The $800 Billion Covid Protection Program Failed Miserably

From the "stop me if you've heard this before" file, but the Federal Government totally screwed up the Paycheck Protection Program started in response to Covid in 2020.  It was part of the CARES Act, or the Coronavirus Aid, Relief, and Economic Security Act to help some select businesses, self-employed workers, sole proprietors, select nonprofit organizations, and tribal businesses continue paying their workers.  

This week, Brad Palumbo at FEE excerpted and summarized a new study (47 page .pdf) from MIT economist David Autor and nine coauthors for the National Bureau of Economic Research.  The researchers tracked where the money went and the results weren't pretty.

The analysis shows that even though 93 percent of small businesses received loans from the program, only between 2-3 million jobs were preserved. The program spent an astounding $170,000-$257,000 for each job it helped preserve! That’s, erm, a lot more than most of those jobs even pay.

Moreover, the study finds that only 23 to 34 percent of the program’s dollars went to workers who would’ve otherwise lost their jobs—meaning the vast majority went to “business owners and shareholders.” (Oh, and a whole bunch was lost to fraud, too).

We've all seen headlines about fraud in the program over the last couple of years, and the study confirms fraud really was going on.  Another clue about fraud was spending about $215,000 to preserve a job for someone who's making $25,000/year, which raises the possibility that money didn't actually go to the person making $25k/year.  The whole thing doesn't seem like a smart use of tax money.

The study also concluded that the benefits were highly regressive; that is, they paid more benefits to higher income people.  The program was intended to provide a temporary paycheck for low income workers affected by the "14 days to flatten the curve lockdowns."  Instead, it paid

...about 75 percent of the benefits flowing to the top 20 percent of earners.

I'm not sure that's as big a problem as they imply.  Consider a mom and pop pizza shop.  The way I read that, it could mean they paid something like the server's wages to the server, but paid something like the owner's portion to them.  If all you record is the number of dollars, of course the higher paid person got more dollars.  That's kind of the definition of being higher paid. 

The plot shows what I'm talking about.  Viewed simply as the total number of dollars, the fifth quintile of incomes has the biggest chunk of PPP benefits - black is PPP compensation and the gray squiggly lined block is PPP capital income while red dots are unemployment compensation and blue Xs are Stimulus checks.  Viewed as a percentage of their annual income, it's  rather different.  The percentage of annual income from the Stimulus checks and Unemployment insurance of the lowest quintile is highest and the goes down for the higher quintiles.  The percentage contribution of the PPP checks and Capital Income is bigger than the lower quintiles but doesn't seem as outsized as the raw number of dollars. 

All of this is secondary to the real questions about whether the lockdowns, the CARES program and PPP were reasonable in any way at all.