| This dissertation investigates the consumption-output anomaly, the phenomenon that the observed correlations of the growth rates of output are consistently higher than the correlations of the growth rates of consumption among countries. Under the assumption of perfect trading and risk-sharing, two identical business cycle economies should have the opposite implication. Using different trading and market structures (including the extreme assumption of autarky), the dissertation identifies the underlying reasons why these theoretical models predict that the correlations of consumption growth should be higher than the correlations of output growth. Because the models consistently yield the same fundamental implications regarding the co-movements of consumption and output growth, this fact points to the direction of re-interpreting the data as the source of a potential solution to the anomaly. The paper re-evaluates the growth rates of the relevant variables by showing systematic errors in the constant price method used to price commodities over time. When an alternative and superior method is used to compare cross-country correlations of consumption and output growth, the results are more in line with economic theory. |
