Don’t Muddy Social Security Debate

As a person who has observed the Social Security program and its staff over the last 40 years, I was stunned by the NY Times op-ed “Social Security: It’s Worse than You Think,” by Gary King of Harvard and Samir S. Soneji of Dartmouth (1-5-13).  It is not uncommon for experts to disagree about the life expectancy estimates that underlie the actuaries’ long-run cost projections; these debates have been going on for decades.  But the allegations in the op-ed were that the actuaries methods were: 1) “antiquated, subjective, and needlessly complicated;” 2) “omitted crucial health and demographic factors” (specifically smoking and obesity); and 3) “the projections were “prone to…potential interference from political appointees.”  Those allegations go way beyond intellectual differences among experts.  And the implications are larger than professional in-fighting, because they call into question the future of the program that serves as the backbone of the nation’s retirement system.  

First, it would be difficult for the methods to be “antiqued, subjective, and needlessly complicated” given the level of interaction between the Office of the Actuary and the country’s most able demographers, economists, and health experts.  Every four years, the independent Social Security Advisory Board (before 1994, the Quadrennial Advisory Council) has established technical Panels that include experts from a variety of relevant fields.  I served on one of these panels in 1990.  After six months of review, the panel concluded that the work of Social Security’s Office of the Actuary and Office of Research and Statistics was “professional and highly competent.”  I might be an easy touch, but the panel included, among others, Larry Summers, former Secretary of the Treasury and President of Harvard, who is well known for not suffering fools, and Peter Diamond, who was awarded the 2010 Nobel Prize in economics.  And every subsequent panel has come to a similar conclusion, even as they recommended changes in solvency measures, economic and demographic assumptions, and presentation.  

Second, King and Soneji are simply incorrect about the Office of the Actuary ignoring the impact of obesity and smoking.  Of course, they take these well-known factors into account.  The Office of the Actuary looks at five well-defined groups of causes of death; these groups include cancer and respiratory disease, which are affected by smoking, and cardiovascular disease, which is associated with obesity.  The fact that these factors are already included explains why when King and Soneji overlay the projections with their assessment of smoking or obesity, they double count the already embedded assumptions and do indeed produce “crazy” results.

Finally, the projections are protected from political interference.   The Social Security projections are reviewed by a Board of Trustees, which consists of six political appointees and two public appointees from different political parties.  Even if the political appointees were tempted, the public trustees would never let it happen.  And my experience is that the political trustees have a strict policy of not interfering.  As assistant secretary of the Treasury for economic policy in 1994, I was the Treasury’s contact with the Office of the Actuary regarding their projections.  I was dismayed to see the long-tern deficit jump sharply on my watch.  I really pushed to see whether the projected increase was called for.  The actuaries were happy to answer any question and to show the data underlying their assessment.  But it was clear that they were not cowed by questioning from a political appointee, and the Secretary made it very clear that the Office of the Actuary had the last word on the numbers. 

So it is perfectly possible that the Office of the Actuary’s projections of life expectancy may be too low.  Reasonable people can disagree on this point.  But none of the op-ed allegations about methodology or political interference hold any water.