Comparing forecasts from fixed and variable coefficient models: The case of money demand

In this paper we introduce a class of tentatively plausible, fixed-coefficient models of money demand and evaluate their forecast performance. When these models are reestimated allowing all coefficients to vary over time, the forecasting performance improves dramatically. Aside from offering insights about improved methods of analyzing time series data, the most promising direct use for point estimates derived from time-varying coefficients is as an aid in calibrating proposed models of the kind discussed here.     

HOUSEHOLD SAVING AND PORTFOLIO CHANGE: EVIDENCE FROM THE 1983–89 SCF PANEL

There are few sources of high-quality data on the dynamics of wealth accumulation. This paper uses newly available data from the 1983 89 panel of the Survey of Consumer Finances to examine household saving and portfolio change over the 1980s. Our main findings are as follows. First, median household wealth rose modestly over the period. Second, while overall wealth inequality increased, households in the top 1 percent of the wealth distribution in 1983 saw their share of total wealth decline, probably reflecting turnover among the very wealthy. Third, although age, income, and initial wealth had significant effects in regressions on household saving, a large part of the variation in saving was unexplained. Finally, there were clear life-cycle patterns in the portfolios of assets and liabilities held by households, with younger households acquiring homes, businesses and all types of debts, and older households divesting themselves of these assets and debts.     

CONSISTENT WEIGHT DESIGN FOR THE 1989, 1992 AND 1995 SCFs, AND THE DISTRIBUTION OF WEALTH

Estimates using survey data are determined by two factors: the data collected and the survey weights. This paper discusses the design and calculation of a set of consistent weights for the Surveys of Consumer Finances. Taking both these weights and the multiply-imputed survey data, we look at estimates of changes in the distribution of wealth over the first half of the 1990s.     

HOUSEHOLD SAVING IN THE U.S.

In this paper the authors present evidence on household saving in the U.S. based on the panel data from the 1983 and 1986 waves of the Survey of Consumer Finances. Saving is measured in these surveys as the change in wealth over the three-year period. Using a variety of models, we are able to explain only about 7 percent of the variation in the level of saving. Demographic factors appear to be modestly useful in explaining saving. However, one fact is very clear from the patterns of correlation extracted so far: either the measurement error in the data is quite large, or idiosyncratic factors are very important in explaining saving behavior, or both.     

Some results on analytic optimal control solution of a simple rational expectations model

In this note, we correct a technical error in a well-known stabilization analysis in the context of a simple macroeconomic model, formulated as an optimal control problem. This error originated in a slip in an unpublished but widely circulated paper by D.A. Livesey, suggesting a procedure for obtaining a solution to a class of optimal control problems frequently encountered in a stabilization literature. We then propose an ad hoc solution method which makes clear the intuition of the solution to the particular model to which the Livesey method was applied, and of the larger class of models to which it belongs. Our solution is so formulated that the contrast between it and the Livesey result is made transparent.     

MEASURING WEALTH WITH SURVEY DATA: AN EVALUATION OF THE 1983 SURVEY OF CONSUMER FINANCES

Because wealth estimates from survey data have usually fallen substantially short of independent aggregate estimates, survey data have not been seen as adequate for assessing questions dependent on a good representation of the entire distribution of wealth, such as estimates of wealth concentration. The 1983 Survey of Consumer Finances (SCF), which contains a supplementary sample of very high income households drawn from a tax-file sample frame, is the first U.S. survey since the 1963 Survey of Financial Characteristics of Consumers that offers hope of accurately measuring the entire wealth distribution. In this paper, we discuss the design of the survey, the critical issue of proper weighting to merge the supplementary sample with an area probability sample, and the role of imputation. We show that the use of ordinary area probability samples alone leads to probable bias in the measurement of highly concentrated assets such as stocks and bonds. We compare the SCF data with aggregates derived from the flow-of-funds accounts of the Federal Reserve Board. While methodological issues cloud exact comparisons, it appears overall that the SCF estimates are at least as credible as other aggregate measurements. Finally, we use the data to assess the change in concentration of wealth from 1963 to 1983. We estimate that the concentration of wealth in terms of households did not change significantly over this period.