In the debate over public-sector pay, statistics are plentiful but seemingly contradictory. Sometimes we are told that public workers are grossly overpaid, sometimes we hear they are actually underpaid, and sometimes we are assured that no one really knows the truth.
How can the average American absorb these arguments and rebuttals and counter-rebuttals? How can we make sense of what appear to be obscure statistical disagreements? Our goal here is to cut through the noise and explain the essentials.
First, to properly compare employee compensation and the public and private sectors, it is necessary to compare compensation on an apples-to-apples basis, by adjusting for differences in worker characteristics. Since public workers tend to be more educated and experienced than private workers, raw salary comparisons make the public sector look more overpaid than it really is. Every factoid, statistic, or even systematic study that does not involve apples-to-apples comparisons can be safely ignored.
It is also important that federal workers not be lumped with state and local. At the federal level, public employees enjoy both higher wages and higher benefits than comparably-skilled private workers. As we've shown here, federal workers are unambiguously overpaid.
At the state and local level, however, public workers usually receive below-market salaries in exchange for higher benefits and better job security. Whether total compensation for state and local workers is excessive depends on the salary-benefit balance.
In the past year or so, four different reports from left-leaning think tanks have argued that the balance tips against state and local workers. These studies, including a series of state-specific studies from the Economic Policy Institute (EPI), are extremely popular among unions and liberal commentators, who cite one or more in almost every debate or discussion on government pay. In fact, some have suggested that disputing the "facts" established in those papers is akin to science denial or know-nothingism.
Before we explain the major shortcomings of those studies, please note again that they apply only to state and local workers, not to federal. If the authors of these studies applied their same methodology to federal pay, they would necessarily conclude that federal workers are overpaid.
On the state and local level, readers may expect that we have obscure technical complaints about the existing studies that will lead to endless back-and-forth debate. Not at all. In fact, the statistical methodology in each paper is basically sound. The problem is that every existing study simply omits large portions of public-sector compensation. State and local workers appear underpaid in these studies only because their pay is undercounted.
Here are the details. Each existing paper is divided into a wage section and a benefits section. In the wage section, the authors create apples-to-apples comparisons using control variables such as education and experience from the Census Bureau's Current Population Survey. The authors find that state and local workers make wages that are somewhere around 10 percent lower than private workers with the same basic attributes. We have little to criticize about this method or its conclusion. (Some of the papers control for union status, which we feel is inappropriate, but it doesn't make a big difference in the results.)
In the benefits sections, the authors use data from the Bureau of Labor Statistics (BLS) on employer contributions to fund employee benefits. The BLS publishes the value of certain benefits--paid leave, supplemental pay, health insurance, retirement savings, etc.--as a percentage of worker wages in the public and private sectors. According to the BLS, state and local benefits are more generous than private benefits, but not so much that they overcome the wage penalty. Depending on the study, benefits either draw state and local workers even with private workers or still leave them slightly behind.
The major problem with measuring benefits this way is that the BLS list is incomplete. In fact, the only benefits it fully counts are the present-day, direct costs to employers of worker compensation. This leaves three main benefits unaccounted for--retiree health coverage, the guaranteed high returns on public pensions, and job security.
Retiree health coverage generally aims to bridge the healthcare gap between retirement (often before age 60 in the public sector) and Medicare eligibility at 65. In many states, coverage continues even after age 65--providing so-called "Medigap" insurance to cover costs not paid for by Medicare, and in some cases even paying Medicare premiums. Nearly every state employee, and most local government employees as well, receive this coverage. Though its value will vary from state to state, it is worth about 12 percent of salaries in California, the state we have studied most extensively. For every dollar a California public employee earns in wages, he receives future retiree health benefits with discounted value of around 12 cents, but the existing studies don't count that as compensation.
Also undercounted in the existing studies are pension contributions. Public-sector workers receive defined-benefit pensions, which work like annuities--guaranteed, fixed payments throughout retirement. Most private workers have defined-contribution plans, like a 401k, in which workers invest money in personal accounts and bear the risk themselves.
EPI's reports and similar studies do count government employers' annual contributions to these plans, but they do not consider the different funding standards between the public and private sectors. State and local pensions fund their plans much more aggressively than private pensions, meaning that they make smaller contributions invested in riskier assets.
Of course, the risk is borne by taxpayers, not retirees, so we need to adjust for these funding differences to calculate how much employees actually will receive in retirement. EPI's method would claim that a public employee receives lower pension benefits than a private employee even if the actual benefits he receives are exactly the same. Obviously, this is wrong. Correcting for this error increases total compensation for California government employees by around 4 percent. In addition, some states have not been meeting even the minimum payments required under current lax pension accounting standards, Since future benefits will be paid regardless of current funding adequacy, this further understates compensation through pension benefits.
Finally there is job security. It is well known that public workers, particularly after they pass a short "probationary" period, are rarely fired or laid off. In fact, public employees overall are discharged at just one third the rate of private sector workers. This kind of job security has monetary value.
Using the tools of financial economics, we have calculated that the job security enjoyed by California public workers is worth about 15 percent of their pay, and this bonus comes at a real cost to taxpayers. Since governments struggle to remove unproductive workers, it becomes very difficult to adjust the skill profile of the public workforce to handle changing needs and responsibilities over time. (See more on this topic here.)
Overall, the omission of retiree health benefits, the guaranteed high rate of return on public pensions, and job security mean that the existing state and local studies substantially undercount state and local compensation. To illustrate, a report from the Center for Wage and Employment Dynamics concluded that public workers in California earn about the same total compensations as private workers after adjusting for skill differences.
In a new Heritage Foundation report, we used much the same methodology but added in the missing benefits. Properly measured, compensation is about 30 percent higher for public workers in California than similar private workers.
Despite all the numbers thrown around by both sides, the facts on public pay are clear. Federal workers receive both a wage premium and a benefits premium over similar private workers. State and local workers see a wage penalty, but the penalty is usually more than made up for in higher benefits.
Pay reformers, like Gov. Scott Walker in Wisconsin, are making a mistake by focusing only on the budget crisis to justify reducing public sector benefits. The easy response from apologists, who seem to be winning in opinion polls, is that public workers are being "scapegoated," and that budget savings can be found elsewhere. The better argument for reformers is that reducing public compensation is both a budgetary necessity and a matter of fundamental fairness. The fiscal mess that many states find themselves in only adds urgency to equalizing public and private compensation.Andrew G. Biggs is a resident scholar at AEI. Jason Richwine is a senior policy analyst at the Heritage Foundation.