−.182	−.234
Black	(.023)	(.342)	(.323)	(.187)	(.247)	(.240)	(.237)
White	−.311	.415	.026	−.562***	−.411**	−.485**	−.311
White	(.212)	(.302)	(.257)	(.146)	(.208)	(.201)	(.193)
Income	.006**	.000	.003	.003	.005	.002	.006*
Income	(.003)	(.005)	(.004)	(.003)	(.004)	(.003)	(.003)
Religious attendance	−.003	.062**	.109***	.058***	.046**	.000	.023
Religious attendance	(.018)	(.026)	(.027)	(.017)	(.021)	(.021)	(.021)
n	1,624	831	797	1,798	1,224	1,212	1,182
χ²	89.82**	90.32*	145.31***	307.54***	100.86***	176.63***	176.80***

Note. Entries are ordinal logistic regression coefficients. Numbers in parentheses are standard errors.

p < .1. ***p < .05. ***p < .01.

Table 3.

Explaining Confidence in the Legislative Branch.

	2000	2002	2004	2006	2008	2010	2012
Republican Party ID	.019	.009	.048	.111***	−.045	−.223***	−.002
Republican Party ID	(.027)	(.038)	(.037)	(.025)	(.029)	(.032)	(.033)
Government worker	.010	−.085	.197	−.041	.064	−.046	.224
Government worker	(.140)	(.182)	(.182)	(.122)	(.146)	(.148)	(.153)
Sex	.296***	.137	.161	.116	.377***	.243**	.400***
Sex	(.101)	(.143)	(.143)	(.094)	(.114)	(.118)	(.120)
Age	−.301***	−.011***	−0.016***	−.014***	−.021***	−.023***	−.024***
Age	(.003)	(.004)	(.004)	(.003)	(.003)	(.004)	(.004)
Education	−.007	.007	−.049*	−.006	.005	−.059***	−.068***
Education	(.010)	(.016)	(.026)	(.010)	(.013)	(.019)	(.020)
Black	−.041	.140	−.632*	−.635***	−.008	.084	−.548**
Black	(.248)	(.366)	(.342)	(.190)	(.246)	(.253)	(.243)
White	−.369*	.343	−.385	−.588***	−.157	−.229	−.652***
White	(.216)	(.324)	(.275)	(.151)	(.207)	(.212)	(.198)
Income	.002	−.003	.008**	.003	.003	.001	.006**
Income	(.003)	(.005)	(.004)	(.003)	(.004)	(.003)	(.003)
Religious attendance	.061***	.047*	.065**	.053***	.051**	.052**	.065***
Religious attendance	(.089)	(.027)	(.027)	(.017)	(.021)	(.022)	(.022)
n	1,631	834	798	1,793	1,230	1,210	1,174
χ²	47.98***	12.16	31.80	81.00	61.93***	144.43***	98.18

Note. Entries are ordinal logistic regression coefficients. Numbers in parentheses are standard errors.

p < .1. ***p < .05. ***p < .01.

Table 4.

Explaining Confidence in the Judicial Branch.

	2000	2002	2004	2006	2008	2010	2012
Republican Party ID	−.050*	.095	.029	.136***	.093***	−.044	.001
Republican Party ID	(.026)	(.036)	(.036)	(.025)	(.029)	(.030)	(.031)
Government worker	.290**	.015	.273	.006	.019	.011	.322**
Government worker	(.139)	(.175)	(.179)	(.120)	(.145)	(.139)	(.149)
Sex	−.174*	.048	−.070	−.090	−.116	−.223*	−.137
Sex	(.099)	(.139)	(.139)	(.093)	(.113)	(.115)	(.116)
Age	−.011***	−.003	−.011***	−.003	−.004	−.005	−.005
Age	(.003)	(.004)	(.004)	(.003)	(.003)	(.003)	(.003)
Education	.059***	.009	.050**	.045***	.020	.005	.037*
Education	(.017)	(.016)	(.025)	(.014)	(.018)	(.011)	(.020)
Black	−.403	−1.03	−.708**	−.948***	−.717	−.291	−.898***
Black	(.247)	(.358)	(.335)	(.186)	(.243)	(.240)	(.238)
White	−.210	−.390	−.013	−.330**	−.359*	−.313	−.568***
White	(.215)	(.301)	(.267)	(.148)	(.203)	(.120)	(.192)
Income	.005	−.004	.003	.001	−.002	.002	−.001
Income	(.003)	(.005)	(.004)	(.002)	(.004)	(.003)	(.003)
Religious attendance	−.024	−.012	.012	.010	−.020	.016	.022
Religious attendance	(.019)	(.026)	(.027)	(.017)	(.021)	(.021)	(.021)
n	1,601	824	791	1,777	1,220	1,198	1,171
χ²	54.90***	25.05***	28.03***	97.20***	30.31***	11.78	28.99***

Note. Entries are ordinal logistic regression coefficients. Numbers in parentheses are standard errors.

p < .1. ***p < .05. ***p < .01.

Table 2 largely confirms the story we previously reviewed in Figure 2. Government workers are no more likely to express confidence from 2000 to 2008. The year 2010, however, marked a significant shift with government works suddenly expressing significantly more confidence in executive branch leaders. It is important to note that this relationship holds even controlling for political partisanship, sex, age, education, race, income, and religious attendance. Something clearly happened between 2008 and 2010 to change the story.

Table 3 presents the results of a series of models predicting confidence in the legislative branch. Here, the story differs a bit from Figure 2. Once controlling for other factors, there is no independent effect of working for government on confidence in the legislative branch of government. This runs counter to expectations from the PSM literature and provides further evidence that government employees are not inherently more confident in government institutions than their non-public-sector counterparts.

Table 4 presents multivariate models explaining confidence in the judicial branch. Recall that the bivariate data presented in Figure 3 presented no clear pattern about confidence in the judicial among public and non-public employees. Table 4 reveals a similarly murky picture—although government employees expressed more confidence in the judicial branch in 2000 and 2012, they did not in the rest of the time series.

Once controlling for other factors, it appears that public and non-public employees currently express different levels of confidence in the executive branch, identical levels of confidence in the legislative branch, and different levels of confidence in the judicial branch. To learn more about the substantive impact of our executive branch findings, we used CLARIFY (Tomz, Wittenberg, & King, 2003) to estimate the probability of confidence in the executive branch by year and employment status. For this analysis (presented in Table 5), we held all other variables at their sample means.

Table 5.

Predicted Probability of Confidence in the Executive Branch by Year and Public/Non-Public Employer.

	2010		2012
	Non-public employee	Public employee	Non-public employee	Public employee
Hardly any	.37	.26	.37	.30
Only some	.49	.53	.51	.54
A great deal	.13	.21	.12	.16

Note. Estimates created using CLARIFY. All other variables were held at their sample means.

The results in Table 5 clearly indicate that employment status has an important but not overwhelming influence on confidence in government. A non-public employee has a .13 probability of expressing a “great deal” of confidence in the executive branch in 2010 vs. a .21 probability for a public employee. In 2012, the substantive impact is a little smaller—amounting to a difference of .12 versus .16.

Conclusion

This article began with a deceptively simple question: Do government employees have more confidence in government institutions than their counterparts in other sectors? We believed that the answer would be yes. After all, it stands to reason that public servants would support the very institutions they work for. Furthermore, the PSM literature suggests that government employees are generally more engaged with political activities (Ertas, 2014a) and trusting (G. A. Brewer, 2003) than their private-sector counterparts. We had every expectation that this would translate to confidence and trust in government institutions.

Our simple question and fairly straightforward hypothesis was not borne out by the data. After estimating multivariate models with a host of controls, it appears that government employees have no more confidence in the legislative branch than their non-public-sector counterparts. Government employees are a part of the legion of Americans who “hate Congress.” There is nothing inherent in the public servant experience that gives our respondents any more confidence in this core institution of American government.

The results surrounding the executive branch are similarly important, if a little more conditional. After controlling for other factors, government employees appear to have expressed no more confidence in the executive branch from 2000 to 2008. In 2010 and 2012, however, public employees expressed substantively more confidence in the executive branch. By examining the bivariate data, we see that the average American expressed more confidence in the executive branch in 2010, but the slope was steeper for public than non-public employees. The question, of course, is why this occurred.

The results for the judicial branch demonstrate that, for the most part, government employees are no more confident in the third branch of government than are their counterparts in other sectors. Government workers did express more confidence in 2000 and 2012, but this effect is not found elsewhere and clearly does not support the model of generalized confidence we had expected at the outset.

In total, our results demonstrate that although public employees may in fact have different values, motivations, and experiences than their non-public-sector counterparts (Perry & Wise, 1990), they do not express more civic minded-attitudes toward government institutions. That is, there is not a substantive difference in the public employee’s perceived investment in, or benefit from government attributable to their sector of employment. As the title of this article suggests, these data demonstrate that there is a limit of PSM. Public servants do not inherently have more confidence in the institutions of government than their non-public-sector counterparts. Instead, their confidence waxes and wanes and is often indistinguishable from their counterparts in other sectors.

We would be remiss if we did not note that our article raises as many questions as it answers. In particular, we have very little understanding for why government employees tended to have more confidence in the executive branch beginning in 2010. It will be important to test this finding in the future with data from Republican presidencies to see whether this is a pattern that is contingent on partisanship, or whether it represents a more durable shift.

Our data are also inherently limited—we rely on a very narrow measure of confidence in government using GSS data. As T. E. Cook and Gronke (2004) explicate, the GSS measure is not without its problems, and it will be important to see whether these results can be replicated with other measures of confidence and contrasting them empirically with measures of the related concept of trust. It is also noteworthy that the GSS measure of trusts asks specifically about confidence in the leaders of these institutions, rather than the institutions themselves. Although we cannot be certain whether respondents comprehend this distinction, future studies should attempt to untangle these measurement problems.

Like many other studies in the literature, we use a dummy variable to indicate whether a person is a public-sector employee, but, as all scholars of public administration know, all public-sector jobs are not created equally. In addition, our measure lumps all non-public service jobs together, leaving private-sector and non-profit jobs in the same category. The next step in this research agenda could be to compare public, private, and non-profit sectors, or to see whether employment sector interacts with job type. It would not be surprising if white-collar public employees have a closer tie to the idea of public service than their blue-collar counterparts in the public service. This might very well translate into different patterns of support for government institutions.

We have also focused exclusively on federal branches of government, but there is an important and growing body of literature examining confidence and trust in state (Kelleher & Wolak, 2007) and local (Rahn & Rudolph, 2005) institutions of government. Many of our respondents work for these subnational governments, and it could be that it is at the subnational level that we witness the greatest differences between confidence in public institutions among public and non-public employees.

Despite these limitations, these results suggest an important caution to the PSM literature and suggest that public employees are not inherently any more trusting of the institutions of government than their counterparts in other sectors. There are, indeed, limits of PSM.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Christopher A. Cooper is a professor and Department Head in the Department of Political Science and Public Affairs and Professor by Courtesy in the Department of Psychology at Western Carolina University.

Tyler Reinagel is an assistant professor in the Department of Political Science and Public Affairs at Western Carolina University.

References

Almond

Verba

(1963). The civic culture: Political attitudes and democracy in five nations. Thousand Oaks, CA: Sage.

Bellante

Link

A. N.

(1981). Are public sector workers more risk averse than private sector workers? Industrial and Labor Relations Review, 34, 408-412.

Boyne

G. A.

(2002). Public and private management: What’s the difference? Journal of Management Studies, 389, 97-122.

Bozeman

(2014). Public service motivation concepts and theory: A critique. Public Administration Review. Advance online publication. doi:10.1111/puar.12248

Brewer

G. A.

(2003). Building social capital: Civic attitudes and behavior of public servants. Journal of Public Administration Research and Theory, 13, 5-26.

Brewer

G. A.

Brewer

G. A.

Jr. (2011). Parsing public/private differences in work motivation and performance: An experimental study. Journal of Public Administration Research and Theory, 21, 347-362.

Brewer

G. A.

Selden

S. C.

Facer

R. L.

(2000). Individual conceptions of public service motivation. Public Administration Review, 60, 254-264.

Brewer

P. R.

Sigelman

(2002). Trust in government: Personal tie that bind? Social Science Quarterly, 83, 624-631.

Brief

A. P.

Motowidlo

S. J.

(1986). Prosocial organizational behaviors. Academy of Management Review, 11, 710-725.

10.

Chanley

V. A.

(2002). Trust in government in the aftermath of 9/11: Determinants and consequences. Political Psychology, 23, 469-483.

11.

Christensen

Lægreid

(2005). Trust in government: The relative importance of service satisfaction, political factors, and demography. Public Performance & Management Review, 28, 487-511.

12.

Citrin

Green

D.L.

(1986). Political leadership and the resurgence of trust in government. British Journal of Political Science, 16, 431-53.

13.

Cook

F. L.

Jacobs

L. R.

Kim

(2010). Trusting what you know: Information, knowledge and confidence in social security. Journal of Politics, 72, 397-412.

14.

Cook

T. E.

Gronke

(2005). The skeptical American: Revisiting the meanings of trust in government and confidence in institutions. Journal of Politics, 67, 784-803.

15.

Cooper

C. A.

Knotts

H. G.

Brennan

(2008). The importance of trust in government in public administration. Public Administration Review, 68, 459-468.

16.

Ertas

(2014a). Political voice and civic attentiveness of public and nonprofit employees. The American Review of Public Administration, 44, 1-20.

17.

Ertas

(2014b). Public service motivation theory and voluntary organizations: Do government employees volunteer more? Nonprofit and Voluntary Sector Quarterly, 43, 254-271.

18.

Glaser

Hildreth

W. B.

(1999). Service delivery satisfaction and willingness to pay taxes. Public Productivity & Management Review, 23, 48-67.

19.

Herian

M. N.

Shank

N. C.

Abdel-Monem

R. L.

(2014). Trust in government and support for governmental regulation: The case of electronic health records. Health Expectations, 17, 784-794. doi:10.1111/j.1369-7625.2012.00803.x

20.

Hetherington

M. J.

(1998). The political relevance of political trust. American Political Science Review, 92, 791-808.

21.

Hetherington

M. J.

(2005). Why trust in matters: Declining political trust and the demise of American liberalism. Princeton, NJ: Princeton University Press.

22.

Hetherington

M. J.

Rudolph

T. J.

(2008). Priming, performance, and the dynamics of political trust. Journal of Politics, 70, 498-512.

23.

Hibbing

(2002). How to make Congress popular. Legislative Studies Quarterly, 27, 219-244.

24.

Hibbing

Theiss-Morse

(1995). Congress as public enemy: Public attitudes toward American political institutions. New York, NY: Cambridge University Press.

25.

Hibbing

Theiss-Morse

(2002). Stealth democracy: Americans’ belief about how goverrnment should work. New York, NY: Cambridge University Press.

26.

Houston

D. J.

(2011). Implications of occupational locus and focus for public service motivation: Attitudes toward work motives across nations. Public Administration Review, 71, 761-771.

27.

Kelleher

Wolak

(2007). Explaining public confidence in the branches of state government. Political Research Quarterly, 60, 707-721.

28.

Lyons

S. T.

Duxbury

L. T.

Higgins

C. T.

(2006). A comparison of the values and commitment of private sector, public sector, and parapublic sector employees. Public Administration Review, 66, 605-618.

29.

Miller

A. H.

(1974). Political issues and trust in government: 1964-1970. American Political Science Review, 68, 951-972.

30.

Perry

J. L.

Wise

L. R.

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

31.

Pew Center for the People and the Press. (2013, October 18). Public trust in government: 1958–2013. Retrieved from http://www.people-press.org/2013/10/18/trust-in-government-interactive/

32.

Putnam

R. D.

(1971). Studying elite political culture: The case of “Ideology.” The American Political Science Review, 65, 651-681.

33.

Putnam

R. D.

(1973). The political attitudes of senior civil servants in Western Europe: A preliminary report. British Journal of Political Science, 3, 257-290.

34.

Putnam

R. D.

(2000). Bowling alone. New York, NY: Simon & Schuster.

35.

Rahn

W. M.

Rudolph

T. J.

(2005). A tale of political trust in American cities. Public Opinion Quarterly, 69, 530-560.

36.

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.

37.

Richardson

L. E.

Houston

D. J.

Hadjiharalambous

C. S.

(2001). Public confidence in the leaders of American governmental institutions. In Hibbing

Theiss-Morse

(Eds.), What is it about government that Americans dislike? (pp. 83-97). Cambridge, UK: Cambridge University Press.

38.

Ruscio

K.P.

(1996). Trust, democracy, and public management: A Theoretical Argument. Journal of Public Administration Research and Theory, 6(3), 431-477.

39.

Schoen

D. E.

Caddell

P. H.

(2013, December 1). America’s crisis of trust. Politico. Retrieved from http://www.politico.com/magazine/story/2013/12/americas-crisis-of-trust-100504.html#.UugSOfYo69Y

40.

Terry

L. D.

(1998). Administrative leadership, neo-managerialism, and the public management movement. Public Administration Review, 58, 194-200.

41.

Tomz

Wittenberg

King

(2003). CLARIFY: Software for interpreting and presenting statistical result. Journal of Statistical Software, 8.

42.

Tyler

T. R.

(1990). Why people obey the law. New Haven, CT: Yale University Press.

43.

Uslaner

(2002). The moral foundations of trust. New York, NY: Cambridge University Press.

44.

Van

Walle

Bouckaert

(2003). Public service performance and trust in government: The problem of causality. International Journal of Public Administration, 26, 891-913.

45.

Van Ryzin

G. G.

(2011). Outcomes, process, and trust of civil servants. Journal of Public Administration Research and Theory, 21, 745-760.

46.

Williams

(2006). Generalized ordered logit/partial proportional odds models for ordinal dependent variables. The Stata Journal, 6, 58-82.

47.

Yang

Holzer

(2006). The performance-trust link: Implications for performance measurement. Public Administration Review, 66, 114-126.