In March, USA Todayreported that federal employees were heavily overpaid, Peter Orszag countered with some apples-to-apples econometrics, and I objected that Orszag (a) seemed to ignore exceptionally generous federal benefits, and that (b) official benefit statistics fail to count a major perk of federal employment - job security. This interview with Andrew Biggs offers further insight on the topic:
Does the public sector get pampered?
you control for the differences in characteristics like education and
age, the federal government has salaries that are 12 percent higher
than similar private sector workers.
You're aware that some studies have come the opposite conclusion: that federal workers are underpaid. Why are they wrong?
difference is in the methodology. We looked at the same person -- based
on age, education, all those things -- and put him in the federal
government or the private sector. The government looked at the same job
in the federal government vs. private sector.
They looked at jobs. You looked at people. Why does the distinction matter?
government promotes faster, and at a younger age, than the private
Some government employees don't participate in Social Security. How does that change the benefits picture?
[T]hat's irrelevant because they're
neither paying nor receiving benefits. If you follow Social Security,
you know it pays a low rate of return... [N]ot to participate in
Social Security is actually a benefit, because they're keeping more.
assume that there is a public/private pay gap. So what? Money attracts
talent and we want a talented federal work force...
...The pay gap difference between federal and
private jobs is not uniform. In general, the pay premium is very large
for low skilled workers, and small or negative for highly skilled
workers... But most people don't have MDs or PhDs. Up through a Masters, you'll
find on average that people make more in the federal government,
especially at the low end with folks like paper clerks.
Striking to me: If the public knew that paper clerks were getting a big premium - and Ph.D.s weren't - they might actually favor the status quo on populist grounds.
Update: Adam Ozimek writes me:
Re: your post on public sector pay, I've crunched the numbers using the
same dataset and methodology used by studies that liberals cite to show
that local and state employees are underpaid. I found that the federal
pay premium was from 17% to 24% depending on how much you controlled
for geography. The wage premium goes as high as 30% if you include the
union wage premium, i.e. union wage premium + public sector wage
Here is my post on it, and here's another using the same data on state and local public sector wages.