In the News

US – Using Twitter to forecast jobless claims

A surprising number of people filed initial claims for unemployment benefits last week, at least if you believe an index based off Twitter.

Economists at the University of Michigan have developed a technique that scans billions of tweets, looks for people tweeting about losing their jobs and then creates a prediction for the Labor Department’s weekly report on initial filings. Their prediction: 342,000 people filed new claims for jobless benefits last week.

Capture d’écran 2014-04-03 à 08.23.36

Their algorithm analyzes the vast fire hose of Twitter and can pick out tweets like this: “2011 was interesting. I ended an engagement, got laid off, started a small biz, and it looks like I’ll be moving this year too. Whew!” (That’s an example the economists used in their paper, among those identified as “job loss” tweets.) Add all those up, figure out how tweets translate to jobless filings, and you have a formula for making a prediction.

Twitter is somewhat more pessimistic than economists, who predict 320,000 people will file. About 311,000 filed the previous week.

Chosen excerpts by Job Market Monitor. Read the whole story at Using Twitter to Forecast New Applications for Unemployment – Real Time Economics – WSJ.

Related Posts


No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Jobs – Offres d’emploi – US & Canada (Eng. & Fr.)

The Most Popular Job Search Tools

Even More Objectives Statements to customize

Cover Letters – Tools, Tips and Free Cover Letter Templates for Microsoft Office

Follow Job Market Monitor on

Enter your email address to follow this blog and receive notifications of new posts by email.

Follow Job Market Monitor via Twitter



%d bloggers like this: