Do Consumers React to Anticipated Income Changes? Evidence from the Alaska Permanent Fund

Abstract

A central implication of the life-cycle/permanentincome hypothesis (LC/PIH) is that consumers should not respond to predictable changes in their income. To test this hypothesis, a number of recent papers have exploited natural experiments to identify anticipated income changes. In particular, recent work by Parker (1999) uses the change in after-tax income due to the cap on earnings subject to the Social Security tax, and a related paper by Souleles (1999) examines the response of consumption to income tax refunds. Surprisingly, Parker and Souleles find that even when income is expected to change within the year, expenditure is excessively sensitive to the timing of the income change. While their results can be interpreted as evidence that our canonical model of consumption is inadequate, an alternative explanation is that the anticipated income changes they exploit are small and irregular, and that households will not bother to change their consumption paths when the computational costs involved are large relative to the utility gains. In support of this interpretation, Browning and Collado (2001) find that the seasonal consumption patterns of Spanish households that work in sectors that provide regular bonus payments do not differ from those of households that do not receive bonus payments. This paper adds to this evidence by exploiting a natural experiment provided by annual payments from the state of Alaska’s Permanent Fund to every resident in the state of Alaska that should yield an unusually powerful test of the LC/PIH. These payments are large and clearly anticipated by Alaskan residents. Using the variation in the size of the payments over time and in the amount received by families of different sizes to identify the response of consumption to payments from the Permanent Fund, I find no evidence that the consumption of Alaskan households reacts to these payments. In addition, I find no evidence that the seasonal pattern of consumption in Alaska differs from that in the other 49 states or that households in Alaska are subject to fewer liquidity constraints, engage in less buffer-stock saving, or spend a smaller fraction of their income on semidurable goods than households in the rest of the United States. However, although households in Alaska do not overreact to payments from the Permanent Fund, I find that the consumption of the very same households is excessively sensitive to their income tax refunds. This evidence suggest that households will take anticipated income changes into account in their consumption decisions when the income changes are large, regular, and easy to predict, but will not do so when they are small and irregular. The paper proceeds as follows. The next section of the paper presents details on the * Department of Economics, Princeton University, Princeton, NJ 08544, and National Bureau of Economic Research (e-mail: chsieh@princeton.edu). I am very grateful to the Bureau of Labor Statistics (BLS) for providing access to the data from the Consumer Expenditure Survey and to its staff, particularly Steven Henderson, Walter VandeHeide, and Wolf Weber, for their generous assistance during my visits to the BLS. I also thank Nanci Jones for providing data and patiently answering questions on the operation of the Alaska Permanent Fund. Anne Case, Angus Deaton, Jonathan Parker, and two referees provided useful comments. The views expressed in this paper are personal and should not be attributed to the BLS. 1 I refer to the certainty-equivalent version of the LC/ PIH, or one in which the expected variance of consumption is constant. As is well known, without these assumptions, the LC/PIH only implies smoothing of marginal utility, not necessarily of consumption. 2 See Christina H. Paxson (1992); John Shea (1995); Jonathan A. Parker (1999); Nicholas S. Souleles (1999); Martin Browning and M. Dolores Collado (2001). Also see Ronald A. Bodkin (1959) for an early example of the use of a natural experiment to test for excess sensitivity. For comprehensive reviews of the large literature on empirical tests of the LC/PIH, see Angus Deaton (1992), Browning and Annamaria Lusardi (1996), and Browning and Thomas F. Crossley (2001). 3 In 1998, for example, the Permanent Fund paid $1,541 to every resident of the state of Alaska.