Do monetary rewards bring happiness? Comparing the impacts of pay-for-performance in the public and private sectors

Abstract

Does money make people happy? Empirical studies have investigated how personal income influences happiness, but the relationship between pay-for-performance (PFP) and happiness remains unknown. To address this gap, we examine the impact of PFP on worker happiness. In particular, we compare the impact between the public and private sectors, because scholars argue that PFP in the public sector often fails to deliver its intended benefits. Using data from the 2002 and 2006 General Social Surveys (GSS), we find that PFP only enhances worker happiness in the private sector, not the public sector. We also find that PFP is negatively related to the perceived relatedness and perceived organizational effectiveness in the public sector.

Keywords:

• Happiness
• pay-for-performance
• perceived organizational support
• perceived organizational relatedness
• perceived organizational effectiveness

Notes

1. Demographic description.

2. Employees’ actual income is a continuous variable, where less than 1000 = 1, 1000–2999 = 2, 3000–3999 = 3, 130,000–149,999 = 24, and 150,000 or over = 25.

3. Occupational prestige is measured by asking respondents to rank a sample of occupational titles with respect to the occupation’s ‘“general standing’.” The occupational prestige scores in the General Social Survey rely on the NORC prestige scale by combining occupational prestige ratings from several prestige surveys (Stevens & Hoisington, 1987). For example, the prestige score is 60.30 for legislators; 55.88 for accountants and auditors; 73.80 for physicians and astronomers; and 78.30 for post-secondary science teachers. In relation to low-end jobs, for example, the score is 29.38 for sales workers; 15.84 for news vendors; and 18.67 for protective service providers (guards).

4. Education is a continuous variable where no formal schooling=0, 1st grade=1, 2nd grade=2, 12th grade=12, 1 year of college=13, 2 years of college=14, 8 years of college=20.

5. As we used data from a single source, we tried to examine for common method bias (CMB) and show this to be of minor or insignificant concern only. To test on the common method variance, we employed an analysis exploring the common latent factors (CLF) as posited in the Harman’s single-factor test. The test includes all items from all the constructs in the study into a factor analysis to determine whether the majority of the variance can be accounted for by one general factor. From the CLF test, we found that the common method variance factor was .33, below the suggested threshold of .50. Hence, we are convinced that CMB is not a major methodological issue in this study (Fornell & Larcker, 1981).