The Shape of Temptation: Implications for the Economic Lives of the Poor
Abstract (Via MIT & Harvard)
The poor often behave as if they are very myopic. For example, consider the stylized fact that they borrow repeatedly (often continuously) at very high interest rates.1 Under standard consumption theory, borrowing at high rates reveals a preference for consumption today since one could always maintain the same investment level by taking a marginally smaller loan and consuming less. These borrowers must value a dollar of consumption today at least as much as R dollars of consumption tomorrow, where R is the gross rate of borrowing. If individuals are in steady state, the Euler inequality In other words, such a commonplace phenomenon continuous high interest rate borrowing directly implies stark myopia is particularly low).3 While certainly the easiest to document, high interest rate borrowing is only one of many behaviors that the standard model would interpret as extreme myopia. A second stark example is from the investment domain. A growing literature suggests that the poor fail to self- nance what appear to be high return divisible investments or to accumulate enough to buy non-divisible investments. This has been found for working capital amongst micro-entrepreneurs and for fertilizer amongst fertilizers (see de Mel, McKenzie, and Woodru¤ (2008) on micro-enterprise returns; Lee, Kremer and Robinson (2009) on inventories; Du- o, Kremer and Robinson (2009) on fertilizer; and Udry and Anagol (2006) for a variety of evidence).
Introduction (Via MIT & Harvard)
In this paper, we argue that a broader frameworknot just low is necessary to understand the variety of behaviors. We build on a growing body of research on time inconsistency (Shefrin and Thaler (1981), Laibson (1997), ODonoghue and Rabin (1999)): individuals are myopic about some decisions and far-sighted about others. The poor are no di¤erent in this regard. The same populations that borrow at very high rates also engage in various far-sighted behaviors from purchase of burial insurance to participation in ROSCAs (Collins et. al. (2009)). Our model focuses on the interaction between time inconsistency problems and poverty. Focusing on this interaction requires reformulating existing time incon- sistency models of consumption. Most models focus on the level of consumption: todays self is tempted to consume more than the long-run self would want. We focus on the composition of consumption: today’s self is tempted to consume more today of certain goods than the long-run self would want.