PensionReforms summary and comments
For more than 25 years, US macro-economic data have shown large decreases in measured 'saving' levels.  No-one really seems to know why and, in the meantime, both household wealth and the wealth to income ratio has grown.

"For many purposes, researchers are content to use estimates of the stock of wealth from household surveys.  By enabling greater future consumption, wealth is an important element of economic well-being, but it may matter little how the households accumulated it.  However, in order to understand the process of wealth accumulation it is critical to be able to distinguish between wealth changes due to decisions to forego consumption and changes that result from externally-driven changes in asset values.  The distinction between price changes and saving is particularly important at the level of the aggregate economy because of the role of saving in the financing of the nation's investment and the link between the saving-investment balance and external trade.  In understanding the causes and consequences of reduced saving, it would be useful to know whose saving has fallen and is it a general phenomenon or one limited to a few socio-economic groups?

This report suggests that we are unlikely ever to find out why the reason for the divergence in macro and micro-economic measures, at least not in the meantime.  It then tries to match the numbers using wealth data from three micro-level surveys: the Survey of Consumer Finances (SCF), Panel Study of Income Dynamics (PSID), and Health and Retirement Study (HRS). 

"We provide comparisons to the macroeconomic estimates of wealth accumulation and saving, explore problems in constructing household-level valuations of wealth, and assess the value of using household-level datasets to examine wealth accumulation and saving behavior in the United States."

The three sets of data produce reasonably consistent results:
"Our first analysis compares the macroeconomic estimates of wealth from the Flow of Funds to comparable measures from the SCF, PSID and HRS.  The Flow of Funds and SCF valuations of net worth correspond closely up to 1998.  Yet, after 1998, the SCF reports a much more rapid acceleration of wealth, concentrated in equity-type assets.  The estimates of wealth in the PSID and HRS are very similar to the SCF for the bottom 95 percent of the wealth distribution, diverging only for the top five percent of households."

One of the problems with longitudinal surveys like the PSID and HRS is that there is a natural attrition in the participants.  That might produce a bias in data amongst the survivors:
"We conclude that both surveys remain very representative of the underlying population as judged by a comparison with the lower 95 percent of households in the SCF."

The report also tried to discover how much of the increased wealth was attributable to new saving and how much to "passive" capital gains.  The SCF data didn't allow that analysis.  For the two longitudinal surveys, the outcome seemed unsatisfactory:
"The primary problems with the estimates of saving and wealth change from the surveys center around serious problems with measurement error.  While measurement errors in the cross-section estimates of wealth can be assumed to be random, they take on a much larger role in the estimate of the change in asset values.  The indirect estimation of active and passive saving as the difference between two independently obtained measures of wealth yields estimates of saving with very high noise-to-signal ratios.  The estimates of saving could be greatly improved by developing methods to reduce the variability of the responses across waves of the survey."

The report concluded on this that:
"The overall changes in wealth match the macroeconomic data closely, showing a secular rise in wealth-income ratios.  Although the measures of saving do demonstrate consistent differences in saving among major socio-economic groups, they do not reflect the general decline in saving rates that is apparent in the aggregate data for the past two decades."

Finally, the report looked at relationships between wealth and the age of respondents in the surveys: "The result is greater evidence of wealth decumulation at older ages."

PensionReforms supports the report's call for higher quality information.  It seems difficult to understand why, for example, the PSID collects "no significant information on the value of employer-provided pensions"; nor that the SCF has a new sample for each survey with no retrospective data collected for each new group.  As the report itself says "We conclude that reliable estimates of saving and wealth changes cannot be derived from a series of independent wealth surveys." 

Without quality data, it is not possible to make quality public policy decisions.

And it will be interesting to see what the recent economic crisis has done to the numbers.  (File size 406 KB; 66 pp)  299