In a recent report, the Center for Economic and Policy Research (CEPR) found that state and local government employees pay a wage penalty of about four percent for working in the public sector, relative to those at the same age and education level in the private sector.1 In the “raw” data — that is before controlling for age and education — state and local government workers appeared to earn about 13 percent more than private-sector workers. But, state and local workers are, on average, about four years older than private-sector workers and half have a four-year college-degree or more, compared to less than 30 percent in the private sector. Once these differences in age and education are factored in, state and local public workers earn less, not more, than their private-sector counterparts.
But what about benefits? The CEPR report on wages used a large nationally representative data set that had detailed information on workers’ wages, personal characteristics, and the sector in which they worked. Unfortunately, no similar data exist that include both information on the value of employee benefits and on workers’ personal characteristics. Nevertheless, some evidence in the CEPR report, as well as separate evidence from government-administered employee benefits surveys (which don’t include information on workers’ characteristics) can still give us some idea of how incorporating benefits would affect the conclusions of the CEPR report.
First, as the CEPR report demonstrated, state and local government workers are older and substantially better educated than private-sector workers. Showing that state and local governments spend more per worker on benefits than private-sector employers do, therefore, does not demonstrate that state and local worker benefits are “too high.” To demonstrate that state and local employee benefits are too high would require, at least, showing that state and local workers have better benefits than a private-sector worker matched by age and education. The CEPR report showed that once workers are matched on this basis, the state and local wage premium of about 13 percent becomes a state and local wage penalty of about 4 percent — a swing of 17 percentage points. While there are no data that allow a corresponding analysis of benefits, the impact of basic controls for age and education on the apparent wage differentials between the two sectors suggests that relatively large differences in benefits per worker may simply reflect the older, better educated, state and local government workforce.
Second, the limited survey data that are available suggest that when workers are compared on a roughly comparable basis, total compensation (wages and benefits) in the state-and-local and private sectors is roughly comparable. The government regularly collects data on employer costs for wages and benefits and publishes breakdowns for several broad categories of workers. None of the breakdowns controls for workers’ characteristics such as age and education, but one of the published categories — managers and professionals — does compare workers who have broadly comparable sets of responsibilities in both the state-and-local and private sectors. In the most recent data (covering December 2009), these government data find that “management, professional, and related” workers in the private sector receive, on average, $48.19 per hour in total compensation, almost exactly the same as the $48.15 per hour in the state-and-local sector. The slightly higher benefit costs for the state-and-local sector ($15.19 per hour compared to $14.07 in the private sector) just offset their slightly lower wage costs ($32.96 per hour compared to $34.12 in the private sector).2
Third, the large share of small businesses in the private sector — which typically offer lower levels of both pay and benefits than larger firms — also distorts comparisons. State and local employee benefit levels are similar to those in large private establishments, which are the more natural comparison group. Data from the same national survey for 2009, for example, show that average hourly benefit costs for state and local government workers were, on average, $13.49 per hour, only about four percent higher than average benefit costs of $12.95 per hour for private-sector establishments with 500 or more workers.3
Fourth, an important share of state and local pension expenditures simply make up for the exclusion of about 30 percent of all state and local workers from Social Security.4 Reported private-sector pension costs, such as those in the survey data cited above, are on top of employer Social Security contributions. For about 30 percent of state and local workers, however, a substantial portion of government pension expenditures only offset lost Social Security earnings after retirement.5
Finally, the CEPR report emphasized that the pay penalty for state and local workers increases with workers’ wage levels. The highest paid workers — typically the older and better-educated workers that are disproportionately employed in the public sector — pay the biggest wage penalty in state and local jobs. Higher levels of job security and good benefits are a key tool for recruiting and retaining these skilled public-sector workers.
1 John Schmitt. 2010. “The Wage Penalty for State and Local Government Workers.” Washington, DC: Center for Economic and Policy Research. See also Bender, Keith A. and John S. Heywood. 2010. “Out of Balance? Comparing Public and Private Sector Compensation Over 20 Years.” Washington, DC: Center for State and Local Government Excellence and National Institute on Retirement Security.
3 Ibid, Table 8.
4 Government Accountability Office. 2007. “State and Local Government Retirement Benefits: Current Status of Benefit Structures, Protections, and Fiscal Outlook for Funding Future Costs.” September, Revised November 15, 2007, p. 7 (fn. 4). www.gao.gov/new.items/d071156.pdf
5 State and local employees were initially excluded from participation in the Social Security system.
John Schmitt is a Senior Economist at the Center for Economic and Policy Research in Washington, D.C. This article was published by CEPR in May 2010 under a Creative Commons license.