Abstract |
This paper explores whether Google Trends, Internet browsing statistics, can capture the behavior of macroeconomic variables in Peru. In particular, we analyze the employment index in companies of 100 or more employees in Lima (EI100). Using Google Trends data, we built an index that captures the behavior of the population seeking employment, which was dubbed Google unemployment index (GUI). The results show that the model that includes the GUI is better at forecasting the IE100 than the model that does not include the GUI. Furthermore, it was found that the GUI is useful for nowcasting and one-period forecasts. However, we found that the GUI is not useful for forecasting EI100 further ahead in the future. The relevance of these findings reside in the fact that Google Trends data is available on a weekly basis, around four months before data from official sources. Google Trends is a powerful tool for policy makers, particularly during times of crisis where following the economic activity in real time is crucial. |