Thursday, January 20, 2011

WEITW Part II:Paul Krugman, Worst Economist in the World {WEOTW}

Paul Kruman is at it again with a post entitled Early Social Security Projections. This got me to exploring some of the issues about Social Security also, and for that I am grateful. This is his attempt at setting up his straw man argument.
You see, in discussions of Social Security it’s often argued that in the program’s early years, nobody could have imagined the increases in life expectancy that have actually occurred, so nobody could have imagined that we’d have as many beneficiaries relative to the number of people of working age. And I thought I knew that this was wrong — that people in the 30s and 40s did know about rising life expectancy, and expected it to continue.

When declaring that "nobody could have imagined", then it seems pretty easy to refute such comments as someone, somewhere must of at least thought about it. And sure enough he uses the 1945 Social Security Trustees’ report {PDF}. But a bigger question is whether the people (Congress and President) that were suppose to make decisions about how to maintain the fund and thus benefits for future generations were even aware of such information (or even read it) and whether they made any corrective action early enough. Several passages of the report paints a different reaction than what Krugman could be imagining.
Two estimates, based on lower and higher assumptions, of the level premium cost of the benefits now provided by the system are 4 percent and 7 percent of covered pay rolls. This means that if pay-roll taxes of this magnitude (employer tax and employee tax combined) had been levied from the beginning of the program and were continued indefinitely, the system as a whole would be self-supporting if the assumptions eventuate. But Congress has now maintained old-age and survivors insurance tax rates at 1 percent each on employees and employers for 9 years instead of permitting the scheduled increases of the act to become effective, although in the war years economic conditions have imposed little obstacle to such increases. {Page 28}

The report's conclusion goes on to ask for clarification on "its short-run and long-run financial policies for the program" considering the 1 percent rate then currently used and whether the fund should be decreased until it is "three times annual benefits". That last portion seems like not much financial backup in the trust fund if allowed to drop to that low. Even if the rates were to increase to 6% total of taxable pay rolls that still only provides positive cash flows during the 1950s and 1960s.

Rising Life Expectancies?

This is Table 9 that Krugman references in his post. It is interesting to note that while the populations of those 65 and older is increasing every time period including for both men and women, adults in the 20 to 64 age group decreased for both low assumptions and high assumptions over the period 1980 to 2000. The low assumptions table even has a decrease in this age group in 2000 that is below even the 1955 estimate with men sliding back to the same spot and women decreasing more in numbers. The two estimated sequences are derived from two reports and not from the SSA/Trust Fund. The two reports are: 1. the 1935 report by the Committee on Economic Security, and 2. the 1938 report by the National Resource Committee.

The low assumptions ratio result as Krugman pointed out was amazingly close considering all the other factors that turned out wrong like total population and immigration patterns. His ratio of adults and the elderly of the low assumptions estimate was 20.8% and the 2000 census data showed 21.1%. But he should have read the report to see that the high assumptions were not given equal consideration to the low assumptions results. The following is from the report on page 20.
(b) Mortality --- Mortality rates by sex and age have been steadily improving since the turn of the century for both sexes and virtually all ages up to 60, with very little change at ages above 60. ... In the low-cost assumptions discussed in this section, very little improvement in mortality rates is assumed. In the high-cost assumptions some improvement is assumed but their assumption of improvement beyond age 65 is believed by many to be too optimistic.

Thus, the report is saying that it is better to go with the low-cost assumptions and only to use the high-cost for the extreme possibility. It was not, like what Krugman is assuming, that the low and high are of equal value and came from the same set of data points and assumptions. The ironic aspect is that Krugman already knows that longevity is not increasing that much for the older population than it is for the younger living longer.
Live Long And Prosper
While life expectancy is rising, life expectancy at age 65 — which is what is relevant here — isn’t rising nearly as fast.


"Damn it, it is not electron shells, we are talking about electron configuration!"
That small quote was from my college professor in Chemistry, but for Krugman it should not be the demographics of age groups that matter as much as participation rates. Table 10 below shows the expected number of recipients of "old-age insurance" up to the year 2000. The difference to note from Table 9 is that even adding up all the columns does not equal the total for the population over 65. The participation rate from these two tables is just over 58% for the low assumptions scenario and 71% for the high assumptions scenario. According to the Social Security Administration, "Nine out of ten individuals age 65 and older receive Social Security benefits." That means as much as a fifth to one-third of the elderly were not suppose to receive the benefits according to these scenarios. Even with this data, it does not tell us how many of the "wives of primary beneficiaries" are 65 or older. Since husbands still tend to be older than their wives, then at least some of that pool of recipients was under 65 making the non-participation rate lower than what I calculated.

It looks like a variety of social attitudes changed since the 1945 report and possibly work/leisure incentives. Although the following quote is related more to the "reconversion" of a war time economy to a consumer economy, it does point out the writers' attitude in the report, "It is assumed, nevertheless, that a rather large number of persons eligible for retirement benefits remain in employment." Another indirect demographic trend that they completely missed is women participation in the work force. On page 23 of the report they discuss greater woman participation because of the depression and subsequent war, "Moreover, the labor market was increased by many married women seeking employment to reinforce what they hoped might be only a temporary inadequacy in their husband's income."

Confirmation Bias
From the analysis above, it shows that Paul Krugman found some shred of evidence that confirmed his political bias and then failed to fully understand the information he was sharing. He, of course, has a great burden defending big government entitlement programs and Keynesianism at all economic and political cost. I close with another quote from the report that shows they understood that the program as designed could have some major flaws that will have to be addressed at some time in the future.
In fact, for demographic reasons alone, as discussed earlier in this section, the system cannot be expected even eventually to level out to a fixed relationship between contributions and benefits.

Labels: , ,

0 Comments:

Post a Comment

<< Home