Are Government Workers Overpaid, or Underpaid?

There has been a lot of back and forth between liberals and conservatives over the last year or so about whether government workers are overpaid, or underpaid.  There are a bunch of problems trying to get a good comparison.  For example, the government does a lot of stuff that the private sector doesn't--there are no "private sector infantrymen" to compare to the ones who work for the federal government.  

Andrew Biggs and Jason Richwine of AEI have hit on the best method so far: they looked at what happens to government employees who leave to work for the private sector, and vice versa.  Their finding:  

According to the SIPP data, the average federal worker shifting to a private job actually accepts a small salary reduction of around 3 percent. Similarly, private sector workers who move to federal jobs don't take a pay cut. They get a first-year raise averaging 9 percent, well above the raise other workers get when they switch jobs within the private sector.

. . . Nationwide, non-teachers who move into teaching receive an average raise of around 8 percent, according to SIPP data, while teachers who leave the profession take an average salary cut of around 3 percent. Similarly, three recent state-level studies (in Florida, Missouri and Georgia) using administrative records found no average wage increase for ex-teachers.  

Of course, it's possible that only the worst government workers leave.  Nonetheless, the best read of the data seems to be that we are, if anything, overpaying for the government workers we hire.  

That does leave open a question: could we, or should we, pay more and improve the quality of the government workforce?  All this tells us is that we are paying most government workers as much as, or more than, they're worth.  It doesn't tell us that we might not want workers who are worth more in the private market.