Interpreting Sensitivity

Sammanfattning: The standard interpretation of aggregate excess sensitivity is that it represents the economy share of non-Ricardian households. However, household level evidence suggests that excess sensitivity can also be explained by the comovement of consumption and income over the lifetime, which occurs due to retirement and changes in household size. If aggregate estimates of excess sensitivity can in part be attributed to the accumulated effect of comovements at the household level, it would significantly alter the policy implications of these estimates. The reason is that unlike under the non-Ricardian interpretation, the comovement interpretation of excess sensitivity does not imply that households would raise consumption due to temporary increases in income. This study is a first attempt to investigate whether aggregate excess sensitivity is in part determined by changes in the national age distribution. I estimate an aggregate intertemporal consumption function and investigate the degree to which excess sensitivity varies with growth in various age cohorts by means of interactions. My results indicate that there is cross-country heterogeneity in what drives aggregate excess sensitivity, and that retirement and shrinking household sizes are significant factors in some countries.