Tuesday, February 01, 2005

Donald, Meet Kevin

Luskin today:

And in the column today, Krugman betrays a fundamental misunderstanding of the economics of Social Security itself. He write, "we don't need to worry about Social Security's future: if the economy grows fast enough to generate a rate of return that makes privatization work, it will also yield a bonanza of payroll tax revenue that will keep the current system sound for generations to come." Krugman has forgotten -- or chosen to ignore -- that under current law Social Security benefits are indexed to wage growth. If the economy grows like Krugman is talking about, yes, payroll tax revenues will grow too -- but so will benefits, nearly perfectly proportionately.

Sadly, No! (trademark Sadly, No! productions) Kevin Drum explains.

Luskin tries to justify saying this by adding:

The sensitivity tables given by the Trustees of the Social Security Trust Funds don't show this -- because they arbitrarily cut off the calculation after 75 years. But the reality is that the early benefits of increased tax revenues are eventually offset by the higher cost of benefits. Gee -- think how good Krugman could make that look if he reduced the window of analysis to just 10 years.

Really, Krugman only focused on 75 years! How dare he! The Wanker! But, look, Luskins's just wrong here. As the SSA explains:

The cost rate decreases with increasing real-wage differentials, because, higher wages affect the taxable payroll immediately but increase benefit levels only gradually as new beneficiaries become entitled. In addition, cost-of-living adjustments (COLAs) to benefits are not affected by changes in wages, but only in prices. Each 0.5-percentage-point increase in the assumed real-wage differential increases the long-range actuarial balance by about 0.54 percent of taxable payroll.

To Luskin, the fact that benefits eventually catch up with increased revenues mean that in the end it's a wash and Social Security is DOOOOOOMED at some unspecified date even longer than 75 years from now. But, that's because Luskin seems to not understand why there's any sort of wee problem with the long room solvency of Social Security -- a presumed unique demographic phenomenon known as The Baby Boom which is going to begin skewing the ratio of retired/workers. This is to a great degree a temporary problem. When all of those people die, and the age distribution of the population is no longer affected by them, then that problem will largely go away. So, if 20 year olds start earning more/paying more in taxes now, they don't start getting juicier benefits for another 47 years, while that additional revenue is in place to take care of "current" obligations.

That doesn't mean that revenue/benefit ratios won't ever needed to be tweaked to keep the program in rough balance, but jeebus on a cracker, how many years of guaranteed solvency do we need to satisfy these people? 5000?

...okay, I'm not exactly right here. Depending on the assumptions made, the retiree/worker ratio problem doesn't go away as the boomers die. So, Luskin is correct when he suggest we should perhaps be worried about a disaster in the system 75+ years from now.* I do think the fertility, mortality, and immigration assumptions used to derive these numbers are pretty silly (mostly the immigration), but that's a separate argument.

*That's sarcasm, of course. The fact remains that if long run wage growth remains high, then it's always the case that higher revenues are received "now" and higher benefits aren't paid until "later." That process doesn't magically stop 75 years from now, but for a precise accounting we'd need a sensitivity analysis result from the "infinite horizon" mode...