Part of the Coronavirus Aid, Relief, and Economic Security (CARES) Act passed by US lawmakers in the early days of the COVID-19 pandemic, the Paycheck Protection Program was designed to help small businesses retain their employees during COVID-induced economic shutdowns. Employers could apply through banks for loans from the Small Business Administration, some of which could be forgiven, depending on how the funds were used.
Following the rollout of the program, some have questioned the distribution of those funds—Chicago Booth’s João Granja, Constantine Yannelis, and Eric Zwick and MIT’s Christos Makridis, for instance, find that businesses in areas hardest hit by the pandemic and its economic fallout were actually less likely to receive PPP funds than businesses elsewhere. But Washington University’s John Barrios says that the program was designed to mirror the distribution of payroll across the country, rather than anticipating which areas would need the most help. And while it’s fair to scrutinize, question, or criticize the design and goals of the program, Barrios says, the PPP does appear to have been implemented as intended, with PPP funds distributed broadly in line with payroll levels throughout the US.
As a result of COVID-19 and the economic stress of closing down an economy, you get these government responses. How are we going to maintain some viability in the economy as you have basically shut down? You know, 120 million people have now stopped working. All these businesses have to close down. And we see policy responses of, “Press the pause button.”
How do we maintain that? It’s not a car. You can’t just shut off the engine and then turn it back on and it’s going to keep on running. But policy makers try to say, “How do we prevent people from separating from firms so that when we start up again, we don’t have to do this reallocation and try to get individuals to match up?”
The initial view of the Paycheck Protection Program, by its definition, was that we want to keep employees with firms, so we’re going to basically subsidize the firm by providing them with money to maintain the payroll for these two or three months that we were going to freeze the economy so that we didn’t have to worry about these individuals going on unemployment, even though we had unemployment programs, as well as subsidies to the unemployment program.
And we wrote up an op-ed—me and a colleague of mine, [Chicago Booth’s] Mike Minnis—we wrote up an op-ed thinking that timeliness versus targeting would be the issue here, because we wanted the money to get there fast, but they wanted it to get to firms that needed it. So they basically gave a very general 500-employee cap, for example, and spread the money.
And what we were going to say was, given that they were utilizing the banks to distribute these funds, we said, well, the problem is, we know from a lot of our research that relationships between banks and lenders would mean that certain companies would get access to the bank lending first because if I’ve been borrowing money from the bank and I’m a small business, I’m more likely to get the first money than John Smith who comes in and says, “Oh, I need . . .” and he’s never met this bank before.
So back before it was even passed and we had the op-ed in Barron’s, we were saying, “Look, these are going to be issues that are going to be affecting how this money gets allocated.” And then, moreover, we said, “How much money would we need?”
Back then, a lot of the debate was about how big should this program be? People throwing out $1 trillion here, $2 trillion there, and these numbers were all over the place. Well, theoretically, how big could this be?
And we did a simple exercise in terms of estimating the total amount by basically going into public data from the census and saying, “Look, if we take the distribution of payroll for these firms that are small businesses, we make some adjustments for the cap on $100,000 salary per employee. We estimate, then, by each state, how much money would need to be doled out?” We basically came up with around $750 million would be required in order to cover all these small businesses for three months to cover payroll. Because initially that was the response.
Eventually—this was during the debate where they had originally funded it—the funding dried up after less than a week. So then they added an incremental $200 million or $300 million, which ended up being close to around $750 million, our estimate in terms of the amount. And then we decided, well, let’s evaluate how this money was doled out.
And what you ended up seeing with the timing and the bank arguments that I made was that if you look at the first allocation, that first allocation was more skewed toward areas where there were banking relationships. Certain firms, certain states were getting more of the money earlier on. In the second allocation, that reversed. We saw now the money going to the areas that were getting less than the first. It was more of a timing issue. In that sense, you were now spreading it to these other areas as well.
And we could use our predicted number, which is what we use to judge the implementation of the actual program, and said, this is a predicted payroll that needed to get there, and this was the actual allocation of the two rounds.
And what you ended up seeing is that the R-squared was basically 0.99. So if you were to plot the predicted and the actual [allocations], which we do in the paper, it’s a 45-degree angle. Most of the states are in there and most of the industries are being allocated, some a little bit more or less, but it’s really pretty close to the distribution of payroll. So if you think of regressing the actual allocation versus the predicted, the coefficient was basically 0.98 or 0.97. So that means for every predicted dollar, the actual allocation was 97 cents. So that seemed to be in the spirit of the design of the program.
Now, in terms of for economic efficiency and allocations and the effectiveness, was this targeted to areas that needed it most? I think it’s an important question. We care about it from a welfare perspective.
But we also have to be a little bit realistic that back then, back in March, when this program was being developed, the connotation was we were in this massive pandemic that was going to come in. There was this huge wave, where numbers about death tolls were in the millions in some of these estimates. So it’s easy ex post to say, “Well, why didn’t they target it to the areas where COVID-19 was going to be highest?” And, well, imagine being the politician back then that says, “Well we should only give the money to New York City areas, LA areas, the larger cities.” You could really imagine that that was going to go well in Congress and in the news from a lot of these other states. The money didn’t seem to be going, at least in the first allocation, into areas where there was higher exposure to COVID-19, where you had more business shutdowns. From my perspective, it still was allocated based on payroll, which was the design back then.
I think it’s right to criticize the goal, the policy goal of covering all payroll, but I don’t think that the implementation was wrong. I think a lot of people now say, “Oh, the banks didn’t give the money right. The money wasn’t given out right.” Well no, it was. It was called the Paycheck Protection Program. It was meant to trace payroll, and what we see is that it traced payroll. It’s just that the tracing of payroll wasn’t that correlated to where COVID-19 was the highest.
And I think in that sense, we need to kind of step back and think about, for the future, what are the issues we need to deal with, and I think some of these issues were, for instance, we didn’t really have tracing. So we couldn’t really target where exactly this was going to hit because we really didn’t know. It’s hard to, ex ante, without any information about today, say, “Oh, yeah, we should just be allocating the money to x and y spots.” It’s still important to see: The money didn’t get to certain areas, what happened? And I think that should be addressed because we still need to make sure that we can come back from this in an efficient way.