Here's how academics are predicting job losses using Twitter

 
Peter Spence
Follow Peter

Estimating economic data accurately is notoriously hard to do, but a team of economists may be making that process easier, using Twitter.

They've developed a helpful new forecasting method, as described in their new paper, "Using social media to measure labor market flows".

The study does what it says on the tin, examining the usefulness of Twitter in predicting labour market trends.

Using data from the social network, the authors construct indexes of job loss, job search, and job posting in the US. Data is gathered by searching through a dataset of 19.3bn tweets for phrases like "lost my job".

Their index is updated monthly, and the authors say that it accounts for 15 to 20 per cent of the gap between consensus forecasts and official labour market estimates.

The study may seem a bit gimmicky - and we've become wary of economists trying to make sense of Twitter in a financially useful way - but it's also a neat way of providing live assessments of key economic events.

The authors highlight how Twitter can be used to give a real-time impression of shocks like Hurricane Sandy in 2012, and last October's partial shutdown of the US government, and how they can impact labour markets.

Such "timely and high-frequency data may be useful to policymakers and market participants", and at very low-cost compared to traditional data collection, which usually involves relatively expensive surveys.

Here's how the University of Michigan Social Media Job Loss Index predictions fare against the official initial claims for unemployment insurance numbers (they update with the latest data here):