Using Managers' Expectations for Ex-Ante Policy Evaluation: Evidence from the COVID-19 Crisis

Abstract

Evaluation of the impacts of government policies during an economic crisis is often delayed until the outcomes are realized. Policies can be better guided if they can be evaluated amid a crisis, before the realization of outcomes. This study examines whether survey data on the expectations of small business managers can be used to evaluate two high-stake subsidies for firms amid the COVID-19 crisis in Japan, namely, Subsidy Program for Sustaining Businesses (SPSB) and Employment Adjustment Subsidy (EAS). Thus, we evaluate the accuracy of managers' expectations, estimate the impact of subsidies on the expected firm survival, and compare it with the estimated impact on realized survival. We find that the managers' expectations on their future sales, survival rate, and the possibility of receiving these subsidies predict the realized outcomes, although they were highly pessimistic about their survival rates. We find that managers adjust their expectations by learning from past forecasting errors throughout the crisis. We find that the estimated impacts of the SPSB on the expected survival rates have the same sign as the estimated impact on the realized survival rates, but the size is more than twice. The estimated impacts of the EAS are both insignificant. Therefore, although its impact may be overestimated, managers' expectations are useful for selecting an effective policy.

Keywords: Firm forecast, Policy evaluation, Small business, COVID-19, Subsidies

JEL Classification: D80, E66, L50, H20