For decades, policymakers have debated whether unemployment insurance provides a critical safety net in tough times or whether it prolongs unemployment by reducing people’s incentives to find new jobs.
But to answer this question, researchers need a clear measure of how much effort people actually put into looking for jobs. And this activity has proven difficult to monitor. For this Scott R. Bakerassociate professor of economics at the Kellogg School, turned to Google—specifically its search traffic data.
Baker and his associate, Andrei Fradkin of Boston University, used data from Google Trends, the free site that analyzes the popularity of various search queries across regions and languages. By tracking queries containing the word “jobs,” they were able to create a new system to measure this activity, which they called the Google Job Search Index.
“We realized we could look at people’s job search habits in a way that traditional government data sets had a hard time doing,” says Baker.
Job search data analysis
Pulling job search data from Google Trends isn’t as simple as Google makes it out to be.
Since Google keeps the exact volume of searches for any given term confidential, Google Trends results represent only the relative frequency of Google searches, on a scale of 1 to 100.
In order to estimate the raw number of work-related searches and ensure they were large enough to use as the basis of their study, Baker and Fradkin used other Google tools. For example, Baker says, Google’s online advertising tool, Adwords, showed that there were 68 million monthly searches for the term “jobs” in the year before April 2013, when they started this project.
Having validated the weight of the data set, the researchers began to analyze it. In doing so, they knew that the GJSI offered several advantages over previous methods for tracking the job search.
First, the most commonly used survey to track time spent job hunting is the American Time Use Survey, a report published once a year by the US Department of Labor. Although meticulously detailed, the survey — which interviews about 26,400 households by telephone throughout the year — does not focus on people who are not working, so it often includes fewer than five unemployed respondents from each state per month. Google, meanwhile, provides real-time access to many millions of searches that can be aggregated across geographies. This big data push allows policymakers to measure changes in local labor market conditions in a way that was previously not possible.
Second, the search engine’s ability to track searches for more specific queries—eg. “marketing jobs,” “New York jobs,” or “Walmart jobs”—means researchers could use it to study how people search across sectors, regions and employers.
And third, while other online platforms like CareerBuilder offer proprietary data that can be useful, that information remains much harder to access and smaller in volume than an open data source like Google Trends.
Evaluating the Google Job Search Index
But, clearly, the attractive qualities of the GJSI would not matter if the researchers could not prove that the index was a true reflection of the volume of people looking for work.
“We have this measure of how often people search for something that contains the word ‘jobs,’ but does that really say anything about how people are searching for jobs in the economy?” says Baker.
To find out, the researchers compared its results with those of several other data sources. This included comScore, which tracks the web browsing habits of 100,000 consenting Americans. They found GJSI to be a good proxy for overall online job search effort.
In addition, comparing their results with those of the American Time Use Survey and other data, they found that Google job searches fluctuated in the same way as those reported in the survey, and that a higher unemployment rate corresponded to a higher GJSI.
“This showed us that the GJSI is telling us something important about the time people spend looking for jobs,” says Baker.
With its importance confirmed, the GJSI could now be used to address a fundamental economic issue—one that was hotly debated in the wake of the Great Depression.
Do unemployment benefits present moral hazard?
Economists have argued the costs and benefits of unemployment insurance since its inception during the Great Depression as part of the Social Security Act of 1935.
“There’s always a question in economic policy that comes down to the debate between two competing forces,” Baker says. “One is the liquidity phenomenon, which serves to improve the situation of the unemployed by redistributing money to people when they need it. The other is moral hazard. It’s the idea that the more you support these programs, the more people will take advantage of them and not find new work.”
After the Great Recession, most states saw their unemployment benefits increase from a maximum of 26 weeks to a maximum of 99 weeks. Did these more generous benefits deter people from working?
Through the Texas Workforce Commission, researchers obtained detailed weekly records of the number of Texans out of work in each metropolitan area between 2005 and 2014, as well as the number of weeks of benefits people had left. They then compared the weekly GJSI results for each of these metropolitan areas with the percentage of unemployed residents at each benefit level. Their results were confirmed earlier research showing that people look for jobs more when they have fewer weeks of benefits left.
In fact, they found that in areas where the average unemployed person had less than 10 weeks of unemployment insurance, the index revealed 66 percent more search activity than in nearby areas where people had an average of 10 to 20 weeks, and a whopping 108 percent more than in metro areas where the unemployed had an average of 30 weeks left.
“This makes sense: If you’re getting close to the end of your unemployment benefits, you’ll feel more pressure to look for work,” says Baker.
The duo also compiled data from all 50 states, drawing on the Current Population Survey, produced each month by the Bureau of Labor Statistics, to track how long residents were on unemployment benefits between 2008 and 2014. Then, using the same method they did in Texas, researchers found that each 10-week extension reduced overall search activity — by both employed and unemployed — by only 1.5 to 3 percent. (After all, people with jobs sometimes look for new ones.)
Translated, that means “there’s a moral hazard story to be told here, but we’re also showing that the overall effect appears to be relatively small,” Baker explains. “If you’re looking to compare how much the overall increase in unemployment is [during the Great Recession] could be due to the increase in unemployment benefits, the answer is very little.”
In addition to shedding light on this financial question, Baker says the value of the GJSI extends far beyond job search tracking. In fact, since the research was published, it has been cited by other academics “who don’t look at job search data at all,” he says.
Instead, they’re trying to use the same toolkit to develop other Google-based economic indicators, including new ways to predict consumer sentiment and economic uncertainty tracking search terms like “recession” and “bankruptcy.”
“This approach can be really useful for policymakers because it can be much more local and granular,” notes Baker, “and they can access it without a six-month or a year delay.”