Risk aversion, wealth, and the DARA hypothesis: A new test

The Pratt-Arrow hypothesis of decreasing absolute risk aversion (DARA) is widely invoked in economic models of uncertainty but empirical tests, especially those using nonexperimental data, have yielded mixed results. This paper reexamines the DARA hypothesis using an expected utility model of life insurance demand and a 23-year sample of aggregate time series data from the U.S. After controlling for household size, age, income, loss probabilities, premium expenses, and inflation, the effect of wealth on insurance demand is found to be positive and statistically significant in linear and loglinear regressions, explaining more than 93 percent of the variation in coverage. Thus, in contrast to the prevailing theory, this empirical test presents evidence of increasing absolute risk aversion. This research was completed during a sabbatical at the University of Pennsylvania's Wharton School of Business. The financial support of the S. S. Huebner Foundation for Insurance Education is greatly appreciated. The author is also indebted to participants at the Forty-Second International Atlantic Economic Conference in Washington, D.C. for helpful comments. Any errors are the responsibility of the author.