Abstract
The lack of perfect rationality among economic agents is an issue economists have been struggling with since the 1950s. While much progress has been made in understanding behaviors that deviate from perfect rationality, there remains a lack of empirical studies explicitly testing for bounded rationality in the uncontrolled marketplace. This thesis uses data on gasoline prices, quantity consumed, and income to create a baseline regression controlling for state and month fixed effects. In addition to this baseline regression, six separate heuristics are chosen to represent a wide variety of predictive power for future gasoline prices (prices in the next month). Combining these two data sources, each heuristic is added independently to the baseline regression and the distribution of the residual is measured using a Jarque-Bera test statistic. Finally, each of the test statistics is compared to the baseline regression to test for the amount of relevant information added to the overall regression. The results of this test show that bounded rationality with respect to future gasoline prices is not a factor in gasoline purchase decisions.