The presence of liquidity constraints will alter agents' consumption-leisure choices. In this paper we test this theoretical prediction by relaxing the usual assumption that labor income is exogenously fixed and allowing labor supply to respond to the presence of borrowing constraints. Agents' consumption and leisure choices are characterised by two regimes - depending on whether the credit constraints are binding - and the determination of the prevailing regime is endogenous.
The effects of liquidity constraints are consequently analysed in the context of endogenous switching regressions models. This framework - in particular the endogeneity of regime determination together with the joint consideration of labor supply and consumption choices - permits us to disentangle the credit demand and supply factors that affect the probability of being liquidity constrained.
Our analysis is based on a data set (the 1987 Survey of Italian households) in which liquidity constrained households are readily observable without having to impose arbitrary conditions on the level of savings, wealth, etc., in order to identify which households face (do not face) liquidity constraints. Even though we have this detailed information, we consider the possibility that mistakes are made when classifying agents and incorporate this into our estimation procedure.
Our results support the idea that borrowing constraints have a significant effect on agents' labor supply decisions while the direct effect on consumption is found to be somewhat smaller than in previous research in the area; the total welfare loss caused by liquidity constraints is found to be approximately equally split between a reduction in consumption and a reduction in leisure time.
