Motivated forecasts : experimental evidence from the presidential elections in Argentina

The growing political polarization may influence a critical input for policymaking: people's economic expectations. This study examines whether political preferences shape individuals' forecasts for key economic indicators (using a preregistered online experiment in the context of Argentina's 2023 election). The experiment (N=1,162) exogenously manipulates (a) the incentives to report accurate forecasts and, (b) the information about current indicators. The results show that providing incentives for accuracy reduces the gap between subjects' forecasts regarding different candidates' performance. Providing information regarding the current economic indicators reduces the variance of the forecasts but not the gaps. These findings are relevant for survey design.