Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I develop an adaptive learning model to study the welfare effects of Social Security policy uncertainty in an aging economy. Agents combine full knowledge of the political process (which Social Security reforms are possible and when they could occur) with limited knowledge about the structure of the economy. The adaptive learning amplifies cyclical dynamics along the transition path to the new steady state. This magnifies the welfare effects of policy uncertainty, compared to a standard rational expectations model. The ex ante consumption equivalent variation that equates the expected utility of consumption (with policy uncertainty) to the utility of expected consumption (across the possible policies without uncertainty) ranges between −0.29% and 0.21% of lifetime consumption in the adaptive learning model compared to −0.012% to −0.018% in the standard model. The welfare cost to future generations is also larger in the adaptive learning model compared to the rational model.