Thursday, January 21, 2016

Are the CBO's Social Security projections unreasonable?

Alicia Munnell of the Center for Retirement Research at Boston College has a Marketplace column that's critical of the Congressional Budget Office's projections for Social Security's financing, so much so that she places the CBO estimates outside of the "reasonable" range that policymakers should focus on. Is she right?

Social Security's Trustees project that the program faces a 75-year actuarial deficit of 2.68% of taxable payroll. For simplicity, that implies that a 2.68 percentage point increase in the Social Security payroll tax -- from the current 12.4% to 15.08% -- would be sufficient to keep the program's trust funds solvent through 75 years. The CBO projects a 75-year shortfall of 4.37% of payroll, which obviously is substantially larger than the Trustees' forecast.

Munnell points to her recent experience heading the Social Security Advisory Board's Technical Panel on Assumptions and Methods. The Tech Panel makes its own recommendations with regard to the main demographic and economic assumptions that drive Social Security financing. Based on the Tech Panel's recommendations, SSA's actuaries scores a 75-year shortfall of of 3.42% of payroll. So Munnell isn't saying that the Trustees projected deficit might not be too low, but rather that the CBO forecast is simply too high to be considered reasonable.

Part of this difference of opinion might come about from a misunderstanding: Munnell points to three areas in which the CBO makes different assumptions than the Trustees: first, CBO assumes greater longevity increases than the Trustees, something which Tech Panels have been recommending for as long as I can remember; second, CBO assumes lower interest rates, which reduce interest earned on the Social Security trust funds. This also is a common area of disagreement, given that current interest rates are so low. Third, CBO assumes that disability benefit claims will continue at higher rates than do the Trustees. I have some sympathy with this argument, given -- as Corner University economist Richard Burkhauser has noted -- the Trustees/actuaries have in the past assumed that relief from rising disability claims was around the corner, but disability continued to increase. While it's not clear by how much CBO's assumptions differ from the Trustees, no one familiar with the process and debates would argue that reasonable people can't disagree on these issues. Having been involved with the Trustees Reports during my time at SSA, there were often people in those meetings who disagreed with where the Trustees ultimately ended up, in ways that would increase or decrease the funding shortfalls.

But Munnell appears to miss another difference of opinion between the CBO and the Social Security Trustees: CBO assumes that earnings inequality will continue to rise, whereas the Trustees assume that after 10 years there won't be any additional changes. Increasing earnings inequality reduces the share of wages covered by Social Security taxes, as well as reducing benefits paid on those wages, but the net effect is to worsen Social Security's long-term deficit. Again, this is an issue on which reasonable people can disagree: my gut is that if health cost increases continue to remain slow, we may see a reduction in earnings inequality. (See this WSJ piece for why.) On the other hard, the Trustees tend to flat-line trends after 10 years due to how SSA's actuarial models are built, which shouldn't be the deciding factor in what's reasonable or not.

It would be nice to see greater detail on CBO's projections, from which we might be able to figure out more precisely what is driving the differences between CBO and the Social Security Trustees. But for now, I don't see any evidence that CBO's projections are unreasonable. The CBO has the more sophisticated model for projecting Social Security's finances and my own view is that the CBO group is more willing to embrace new evidence or thinking as it comes up. That said, as Munnell points out, only time will tell which long-term assumptions turn out to be correct.

If the CBO is right, though, we face a large problem that our political processes are very ill-equipped to address prior to a crisis coming about. That's worrying.


JoeTheEconomist said...

Solid close...

WilliamLarsen said...

Past performance is no indication of future performance. A common phrase when speaking with financial planners. However, having been looking at the SS-OASI problem from the early 70's, I can say that the Trustees have never projected any where close to to reality. The projection of 1983 thought 2064 was the year SS-OASI would run short. Over the course of two decades the date moved forward as early as 2029 and now recently is 2034.

The CBO likewise does not have a good track record either. The last general budget surplus was in 1957, though they routinely project numerous scenarios for proposed congressional legislation.

IT was refreshing to read that "the Trustees tend to flat-line trends after 10 years due to how SSA's actuarial models are built". This has always been a question I have had. My analysis revealed a linear symmetry to the SSA Trustees output. With changes in birth rate, labor participation rates, inflation and wage growth, how did the Trustees seem to be so linear?

Just one more bit of information that supports there is no design for Social Security. No criteria was ever created to measure the success or failure of Social Security. It is simply amazing it has survived scrutiny all these years.

My guess; as long as social security pays checks on time, no one will look behind the curtain and ask the important questions. I think we are about to see behind the curtain.