The Psychology of Job Loss: Using Social Media Data to Characterize and Predict Unemployment

10 Pages Posted: 25 May 2016 Last revised: 2 Jul 2016

See all articles by Davide Proserpio

Davide Proserpio

Marshall School of Business - University of Southern California

Scott Counts

Microsoft Corporation

Apurv Jain

MacroXStudio; Harvard Business School; Microsoft Corporation - Microsoft Research - Redmond

Date Written: May 23, 2016

Abstract

Using data from social media, we study the relationship between the macroeconomic shock of employment instability and psychological well-being. We analyze more than 1.2B Twitter posts from over 230,000 U.S. users who either lost a job or gained a new job over a period spanning five years, from 2010 to 2015. First we quantify the magnitude and length of effects of job loss/gain on psychological variables such as anxiety, sadness, and anger. We then define a behavioral macroeconomic model that leverages these changes in psychological state to predict levels of unemployment in the U.S. Our results show that our psychological well-being measures are leading indicators, predicting economic indices weeks in advance with higher accuracy than traditional forecast techniques. Taken together, these findings suggest that by capturing the human experience of a shock like job loss, social media data can augment current economic models to generate a better understanding of the overall causes and consequences of macroeconomic performance.

Keywords: Big Data, Social Media, Unemployment, Micro Foundation, Psychology, Job Market

Suggested Citation

Proserpio, Davide and Counts, Scott and Jain, Apurv, The Psychology of Job Loss: Using Social Media Data to Characterize and Predict Unemployment (May 23, 2016). Available at SSRN: https://ssrn.com/abstract=2783520 or http://dx.doi.org/10.2139/ssrn.2783520

Davide Proserpio

Marshall School of Business - University of Southern California ( email )

701 Exposition Blvd
Los Angeles, CA 90089
United States

HOME PAGE: http://dadepro.github.io/

Scott Counts

Microsoft Corporation ( email )

One Microsoft Way
Redmond, WA 98052
United States

Apurv Jain (Contact Author)

MacroXStudio

981 Mission st.
San Francisco, CA 94103
United States

HOME PAGE: http://www.macroxstudio.com

Harvard Business School ( email )

Soldiers Field Road
Morgan 270C
Boston, MA 02163
United States

Microsoft Corporation - Microsoft Research - Redmond ( email )

Building 99
Redmond, WA
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
288
Abstract Views
1,920
Rank
192,078
PlumX Metrics