Are we undermeasuring U.S. savings?

Savings are hard to measure; the government's Bureau of Economic Analysis almost surely understates the true figure...

Perhaps the biggest problem with the way the government measures savings is its failure to take account of changing asset values like rising home prices. If we looked at the balance sheets of American families, instead of just subtracting consumption from income, we would find a more favorable picture, despite some short-term volatility.

In a speech last fall, Roger W. Ferguson Jr., vice chairman of the Federal Reserve, expressed concern that households weren't saving more, but he acknowledged that the ratio of household net worth to disposable income "has been essentially trendless over the past two decades," adding in a footnote that "this alternative concept of the personal saving rate has, in fact, shown a slight positive trend since the early 1950's."

The treatment of capital gains in the widely cited saving figure leads to a number of problems. Consider pensions. Employer contributions are counted as personal income, but later - and larger - payouts to retirees register as consumption, while the capital appreciation that substantially finances these payouts registers nowhere.

That's according to a Dartmouth economist, Steven F. Venti, who contended in a 2003 report written with two Dartmouth colleagues, Annamaria Lusardi and Jonathan Skinner, that the treatment of pensions had played a large role in dragging down savings to record levels, "accounting for over 40 percent of the total decline in the personal saving rate from 1988 through the turn of the century."

Other measurement issues also play a role. In a paper on alternative measures of personal saving, Marshall B. Reinsdorf, a research economist at the Bureau of Economic Analysis, asserts that adjusting the calculation for several such factors, including the treatment of spending on durable goods and the effect of inflation on interest rates, can account for much of the apparent decline in personal saving in the 1990's.

Daniel Akst at The New York Times offers more; I've added the links in the above quotation.

Note also that the U.S. spends more on education than most countries, but this is counted as consumption not savings (admittedly we wish to measure outputs, not inputs, but U.S. higher education is excellent).  R&D investments do not count toward national savings figures either; on these facts see the Michael Mandel article in 17 January Business Week.