In the Jefferson Policy Journal two weeks ago, we reported the results from a recent study of public-sector versus private-sector compensation. This time, we respond to a presentation by state officials that gives a different perspective.
The study we cited, by economists William Even of Miami University and David Macpherson of Trinity University, found that public employees in Virginia have a modest but statistically significant advantage over private-sector employees in overall compensation—despite a modest disadvantage in pay. The generosity of public-sector benefits explains the difference.
In contrast, one of the messages in a report from the Department of Human Resource Management (DHRM), presented on September 19 to a joint meeting of two General Assembly subcommittees in Richmond, was that state employees as a whole are substantially underpaid.
The presentation did not take account of our public sector’s more generous benefits, especially in terms of pensions and health insurance—a major omission when considering government pay. Furthermore, the top-line descriptive statistics in the DHRM report don’t rise to the level of analytical rigor required to answer the question of how public employees are compensated as compared to private employees.
The DHRM does a direct comparison of state salaries in a list of selected occupations, which it says are, on average, 16.62 percent below those in private industry. In terms of specific occupations, accountants make 11.9 percent less than in the private sector, electricians make 22.38 percent less, and so on.
But that doesn’t tell us whether the public employees in these occupations are more or less qualified than their private-sector counterparts. For instance, what if the typical private-sector accountant is more highly educated than a public sector one? The DHRM study misses this distinction completely.
In contrast, the Even-Macpherson study used a more empirically precise “human capital” approach that utilized Census Bureau data to compare employees of comparable education and years of work experience. It also controlled for the number of hours worked per year, for broad occupational groups, and—as is normal in such research—gender, race, ethnicity, and whether a person was disabled.
In their regression analysis, Drs. Even and Macpherson found that the “premium”—the advantage in total compensation for being a public (government) rather than private-sector employee—is 4.95 percent in Virginia. That figure, while lower than the 14.91 percent for public employees nationwide, still suggests that the DHRM analysis is flawed.
As Dr. Jeffrey Keefe acknowledged in a study for the union-backed Economic Policy Institute, “comparing workers of similar ‘human capital’ or personal productive characteristics and labor market skills is considered the best alternative, and well accepted by labor economists.”
(That’s not to say Keefe’s study reached an accurate conclusion. Methodological problems identified by the Center for Union Facts biased the study’s results downward. When these are corrected for, a public-sector “penalty” flips and becomes a premium.)
The data in the study by Even and Macpherson combine state and local government employees; results specifically for state workers were not available. On earnings, they agree with DHRM that there is a penalty—but the magnitude is far smaller than the department suggests when it says average state pay in Virginia is 10 percent below private-sector pay. Moreover, the addition of pensions and benefits—missing in the DHRM presentation—makes it clear that Virginia public employees are doing well relative to the private sector.
The bottom line is that the DHRM data are an insufficient basis for legislators and others to use in judging whether current state salaries are appropriate. Top-line figures aren’t informative when it comes to answering the question of how well a public employee is compensated relative to a comparable employee in the private sector. The moral of the story is clear: When it comes to public-sector compensation or public-sector pay, don’t be impressed by simple presentations. They may not show the most useful data.