Reopening Businesses Doesn’t Mean People Will Start Spending

CHRISTIANE AMANPOUR: Now, since the outbreak of COVID-19, we have been bombarded, of course, with stats and facts of the economic fallout. Like this one. At 13.3 percent, the unemployment rate in the United States is at near historic highs. But what does that actually mean? Well, our next guest is trying to make sense of it all, and he is Raj Chetty. He is an economics professor at Harvard University, and here he is talking to our Hari Sreenivasan about his research, about where the economic fallout lands.

HARI SREENIVASAN: Thanks, Christiane. Raj Chetty, thanks for joining us. So, Raj, people are generally familiar with the big 30,000-foot view of what has happened to the economy thanks to COVID. We know that there is massive unemployment, we know that spending has decreased. What’s different about the index you have built?

RAJ CHETTY, PROFESSOR OF ECONOMICS, HARVARD UNIVERSITY: Yes, Hari, I think the data people have seen in the newspaper is coming from surveys that the government conducts on a regular basis of businesses and households across the United States. Those are the surveys that are used to construct GDP, unemployment rates, the big numbers that you see in the headlines. What we are missing in that national picture is exactly what’s going on at the ground level, where are jobs being lost, in what particular businesses, for what income groups? Why exactly has spending fallen so much, and what does that mean in terms of which kinds of folks have lost their jobs as a result? And so, what we’re trying to do with the data that we’ve constructed using information from private companies is really grill down and show neighborhood by neighborhood, income group by income group, how is this crisis unfolding, who needs help and what should we be doing from a policy perspective.

SREENIVASAN: So, who stopped spending the most?

CHETTY: Yes. So, when we look at this data, start with the spending picture, which is where I think this crisis really began, you see that spending in the U.S. as measured by credit cards has fallen by about 30 percent from where it was in February to essentially a couple weeks after the COVID crisis began in the U.S. in the middle of March. We saw a 30 percent reduction in overall consumer spending, which, just to be clear, you know, is the biggest reduction in spending we’ve ever seen in the United States and recorded data, right. So, unprecedented in scale. Now, here’s what’s interesting, when you break that overall data up, you find that the vast majority of that reduction in spending is coming from high-income households rather than low-income households. So, you might have had the intuition that, you know, the folks who are going to bear the branch (ph) of this crisis, people who might have lost their jobs and so forth, are lower income folks, and they would be the ones cutting back on their spending. But, in fact, you see that, you know, by the end of May, two-thirds of the spending reduction we’re seeing in the U.S. is coming from people in the top quarter of the income distribution. So, it’s really the rich who have retracted their spending the most.

SREENIVASAN: So, then that has a ripple effect? The first being that where they spend their money, their neighborhoods are going to feel the economic pain.

CHETTY: That’s exactly right. So, you know, thinking about the mechanics of this, why have the rich reduced spending the most? If you look at data on just the amount of time people are spending outside, you see a pattern where high-income folks have cut the amount of time they’re spending outside their homes much more than low-income folks. So, what does that do? Local businesses that rely on foot traffic and particular businesses that are providing services in person, think of your local restaurants, retail shops and so forth, they have, you know, as you can just see walking around on the streets, they’ve lost a tremendous amount of business revenue. And in particular, they’ve lost a tremendous amount of revenue if they’re located in very affluent neighborhoods. So, think of the upper east side of Manhattan, for example. There you’re seeing business, small businesses, there losing 70 or 80 percent of their revenue relative to baseline, relevant to January or February. In contrast, you look at — another place in New York, look at the Bronx, for example, you’re seeing 30 percent reductions in revenue from local businesses, and that’s coming from the fact that, as I was saying at the beginning, high- income folks are cutting spending the most.

SREENIVASAN: So, in these areas, then, does that mean there are more layoffs or fewer job openings?

CHETTY: Exactly. So, the third piece of kind of this chain of dominoes falling, so high-income folks spends less, local businesses revenues fall, especially in affluent areas. Who basically takes the hit from that? It’s predominantly lower-income workers at these businesses in affluent areas. So, the restaurants in the upper east side that employ lots of folks that live in other parts of New York, in lower income neighborhoods, those folks tend to be the ones who have lost their jobs at much higher rates than people who happen to be working in similar restaurants in low-income areas or people who happen to be working for Amazon which has experienced a boon in business because of the nature of this shock.

SREENIVASAN: So, are there particular stats on low wage workers that are in affluent areas?

CHETTY: Yes. So, low wage workers in affluent areas, 70 percent of them who are working in small businesses have lost their jobs, 70 percent. So, you know, the odds are you lost your job if you’re working in a restaurant, working in a retail store in an affluent area. And so, those folks are just getting by, at this point, because of the unemployment benefit system. In less affluent areas, that number is 30 percent. So, there is a big difference across these places.

SREENIVASAN: Where this is happening and who this is happening to, how does that impact the recovery? Is that going to make it take longer?

CHETTY: And so — yes. So, in addition to seeing higher job losses in these particular areas — give you another example, you know, San Francisco — the center of San Francisco, very affluent neighborhoods, again, tremendous number of layoffs there at the heart of Silicon Valley. In addition to seeing higher rates of job loss there for the low-income workers, and I want to stress, it’s the low-income workers there, not the high-income workers there who lost their jobs. We’re also seeing fewer job ads being posted in those areas, and if you look at jobs that don’t require a college degree, for example, you see that those job postings have fallen the most in places like Downtown San Francisco in Silicon Valley. So, it’s not just that there is greater job loss there to begin with, we’re also seeing ominous signs that the recovery might take longer in many places, in particular for less educated workers because we’re not seeing much of a recovery in spending yet for more affluent households. We’re still able to self-isolate while the threat of COVID is present.

SREENIVASAN: Is that changing as states open back up?

CHETTY: You can ask. You know, do we actually see more spending, more employment in the states that opened earlier? Does this seem to have an effect? And the answer, unfortunately, is really no. Especially if you look at higher-income folks where spending has fallen the most, you look at the businesses that have been hurt the most, we’re not seeing much of an accelerated recovery at all when you have an earlier reopening of a state. And I think that’s fundamentally because the reason that spending has fallen is not because of shutdown orders or because, you know, certain businesses were ordered to be closed, it’s because of people’s choices. If you look at the data, you see that spending actually fell before a lot of the shutdown orders went into place. And what that’s telling you is, people are cutting back on their spending because they themselves are worried about their health. And you’re not going to be able to undo that fully by just issuing an order that you’re reopening businesses.

SREENIVASAN: So, were you able to compare states, perhaps neighboring states, on kind of both sides before the orders came down? Were you able to see that spending habits changed, vs. an order that comes down to order back up, and that says, fine, we can go ahead and open businesses again, but does that change the spending habits?

CHETTY: Yes. So, to give you a concrete example, Hari, of the type of analysis we did, let’s take Minnesota and Wisconsin. So, Minnesota and Wisconsin shut down at roughly the same point. They issued stay-at-home orders. And what we saw in both states is that spending had essentially fallen by about 30 percent even before the shutdown orders went into place. So what that shows you is, people were making choices on their own, and it was not fundamentally the states’ decisions that drove down that change in spending behavior. Minnesota then reopened two weeks before Wisconsin did, partial reopening of its businesses. But if you look at the data, you can track spending patterns in Minnesota and Wisconsin over time, and you see that they basically perfectly match each other, even though Minnesota opened two weeks before Wisconsin did. And so it’s that type of analysis, not just done with those two examples, but comparing states across the U.S., you can see that we can’t count on reopenings by themselves to revive our economy.

SREENIVASAN: So, let’s look a little bit at the measures that we took as a government. So, first, let’s talk about the CARES Act. How do you measure the impact of the CARES Act?

CHETTY: Yes, so the CARES Act had two central elements. The first was stimulus payments to households, and the second was loans to small businesses. So, let me start with the stimulus payments to the households. So, these are the typically $1,200 checks that lots of folks received in the mail around April 15 to try to support spending, increase spending in the U.S. And so what we can do to these data, coming back to the credit card spending data, you can literally look day by day and ask, the day that people got these checks, we were able to see in bank accounts when many of these deposits went in, literally the day of April 15. And so we can see, are people starting to spend more on April 16, 17, 18, once they have this extra $1,200 in their checking account? And, sure enough, you do see that spending goes up quite significantly immediately afterward. So, at that level, the policy seems to have worked. But here’s the wrinkle there. Spending went up primarily among lower-income families who, perhaps because they have less savings, have a tendency to spend more when they get a check in the mail. But if you look at the high-income families — and, remember, that’s where most of the spending reduction came from to begin with — we see relatively little impact. If they got a stimulus check, there wasn’t that much of a propensity to spend it.

SREENIVASAN: The Paycheck Protection Program was supposed to be something targeted exactly toward businesses. How do you measure the impact of that? What did employers do with those funds, if they applied for and got it?

CHETTY: So, it turns out that, for the most part, you were eligible for the Paycheck Protection Program if you had fewer than 500 employees, but if you had more than 500 employees, you weren’t eligible. And so what you can do is compare firms that happened to have less than 500 employees and would have been eligible for the PPP program vs. firms that had more than 500 employees and ask, did the firms that had less than 500 employees, did they actually layoff fewer workers once the program began on April 3? So, that’s kind of a way to — sort of like an experiment that you can analyze. And what you end up finding in the data — and we were quite surprised by this — is, there’s actually not that much of a differential effect. There might have been a modest effect of the paycheck Protection Program on increasing employment in those firms with fewer than 500 employees. But, overall, you see an enormous decline in the fraction of workers those firms were employing, just like the firms that were larger and would not have gotten assistance from the PPP.

SREENIVASAN: So, the $300 billion that went through to consumers didn’t have the effect that we thought it was going to have, at least not evenly. The $500 billion that went to businesses, you’re saying, didn’t have a measurable impact on whether it actually kept more people employed or not. So, now I have got to ask the other question. This is, based on all this other information you’re looking at, what should we be doing?

CHETTY: Yes, so I think this is a very unusual crisis. I think the policy responses that were undertaken, I don’t want to say they had no effect. I think some of what we have done — there’s one other important piece, Hari, that we haven’t touched on yet, which was the extension of the unemployment insurance benefits, in particular for people who are unemployed, sustaining their incomes. That, I think, has been incredibly important. Had we not taken those steps early on, which I think the government did very effectively, we would have an even bigger crisis on our hands, because lots of people wouldn’t be able to pay rent, would have no money to spend on food and so forth. Everyone has lost their jobs. And that would be an even bigger crisis than what we have. So, some of the steps the government has taken, I think, have been incredibly important, in particular that expansion of the unemployment insurance benefits.

SREENIVASAN: So, if unemployment insurance benefit programs were working, and they actually had a measurable impact, according to all the information, would the money that we deployed have been better spent if we actually just gave unemployed people more money to spend and, well, boost the economy?

CHETTY: Yes, and so that’s exactly where I was going with this, as you anticipated. I think targeting our assistance — there is only a finite amount of resources the government can spend at the end of the day. I think targeting that assistance to folks who have lost their job and are in the greatest need of help at the moment is the smartest strategy, as opposed to the stimulus, which also, I think, had that sort of intention, but also gave these $1,200 to people who have not lost their jobs, right? So, the distinction between the unemployment benefits and the stimulus is that the unemployment benefits specifically go to people who have lost their income. And that, I think, is critical, because those folks, had we not given them unemployment benefits, they would have lost purchasing power, and you would have seen their spending fall for lower-income folks even further. And that would have had a further ripple effect and so on. And so the way I look at it is, the deep problem that overall spending in the U.S. has fallen, and that that’s having this cascading effect of lower employment and so forth, that can only be solved by addressing the fundamental health concern, so, the public health issues. How do you get people to feel confident about coming back out of their houses, to going and working in the neighborhoods where they usually worked? Of course, the long-term solution here is a vaccine, but, in the interim, are there public health measures we can figure out through social distancing or contact tracing, et cetera, that can restore consumer confidence? That, I think, is the only fundamental solution. In the meantime, there are lots of people who are going to be suffering as a result of this crisis, in particular what we see in the data, low-income workers. I think we need to think hard about how we can use our limited resources to support those households while they’re out of a job.

SREENIVASAN: You have done a lot of research about kind of the rungs of opportunity and how some disappear or how people slip back and forth. We saw a tremendous amount, for example, of wealth lost for African- American households after the real estate and the financial crisis 10, 12 years ago. When you see this crisis playing out, and you see communities of color who might be low-income disproportionately affected here, are you concerned about a longer-term impact on social mobility?

CHETTY: I’m very concerned, and I’m glad you raise that. I actually think one of the things we need to keep our eye on in this crisis — we’re all very focused on the immediate current jobs numbers — 2.5 million jobs added last month, that’s great news. But what I think we really need to be keeping our eye on is the potential long-term impacts of this crisis. And so let me give you one piece of data that we have been tracking that we think illuminates why this crisis could have very long-lasting impacts. One of the things we have been looking at is how much kids are learning in school. So, as you know, most schools have essentially shut down to instruction online. And we have been looking data from a platform called Zearn, which provides a math curriculum that many students use in school online, so this is essentially doing lessons, math lessons, online, which they were doing in school before, and now they’re doing remotely. For high-income families, you see a sharp dip in the amount of time that kids are spending on the Web site, the amount of progress they’re making in these lessons. But, within a week or two, that bounces back, and they’re basically back to their baseline levels of learning or exceeding where they were before. For low-income families, you see something like a 60 percent reduction, and it essentially never comes back. What I think it illustrates is a broader issue, which is, folks that are in lower-income backgrounds, you know, people of color, I think, in this crisis, are going to experience shocks, not just in terms of their current economic situation, but also potentially adverse impacts that are going to last into the next generation, because we have seen from our prior research that what is going on in the early part of childhood in particular has dramatic effects on social mobility. The environment that you’re in as you’re growing up, your educational opportunities and so forth, are a key driver of how much you’re earning when you’re 30 years old.

SREENIVASAN: Raj Chetty, thanks so much for joining us.

CHETTY: My pleasure. Thank you.