The budget impasse in Wisconsin has caused reporters and pundits to bombard Americans with statistics purporting to compare public- and private-sector pay. Left-leaning think tanks, including the Economic Policy Institute (EPI), have promulgated studies claiming that public-sector employees in any number of states are underpaid.

EPI has issued a series of papers on state- and local-government employees — covering Wisconsin, Ohio, Michigan, New Jersey, and Minnesota, with others surely to come — all concluding that public employees earn less than similar private-sector workers, even after accounting for their more generous benefits. From the Washington Post’s Ezra Klein, to Paul Krugman of the New York Times, to AFL-CIO chief Richard Trumka, left-leaning commentators have used these studies as talking points.

But those talking points are wrong. As we showed in a recent paper, the EPI studies substantially underestimate public-sector compensation. (EPI disagreed, and we recently published a response.) Here’s what they leave out:

Overall benefits: EPI uses Bureau of Labor Statistics data on benefits, but the BLS doesn’t publish benefit data by state; rather, it groups them together by region. The EPI study on Wisconsin workers, for instance, uses benefits data for the East North Central Census Region, which includes Illinois, Indiana, Michigan, Ohio, and Wisconsin. So if Wisconsin public employees receive more generous benefits than workers in nearby states, which was part of Governor Walker’s rationale for increasing employee contributions, we wouldn’t know it from EPI’s data.

This study by HCTrends, by contrast, shows that Wisconsin workers paid 40 percent less for health coverage than do public employees in Michigan and 60 percent less than public employees in Indiana or Illinois. Likewise, Wisconsin workers contribute less toward their pensions than other public employees. The bias caused by regional grouping cuts both ways, of course, but in Wisconsin — ground zero for our nation’s labor disputes — it almost surely understates the value of public employees’ health and pension coverage.

Pensions: EPI’s approach underestimates public-employee pension benefits by around one-third, because it counts only what employers contribute today, not what employees will receive when they retire. State and local pensions invest far more aggressively than private plans, meaning that for a given level of guaranteed retirement benefits, state and local pensions put aside significantly less money today. Unlike a 401(k), of course, if the investments go sour, it’s the government — meaning, taxpayers — who bear the cost.

Simply put, if a public and a private employee received the exact same benefit, EPI’s approach would report that the public employee received significantly less.

Retiree health care: Four out of five public employees receive subsidized retiree health coverage, which allows them to retire in their 50s without having to buy insurance in the more expensive individual market. The California State Department of Personnel Administration states that “a career State employee . . . secures an additional $493,851 worth of compensation during the first 20 years of retirement.” For a Milwaukee public-school teacher, retiree health coverage is equivalent to a career-long salary increase of over 20 percent. The EPI studies treat retiree health care as if it does not exist: They simply omit any reference to it.

Job security: It’s well known that public-sector jobs offer far greater job security than the private sector, with less than one-third the chance of being fired or laid off. Economists going back to Adam Smith have believed workers will trade off compensation against job security, such that more secure employment should pay less. For Wisconsin workers, we calculated that job security was worth about an extra 9 percent of pay. For California workers, job security is worth about an extra 15 percent of pay.

In short, the EPI studies undercount pension benefits, omit retiree health coverage, and ignore the value of job security. The errors can really add up — to 30 percent more compensation, in the case of California.

Sadly, it’s easier to put out a dozen poor studies than to get a single analysis right. But many fights on public-sector pay are yet to come in states around the country. Taxpayers and their advocates need to be ready to counter false claims about government pay.

— Andrew G. Biggs is a resident scholar at the American Enterprise Institute. Jason Richwine is a senior policy analyst at the Heritage Foundation.