Related Links

• Disincentive Effects of Unemployment Insurance Benefits
• How Did Pandemic UI Benefits Affect Employment Recovery in Local Industry Markets?

Transcript

Tim Sablik: My guests today are Andreas Hornstein and Marios Karabarbounis. Andreas is a senior advisor and Marios is a senior economist, both in the Research department of the Richmond Fed. Andreas and Marios, thanks for joining me.

Andreas Hornstein: Thanks for having us, Tim. Glad to be here.

Marios Karabarbounis: Happy to be here.

Sablik: Today, we're going to be talking about the different effects of unemployment insurance benefits. Unemployment insurance, or UI for short, allows qualified workers who have lost their job to continue receiving income while they look for new work.

During a crisis, this can help stimulate demand because it enables unemployed workers to continue spending. But if unemployment benefits are too generous, they could act as a disincentive for recipients to reenter the labor force, potentially slowing down recovery.

In a working paper you published last November along with several co-authors, you both looked specifically at these disincentive effects. In the paper, you noted that it can be challenging for economists to separate the competing effects of UI benefits. What are some reasons for that?

Hornstein: Let's step back for a moment.

We model unemployment as a choice problem. It's true that somebody that gets laid off may not have much of a choice. But once you're unemployed and you are looking for work, there are choices to be made — how much time and how much effort do you spend in the search process, where to look for work, what kind of work to look for and, once you're successful and have generated a job offer, whether to accept that offer or continue to search for a better offer. So, when we talk about incentive effects of UI, we think of how UI affects these decisions.

UI isn't simply insurance like property insurance, where you get a lump sum payment in an insurance event. Rather, you receive a payment for as long as you remain unemployed, subject to some bounds on the duration. This UI makes being unemployed relatively less costly. You may devote relatively less effort to generate an offer, or you may require a higher wage to accept that offer. Economists call that a reservation wage.

But, as you said, UI payments may also stimulate local demand. Many unemployed do not have the financial resources to weather a jobless spell easily. UI will be used mostly for contemporaneous spending and without UI being there, you will see a decline in demand. So, demand coming from UI shows up in an increased availability of jobs because more employers out there who want to hire you because the demand is there.

So, we have two countervailing effects of UI for the unemployed. It may reduce the incentive to look for a job, but it also increases the availability of jobs. The net impact on unemployment may well be ambiguous.

Sablik: Why did you decide to examine the disincentive effects of UI benefits specifically?

Karabarbounis: The reason we focus on the disincentive effect is because there is a big disagreement about it among economists and policymakers. For example, during the Great Recession, unemployed workers would keep collecting their benefits for a longer period of time compared to normal times. Economists are still debating whether this extension discouraged workers from searching for a job and slowed down the employment recovery.

There is a similar disagreement about the pandemic unemployment benefits. As we all know, the pandemic benefits were quite substantial. However, many researchers documented that they didn't seem to have much of an impact on the labor market. We found it hard to believe that such substantial provisions didn't have at least some negative effect on the labor market. So, we decided to study this problem, having in mind that this disincentive effect might be counteracted by a positive demand effect.

Sablik: A good opportunity to study this type of question is when you have a big change in policy and, as you mentioned, UI benefits were substantially enhanced during the pandemic. Andreas, can you give listeners a quick overview of what some of those changes were?

Hornstein: Sure.

We mainly study the effects of the pandemic unemployment insurance assistance program. This program increased UI benefits, it increased the duration for which you will be able to receive benefits and it lowered requirements to get UI benefits. That more or less applied for the months of April through July 2020. This period covers the widespread pandemic shutdowns and the early recovery period.

Parts of this program are not that different from UI modifications you see in any other recession, for example, the extension of duration for which one would receive UI. But the increase of benefits and eligibility was unprecedented.

First, everybody who qualified for UI benefits received an additional $600 of weekly benefits on top of normal UI benefits. The estimates are that with this additional payment, total UI payments exceeded pre-UI work income by about 50 percent for the average UI recipient. In normal times, UI payments are about 40 percent or less of previous work income.

Second, UI eligibility was extended to self-employed and to those not meeting the usual requirements of previous work experience. We estimate that for the low-income groups in our sample, this meant that the fraction of unemployed receiving UI benefits tripled from about 20 percent pre-pandemic to about 60 percent during our sample period.

So, the increased UI eligibility and benefits resulted in total UI payments in 2020 that were about three times those paid out in 2010 during the Great Recession. So, yes, the changes to UI in the pandemic were large.

Sablik: Once you have the data on this pandemic episode, how did you then go about isolating the disincentive effects of UI benefits?

Karabarbounis: Our research starts from the observation that average wages are widely different across labor markets in the United States. Take for example a restaurant that is located in upstate New York. This restaurant is paying lower wages compared to a restaurant that is located in Manhattan. During the pandemic, all eligible unemployed received $600 per week on top of the normal benefits. This means that the restaurant that is located in upstate New York and is paying lower wages likely found it more difficult to attract workers compared to the restaurant in Manhattan that is paying higher wages. This is the disincentive effect of the benefits. And this is what we try to estimate.

But at the same time, for workers who are located in upstate New York, this $600 replaces a large fraction of their typical weekly paycheck. So, their purchasing power increases, which implies more demand for local goods and services and, at the end of the day, more demand for workers. For the unemployed workers in Manhattan, this $600 replaces a smaller fraction of their typical paycheck, so the demand effects are not as strong in this case.

So, we have two opposing effects, a demand effect and a disincentive effect. In our research, the way we isolate the demand from the disincentive effect is to examine low- and high-wage businesses that are located close to each other that belong to the same geographical area. The idea is that low- and high-wage businesses that are located close to each other both benefit from the stimulative effects of unemployment insurance benefits. When we compare them, the demand effect is eliminated because it's common. This allows us to exclusively isolate and identify the disincentive effect of the benefit.

Sablik: What were your key findings? You mentioned that there were a couple of different changes to UI benefits in terms of the amount of benefit the duration and who was eligible. Did you find anything about which of these changes had the biggest impact on the incentives of unemployed workers?

Hornstein: As mentioned, we expect that the disincentive effects of UI should have slowed down reemployment recovery of low-paying businesses relative to adjacent high-paying businesses. And that is what we find. When the $600 UI supplement was in effect, low-paying businesses recovered more slowly. Once the $600 UI supplement expired, low-paying businesses started to catch up. So, we estimate that the changes to UI during the pandemic delayed the employment recovery by about four and a half percentage points between April and December 2020. We call that a recovery gap.

Second, for each of the UI changes in isolation, they had only limited impact on overall employment losses. But jointly, they had a large impact. Somewhat to our surprise, UI eligibility expansion seems to have had the biggest impact.

Sablik: Based on all your results, do you have any recommendations for policymakers who might contemplate changing UI benefits in a future crisis?

Karabarbounis: Although we find sizable disincentive effects of the pandemic benefits, this does not mean that we advocate against using these important tools, especially during times of distress. We recognize the multiple benefits that these programs carry. As we mentioned, the demand effects from the benefits help the economy recover faster and benefits act as insurance against job loss. However, we would like policymakers to have in mind that unemployment benefits can make it harder for businesses to find workers, especially when these businesses are typically paying lower wages.

Sablik: Andreas and Marios thank you so much for coming on the show to talk about this research.

Hornstein: Thanks, you're welcome.

Karabarbounis: Thank you.