AMERICAN.COM

In the wake of CBO’s recent report finding that federal employees are overcompensated by an average of 16 percent, public employee unions and members of Congress who support them had two main reactions.

First, they say, the Bureau of Labor Statistics—not the CBO—are the real experts on pay. Second, they argue, pay studies should compare jobs, not the education or experience of the people who fill those jobs. Well, they’ve got what they want.

The Journal of Economic Perspectives yesterday published a new study by Brooks Pierce and Maury Gittleman of the Bureau of Labor Statistics which uses BLS estimates of the skills required in different occupations to compare public and private sector pay.

Brooks and Gittleman restrict their analysis to state and local government employees, but their findings are striking: state government workers receive salaries about even with private sector levels, while local government workers receive salaries around 9 percent above private sector levels. Once you include benefits, state employees are overpaid by around 9 percent and local government workers by around 18 percent. Moreover, there’s good reason to believe these estimates are conservative, since they exclude the value of retiree health coverage for public employees and understate the true value of defined benefit pensions.

You can’t say for sure what this implies for federal government employee pay, but they’re generally regarded as better paid than state and local government workers.

Writing for the Washington Post, Colleen Kelley of the National Treasury Employees Union pushes back against the recent CBO report finding that federal employees are, on average, overpaid by around 16 percent.

Kelley notes that the CBO “compared characteristics of the federal and private-sector workforces, rather than comparing jobs done in each sector. The latter is the approach of the Bureau of Labor Statistics, which has consistently found a pay gap in favor of the private sector. The latest BLS report shows that gap to be an average of 26 percent.”

One problem with the BLS report, that CBO noted over 25 years ago and that more recent academic literature has confirmed, is that the federal government may not fill the same jobs with the same employees. A study of BLS occupational data by Melissa Famulari of the University of California, San Diego, found that, “Federal workers have significantly fewer years of education and experience than private sector workers in the same level of responsibility in an occupation.”^[1] Famulari finds that these differences play out through federal hiring and promotion practices:

The Federal government, particularly in Washington, DC, hires workers at initially higher levels of work. These differentials are so large that, even after a number of years on the job, private sector workers are employed at substantially lower levels of responsibility than the starting levels of responsibility for DC Federal government workers. In addition, the Federal government, particularly in DC, promotes workers more quickly than in the private sector, conditional on observed worker characteristics.

Famulari concludes: “The large private sector premium paid to workers in an occupation and level is largely explained by the more valuable skills of private sector workers within an occupation and level.”

Most economists agree that the main determinants of pay are the skills and abilities that workers bring to the job. If these differ between the federal government and the private sector, as Jason Richwine and I explained in our AEI paper on federal pay, then it is possible for federal jobs to be underpaid while federal workers are overpaid. But it’s pay to workers that we care about.

[1] Famulari, M. “What’s in a Name? Title Inflation in the US Federal Government.” Working paper. 2002. Revision requested by Industrial and Labor Relations Review.

The Congressional Budget Office’s recent study on federal employee pay found that federal retirement benefits were about 3.5 times more generous than those paid to similar workers in the private sector, helping drive an overall compensation premium of around 16 percent over private sector levels. But the CBO study missed an important benefit offered to federal employees that Jason Richwine and I caught in our own analysis of federal pay.

In addition to a traditional defined benefit pension plan, federal employees participate in the defined contribution Thrift Savings Plan. There is a little-known but generous subsidy to the largest investment fund in the TSP, the so-called G fund, which invests in special-issue U.S. Treasury securities. What most people don’t know is that the G fund pays significantly higher interest rates than similar Treasuries available to private sector workers with 401(k) pension plans.

The TSP states: “Although the securities in the G Fund earn a long-term interest rate, the Board’s investment in the G Fund is redeemable on any business day with no risk to principal. The value of G fund securities does not fluctuate; only the interest rate changes.” This is unlike marketable Treasuries, which like any bond will rise or fall in value as interest rates change.

The TSP advertizes this as “The G Fund Advantage.”

“The G fund rate calculation described above results in a long-term rate being earned on short-term securities. Because long-term interest rates are generally higher than short-term rates, G fund securities usually earn a higher rate of return than do short-term marketable treasury securities…. From January 1988 through December 2010, the G fund rate was, on average, 1.77 percentage points higher per year than the three month T-bill rate.”

In effect, the G fund receives an interest rate subsidy of around 1.77 percentage points versus marketable investments of similar risk. In other words, if federal employees were offered marketable Treasuries the same as everyone else, they would receive less interest and the federal government would reap the savings. As of December 31, 2010, the G fund held $128.6 billion, making for an annual interest rate subsidy of around $2.3 billion. That’s equivalent to raising average federal salaries by around 2 percent.

The Congressional Budget Office’s finding this week that federal employees receive average compensation 16 percent above that of similar private sector workers builds on work that Jason Richwine and I did in an AEI paper last year. One of our key results was that federal retirement benefits are much more generous than private sector benefits—3.5 times more generous in CBO’s study, and even more so in ours.

One element in this conclusion is in how you “discount” future retirement benefits to calculate their value to workers today. We and CBO both assumed that if a future benefit was guaranteed, you should discount it using the market interest rate on guaranteed assets. You’d use that lower interest rate even if the government funds its benefits using risky assets with higher expected returns, because what you’re valuing is the benefit, not the government’s strategy for financing it.

Our friends at the Economic Policy Institute, however, think that somehow Jason and I duped CBO into using this method:

The experts consulted by CBO for this report include my EPI colleague, Heidi Shierholz, as well as two researchers on the other end of the political spectrum, Andrew Biggs of the American Enterprise Institute and Jason Richwine of the Heritage Foundation. Though it’s nice that CBO considered a range of opinions, it’s a shame that Biggs, Richwine and others have been successful in convincing CBO and others to use low Treasury yields to estimate the cost of future pension benefits.

I’d like to think that our prior work made clear how important it is to correctly value future retirement benefits. But it’s evidence of how out of step EPI is with the rest of the economics profession that we really didn’t have to convince CBO of anything.

If anything, I think CBO may not have gone far enough. The 20-year Treasury yield today is around 2.6 percent. CBO assumed a 4 percent Treasury rate, then added 100 basis points to account for the fact that federal pensions aren’t as guaranteed as Treasuries. I think a 100-basis point risk adjustment is too much for already-accrued federal pensions, which seem almost certain to be paid even if the terms by which future benefits are earned change. The value of benefits is very sensitive to the discount rate, so using plausible assumptions, the federal compensation premium today could be even larger than what CBO found.

The Congressional Budget Office has released a new study showing that federal government employees receive significantly higher compensation than private sector workers with the same levels of education and experience. The CBO report confirms many of the findings of a 2011 study I wrote with Jason Richwine of the Heritage Foundation and helps rebut claims that federal workers are underpaid.

CBO found that federal employees receive average salaries that are about 2 percent higher than those for similar private sector employees and benefits that by 48 percent exceed private sector levels. Total average federal compensation is 16 percent above private sector levels. With federal employee compensation totaling $200 billion per year, a 16 percent pay premium is big money.

CBO’s methods are broadly consistent with the 2011 AEI study, although we found a larger federal pay premium because we sought to capture a broader range of federal compensation—including the implicit value of federal workers’ near-total job security—and because of somewhat different economic assumptions. Nevertheless, the CBO report serves as a valuable contrast to figures generated by the federal Office of Personnel Management claiming that federal employees are underpaid by 26 percent relative to private sector jobs.