Can consumer confidence data predict real variables? Evidence from Croatia

This paper uses monthly data to examine the links between consumer confidence and real economic variables in Croatia, and it tests whether movements in the former contain predictive power for the latter. The results suggest that changes in consumer confidence help to explain retail turnover and imports, while expectations about forthcoming major purchases have particularly strong predictive power for retail turnover. We also find that the inclusion of confidence on the right-hand side improves the fit of simple models of retail turnover, a variable that is highly correlated with quarterly GDP. The results therefore highlight the usefulness of these survey data in helping to explain and forecast the real economy.