MATH GURU ONLY

Review of Public Personnel Administration 31(2) 111 –127

© 2011 SAGE Publications Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

DOI: 10.1177/0734371X10386184 http://roppa.sagepub.com

ROP386184ROP31210.1177/0734371X10386184Llorens and StazykReview of Public Personnel Administration

1Louisiana State University, Baton Rouge 2American University, Washington, DC

Corresponding Author: Jared J. Llorens, Assistant Professor, Louisiana State University, Public Administration Institute 3200 Patrick F. Taylor Hall, Baton Rouge, Louisiana 70803 Email: [email protected]

How Important Are Competitive Wages? Exploring the Impact of Relative Wage Rates on Employee Turnover in State Government

Jared J. Llorens1 and Edmund C. Stazyk2

Abstract

In recent years, public management research has made great strides in explaining the drivers of employee turnover in the public sector, with key findings related to the role of employee loyalty, organizational satisfaction, person-organization fit, and compensation. This article contributes to this growing body of literature by assessing the influence of a previously untested driver of employee turnover at the state level of government: public–private wage equity. Contrary to conventional wisdom, results suggest that public–private wage equity does not significantly influence voluntary separation rates, whereas state government unionization and the average age of state government employees are found to be indirectly related to voluntary separation. Results also point to the potential implications of ethnicity, gender, and public service motivation in state government employee turnover and provide key insights for those seeking to further understand the impact of reduced expenditures on public sector wages and shifting age distributions in public sector employment.

Keywords

employee turnover, compensation, public service motivation

Articles

112 Review of Public Personnel Administration 31(2)

Introduction

In recent years, public management scholarship has devoted a great deal of effort toward explaining the drivers and extent of turnover in the public sector. At its heart, turnover is a topic of great concern for public managers since excessive turnover rates may substantially limit organizational productivity. Concerns about the potentially harmful effects of excessive turnover are further heightened when viewed in light of the relatively lengthy and complex hiring processes of most public sector organiza- tions as well as the unique qualifications required of many public sector occupations. Whereas private sector organizations generally operate at a competitive advantage in terms of the ability to hire new candidates quickly and “on the spot,” public organiza- tions and their managers routinely face hiring cycles of 3 months or greater.

This article contributes to the growing literature on public employee turnover by pro- viding an explanatory model of turnover that includes a previously untested measure of state government compensation: public–private wage equity. Prior research has found compensation to be a significant determinant of public employee turnover (Lee & Whitford, 2008; Selden & Moynihan, 2000) but has yet to explore the distinct role relative wage rates play in the turnover process. Although overall compensation may affect turn- over, we seek to assess the unique role of public sector wage rates, when compared to those offered in the private sector, on employee turnover. Conventional wisdom in the field of public sector compensation has held that when public sector wage rates fall below comparable private sector wage rates, turnover in favor of private sector employ- ment is to be expected; however, to date, this assumption has yet to be fully tested. We begin with a review of the relevant literature on turnover and then describe the determi- nants of our explanatory model of state government voluntary separation. Last, we pres- ent the results of our analysis and discuss implications for future research.

Public Employee Turnover: An Overview Employee turnover is a topic of immense importance to public and private sector organizations. In part, this importance reflects the tremendous costs—financial and otherwise—often associated with turnover (Balfour & Neff 1993; Staw, 1980). Financially, turnover may lead to increased personnel expenses, particularly in the areas of recruitment and training (Balfour & Neff, 1993; Staw, 1980). However, employee turnover can also bring about a loss of organizational knowledge, history, and memory (Moynihan & Pandey, 2008; Staw, 1980). Scholars and practitioners have long recognized such costs may reduce overall organizational performance and often impose substantial burdens on organizations and managers (Balfour & Neff, 1993; Bertelli, 2007; Mobley, Griffeth, Hand, & Meglino, 1979; Moynihan & Pandey, 2008; Selden & Moynihan, 2000; Staw, 1980). Given the potential implications of these costs for organizations, a great deal of research on the causes, or drivers, of public and pri- vate sector turnover exists. Interestingly, there is substantial overlap between the theo- retical context of turnover in the public and private sectors (Moynihan & Pandey, 2008), with much of the literature pointing to employee satisfaction—or, more specifically,

Llorens and Stazyk 113

dissatisfaction—as the primary, overarching determinant of turnover (Bright, 2008; Lee, Gerhart, & Trevor, 2008; Moynihan & Pandey, 2008).

Employee job satisfaction is the most frequently used and single most reliable pre- dictor of turnover (Moynihan & Pandey, 2008). In this case, employees expressing high levels of job satisfaction seem less likely to leave their organizations. The bene- fits of using job satisfaction as a predictor of turnover trace in part to the concept’s overlap with several other important organizational factors typically associated with separation, including job routineness, pay and promotion, goal and role clarity/conflict, procedural constraints, organizational involvement, supervisory style, promotional opportunities, and employee burnout (see Bertelli, 2007; Kim, 2005; Moynihan & Pandey, 2008, p. 208; Rubin, 2008). Building on the apparent relationship between job satisfaction and turnover, other scholars have suggested an employee’s level of job involvement (Bertelli, 2007; Felps et al., 2009), intrinsic motivation (Bertelli, 2007; Bright, 2008), and overall organizational satisfaction (Bright, 2008; Lee & Whitford, 2008; Moynihan & Pandey, 2008) may also provide insight into an employee’s turn- over decision. That said, job satisfaction alone may not adequately capture turnover or turnover intentions. For instance, other factors like organizational commitment and adequate pay may offset job satisfaction or more directly influence turnover decisions—thereby limiting the concept’s usefulness as a single predictor of turnover (e.g., Felps et al., 2009; Kim, 2005; Mobley et al., 1979; Marsh & Mannari, 1977; Mobley, Horner, and Hollingsworth, 1978; Moynihan & Pandey, 2008). Fortunately, research suggests a variety of other drivers of turnover exist and that these drivers tend to fall into three distinct categories: (a) external (to the organization) environmental factors, (b) individual factors, and (c) organizational factors (e.g., Mobley et al., 1979; Moynihan & Pandey, 2008; Selden & Moynihan, 2000).

External, environmental factors tend to account for the role economic conditions play in shaping or driving turnover and turnover intentions (McCabe, Feiock, Clingermayer, & Stream 2008; Mobley et al., 1979; Moynihan & Pandey, 2008). For instance, research has long recognized the perceived availability and evaluation of alternative job opportunities by employees influences turnover (Cotton & Tuttle 1986; Forrest, Cummings, & Johnson, 1977; Lee et al., 2008; Locke, 1976; March & Simon, 1958; Mobley et al., 1979; Park, Ofori-Dankwa, & Bishop, 1994; Price, 1977). Workers employed in areas with strong, positive economic conditions and a large number of attractive job opportunities are more likely to leave their positions. These findings provide strong evidence geographical economic variations play a powerful role in influencing turnover and turnover intentions.

In addition to environmental forces, scholars have found numerous individual fac- tors also drive turnover decisions. Early research suggested age, tenure (or length of service with an organization), sex, race, family responsibilities, education, personality, and other personal considerations (like number of previous jobs held) often affect turn- over (see e.g., Mobley et al., 1979; Moynihan & Pandey 2008; Selden & Moynihan, 2000). More specifically, findings suggested those employees who were older, had longer tenure with an organization, were White males, were married, and were highly educated tended to be more likely to remain with their organizations. Later research

114 Review of Public Personnel Administration 31(2)

has confirmed the relevance of many of these factors on both turnover intentions and actual turnover (see e.g., Moynihan & Pandey, 2008). However, some important exceptions exist. For instance, recent research finds little evidence women and minori- ties leave organizations more frequently than White men (Meier, Mastracci, & Wilson, 2006; Moynihan & Pandey, 2008). Nevertheless, research clearly demonstrates indi- vidual factors may influence employee turnover and turnover intentions.

Finally, prior research also indicates organizational factors are frequently important pre- dictors of voluntary turnover (Mobley et al., 1979; Moynihan & Pandey, 2008; Selden & Moynihan, 2000). As Moynihan and Pandey (2008) point out, research in this area more accurately examines how the “interaction of individual employees and the characteristics of their organizations” drive turnover decisions (p. 207). Consequently, organizational fac- tors are regularly linked to issues of job satisfaction, motivation, and organizational com- mitment (Bertelli, 2007; Moynihan & Pandey, 2008). However, research in this area also tends to examine more fully the role pay and compensation, promotion, training, supervi- sion, organizational and group culture, and human resource policies assume in shaping an employee’s turnover decision (Bertelli, 2007; Felps et al., 2009; Kim, 2005; Mossholder, Settoon, & Henagan, 2005; Moynihan & Pandey, 2008; Rubin, 2008).

Generally, findings suggest opportunities for advancement (Kim, 2005; Lee & Whitford, 2008; Selden & Moynihan, 2008), supportive supervision and management (Bertelli, 2007; Chang, 2009) and human resource policies (Rubin, 2008), and positive organizational and group cultures (Bertelli, 2007; Felps et al., 2009; Mossholder et al., 2005; Moynihan & Pandey, 2008) tend to decrease turnover. Results examining the role of training and turnover have been mixed (Moynihan & Pandey, 2008). In some cases, training seems to promote retention (Curry, McCarragher, & Dellmann-Jenkins, 2005; Moynihan & Pandey, 2008); in others, the increased marketability associated with the training process may make employees more likely to leave their organizations (Ito, 2003; Moynihan & Pandey, 2008). Finally, prior research has found average pay rates (Selden & Moynihan, 2000) and employee satisfaction with pay (Bertelli, 2007; Lee & Whitford, 2008) are both significant predictors of voluntary turnover and turnover intent. Interestingly, the unique role competitive wage rates may play on voluntary employee turnover remains unexplored.1 Consequently, this article seeks to examine whether public sector wage rates, when compared to those offered in the private sector, affect employee turnover. Testing this unique aspect of public sector compensation will provide a better understanding of the relationship between compensation and employee turnover, a relationship that has only grown in importance in light of the current eco- nomic crisis gripping state and local governments across the United States.

Data and Method Measuring Employee Turnover

As Selden and Moynihan (2000) note, employee turnover may be either voluntary or involuntary. Voluntary turnover reflects instances in which employees leave an orga- nization of their own volition, whereas involuntary turnover represents cases where

Llorens and Stazyk 115

employees are involuntarily released by their employers. Although involuntary turn- over rates have become increasingly relevant in the current labor market, we choose to examine the phenomenon of voluntary employee separation, which is arguably influenced by factors distinctive from those affecting involuntary employee separa- tions. In particular, we use 2007 voluntary separation rates for classified state govern- ment employment provided by the National Association of State Personnel Executives (NASPE).2 As the professional association for state government personnel executives, NASPE regularly conducts a number of member surveys on key indicators of state government personnel performance of which state government employee turnover is a component. Table 1, above, lists voluntary separations rates for 2007.3

Determinants of Employee Turnover Public–private wage equity. Prior research has found employee compensation to be a

significant predictor of employee turnover in the case of state and federal government

Table 1. Voluntary Separation Rates by State, 2007

State 2007 State 2007

Alabama 6.9 Montana — Alaska — Nebraska 8.0 Arizona 15.5 Nevada 8.0 Arkansas 12.2 New Hampshire 6.3 California 7.5 New Jersey 2.6 Colorado 9.1 New Mexico 10.2 Connecticut 2.4 New York 2.9 Delaware 3.7 North Carolina 9.1 Florida — North Dakota 6.6 Georgia 2.5 Ohio 2.0 Hawaii — Oklahoma 9.3 Idaho 9.3 Oregon 5.9 Illinois 1.9 Pennsylvania 2.1 Indiana 13.0 Rhode Island — Iowa 2.5 South Carolina 8.2 Kansas 9.0 South Dakota — Kentucky — Tennessee 6.6 Louisiana 10.9 Texas 10.8 Maine 6.7 Utah 7.7 Maryland 7.1 Vermont 5.2 Massachusetts 4.6 Virginia 9.2 Michigan 2.0 Washington 4.1 Minnesota 6.1 West Virginia 6.9 Mississippi 10.5 Wisconsin 4.3 Missouri 8.4 Wyoming 9.3

116 Review of Public Personnel Administration 31(2)

employment (Lee & Whitford, 2008; Selden & Moynihan, 2000). However, to date, this predictor has yet to be operationalized in a manner that fully assesses public employee wage rates relative to those rates offered in the private sector. In many respects, it is this comparison that has functioned as the primary driver of discussions concerning public– private wage equity, with public managers at all levels of government commonly professing their inability to attract and/or retain high quality candidates because of noncompetitive wage rates (e.g., Partnership for Public Service, 2009). Quite often, discussions centered on pay equity or pay comparability have taken on political tones with advocates of smaller government asserting that government employees are simply paid too much and supporters of public employment asserting that public employees are not sufficiently compensated given their duties and responsibilities.

Given the controversy that the topic of public sector pay can engender, a number of methodological approaches have been employed to estimate the extent of pay equity between the public and private sectors and, as Miller (1996) notes, even one’s choice of methodology can often be guided by ideological biases. One common, base-level approach is to compare the average salary rates of public and private sector employees for a specific geographic region or level of government. While appealing on the basis of its conceptual conciseness, this approach has often been criticized for providing a misleading depiction of pay equity since it does not take into account the substantial variation in occupations that exist between the public and private sectors. In particular, this approach includes the entire segment of service industry employment in its evalu- ation of private sector pay rates, which has the effect of driving down average private sector pay rates because of the relatively low wage structure of the service industry. Since most public sector organizations have no occupational equivalent to those jobs within the private sector service industry, their average pay rates would appear to be relatively higher. However, when disaggregated by occupation one might find that, in fact, public sector employees experience pay rates actually lower than those offered in the private sector. For example, highly skilled occupations in medicine and the sci- ences are compensated at rates in the private sector labor market that simply cannot be matched by most public sector organizations.

To address this methodological shortcoming and provide more accurate assess- ments of public–private wage equity, academic research on public–private pay equity has sought to control for factors commonly thought to influence public and private sector pay rates to provide more accurate assessments of public-private wage equity. Furthermore, prior research has commonly employed two distinct, but related, meth- odologies for estimating the presence of equitable wage rates within the labor market. The first method consists of estimating a wage equation with employee wage as the dependent variable and factors thought to influence wage rates as independent vari- ables. Such factors generally include education, occupation, industry, and age. In addi- tion, dummy variables can be included into the wage equation to capture the impact of select characteristics on wages. In the case of research on public–private wage equity, a dummy variable for sector of employment can be added to measure the impact of sector of employment on wages.

Llorens and Stazyk 117

The second common methodological approach for estimating wage equity is the Blinder–Oaxaca wage decomposition procedure (Blinder, 1973; Oaxaca, 1973). This approach consists of estimating two separate wage equations, one for a base group and one for a comparison group. For example, when estimating male/female wage gaps, males could serve as the base group and females could serve as the comparison group. Next, a projected wage is estimated using the mean characteristics, or human capital endowments, of the comparison group and the coefficient values estimated in the wage equation for the base group. Any resulting differential in wages not attributable to dif- ferences in human capital endowments is said to be evidence of a wage differential based on the characteristic that distinguishes the base group from the comparison group. The benefit of this approach is that it takes into account the differences in endowments between the two groups and allows researchers to estimate the wage that the comparison group would expect to receive if the market valued its human capital endowments in the same manner as the base group.

For this analysis, we have chosen to utilize the Blinder–Oaxaca wage decomposi- tion procedure to estimate disaggregated, public–private wage differentials on a state- by-state basis for year 2007.4 In particular, we use wage data from the U.S. Bureau of Labor Statistics’ Current Population Survey (CPS, 2007a). The CPS draws on a monthly sample of the U.S. labor force, and, along with a host of other survey items on respondent wages and sector of employment, it also captures a number of key human capital characteristics such as education, age, and marital status.5 Using private sector employees as a base group and state government employees as a comparison group, the decomposition procedure, shown below, calls for the estimation of two separate log wage equations, where H represents the following human capital characteristics commonly thought to influence wage: age, age-squared, full-time employment status, gender, race, educational attainment, occupation, marital status, and location in a rural or urban working environment.6

Base Group—Private (In wage i ) = αi (base) + βHi + ui

Comparison Group—Public (In wage i ) = αi (comparison) + βHi + ui

Next, an overall wage differential is projected using estimates provided by the wage equations in Step 1. The first component of the overall wage differential is composed of the difference in the constants between the base and comparison wage regressions, αi (base)-αi (comparison). The second component, as Blinder (1973) explains, is the “difference between how the high-wage [base] equation would value the characteristics of the low-wage [comparison] group and how the low-wage equation actually values them” (p. 438). This portion of the wage differential is defined as Σ

i x-

i (comparison)

(βHi (base)-βHi (comparison) ). These two portions are summed to obtain an overall wage differential between public and private sector employment, which is expressed as a percentage difference. For ease of interpretation, we transform this differential into an earnings ratio. For example, if it is found that in a given state there is 3% wage differential for working in state government, when holding all other human capital

118 Review of Public Personnel Administration 31(2)

characteristics constant; this can be expressed as an earnings ratio of 97% since state government employees would be found to earn approximately 97% of the wages of a comparable private sector employee. To provide for added consistency, wage differen- tials for each state are estimated on a 3-year basis such that the estimates of wage dif- ferentials for 2007 are based on data from years 2005-2007. Table 2, above, lists state government earnings ratio data for 2007. Given prior research linking public employee pay rates to turnover, we predict that state government earnings ratios will be indi- rectly related to voluntary separation rates.

Political ideology. Existing research on state government employment has com- monly tested the impact of a state’s political ideology on key metrics such as bureau- cratic representation and wage equity (Brewer & Selden, 2003; Lewis & Nice, 1994; Llorens, 2008; Llorens, Wenger, & Kellough, 2008). The intent of incorporating this measure has been to assess the impact of state attitudes toward state government employ- ment, with the assumption being that more liberal states tend to view government

Table 2. Public/Private Wage Differential (Earnings Ratio %), 2005-2007

State % State %

Alabama 96.1 Montana — Alaska — Nebraska 90.0 Arizona 97.4 Nevada 99.0 Arkansas 92.3 New Hampshire 93.2 California 101.5 New Jersey 104.4 Colorado 97.5 New Mexico 90.7 Connecticut 102.7 New York 94.9 Delaware 101.0 North Carolina 91.9 Florida — North Dakota 97.3 Georgia 86.8 Ohio 102.0 Hawaii — Oklahoma 91.9 Idaho 85.8 Oregon 94.7 Illinois 95.8 Pennsylvania 98.7 Indiana 91.1 Rhode Island — Iowa 101.5 South Carolina 93.3 Kansas 90.4 South Dakota — Kentucky — Tennessee 87.9 Louisiana 95.8 Texas 90.9 Maine 91.8 Utah 89.0 Maryland 92.6 Vermont 96.7 Massachusetts 93.5 Virginia 89.1 Michigan 90.8 Washington 93.8 Minnesota 98.0 West Virginia 91.6 Mississippi 93.0 Wisconsin 100.1 Missouri 81.7 Wyoming 95.6

Llorens and Stazyk 119

employment in a more favorable light than more conservative states. To test the impact of political ideology on the extent of state government turnover, we include a measure of state institutional ideology developed by Berry, Ringquist, Fording, and Hanson (1998).7 This measure ranges from 0 to 100, conservative to liberal, and is based on ideology rankings of state governors and the Republican and Democratic parties in each chamber of state legislatures. As such, we predict that this measure of ideology will be indirectly related to voluntary separation rates or that states with more liberal institutional leanings will experience lower voluntary separation rates in their state workforces.

Economic environment—unemployment and per capita income. It is commonly accepted that the economic environment surrounding a particular labor market significantly affects the flow of labor from sector to sector. For public sector employers, this has generally meant that in relatively prosperous economic times, employees are more likely to depart voluntarily since there is an increased likelihood of more beneficial employment opportunities in the private sector. On the other hand, in lean economic times, one generally expects public sector employees to be less likely to separate because of the decreased likelihood of alternative employment opportunities. We include two variables to capture the impact of the overall economic environment of state govern- ment turnover, state unemployment and per capita income.

With regard to unemployment, Selden and Moynihan (2000) unexpectedly found unemployment to be positively related to state government quit rates; however, in the absence of additional empirical evidence supporting these findings, we predict that unemployment will be inversely related to voluntary state government separation rates. On the other hand, state per capita income has been used as a measure of overall state fiscal health under the assumption that states with higher per capita incomes will be better able to offer competitive wage rates for state government employees (Kearney, 2003; Llorens, 2008). Given our prior assumptions about the impact of relative wage rates on employee turnover, we predict that per capita income will be inversely related to voluntary separation rates; however, we also acknowledge the opposite effect may be anticipated to the extent that per capita income also reflects the health of the com- peting private sector labor market.8

Unionization. Prior research has found that state government unionization significantly influences state government employment along a host of key metrics (e.g., Belman, Heywood, & Lund, 1997; Kearney, 2003; Kearney & Morgan, 1980; Riccucci, 1986). Given the core mission of public employee unions to protect and support public sector employment, much of the existing theory suggests the prevalence of public unions should be inversely related to public employee turnover (Blau & Kahn, 1981; Cotton & Tuttle, 1986). However, the results of prior research on the impact of public unions vary by level of government. For instance, while Selden and Moynihan (2000) find support for the predicted union effect in state government employment turnover, Kellough and Osana (1995) observe the opposite when analyzing federal government employment. Given such results and our focus on state government employment, we predict the extent of state government unionization will be inversely related to voluntary separation rates.9

120 Review of Public Personnel Administration 31(2)

State employee age. Much has been written about the evolving nature of tenure expectations among younger workers, with conventional wisdom holding that younger employees belonging to Generations X and Y are less likely to remain with a sin- gle employer for an extended period of time, whereas older generations tend to remain committed to a single employer. This dynamic would suggest those workforces with younger employee bases would experience higher turnover, whereas those with older employees would experience less turnover. This proposed relationship between age and turnover is substantiated by Moynihan and Pandey (2007) who find employee age to be negatively associated with long-term turnover intent in a survey of public, pri- vate, and nonprofit organizations. In addition, in a survey of Texas state government employees, Moynihan and Landuyt (2008) find employee age—apart from experience— has a negative impact on turnover intent. As such, we predict that the average age of state government employees will be negatively associated with voluntary separation rates.10

State population. Last, we include a measure of state population to control for the potential effect that the size of a state might have on employee turnover. Although existing research does not provide direct evidence for a predicted effect of this variable on employee turnover, we put forth that it does provide a relatively accurate measure of the size of the overall labor market since one would expect that states with higher populations would maintain more robust employment opportunities.

Descriptive statistics for both the dependent variable and independent variables are provided in Table 3 below.

Model Estimation Using the variables described above, we construct a state-level, cross-sectional dataset for 2007 and report the results of an ordinary least squares regression analysis. As can be seen in Table 3, we were unable to obtain turnover data for a total of seven states. We acknowledge that this is not an optimal observation number for OLS regression analysis, but it is unavoidable given the limitations on obtaining alternate measures of employee turnover.11 All results are based on a robust estimation of standard errors. The results of our analysis are found in Table 4.12

Results and Discussion Overall, the results of our analysis provide a number of valuable insights for research on public employee turnover. With regard to the impact of public–private wage equity on voluntary separation rates, we find no statistically significant relationship. This result runs counter to existing research on the effect of average wage rates and pay satisfaction on employee turnover but can possibly be explained by a number of com- peting hypotheses. First, prior research has found that public–private wage equity varies substantially on the basis of gender and ethnicity and that lower relative wage rates for women and minorities in the private sector have the effect of “pushing” these

Llorens and Stazyk 121

groups into the public sector (Bergmann, 1971; Llorens, 2008; Llorens, Wenger, & Kellough, 2008). Although public–private pay equity may not be a significant predic- tor of employee turnover in the aggregate, if disaggregated by ethnicity and gender, results may differ substantially. This alternative explanation is partially supported by Moynihan and Landuyt (2008) who find, all things being equal, “women are signifi- cantly less likely to state intent to quit than their male counterparts” (p. 132). To the extent possible, future research should seek to further explore how turnover and its determinants vary by ethnicity and gender.

Table 3. Descriptive Statistics

Observations M SD Min Max

Dependent variable Voluntary separations

43 6.91 3.32 1.90 15.50

Independent variables Earnings ratio (%)

43 94.27 4.91 81.70 104.40

Institutional ideology

43 50.93 25.72 9.49 94.77

Unemployment rate (%)

43 4.36 0.90 2.60 7.10

Per capita income (US$)

43 37,424.93 6,009.71 28,541.00 54,981.00

Unionization (%) 43 32.61 20.68 3.70 73.73 Average age 43 45.41 1.29 43.40 50.00 State population 43 6,366,544 6,745,522 522,830 36,600,000

Table 4. Explanatory Model of Voluntary Separation

Coefficient Robust SE t p > t β

Earnings ratio (%) -0.06 0.10 -0.61 0.54 -0.09 Institutional ideology 0.01 0.02 0.37 0.71 0.05 Unemployment rate (%) -0.33 0.53 -0.63 0.54 -0.09 Per capita income (US$) 0.00 0.00 -0.14 0.89 -0.02 Unionization (%) -0.09 0.03 -3.08 0.00 -0.57 Average age -0.78 0.39 -1.97 0.06 -0.30 State population 0.00 0.00 0.20 0.84 0.02

Note: Number of observations = 43. F(7, 35) = 8.65. Prob > F = 0.0000. R-squared = .5596. Root MSE = 2.4142.

122 Review of Public Personnel Administration 31(2)

Second, that public employees may be willing to accept lower, less competitive salaries than their private-sector counterparts suggests a range of other factors may also be responsible for employee attraction and retention. One commonly held belief is that nonwage benefits (i.e., health insurance, pensions, paid leave, etc.) in the public sector are, on average, more generous than comparable benefits in the private sector and that these additional benefits ultimately help to retain many public sector employ- ees. This effect is supported by Selden and Moynihan’s (2000) finding that the avail- ability of on-site child care significantly reduces turnover, but more recent research has found pay to be a more powerful predictor of employee satisfaction than benefits (Barrett & Greene, 2008; Ellickson, 2002; Lee & Whitford, 2008; Moynihan & Landuyt, 2008). Ultimately, future analysis should explicitly test the impact of nonwage bene- fits on employee turnover.

Although not explicitly tested in our model, public service motivation also repre- sents one factor that offers a plausible explanation for our findings. Public service motivation research argues certain individuals may be uniquely attracted to public insti- tutions and the opportunities these institutions provide to fulfill a variety of personal— often altruistic—intentions (Pandey & Stazyk, 2008; Perry & Wise, 1990). Among individuals with high levels of public service motivation, the ability to realize these motives and intentions appears to matter more than financial remuneration (Houston, 2000, 2006; Pandey & Stazyk, 2008; Perry & Wise, 1990; Wright & Pandey, 2008). As such, our results are wholly in-line with, and appear to lend additional credence to, findings in public service motivation scholarship.

Neither state institutional ideology, unemployment, per capita income, nor state population is found to have a statistically significant effect on voluntary separation rates. However, unionization and state employee average age are both found to significantly affect separation rates in the manner predicted. In the case of unionization, a 1% increase in state government unionization is predicted to decrease voluntary separation rates by less than a 10th of a percent. Likewise, for every 1-year increase in the average age of a state’s workforce, our model predicts a 0.7% decrease in voluntary separation rates. While the results for both variables comport with existing research, the predicted impact for average employee age is perhaps the most important for scholars and public manag- ers alike. In light of the growing retirement bubble in the public sector, one can reason- ably predict that average age rates will mostly likely drop substantially in the coming years with the entry of younger workers into state workforces and, holding all else equal, one would expect yearly turnover rates to rise substantially as well. For example, if the State of Illinois, whose average state employee age in 2007 was 46.4, were to experience a substantial influx of younger employees such that its average age decreased by 5 years, our model predicts that, ceteris paribus, the state would experience a 3.85% increase in voluntary turnover as a result. With a total of 105,471 full-time employees in 2007, this would result in a loss of approximately 4,060 employees simply due to shifting age demographics (U.S. Census Bureau, 2007a). Addressing this potential challenge to workforce stability will be key to ensuring that public organizations remain productive and capable of meeting their mission requirements.

Llorens and Stazyk 123

In sum, this research effort has sought to further explore the determinants of volun- tary turnover in state government by examining the unique impact of competitive wages. While the results of our analysis with respect to this variable run counter to prior research and conventional wisdom, our findings do raise important questions for future research on public employee turnover. For instance, how do the drivers of voluntary employee turnover vary by demographic characteristics such as race and gender? Similarly, to what extent does public service motivation play a role in the decision of public employees to remain with organizations that provide less than competitive wage rates? Future efforts to address these questions should greatly enhance the field’s understanding of voluntary employee turnover and could help inform human resources policies and procedures aimed at stabilizing its effects.

Declaration of Conflicting Interests

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

Funding

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

Notes

1. For this analysis, competitive wage rates will refer to comparisons between state govern- ment wage rates and private sector wage rates in a given state. However, we acknowledge that to the extent that local government, federal government, and nonprofit employment represent viable employment alternatives in a given state, wage rate comparisons for these sectors might also influence state government turnover.

2. It should be noted that, because of data limitations, our analysis does not address the issue of nonclassified employee turnover. However, we acknowledge that this particular aspect of employee turnover is worthy of future research, especially given the increase in nonclas- sified employment in states like Georgia.

3. A total of 11 states did not fully respond to NASPE’s request for 2007 turnover and age dis- tribution data. Subsequently, we sought to obtain this missing data by analyzing publicly available, individual state workforce reports for those states in which particular data points were missing. Through this method, we were able to obtain turnover and/or age distribu- tion data for the following states: Massachusetts, New York, Texas, and Vermont. Citations for state workforce reports are provided in the References section.

4. We chose to employ the Blinder–Oaxaca estimation technique, rather than the “dummy- variable” approach, for two primary reasons. First, this technique is more consistently used in recent compensation literature and allows our analysis to be more readily compared to related research. Second, this technique provides a more sophisticated measure of estimated wage differentials by distinguishing between portions of a differential attributed to human capital endowments and sector of employment. However, the overall implication of this choice is not as substantial since both methodologies provide comparable results (Darity and Mason, 1998).

5. For a more detailed description of the Current Population Survey, see the Bureau of Labor Statistics’ description of the survey at http://www.census.gov/cps/.

124 Review of Public Personnel Administration 31(2)

6. Age-squared is controlled for to account for the nonlinear effect of age on wage rates. It should also be noted that we exclude teachers from our analysis since they are not tradition- ally considered part of the core state government civil service. This exclusion is consistent with the Equal Employment Opportunity Commission’s reporting requirements (EEO-4) for state government employment characteristics.

7. This measure was obtained from the “revised 1960-2006 government ideology series” and reflects data for 2006, the most recent year for which ideological scores are available.

8. Data for state unemployment rates were obtained from the U.S. Bureau of Labor Statistics (2007b).

9. Data for this measure were derived from the U.S. Bureau of Labor Statistics Current Popu- lation Survey (2007a).

10. Data for this measure were obtained from the National Association of State Personnel Executives (2007) and individual state workforce reports. Average age data for Texas reflect data for Fiscal Year 2006, the only year for which data were publicly available.

11. Although not optimal, this data limitation is consistent with prior research on state govern- ment turnover (Selden & Moynihan, 2000), and those states for which data are missing (Alaska, Florida, Hawaii, Kentucky, Montana, Rhode Island, and South Dakota) do not reflect a systematic geographical bias in our analysis.

12. In analyzing bivariate correlations and variance inflation factors, we found no evidence of multicollinearity in our explanatory model. Specifically, we obtained a mean VIF of 1.74 and no individual VIF of greater than 2.92.

References

Balfour, D. L., & Neff, D. M. (1993). Predicting and managing turnover in human service agen- cies: A case study of an organization in crisis. Public Personnel Management, 22, 473-486.

Barrett, K., & Greene, R. (2008). Measuring performance: The state management report card for 2008. Retrieved August 2, 2009, from http://www.pewcenteronthestates.org/uploadedFiles/ Grading-the-States-2008.pdf

Belman, D., Heywood, J. S., & Lund, J. (1997). Public sector earnings and the extent of union- ization. Industrial and Labor Relations Review, 50, 610-628.

Bergmann, B. R. (1971). The effect on white incomes of discrimination in employment. Journal of Political Economy, 79, 294-313.

Berry, W. D., Ringquist, E. J., Fording, R. C., & Hanson, R. L. (1998). Measuring citizen and government ideology in the American states, 1960-93. American Journal of Political Science, 42, 327-348.

Bertelli, A. M. (2007). Determinants of bureaucratic turnover intention: Evidence for the Depart- ment of the Treasury. Journal of Public Administration Research and Theory, 17, 235-258.

Blau, F. D., & Kahn, L. M. (1981). Race and sex differences in quits by young workers. Indus- trial and Labor Relations Review, 34, 563-577.

Blinder, A. S. (1973). Wage discrimination: Reduced form and structural estimates. Journal of Human Resources, 8, 436-455.

Brewer, G., & Selden, S. C. (2003, August 28-31). Bureaucratic representation in U.S. state governments: Determinants of minority and female employment success. Paper presented at the American Political Science Association, Philadelphia, Pennsylvania.

Llorens and Stazyk 125

Bright, L. A. (2008). Does public service motivation really make a difference on the job satisfac- tion and turnover intentions of public employees? American Review of Public Administra- tion, 38, 149-166.

Chang, C.-H. (2009). The relationship between perceptions of organizational politics and employee attitudes, strain, and behavior: A meta-analytic examination. Academy of Management Jour- nal, 52, 779-801.

Cotton, J. L., & Tuttle, J. M. (1986). Employee turnover: A meta-analysis and review with implications for research. Academy of Management Review, 11, 55-70.

Curry, D., McCarragher, T., & Dellmann-Jenkins, M. (2005). Training, transfer, and turnover: Exploring the relationship among transfer of learning factors and cautionary assessment. Child and Youth Service Review, 27, 931-948.

Darity, W. A., Jr., & Mason, P. L. (1998). Evidence on discrimination in employment: Codes of color, codes of gender. Journal of Economic Perspectives, 12, 63-90.

Ellickson, M. C. (2002). Determinants of job satisfaction of municipal government employees. Public Personnel Management, 31, 343-358.

Felps, W., Mitchell, T. R., Heckman, D. R., Lee, T. W., Holtom, B. C., & Harman, W. S. (2009). Turnover contagion: How coworkers’ job embeddedness and job searching behaviors influ- ence quitting. Academy of Management Journal, 52, 545-561.

Forrest, C. R., Cummings, L. L., & Johnson, A. C. (1977). Organizational participation: A cri- tique and model. Academy of Management Review, 2, 586-601.

Houston, D. J. (2000). Public-service motivation: A multivariate test. Journal of Public Admin- istration Research and Theory, 10, 713-727.

Houston, D. J. (2006). “Walking the walk” of public service motivation: Public employees and charitable gifts of time, blood, and money. Journal of Public Administration Research and Theory, 16(1), 67-86.

Ito, J. K. (2003). Career branding and mobility in the civil service: An empirical study. Public Personnel Management, 32, 1-21.

Kearney, R. C. (2003). The determinants of state employee compensation. Review of Public Personnel Administration, 23, 305-322.

Kearney, R. C., & Morgan, D. R. (1980). Unions and state employee compensation. Review of Public Personnel Administration, 23, 305-322.

Kellough, E. J. & Osuna, W. (1995). Cross-agency comparisons of quit rates in the federal ser- vice: Another look at the evidence. Review of Public Personnel Administration, 15, 58-68.

Kim, S. (2005). Factors affecting state government information technology employee turnover intentions. American Review of Public Administration, 35, 137-156.

Lee, S.-Y., & Whitford, A. B. (2008). Exit, voice, loyalty and pay: Evidence from the public workforce. Journal of Public Administration Research and Theory, 18, 647-671.

Lee, T. H., Gerhart, B., Weller, I., & Trevor, C. O. (2008). Understanding voluntary turnover: Path-specific job satisfaction effects and the importance of unsolicited job offers. Academy of Management Journal, 51, 651-671.

Lewis, G. B., & Nice, D. D. (1994). Race, sex, and occupational segregation in state and local governments. American Review of Public Administration, 24, 393-407.

Llorens, J. J. (2008). Uncovering the determinants of competitive state government wages. Review of Public Personnel Administration, 28, 308-326.

126 Review of Public Personnel Administration 31(2)

Llorens, J., Wenger, J., & Kellough, J. E. (2008). Choosing Public Sector Employment: The Impact of Wages on the Representation of Women and Minorities in State Bureaucracies. Journal of Public Administration Research and Theory, 18, 397-413.

Locke, E. A. (1976). The nature and consequences of job satisfaction. In M. D. Dunnette (Ed.), Handbook of industrial and organizational psychology. Chicago: Rand-McNaly.

March, J. G., & Simon, H. A. (1958). Organizations. New York: Wiley. Marsh, R. M., & Mannari, H. (1977). Organizational commitment and turnover: A predictive

study. Administrative Science Quarterly, 22, 57-75. McCabe, B. C., Feiock, R. C., Clingermayer, J. C., & Stream, C. (2008). Turnover among city man-

agers: The role of political and economic change. Public Administration Review, 68, 380-386. Meier, K. J., Mastracci, S. H., & Wilson, K. (2006). Gender and emotional labor in public

organizations: An empirical examination of the link to performance. Public Administration Review, 66, 899-909.

Miller, M. A. (1996). The public private pay debate: What do the data show? Monthly Labor Review, 119, 18-29.

Mobley, W. H., Griffeth, R. W., Hand, H. H., & Meglino, B. M. (1979). Review and conceptual analysis of the employee turnover process. Psychological Bulletin, 86, 493-522.

Mobley, W. H., Homer, S. O., & Hollingsworth, A. T. (1978). An evaluation of precursors of hospital employee turnover. Journal of Applied Psychology, 63, 408-414.

Mossholder, K. W., Settoon, R. P., & Henagan, S. C. (2005). A relational perspective on turn- over: Examining structural, attitudinal, and behavioral predictors. Academy of Management Journal, 48, 607-618.

Moynihan, D. P. & Landuyt, N. (2008). Explaining turnover intention in state government: Examining the roles of gender, life cycle, and loyalty. Review of Public Personnel Admin- istration, 28, 120-143.

Moynihan, D. P., & Pandey, S. K. (2008). The ties that bind: Social networks, person- organization value fit, and turnover intention. Journal of Public Administration Research and Theory, 18, 205-227.

National Association of State Personnel Executives. (2007). Research and Surveys. Lexington, KY: National Association of State Personnel Executives.

Oaxaca, R. (1973). Male-Female wage differentials in urban labor markets. International Eco- nomic Review, 14, 693-709.

Pandey, S. K., & Stazyk, E. C. (2008). Antecedents and correlates of public service motivation. In J. L. Perry & A. Hondeghem (Eds.), Motivation in public management: The call of public service (pp. 101-117). Oxford, UK: Oxford University Press.

Park, H. A., Ofori-Dankwa, J., & Bishop, D. R. (1994). Organizational and environmental deter- minants of functional and dysfunctional turnover: Practical research implications. Human Relations, 47, 353-367.

Partnership for Public Service. (2009). Cyber in-security: Strengthening the federal cybersecurity workforce. Retrieved March 3, 2010, from http://ourpublicservice.org/OPS/publications/ viewcontentdetails.php?id=135

Perry, J. L., & Wise, L. R. (1990). The motivational basis of public service. Public Administra- tion Review, 50, 367-373.

Price, J. L. (1977). The study of turnover. Ames, IA: Iowa University Press.

Llorens and Stazyk 127

Riccucci, N. M. (1986). Female and minority employment in city government—The role of unions. Policy Studies Journal, 15, 3-15.

Rubin, E. V. (2008). The role of procedural justice in public personnel management: Empirical results from the Department of Defense. Journal of Public Administration Research and Theory, 19, 125-143.

Selden, S. C., & Moynihan. D. P. (2000). A model of voluntary turnover in state government. Review of Public Personnel Administration, 20, 63-74.

State of Massachusetts. (2007). 2007 Annual Report. Retrieved February 5, 2010, from http:// www.mass.gov/?pageID=afsubtopic&L=5&L0=Home&L1=Employment%2c+Equal+Acc ess%2c+Disability&L2=HR+Policies&L3=HRD+Publications&L4=HRD+Annual+Repor t&sid=Eoaf

State of New York. (2008). New York State Workforce Management Report. Retrieved February 5, 2010, from http://www.cs.state.ny.us/hr/docs/2008.pdf

State of Texas. (2007a). A Summary of the State of Texas Workforce for Fiscal Year 2006. Retrieved February 5, 2010, from http://www.sao.state.tx.us/Reports/report.cfm/report/07-707

State of Texas. (2007b). An Annual Report on Classified Employee Turnover for Fiscal Year 2007. Retrieved February 5, 2010 from http://www.sao.state.tx.us/reports/main/ 08-703.pdf

State of Vermont. (2007). State of Vermont Workforce Report Fiscal Year 2007. Retrieved February 5, 2010, from http://humanresources.vermont.gov/sites/dhr/files/pdf/workforce _reports/DHR-WorkForce_Report_FY07.pdf

Staw, B. M. (1980). The consequences of turnover. Journal of Occupational Behavior, 1, 253-273.

U.S. Bureau of Labor Statistics. (2007a). Current population survey. Retrieved July 31, 2009, from http://www.bls.gov/cps/

U.S. Bureau of Labor Statistics. (2007b). Alternative Measures of Labor Underutilization for States. Retrieved July 31, 2009, from http://www.bls.gov/lau/stalt07.htm

U.S. Census Bureau. (2007a). Illinois state government employment data: March 2007. Retrieved March 3, 2010, from http://www2.census.gov/govs/apes/07stil.txt

Wright, B. E., & Pandey, S. K. (2008). Public service motivation and the assumption of person organization fit: Testing the mediating effect of value congruence. Administration & Society, 40, 502-521.

Bios

Jared J. Llorens, PhD, is an assistant professor in the Public Administration Institute at Louisiana State University, Baton Rouge. His research interests include public sector compen- sation, human resources automation, and civil service reform. His work has appeared in the Journal of Public Administration Research and Theory, the Review of Public Personnel Administration, and Public Personnel Management.

Edmund C. Stazyk, PhD, is an assistant professor in the Department of Public Administration & Policy at American University, Washington, DC. His research focuses on the application of organization theory and behavior to public management, public administration theory, and human resources issues. His primary interests are in the areas of organizational and individual perfor- mance with an emphasis on employee and public service motivation.