Explaining consumption: A simple test of alternative hypotheses

A method of testing the relative importance for consumption of risk-sharing behavior and changes in current income is proposed and estimated using data across Canadian provinces. The focus of the estimation is less on whether or not the risk-sharing model can be rejected than on how much each of these hypotheses can contribute to explaining overall variation in consumption. Both types of behavior are found to be statistically significant but the risk-sharing model is found to explain considerably more of the growth in consumption than does changes in income.