ETLA Predicts the Unemployment Rate with Google Searches in Europe

ETLA, The Research Institute of the Finnish Economy, together with 28 leading European research institutions, has constructed a tool for real-time economic forecasting called ETLAnow. It predicts automatically the unemployment rate for the EU-28 countries using Google search data. The model provides daily updates on the current and future unemployment rate.

ETLAnow is the first publicly available economic forecast model that uses web search data. It is available at www.etlanow.eu. It publishes the daily forecasts also on Twitter @ETLAnow. More generally, the aim is to understand whether big data could improve economic forecasts.

The ETLAnow model is based on the finding that Internet search trends on topics that are related to unem-ployment—such as searches for jobs—are correlated with the present and future unemployment rate. ETLA’s research tells that including Google search terms improves short-term unemployment forecasts. In particular, Google search trends improve the prediction accuracy around turning points. Those are often hard to predict. The project and the findings are described in more detail at the ETLAnow website and in a recent ETLA report.

ETLA has collaborated with 28 major European economic research institutions in order to develop the tool. ETLAnow participates in Eurostat’s 2016 Big Data forecast competition.

For Additional information:

Researcher Joonas Tuhkuri (jtu@etla.fi), ETLA
Research Director Petri Rouvinen (pro@etla.fi), ETLA

Download the press release as a pdf file.