Tuesday, May 06, 2003

Social [In]Security: A large spread in the "Ideas" section of the Boston Globe this weekend featured an infotorial (now in the pay-only archives) by Dean Baker and Mark Weisbrot of CEPR in which they argue, in effect, that Social Security reform is a sop by the GOP to financial big boys. (Never you mind that the late Pat Moynahan, who was anything but in bed with Wall Street, was one of the first to point to the coming trust fund deficit.) Beyond that, Baker and Weisbrot's analysis is so full of holes that it's a wonder that it was run by any self-respecting editor. Some examples:

The "war against Social Security" is based on a shortfall, within 75 years, of "three-quarters of one percent of our national income," by which I assume they mean GDP, which shortfall will have to be covered out of general revenue via higher taxes. (Three-quarters of 1% appears to come out to about $100 billion, which is about enough to float 1 million retired baby boomers for one year, at today's rates -- which will be laughable 30 years from now, even at a low rate of inflation.) First of all, that kind of share of GDP is "huge" and "risky" when it's in the form of a tax cut instead of a tax hike. (Plus, run it through some dynamic analysis and you might find it does nearly as much harm as good.) Second, and more to the point, it's a weaselly statistic, since the important relationship is between revenue and shortfall, not GDP and shortfall.

Real-wage growth, they argue, will more than offset any tax bite to cover the trust fund deficit. That's possible, but in the next paragraph the authors steal that possibility when they note that the 30-year U.S. trend of real-wage stagnation makes Social Security all the more necessary. So which is it, boys? Can we grow our way out of the crisis or not?

The authors also aim to debunk a number of "accounting tricks" that reformers use to bolster their arguments. They say that reformers point to the fact that demographics suggest that we will move from an earners-to-recipients ratio of 3-1 today to 2-1 in 2035, then they claim (again) that increases in wages and productivity will cover demographic shifts. First of all, they cherry pick their numbers. Reformers are more likely to point to the 16-1 ratio in 1950 and the 5-1 ratio in 1960 (perhaps you see a pattern emerging?), and to the further demographic trends that threaten Social Security, like retirees living longer after retirement. Baker and Weisbrot's argument seems to be that this ship is seaworthy by redefining the meaning of "hull."

Another "trick," the authors say, is that the trust fund is "disparaged" as a box of IOUs. (It is exactly that.) "The bonds held by the Social Security Trust Fund," they say, "are backed by the full faith and credit of the US government, and it is a bit ridiculous to suggest that our government would default on them." Wow. Straw Man Alert! I don't recall anyone saying that default was the problem. The problem will be the costs associated with not defaulting. The trust fund took in the Social Security withholdings and loaned it to the Treasury, which promptly spent it. So the only accounting "trick" here is in the hands of the authors when they say that the trust fund contains securities "amounting to $1.5 trillion and growing by $200 billion a year." The only thing growing by $200 billion a year is the debt on the trust fund. When those securities are called due to pay benefits, guess where it'll come from. General tax increases. As noted before, the authors call it a "relatively small" tax increase needed to cover the shortfall. But remember that it is a tax in perpetuity, not a one-time hit to cover an anomalous shortfall. And this will not "pay for" Social Security. It will simply be a necessary complement to what is already drained from your paycheck as Social Security contributions.

This is ridiculous stuff, and the Globe should be ashamed for running it. Even the GAO sees the need for reform, and the GAO is not exactly known for being packed with fire-breathing Friedmanites.

No comments: