05/02/2012 01:42 pm ET Updated Jul 02, 2012

Everyone Agrees That the Decline in Private Sector Pay Has Been Understated

Jason Richwine and Andrew Biggs have a piece saying that many public sector workers are overpaid in which they also say that I agree with them in much of their analysis. This is true.

Let me outline what I think are areas of agreement. First, we seem to agree that if we just compare the straight wages paid to public sector and private sector workers, the latter do better. When we adjust for education and experience, private sector workers tend to get higher pay than their counterparts in the public sector.

This is not true across the board. My colleague John Schmitt has found that while workers with college and advanced degrees (e.g. doctors and lawyers) get less in the public sector, less educated workers get paid the same or slightly more than their counterparts in the private sector. In other words, there is less inequality in public sector wages than we see in the private sector, with the average being somewhat lower.

We also agree that the lower wages for public sector workers are largely or completely offset by higher benefits. The key difference here is that public sector workers are far more likely to have a traditional defined benefit pension plan. Most workers in the public sector still have defined benefit pensions, while less than 20 percent of workers in the private sector do. (The difference is considerably less stark if we restrict the comparison to large private firms, where defined benefit plans are still common.)

Richwine and Biggs conclude that public sector workers are overpaid because the two assign a high value to the nature of the guaranteed benefit in traditional pension plans. Their point is that a guaranteed benefit of $1,500 a month in post-retirement is worth much more than a contribution to a 401(k) type plan that would on average provide the same benefit.

Richwine and Biggs are absolutely correct on this point. This guarantee does have considerable value for workers. The point here is straightforward. We may expect the stock market to provide returns that average 10 percent before adjusting for inflation, but if the market happens to be down in the year I retire, then I am out of luck. If the government has assumed this market risk for me, and will simply give me the average return, then it has made me considerably better off.

There are several conclusions that logically follow from Richwine and Biggs' point. First, if a state were to take the exact same money that it is now contributing to a defined pension plan and instead give it to workers in the form of a 401(k)-type plan, as Utah recently did for many state employees, it will not save taxpayers a penny. However, it will eliminate Richwine and Bigg's basis for claiming public sector workers are overpaid. Perhaps this has made taxpayers in Utah happy, but I am not sure how many people around the country view reducing the security of public sector employees as an end in itself.

A second conclusion that follows from Richwine and Biggs's point is that advocates of Social Security privatization, such as those in the Bush administration, were misleading people when they compared the uncertain returns from individual accounts with Social Security's guaranteed benefit. Social Security's guarantee is worth a great deal, so a straight up comparison of dollar values is not accurate. (In addition, the privatizers also hugely overstated what individuals could expect to earn from private accounts.)

A third conclusion is that most analyses of wage growth over the last three decades, which show the relative stagnation in real wages for most of the population, understate the losses to workers. In 1980, close to half of all workers had a defined benefit pension. With the current figure close to 20 percent, there has been an enormous loss in security for a large segment of the workforce. This loss is not picked up in calculations that simply measure payments for wages and benefits.

If we concede that the benefit guarantee has substantial value to workers, the question is how we should think about the cost of the guarantee to governments and ultimately taxpayers. I have argued that, given current market condition, this guarantee carries a trivial cost, essentially because governments will live indefinitely and have little reason to be concerned whether the market is up or down in a specific year. In other words, a state or local government has no reason to care if the market is down the year I retire, as long as they get the average return on stock right.

The last point is key. State and local governments, like private pension plans, got the average return hugely wrong in the late 90s and the last decade. The reason was that the stock market was hugely overpriced. Price to earnings ratios in the stock market were far out of line with historic averages, peaking at over 30 in the 90s stock bubble, compared to a long-term average of less than 15.

The notion that an over-valued stock market could provide the historic average return is laughable on its face. Suppose that a bond pays $5 a year in interest. If the bond is priced at $50 then the return is 10 percent. However if the bond price goes to $100, then the $5 annual interest payment will only imply a return of 5 percent.

This logic is about as simple as it gets, but none of the great minds of the economic or accounting professions were interested in hearing this argument back in 90s stock bubble or during the years of the housing bubble in the last decade, when the stock market was still over-valued. They wanted us to in effect believe that the bond was still providing a 10 percent annual return, even when we were paying $100 for it.

Incredibly, now that the stock market (and pension funds) have taken its hit, they want us to start assuming lower rates of return pension assets held in stock. On this issue, I am afraid that I will again have to side with arithmetic over the leading authorities in the economics profession. Having done the analysis, there is virtually no possibility that pension plans will have markedly worse returns than the approximately 8.0 percent average nominal yield they now project. In other words, the risk to taxpayers from guaranteeing the pensions of public employees is close to zero.

This means the cost to taxpayers of public employees' compensation packages are on average no greater than the cost to private sector employees. The big difference is that public employees can expect a considerably more secure retirement.

Richwine and Biggs conclude their piece by telling us that "basic fairness" dictates that we have to do something about the greater level of retirement security enjoyed by public sector workers. This is an interesting notion of fairness.

In today's economy basic fairness might raise images of Wall Street traders pocketing hundreds of millions a year or CEOs making tens of millions as they push their companies to the brink of ruin. But for Richwine and Biggs, the main concern is schoolteachers living on pension of $2,500 a month.