Discrete breaks in the US savings: evidence and permanent income implications

Abstract

The main purpose of this paper is to investigate whether permanent, discrete breaks in U.S. savings have occurred. Breaks are hypothesized to occur in both a constant (intercept) and a time-trend term. To test for these breaks, I draw upon the recent methods applied to gross national product. Evidence in this paper suggests that permanent breaks in saving occurred during both the early 1970s the middle 1980s. As a related issue, the stationarity of savings is also discussed. Evidence presented within favors stationarity when breaks are included but not when breaks are omitted. implications of breaks and stationarity for the expected long-run value of assets are examined. An additional related issue that is also addressed in this paper regards test of the PIH. The PIH suggests that savings act as a buffer-stock to smooth consumption. One framework which captures this aspect of the PIH is due to Campbell (1987) and Campbell and Deaton (1989). In this paper, modify their framework to include discrete breaks in savings. Evidence presented within suggest that, when breaks are added to their framework, the "rainy day" finding is strengthened