Do Performance Pay Workers Experience Less Job Turnover and Longer Tenures?

9th Global Conference on Business & EconomicsISBN : 978-0-9742114-2-7

Do Performance Pay Workers Experience Less Job Turnover and Longer Tenures?

Patrick L. O'Halloran

Economics, Finance and Real Estate Department, MonmouthUniversity

Abstract

Performance Related Pay (PRP) is a contemporaneous reward and thus may substitute for deferred compensation. Deferred compensation is associated with longer tenures and lower turnover rates. Thus, PRP may be proxy for greater turnover and shorter tenures. Alternatively, a main conclusion of the PRP literature is that PRP can serve as a sorting mechanism that tends to attract higher productivity workers. Furthermore, higher productivity workers experience longer tenures and less job market turnover, especially involuntary layoffs. I find that workers receiving contemporaneous PRP in the form of piece rates, commissions, and tips are no more likely to experience turnover than wage workers. Moreover, those receiving deferred PRP in the form of bonuses, stock options, and profit sharing experience longer tenures and less job turnover in comparison to wage workers. Specifically, profit sharing appears to have a significant impact on job separations and quits supporting the theory that profit sharing enhances employment stability.

JEL Classification: J33, J63, M52

Keywords: Performance Pay, Turnover, Profit Sharing, Bonuses, Deferred Compensation

Corresponding Author:

Patrick L. O'Halloran,

Department of Economics, Finance and Real Estate

LeonHessBusinessSchool

MonmouthUniversity,

400 Cedar Avenue, West Long Branch, New Jersey 07764-1898

e-mail:

ph: 1-732-571-7530

Do Performance Pay Workers Experience Less Job Turnover and Longer Tenures?

Abstract

I. Introduction

There has been a great deal of research on the determinants of turnover and also a vast amount of literature concerning the impact of performance related pay (PRP) on such things as incentive provision and sorting, but relatively few studies explicitly investigate how turnover is affected by various types of PRP. There are two potential relationships between PRP and turnover. PRP may serve as a contemporaneous reward substituting for deferred compensation. Deferred compensation is typically associated with longer tenures and lower turnover rates. Thus, PRP may be proxy for greater turnover and shorter tenures. On the other hand, PRP acts as a sorting mechanism that tends to attract higher productivity workers who experience longer tenures andless job market turnover. Furthermore, PRP is known to result in better matches between jobs and employees. Thus, PRP may be a proxy for lower turnover and longer tenures. Consequently, the impact that PRP has on labor turnover is ambiguous.

This paper seeks to explore the overall impact PRP has on job market turnover as well as addressing what impact specific types of PRP have on labor market turnover in the form of quits and layoffs. Specifically, I test whether workers receiving various types of PRP experience differing rates of job market separations in the form of either quits or layoffs than those receiving standard input based pay. From a policy perspective, if PRP workers do experience differential rates of job separation and tenure in comparison to those receiving straight input pay, firms may be able to use PRP schemes as a mechanism to retain their most valued employees.

PRP can take several forms. PRP can be based on individual performance such as salary/wage pay, piece rates, bonuses, and commissions, or on group performance such as stock options and profit sharing. Furthermore, PRP can be based on measures of input such as wages/salary and bonuses, or based on measures of output such as piece rates, tips, commissions and profit sharing. Moreover, various types of PRP can be either objective or subjective in nature. PRP based on measurable output is theoretically less prone to a supervisor’s subjective performance evaluation thanPRP based on observed performance evaluation (Heywood and O’Halloran, 2004). In what follows, I assume that piece rates, commissions and profit sharing are more objective types of PRP while tips, bonuses and stock options are more subjective types of PRP.

Empirical evidence reveals that those who receive PRP, regardless of type, do experience fewer job separations and hence longer tenures. However, I observe that those receiving pay in the form of piece rates, commissions, tips, and bonuses are no more likely to experience job separations than input pay workers. Conversely, those receiving compensation in the form of stock options and profit sharing experience lower rates of job separation. Therefore, the main driver of the lower separation rates among PRP workers appears to stem from stock options and profit sharing. Furthermore, considering the distinction between quits and layoffs, the data revealthat workers receiving subjective PRP are less likely to be terminated but no less likely to quit while workers receiving objective PRP are less likely to quit but not significantly less likely to be terminated. Also, workers who receive contemporaneous PRP are less likely to be terminated but not significantly less likely to quit while workers who receive deferred PRP are less likely to quit and less likely to be terminated. I also find cursory evidence of stronger separation responses to PRP among women. The result on profit sharing appears especially robust supporting the argument that profit sharing enhances employment stability (Weitzman, 1985). It also tends to support the argument by Azfar and Danninger (2001) that profit sharing allows for increased expected returns on firm-specific human capital investments. In contrast to Azfar and Danninger’s findings, I find that profit sharing tends to reduce the incidence of quits but not the incidence of layoffs. As a result, differentiating between the various types of PRP may be crucial if a firm wants to effectively use PRP as a retention device to retain their most valued employees.

This paper will proceed as follows: Section II will provide background on the two opposing hypotheses concerning the relationship between PRP and turnover; Section III will describe the data and provide some initial findings; Section IV will summarize the empirical results; and Section V will offer conclusions and further recommendations.

II.Exploring the Relationships between PRP and Turnover

One conclusion of the PRP literature is that PRP serves as a sorting mechanism, whereby higher productivity workers self-select into PRP jobs. A second conclusion of this literature is that PRP tends to elicit greater effort on behalf of workers and hence increase worker productivity. Furthermore, PRP workers are believed to receive a compensating differential for the greater risks often associated with various types of PRP in comparison to non-PRP workers. Consequently, PRP workers typically receive higher earnings than time-rate workers due to selection, productivity, and risk effects associated with PRP.

Most prior research has shown that PRP workers typically receive higher earnings compared to non-PRP workers. Seiler (1984) finds that piece rate workers experience an earnings premium of 14% resulting from greater risk inherent in incentive pay and an incentive-effort effect. Brown (1992) further confirms that piece rate workers do experience higher earnings than those paid time rates, but not for bonuses subject to supervisor ratings. Ewing (1996) and Parent (1999) both use data from the National Longitudinal Survey of Youth 1979 (NLSY79) and find that PRP workers earn more than time rate workers. Furthermore, Azfar and Danninger (2001) find that profit sharing is related to higher wage growth than those without profit sharing. Booth and Frank (1999), using data from the British Household Panel Survey, find higher earnings among workers receiving bonuses, profit related pay, or commissions.

Given the sorting function that these types of PRP schemes play, it seems reasonable to assume that jobs incorporating PRP will attract higher ability/higher productivity workers. Indeed, much of the cited research in this paper points to the pivotal role PRP plays in attracting higher ability/higher productivity workers. Specifically, Lazear (1995) finds that individually based PRP attracts workers of higher ability and hence serves as a sorting mechanism, leading to higher productivity. Ewing (1996) examines the effect that PRP has on earnings. He finds that, after controlling for many personal, institutional, firm-specific, and demographic factors available within the NLSY79, that those working in PRP jobs earn a substantial wage premium which he suggests as evidence that more productive workers self-select into jobs with PRP. Furthermore, Blinder (1990) and Ehrenberg (1990) theorize that PRP can influence employee behavior and provide incentives for individuals to increase their productivity. Paarsch and Shearer (2000) find evidence of incentive and productivity effects among piece rates. Moreover, other researchers have found support for the assumption that PRP also provides better incentives alleviating the agency problems inherent where there is asymmetric information between the employers and employees. Lazear (1986) finds that moving from hourly pay to piece rate pay in one particular company results in a 41% increase in productivity for which over half of the gain is attributed to incentive effects and the remainder attributed to sorting effects. Consequently, evidence suggests that the increase in earnings experienced when moving from a straight input payment system to a PRP system can be jointly associated with both sorting and incentive provisions.

According to prior research, what types of jobs typically entail PRP? Lazear (1995) points out that several conditions must be met if PRP is to serve as an effective sorting mechanism. Specifically, Lazear hypothesizes that PRP in the form of piece rates will be more successful when: (a) the cost of measuring output is low (in terms of both time and money); (b) when the value of alternative activities (such as shirking) is high; (c) when the alternative wage is high; (d) when workers are relatively young and haven’t accumulated a great deal of firm specific human capital; and (e) when workers are more heterogeneous. Further, Brown (1992), and MacLeod and Parent (1999) find output based pay to be more prevalent in the absence of substantial team production and when monitoring output is relatively costless. Jirjahn and Stephan (2004), also find that output based pay is common where worker performance is easily measured and tasks tend to be relatively easy. Brown (1990) finds that PRP is more commonly found in larger establishments due to the economies of scale involved in measuring performance. Baker (1992) points out that when individual output is a good metric for performance and asymmetric information exists, one would expect to see many employees receiving pay based on individual output such as piece rates, bonuses, and commissions. Conversely, when there is symmetric information or when no measures of performance exist, input based pay will dominate. If there is asymmetric information and no good measures of performance are available, Baker expects incentive contracts based on firm performance to dominate, such as profit sharing or stock options.

Assuming that PRP attracts higher ability workers and provides better incentives than input pay, one would expect that PRP workers would be less likely to experience job market turnover, especially involuntary layoffs. This will be true if PRP tends to attract or sort workers who can earn more within firms that offer PRP than they could alternatively earn working in firms that base pay solely on input. Many studies find performance gains and selection effects from PRP (Asch, 1990; Fernie and Metcalf, 1999; Parent, 1999; Lazear, 2000; Shearer, 2005; and Freeman and Kleiner, 2005). Furthermore, many prior studies have found that workers receiving higher pay and higher wage growth experience lower rates of job separation (Pencavel, 1972; Jovanovic, 1979; Viscusi, 1980; Bartel and Borjas, 1981; and Topel and Ward, 1992). Accordingly, PRP may be associated with lower turnover rates and higher levels of tenure.

On the other hand, an alternative hypothesis views PRP as a contemporaneous reward for output and hence a substitute for deferred compensation. Deferred compensation, according to agency theory, results in “wage-tilt” or a steep age-earnings profile where a worker is “under-paid” (i.e., paid less than their value of marginal product) early in their tenure with a firm and “over-paid” (i.e., paid more than their value of marginal product) in later years with the employer as a reward for desirable behavior (Salop and Salop, 1976; Lazear, 1979, 1981). This is usually intended to provide an incentive to limit the amount of worker shirking, malfeasance and turnover and hence increase worker productivity. Lazear (1979, 1981) focuses on the role upward sloping age-earnings profiles play in aligning agent’s goals. This approach involves basing future pay increases on current output and effort. Lazear (1995) points out that although roughly equivalent to a piece rate scheme, upward sloping age earnings profiles differ in two ways. First, the evaluation period is typically much longer than under the piece rate scheme. Second, the worker must remain at the firm long enough to reap the reward for good performance. Accordingly, deferred compensation may be associated with longer tenures and thus, PRP may be a proxy for greater turnover and shorter tenure. Consequently, PRP may be associated with greater turnover and shorter tenure instead of lower turnover and longer tenure as hypothesized above. Thus, the effect PRP may have on labor market turnover and tenure is theoretically ambiguous.

There is some evidence for deferred compensation from agency theory. Although Hutchens (1997) finds support for Lazear’s deferred compensation hypothesis using the NLSY, Ippolito (1991) finds that wage tilt has no significant impact on tenure. Frank and Hutchens (1993) observe evidence of “wage-tilt” among airlines pilots which they attribute to deferred compensation, since one wouldn’t expect a pilot’s productivity to rise with experience. Also, several researches have compared wage profiles between the self-employed and employed as evidence of deferred compensation. According to agency theory, the steepness of the age-earnings profile, or the degree of “wage tilt”, reflects the desire of employers to provide incentives to employees to avoid unwanted activities. Since the self-employed are not subject to agency problems, one would expect the self-employed to have flat age-earnings profiles. Lazear and More (1984) and Brown and Sessions (2006) both find relatively flat experience-earnings profiles among the self-employed and steeper profiles among the employed. Additionally, Brown and Sessions observes self-employed workers having the flattest profile, PRP employees having steeper profiles than the self-employed and non-PRP employees having the steepest profile. They conclude that these results provide evidence of the agency theory of deferred compensation. Therefore, the presence of deferred compensation as an incentive device implies that deferred compensation may partially substitute for PRP. Consequently, there appears to be sufficient evidence that PRP may substitute for deferred compensation. Therefore, any conclusion regarding the relationship between PRP and turnover remains ambiguous.

Evidence of PRP workers experiencing shorter tenure and having higher turnover levels is sporadic and typically associated with particular types of PRP. Prior research has revealed that firms typically adopt certain PRP practices such as piece rates if they expect higher quit rates. Goldin (1986) argues that workers who expect lower tenures will be less motivated by upward sloping age earnings profiles. Kelly, et. al. (1998) show that workers with shorter expected tenure will be more likely to seek jobs that reward contemporaneous remuneration. Furthermore, Goldin (1986) and Geddes and Heywood (2003) find that women are more likely to be paid piece rates because of their lower expected tenure but are less likely to receive PRP in the form of commissions or bonuses. These findings further highlight the importance of differentiating between different types of PRP schemes.