Monday, October 27, 2008

Fudging Data to Prove Markets Are Fair

In The Confiscation of American Prosperity, I looked at the career of Martin Feldstein, until recently longtime director of the National Bureau of Economic Research, chief economic advisor during the Reagan administration, and inveterate foe of Social Security.

In the recent issue of the NBER digest, you can read a short discussion of Martin Feldstein's "proof" that wages have kept pace with productivity.

In case you do not want to purchase the article, you can read an earlier version here

The result is a display of incredible virtuosity. By selecting the appropriate deflators [a measure of inflation], including benefits as part of wages, and comparing wages with National Income rather than the Gross Domestic Product, voila, Feldstein manages to get a relatively constant wage share of national income.

Using a different deflator is a common tactic for arguing that wages have kept pace with productivity. At the same time, many fees that ordinary people face are not included in any of the deflators. Also keep in mind that wages include multimillion dollar executive salaries as well as those of minimum-wage workers.

National Income differs from Gross Domestic Product because it includes the depreciation. By subtracting depreciation, the wage share increases.

The nature of depreciation is changed greatly over the. The Feldstein covers, 1970-2006. During this period, the nature of investment changed, with less long-term investment in fixed capital goods at the same time that software, which has a very rapid depreciation, became counted as part of capital expenditure. If a $1000 were invested in a factory building in 1970 and an equivalent thousand dollars were invested software in 2006, a relatively small part of the 1970s investment would be counted as depreciation. In contrast, a large part of the 2006 investment would represent depreciation. So, between 1970, depreciation rose from about 10% of GDP to 13%. Since National Income rose more slowly, that kept the wage share from declining as much.

In so far as benefits are concerned, the two most important are pensions and medical care. Since medical care has experienced rapid inflation over the period covered, a careful study would deflate medical care costs to account for inflation -- especially in a paper that is supposedly paying special attention to the appropriate way to account for inflation. In other words, the value of a dollar's worth of medical care in 2006 was much less than comparable expenditure in 1970.

So, by carefully selecting measures of production, inflation, and wages, Feldstein made the case that wages have kept up with productivity and growth of income. This exercise shows more about the degree to which committed ideologues will go to prove that markets are both efficient and just.


Anonymous said...

Feldstein is examining the proposition that the Wage equals Price time MP. In this case, using the same price index to deflate real output and wages is correct. Calling the results "real compensation" is less correct, but he is careful to amend his statement to say "using the same price deflator." The issue of using GDP or national income depends upon how well depreciation is measured. In any case, a recent article in the Monthly Labor Review covered the many ways wages and compensation are calculated. It is not a simple call.

Michael Perelman said...

I am not sure that I would agree that using the same price index for output and wages is correct. Most of his discussion is about shares.

The appropriate deflator depends upon what is deflated. For example, the cost of living for older people (because of the greater weight of medical care) or rural people is different from young urban people..

Take the example of various fees that are not counted or the higher interest rate that poorer people use.

You are correct that macroeconomic measurement is a can of worms, but Feldstein is very selective about his worms.