THE SAVINGS BEHAVIOR OF RICH AND POOR: A STUDY OF TIME PREFERENCE AND LIQUIDITY CONSTRAINTS

This dissertation studies the savings propensities of rich and poor. In the first Chapter, I present empirical evidence from panel data which suggests that the time preference rate of rich households exceeds that of poor households by 3 or 4 percentage points. These results are obtained by estimating the stochastic first order condition for intertemporal consumption. I also find a negative correlation of lagged income with consumption growth among the poor. This correlation violates the Euler equation and suggests that the poor may be liquidity constrained. Since liquidity constraints steepen the time profile of consumption, the time preference estimate for the poor may reflect only a lower bound.;The third Chapter presents an equilibrium explanation of liquidity constraints which could explain the rejection of the Euler equation for poor households. Since a borrower's current income provides a signal to banks about permanent income, it influences the terms on which he/she may borrow. I show that an income pooling equilibrium may emerge in which banks offer a single low interest rate loan to high current income borrowers and a higher priced contract to low current income borrowers. Within each income class, low permanent income borrowers are credit rationed and therefore exhibit high marginal propensities to consume.;In the final Chapter, I simulate the general equilibrium impact on savings of lump sum transfers from rich to poor using a multi-period life cycle model. Assuming that the poor have a substantially higher rate of time preference and/or are liquidity constrained, I find that long run savings is fairly insensitive to transfers from rich to poor.;The second Chapter examines the implications of income volatility and bankruptcy for the propensity to consume in a two class, two state life cycle model. I show that the marginal propensity to consume increases with the probability of bankruptcy. In this model, the high consumption propensities observed among the poor reflect their rational response to a higher probability of low income and default.