Economists Adopt Social Media Monitoring to Collect More Timely Data

 

The Bank of England is a recent example of organizations monitoring social media to collect information to collect market intelligence. The central bank monitors social media and other online sources to predict Britain’s economic trends.

Official economic figures are typically lagging indicators. By the time data are compiled and reported, the economic activity they describe has already faded into not quite ancient history.

Information gleaned from the Internet can be timelier than traditional sources. Statistics on online job searches can provide insight into employment trends; prices posted online can spot rising inflation before more official sources; consumer comments can provide insights into consumer sentiment and perceptions.

That type of evidence could enable the bank to respond to economic slowdowns or smoldering inflation more quickly, thus reducing economic swings. More timely data might even help the bank better cope with financial emergencies like the 2008 banking crisis.

Information from online sources has already been more reliable, Bank of England Chief Economist Andy Haldane, who’s leading a task force that oversees the media monitoring effort, according to Sky News.

“We have a new advanced analytics team who are constructing little models, algorithms and methods for extracting this data. We have a data lab. This is quite a big strategic change for the bank,” he said. “This is going to be quite a big shift from the past.”

The Twitter indicator

Researchers at Stanford University and the University of Michigan agree that social media monitoring offers a promising new approach to measuring and analyzing economic activity.

In the project, “University of Michigan Economic Indicators From Social Media,” researchers scour Twitter for terms related to job changes, such as “lost my job,” states the research paper, Using Social Media to Measure Labor Market Flows. They then compile and analyze the terms to create a Social Media Index that predicts initial claims for unemployment insurance.

“Based on an admittedly short sample period, a simple statistical analysis suggests that the social media index is a better indicator of the true state of job loss than is initial claims. There is no other official statistic that is available at as high frequency and in close to real time for assessing job flows, researchers explained in their paper.

Although the University of Michigan index typically coincides with unemployment claims, it sometimes predicts unemployment changes before official figures because it provides “real time” information.

During the government shutdown in October 2013, their social media index predicted job losses before official unemployment reports.

“This episode illustrates the usefulness of social media for measuring and analyzing the impact of unexpected events,” the researchers state. “Our social media indexes provide high-frequency and contemporaneous information that is not available in conventional sources.”

The early results from economists’ use of social media monitoring suggests that corporate planning and business development units in corporations may also benefit from monitoring of social media.  Aggregated social media monitoring data could conceivably reveal “intent to purchase” and timeframe of specific big-ticket purchases such as homes or automobiles.  Similarly, economic analysis of social media might well identify the upgrade cycle for electronics and other products.

The application of social media monitoring for economic analysis is now in its infancy.  In the not too distant future, it’s likely to enable economists in government agencies, universities and corporations to better predict consumer activity and economic growth.

Proven tools to aggregate the online data already exist. Measurement tools likely will evolve quickly.

What do you think about the aggregation of social media data for economic analysis?

Please post your comments below.