Motivation

• The official Brazilian measure of unemployment, published by the Brazilian Institute of Geography and Statistics, is released with a lag.

The quarterly average of unemployment rate is disclosed around 1 month later.

• This isn’t good for policymakers, market participants, or the population. Therefore, there is a loss of economic efficiency as well as a waste of resources and efforts.

General Idea

• The main idea of this project is to use Facebook data to measure Brazilian labor market conditions prior to the official one.

• Once Facebook data is available in real time and at a very high frequency, an estimate of unemployment, for example, could be obtained as soon as needed.

• Furthermore, this data would allow us to get measures of unemployment at different levels: Brazil, Regions, States, Cities....

Data Usage Idea

• Posts that contain a combination of the chosen terms would be classified as "job search" or "job posting".  

• Groups of terms that could be used would be: one related to the "job search" (demand), for which phrases such as "find * job", "look * job" would be used, and another directly related to "job posting" (supply) which would consist of phrases such as "hiring *", "job opportunities", etc.

References

[1] Antenucci, D., Cafarella, M., Levenstein, M., Ré, C., & Shapiro, M. D. (2014). Using social media to measure labor market flows (No. w20010). National Bureau of Economic Research.

[2] Daas, P. J., & Puts, M. J. (2014). Social media sentiment and consumer confidence (No. 5). ECB Statistics Paper.