Bryan Caplan  

Two Questions for Orszag

How to Evaluate Government Pay... We Smell a Rat...
I've got two questions for Peter Orszag:

1.  You claim that education and age differences explain the entire difference in average pay between federal employers and other workers.  Does "average pay" include benefits?  The salary gap reported in USA Today was only about $8k, but the reported benefits gap was an additional $31k.  I'm almost sure that you're not counting benefits - a massive oversight.

Here's why.  According to Peter, educational differences account for 82% of the earnings gap, and federal workers are roughly 17 percentage-points more likely to have a college degree.  Even with a 70% return to college, education and age combined would only explain why federal employees' earnings are .17*.7/.82=14.5% higher.  That's just about right for the salary gap, but it appears to leave the massive benefits gap entirely unexplained.  Am I wrong, Peter?

2. Federal workers have a lot more job security than other workers - especially today.  Extra job security is yet another benefit that federal workers enjoy, but since it's intangible, it's almost certainly not included in the official value of their salary or benefits.  So by the logic of compensating differentials, Peter, shouldn't federal workers' measured (salary + benefits) be lower than comparable workers get elsewhere?

Comments and Sharing

COMMENTS (18 to date)
RG writes:

So, how much lower should a tenured professor get paid than someone in the private sector given the additional job security. Looking at the salaries of faculty in my department (Comp Science not Econ), they get more or less whats in line with the industry scaled to 9 months.
Any comments?

jc writes:

Fwiw, professors at private universities often make more than those at public universities (I'm thinking of an AACSB salary survey I recently read). Occupation wise, I suppose that would be an apples to apples comparison.

Tenure is offered in both settings. But before tenure, I've been told that it's easier to get fired at many private schools, where one may be expected to excel not just as a researcher, but also as a teacher. Similar publication requirements for tenure, but if students and/or parents don't feel like they're getting their money's worth, education wise, you're out the door.

I also know of a few more prestigious schools where virtually nobody is granted tenure (it's technically possible, but the requirements are ridiculously high). Compensation does tend to be higher than average.

Felix writes:

Isn't there a second problem? Don't the feds, if not other levels of government, pay for college degrees? That is to say, they pay more as a function of an employee's degree - which is not the same thing as paying for work done.

That's a bubble marker, by the way. That is, something that is valuable because it is valuable. I speak of the value of a degree.

Marcus writes:

Peter hasn't actually explained anything.

We should fully expect that higher paying jobs attract more highly educated workers.

If Walmart paid $60/hour to run a cash register imagine all the engineers who would apply for the jobs. But that wouldn't be a very productive use of engineers.

Since Peter appears to believe he's answered the question, reword the question. Why should government workers be more highly educated than their private counterparts?

Hume writes:


I'm pretty sure that law professor salaries are much less than practicing attorneys with similar credentials.

Ted Craig writes:

He's not talking about federal workers, but former SF Mayor Willie Brown did say:
“The deal used to be that civil servants were paid less than private sector workers in exchange for an understanding that they had job security for life,” Brown asserted. “But we politicians — pushed by our friends in labor — gradually expanded pay and benefits . . . while keeping the job protections and layering on incredibly generous retirement packages. . . . This is politically unpopular and potentially even career suicide . . . but at some point, someone is going to have to get honest about the fact.”

Randy writes:

The education gap is interesting, but I don't believe it means what Peter thinks it means. It doesn't mean that these "educated" individuals have produced anything of greater value by which to earn their higher compensation, but only that the education system is an entitlement path for the political class.

eccdogg writes:

Federal Employees have a great deal and this can easily be seen in the fact that once a federal employee has a job he is 3 times less likely to voluntarily leave his job as a private sector worker is.

Private sector quit rate per year ~25%+
Federal quit rate per year ~9%

It doesn't look like the feds are losing many folks to the private sector.

See the stats here (table 16)

Floccina writes:

Fed Gov. employees are often over qualified educationally because the pay is high. For example a lot of letter carriers have college degrees despite it being a job that should only require a high school degree. If you over pay and for example use a civil service exam you will get employees with higher educational credentials. That does not mean you are not over paying. In fact the worst damage of overpaying for Fed. jobs might be that well qualified people are less productive than they would be otherwise .

Ted Craig writes:


The same is true (or was) for auto assembly workers.

bjk writes:

Employees generally tend to discount the value of benefits, especially if the benefits are 20 years down the road . . . Recruits in the army, for instance, much prefer higher salary to higher benefits. The FG could get more for less by increasing salary and reducing benefits.

David S. writes:

Government employees are both underpaid and overpaid. Unlike in the private sector, Government employees (at all levels) tend to have much more uniform pay schedules tied to experience, not to the market for their skillset.

Someone with a BS in English will be quite overpaid by the Government. A medical doctor or someone with a Ph.D. in a useful field like economics or biology will likely be quite underpaid. Employees with valuable IT skills are underpaid and letter carriers are overpaid. High school teachers as a whole are probably overpaid, but math and science teachers are paid much less than their skills are worth, which is why qualified people in those fields are so difficult to recruit.

I could easily believe that the net of all these inefficiencies was an overpayment, but I think the total errors in both directions are much larger than the net result would indicate.

Randy writes:


I'm familiar with the system and understand your point, but to me there's a bigger point, which is that the amount paid to government workers is entitlement based, not value based. That is, a Ph.D. or technical specialist working for the government is not paid based on any voluntary value for value transaction, but rather, on a political calculation as to who gets what share of the revenue collected from the productive sector.

David Youngberg writes:

@bjk: "Employees generally tend to discount the value of benefits, especially if the benefits are 20 years down the road . . . Recruits in the army, for instance, much prefer higher salary to higher benefits."

How much of that can be explained that army personnel are far more likely to die on the job compared to most other government jobs?

Felix writes:

David Youngberg writes: army personnel are far more likely to die

The words, "danger" and "civilian government worker" go together like stripes and dots, but, I doubt the "far". Army may not even be more likely to die on the job. Hasn't the IRS taken casualties recently? :)

mark writes:

"If Walmart paid $60/hour to run a cash register imagine all the engineers who would apply for the jobs. But that wouldn't be a very productive use of engineers."

I don't know about that. I have a manuel labor position and would absolutely love to have folks from engineering and IT do some of the work. I try to use their programs and the systems they design and I would like to see them use those systems as well. Walmart seems so wildly successful that perhaps their systems work better then the systems I use. A smart engineer figuring out how to make cashiers .01% more efficient would certainly be worth $60/hour.

Randy writes:

Great comment. My job involves working directly with customers and cashiers, and directly with the engineers. The problem is getting the engineers, and marketing, and testing, and production, and management, each of whom have their own agendas, to listen.

mark writes:

hume writes correctly
"I'm pretty sure that law professor salaries are much less than practicing attorneys with similar credentials."

This just proves that there are significant nonmonetary components of employment choice. The main difference is that one group gets to work on topics of their own choosing and mainly on their own schedule. The other has to work for clients both in terms of topics and schedule.

There are also significant differences in how one markets oneself between the two groups.

The point, to me, is that value laded arguments to critique income levels or tax income cannot be sustained because the nonmonetary aspects cannot be quantified.

Comments for this entry have been closed
Return to top