THE 2007–2008 U.S. RECESSION: WHAT DID THE REAL‐TIME GOOGLE TRENDS DATA TELL THE UNITED STATES?

Abstract

In the extant literature of business cycle predictions, the signals for business cycle turning points are generally issued with a lag of at least 5 months. In this paper, we make use of a novel and timely indicator—the Google search volume data—to help to improve the timeliness of business cycle turning point identification. We identify multiple query terms to capture the real‐time public concern on the aggregate economy, the credit market, and the labor market condition. We incorporate the query indices in a Markov‐switching framework and successfully “nowcast” the peak date within a month that the turning occurred. (JEL E37, G17)