How does government pay compare with that in the private sector? The answer is pretty much what you’d expect: Wages and salaries are lower, but benefits are better. Overall compensation is, if anything, slightly lower, though this varies by class of worker; less educated workers are better paid in the public sector and more educated workers are better paid in the private sector. This again is not surprising when you consider that less educated workers in the private sector often earn poverty wages with no health benefits and government employers have little incentive to shift costs onto Medicaid and other government programs.

But anti-government ideologues have deep pockets, so a minor industry has sprung up trying to show that government workers are overpaid. The latest report in this genre comes from Citizens Against Government Waste, and is typical of its kind. The research was done by an outfit called John Dunham and Associates, a.k.a. guerrillaeconomics.com. It shows many of the tell-tale signs of shoddy research:

1. Anecdotal evidence. Somewhere, there’s a government worker making an obscene amount of money and abusing the system. But focusing too much on specific examples is a sure sign they aren’t representative. The CAGW report claims the average San Diego firefighter makes more than $180,000 per year, but the source for this dubious assertion turns out to be a vague quote from the San Diego mayor’s office at a time when the mayor is in a fight against public-sector unions. CAGW actually didn’t need to look hard to find a government employee abusing the system, namely San Jose Councilman Pete Constant, one of the presenters at the CAGW press conference yesterday. As it turns out, Constant is receiving disability pay as a former police officer while also earning a councilmember’s salary. This actually is fairly typical, as public-sector pay scandals usually involve elected officials or senior administrators with connections. In this case, San Jose councilmembers apparently exempted themselves from rules barring double dipping.

2. Ignoring obvious explanations. Much of the CAGW report is a compendium of state and local government budget woes—bankruptcies, layoffs, etc.—which the report blames on public pensions. The recession is mentioned only once, in passing. Likewise, there’s no mention of the fact that pension underfunding is itself a function not just of the 2007–09 stock market collapse, but also the failure of elected officials in many states and localities to make required pension contributions over many years, as Ross Eisenbrey noted yesterday.

3. Dubious sources. The authors claim each additional dollar in wages comes with $1.17 in additional benefits. The expert source? USA Today. The report also warns of the danger that public pensions will trigger a Greek-style fiscal crisis, citing pundit Fareed Zakaria. My personal favorite is the authors citing the Reason Foundation’s Adam Summers on why you should not control for education when analyzing compensation (public-sector workers tend to be better educated, so you can imagine how this biases results): “State and local governments hire more educated people not because the job duties demand more education, but because they have access to the public’s money and government budgets are not as constrained as those of private firms.” In other words, educated workers take government jobs that don’t require their skills because they get to play with a bigger piggy bank. I’m sure public schoolteachers who spend $1.3 billion of their own money to fund classroom supplies would be interested to learn this.

4. Alarmist and paranoid language. The authors darkly warn that if public workers in Michigan have their collective bargaining rights restored, “there would be no ceiling on the amount of money that unions could obtain in labor contracts.” Are the authors actually in favor of imposing legal limits on pay (for CEOs, say)? Now that would be interesting. The authors also repeatedly complain of a “lack of transparency” in government data sources, as if the Bureau of Labor Statistics and other agencies were hiding something. In fact, when BLS releases summary statistics instead of detailed data to the general public, this is to protect the privacy of employers and individuals who respond to government surveys (serious researchers can often access micro-data under strict confidentiality agreements). However, complaining about lack of data is how the authors justify what appears to be a very convoluted methodology, below.

5. Impossible-to-replicate methodology. The authors ignore standard economic models in favor of using “a system of 22 equations with 3,564 variables” to estimate public and private sector employment by major occupational group. It gets even fuzzier from here, but the authors appear to use occupation-weighted averages in the different sectors to compare wages and benefits, finding that state governments pay on average 6.2 percent more per hour in wages and benefits than the private sector. This is described in a “detailed methodology” appendix that is one-and-a-half pages long. Despite this lack of transparency (ahem), the little that is disclosed is enough to set off alarm bells. For starters, they appear to lump together workers in broad occupational categories like “management” or “healthcare practitioners and technical,” implicitly assuming they should be paid the same. Even within smaller occupational groups, like “protective service,” this would not make much sense (should police officers be paid the same as private security guards?) They then compound the problem by hand-waving when it comes to compensation, assuming, for example, that all government employees were assumed to have similar benefit levels.