Google Searches Could Predict Future Traffic Jams, Study Finds
If you stop to quickly Google traffic conditions before you rush out the door, you’re not alone. In fact, searches for “traffic jams” may actually be able to predict congestion before it happens, according to new research in PLOS One.
The study compared German Google searches for stau, or traffic jam, to reports of traffic jams reported by ADAC, a major German auto club. Author Nikolaos Askitas of the Institute for the Study of Labor in Bonn found that Google searches for road conditions peaked two hours before reports of traffic jams, both during morning rush hours (between 7 a.m. and 9 a.m.) and the end-of-day rush between 4 p.m. and 6 p.m. He used a program that scraped traffic data from ADAC every five minutes between September 28, 2015 and November 14, 2015. He linked these to weekly Google searches provided by Google Trends.
After controlling for variations in time and day of the week, Askitas argues that 80 percent of variation in traffic conditions can be linked to preceding Google search spikes.
“A one percent increase in Google [searches for traffic jams] implies a .4 percent increase of traffic jams,” he writes in the study.
The data isn’t comprehensive, but it represents a proof-of-concept for using search engines to examine wide-spread societal behavior. More research might be able to create further predictions of future behavior, and, in the case of traffic, lead to preventative measures.
Know of something you think we should cover? Email us at firstname.lastname@example.org.