Nowcasting Unemployment Rate in Turkey: Let's Ask Google

Abstract

We use linear regression models and Bayesian Model Averaging procedure to investigate whether Google search query data can improve the nowcast performance of the monthly nonagricultural unemployment rate for Turkey for the period from January 2005 to January 2012. We show that Google search query data is successful at nowcasting1 monthly nonagricultural unemployment rate for Turkey both in-sample and out-of-sample. When compared with a benchmark model, where we use only the lag values of the monthly unemployment rate, the best model contains Google search query data and it is 47.8 percent more accurate in-sample and 38.3 percent more accurate for the one month ahead nowcasts in terms of relative root mean square errors (RMSE). We also show via Harvey, Leybourne, and Newbold (1997) modification of the Diebold-Mariano test that models with Google search query data indeed perform statistically better than the benchmark.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 31 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.