Does information about current inflation affect expectations and decisions? Another look at Italian firms

I document the response of the inflation expectations, and pricing and labour demand decisions of Italian firms to randomly provided information about recent inflation and assess the causal effect of the former on firms’ decisions. I use a standard menu cost model to show that conventional IV2SLS estimates based on variation of agents’ inflation expectations generated by experimental manipulation of their information sets are likely devoid of casual content because in such experimental settings some assumptions required for their causal interpretation fail. I discuss alternative estimators based on assumptions more likely to be consistent with the underlying theoretical framework. Empirically, I find that randomly informed firms substantially revise their inflation expectations but do not revise pricing and hiring decisions. Causal inference from appropriate estimators consistently reveals that the lack of reduced form effects reflects absence of statistically significant effects of expected inflation on firms’ decisions rather than offsetting responses. These results cast doubts on the possibility of obtaining substantial real effects through communication strategies that reach the general public more effectively.