Is Google Trends Useful in  Nowcasting Unemployment Rate During the Pandemic at Regional and National Level in Romania?

Mihaela Simionescu, PhD Professor (Full)  (,
Institute for Economic Forecasting, Romanian Academy
Given the role of Covid-19 pandemic in accelerating the digital transformation in Europe, the main aim of this paper is to explain the unemployment rate in Romania based on Internet searches for jobs during the epidemic at national and county level. Google Trends indexes for certain keywords related to jobs in Romanian language (’’locuri de muncă’’ and ’’joburi’’) and the most famous websites with job announcements (’’eJobs’’ and ’’Hipo’’) are considered.  At national level, the unemployment rate in the period February 2020- December 2022 in Romania is explained using an autoregressive distributed lag model (ARDL) based on Google Trends index for ”locuri de munca”,  while for youth unemployment ”joburi” and ”eJobs” are relevant. At county level, a spatial error model based on searches for ” locuri de munca” performs better than OLS regression. The results support the recommendations to improve governmental making-decision process.
Key-words: unemployment; Google Trends; OLS; spatial model; ARDL model