Cutting Off Emergency Unemployment Benefits Hasn’t Pushed People Back to Work

Helene Laurusevage still gets up at 6 a.m. every day and packs lunch for her husband. She still sits down at her computer every morning to hunt for a job, and still updates the meticulous spreadsheet that she uses to document that search. Just one thing has changed: In January, the biweekly checks for $1,126 from the Pennsylvania Department of Labor and Industry stopped showing up in her bank account. Laurusevage, 52, is one of more than a million Americans who lost payments when Congress allowed the Emergency Unemployment Compensation program to expire at the end of last year. The program, which Congress created in 2008, extended jobless benefits beyond the standard 26 weeks provided by most states; at its peak, the federal government provided an unprecedented 6 million workers with up to 73 weeks of benefits.1 The Senate earlier this year voted to renew the program, but House Speaker John Boehner hasn’t allowed the measure to come to a vote in the House. The case against extending unemployment benefits essentially boils down to two arguments. First, the economy has improved, so the unemployed should no longer need extra time to find a new job. Second, extended benefits could lead job seekers either to not search as hard or to become choosier about the kind of job they will accept, ultimately delaying their return to the workforce.2 But the evidence doesn’t support either of those arguments. The economy has indeed improved, but not for the long-term unemployed, whose odds of finding a job are barely higher today than when the recession ended nearly five years ago. And the end of extended benefits hasn’t spurred the unemployed back to work; if anything, it has pushed them out of the labor force altogether. Of the roughly 1.3 million Americans whose benefits disappeared with the end of the program, only about a quarter had found jobs as of March, about the same success rate as when the program was still in effect; roughly another quarter had given up searching. The rest, like Laurusevage, were still looking.3 Laurusevage didn’t expect it to be this hard. She had been her family’s primary breadwinner, earning roughly$60,000 as a health and safety officer for a Philadelphia-area heating and air conditioning company. Her husband, David, earns less than $35,000 a year selling truck parts. When her position was outsourced in April of last year, she thought that as a college graduate with a three-plus-decade history of steady work, she would find a job relatively quickly. But in many ways, her experience is typical. The long-term unemployed — typically defined as those out of work more than six months — are slightly more educated on average than the broader population of job seekers. And older workers like Laurusevage face a particularly tough time: The typical job seeker in her 50s has been out of work 26 weeks, versus 17 weeks for the typical 20-something. Laurusevage also lost her job at a time when the economy, though improving, was still far from fully healed. The unemployment rate in her home state of Pennsylvania was 7.6 percent in April 2013, down from 8.7 percent in the worst of the economic downturn, but still three percentage points above its pre-recession level. That made her much less likely to find a job quickly: The unemployment rate at the time someone loses a job, more than demographics or any other factor, is by far the strongest predictor of how long it will take that person to find a new job.4 Once Laurusevage’s search crossed the six-month mark, her chances of finding a job dropped sharply. Only about 10 percent of the long-term unemployed find jobs each month, a metric known as the job-finding rate. Among those unemployed six months or less, the finding rate is nearly 25 percent. Moreover, the short-term unemployed have seen their chances improve steadily during the recovery, while the finding rate for the long-term unemployed has been nearly flat. In other words, the long-term unemployed have hardly experienced a recovery at all. Some economists have argued that the end of the emergency unemployment benefits program could be a harsh but necessary medicine, prodding the unemployed back to work. But Laurusevage, at least, needed no prodding. The carefully formatted spreadsheet she uses to track her job search is a record of frustration: 246 applications, 24 interviews, no offers. In most entries, the “Outcome” field is blank; many companies these days don’t even bother to send a formal rejection notice. “I can’t even find that crap job. I’ve applied to Wawa and Wawa won’t take me, or I haven’t heard back from them,” she said, referring to the regional convenience store chain. “I will take whatever you have. Temporary, part-time, seasonal, anything.” The government doesn’t directly track what happens to people when their benefits run out. But the Current Population Survey, the monthly survey used to calculate the unemployment rate and related data, does ask people what state they live in and how long they’ve been unemployed. That means we can identify people who are likely to have lost benefits when the emergency program ended and compare their success at finding jobs to people who likely received benefits when the program was still in effect.5 This approach isn’t perfect — for one thing, we can only tell whether people likely qualified for benefits, not whether they actually received them — but past research suggests this methodology has a strong correlation with actual receipt of benefits.6 This analysis suggests the end of benefits has had at most a small impact on the rate at which workers find jobs. In 2013, people who likely qualified for emergency benefits had a monthly job-finding rate of 12 percent. In the four months since the program ended, the job-finding rate for likely cutoff victims — people who would likely have qualified for extended benefits if the program had been renewed — was slightly higher, at 14 percent. But the difference isn’t statistically significant. Even if it holds up as more data comes in, it could be the result of the improving economy rather than the direct impact of the end of the emergency program. The end of the program might have had another effect, however: It may have made people less likely to keep trying to find jobs. Unemployment programs usually require recipients to show that they’re actively applying for jobs; once benefits expire, that incentive goes away. About 19 percent of cutoff victims have dropped out of the labor force each month this year, meaning they stopped actively looking for work. That’s a bit higher than the 16 percent who dropped out each month last year; that difference, too, is at most marginally statistically significant.7 We don’t yet have enough evidence to draw a firm conclusion about what impact the end of emergency benefits has or hasn’t had on the prospects of the long-term unemployed. The Current Population Survey is large, covering about 60,000 households each month, but only a small fraction of the people interviewed each month are unemployed, and even fewer were affected by the benefits cutoff. The total sample for likely cutoff victims over four months is a bit under 1,200 households, enough to draw only preliminary conclusions about a group that totals well over 1 million people in the overall U.S. population. This much is clear, however: There has been no sudden surge of former benefits recipients into jobs. Nor have they abandoned the labor force in droves. Most have done what Laurusevage has done: continued looking for work, but without the lifeline that benefits provided. For Laurusevage, the cutoff has been wrenching. Her husband’s salary, combined with the$563 a week she got in unemployment benefits — the equivalent of about \$30,000 a year — was enough to make ends meet. But once her benefits expired in December, life got harder. Their savings depleted, they scraped together this month’s mortgage payment only by borrowing from David’s 79-year-old mother. They don’t know what they’ll do this month.

“We are about to go under,” Laurusevage said. “My entire savings account is gone. Everything I spent years to save is gone.”

## Footnotes

1. These figures combine the Emergency Unemployment Compensation (EUC) program and another federal program known as Extended Benefits. ^
2. Boehner has also argued against renewing the program on the grounds that a retroactive reinstatement would be difficult to implement, a claim the U.S. Labor Department rejects. ^
3. These figures, based on longitudinally linked Current Population Survey microdata, look at the labor force status in March of people who were unemployed in December and were likely recipients of EUC benefits (as determined using the methodology described later in this article). Because the CPS is conducted in the middle of the month, it isn’t possible to know exactly how many of these people actually saw their benefits expire and how many found jobs or stopped looking in late December, when the program was still running. However, my methodology identifies 1.3 million people as likely EUC recipients in December, almost exactly the figure reported by the Department of Labor. ^
4. Based on my earlier research, which looked at national rather than state-level data, someone who lost a job in Pennsylvania in April 2013 had roughly double the chances of becoming long-term unemployed as someone who lost a job before the recession, when the state unemployment rate was 4.6 percent. ^
5. A bit more on the methodology here: To identify people who likely lost benefits, I found those who reported having been unemployed longer than their state’s maximum duration of benefits (26 weeks in most states) but for less time than their state’s maximum in December, when the EUC program was still in place. I also limited the sample to people who reported losing a job rather than having quit or entered the labor force for the first time. For the comparison group, I used the same basic approach to identify people who likely qualified for EUC benefits in their state.

The two groups are demographically very similar, but not identical. Most important, the median duration of unemployment is slightly lower for the cutoff group than for the comparison group. All else equal, this would tend to make the reported job-finding rate for the cutoff group slightly higher, which is another reason to be cautious about the slightly higher finding rate for this group. ^

6. Another problem: This methodology assumes that people file for benefits immediately upon losing a job and receive them weekly until those benefits expire. Neither assumption entirely reflects reality. Many people delay filing for benefits, in some cases for several weeks or even months. And many people stop receiving benefits for some period during their eligibility. That means my sample likely captures at least some people who were still receiving benefits at the time of their interview. To test whether this would have an effect, I re-ran my calculations excluding anyone unemployed for less than 30 weeks, and again for anyone unemployed less than 35 weeks. Both groups had slightly lower job-finding rates, which is consistent with the notion that the cutoff in benefits had little impact on the probability of returning to work. ^
7. These findings are consistent with past research that found that unemployment benefits have a bigger impact on the dropout rate than on the job-finding rate. That is, benefits make people more likely to keep looking (or at least to say they’re looking) for jobs when they might otherwise have given up, but not much less likely to find or accept jobs. ^

Ben Casselman is FiveThirtyEight’s chief economics writer.