A New Index of Uncertainty Based on Internet Searches: A Friend or Foe of Other Indicators?

Abstract

The preliminary evidence in the literature suggests that changes in uncertainty have a role in shaping the U.S. economic cycle. But what is effectively measured by the different available indicators of uncertainty still remains an "uncertain" issue. This paper has two aims: (i) to introduce a new uncertainty indicator (GT) based on Internet searches; and (ii) to compare the main features and the macroeconomic effects of alternative measures of uncertainty, including our own. Results suggest that GT shocks embody timely information about people's perception of uncertainty and, in some cases, earlier than other indexes. Furthermore, the effect of uncertainty shocks on output is more influenced by parameter breaks due to insample events than by model specification. The consequence is that an all-comprehensive indicator able to weight different sources of uncertainty is preferable.

Google It Up! A Google Trends-based Uncertainty index for the United States and Australia

“…While writing this paper, we became aware of a contribution by Bontempi, Golinelli, and Squadrani (2016). They construct an index of economic uncertainty for the United States using Google Trends data, contrast it with alternative measures of uncertainty, and use their measure in a VAR context to analyze the contribution of uncertainty shocks for the dynamics of employment and industrial production.…”

Immigration, uncertainty and macroeconomic dynamics

“…Moreover, our findings have pointed out that the observed macroeconomic outcomes depend on the way in which immigration-related uncertainty is measured. As opposed to Donadelli (2015), Bontempi et al (2016) and Castelnuovo and Tran (2017) who show that news-based EPU shocks and Google searchbased macroeconomic policy uncertainty shocks produce similar impulse responses of main macroeconomic aggregates, our analysis suggests that the macroeconomic implications of a shock to the MPUI and MFI (i.e. news-based immigration-related sentiment shock) of Baker et al (2015) are different from those generated by a shock to our newly developed GTMU index (i.e., Google search-based immigration-related uncertainty shock).…”

A Language-Independent Measurement of Economic Policy Uncertainty in Eastern European Countries

“…5 With the general usage of Google Trends in academia, research on economic policy uncertainty started to use search volume as well. Existing studies, however, almost exclusively focus on the United States and apply Google search volume quite differently: While Dzielinski (2012) and Donadelli (2015) analyze search volume of only one single word, recent studies by Bontempi, Golinelli, and Squadrani (2016) and Castelnuovo and Tran (2017) pursue a more sophisticated approach. Dzielinski (2012) uses one search term (i.e., "economy") within the entire analysis and links search volume to stock market returns and volatility.…”

“…Notably, the authors select different search terms for the United States and Australia, respectively. Similarly, Bontempi, Golinelli, and Squadrani (2016) select 184 policy-relevant search terms that are closely related to the keywords in Baker, Bloom, and Davis (2016) US index for economic policy uncertainty. Their results show that such an uncertainty indicator tracks macroeconomic effects and provides timely information.…”

“…7 Given this outline on existing research on economic policy uncertainty, we identify the following issues missing in current research: First, existing studies mainly focus on developed countries (or mostly on the United States) and put emerging countries mostly aside. Second, the use of Google search volume to capture economic policy uncertainty is yet rather basic in literature (with the exception of Bontempi, Golinelli, and Squadrani 2016). And third, all existing indices based on search volume require country-specific language proficiency.…”