Two aspects of the Paycheck Protection Program’s loan forgiveness create powerful incentives for employment — perhaps too powerful. They provide more than 100 percent — and in some cases more than 200 percent — government reimbursement for increases in payroll. Is it really possible that an employer can pay a relative $5,000 and get $13,000 reimbursed by the government?
Congress should consider a clean-cut deletion of the provisions limiting loan forgiveness under section 1106(d) of the Coronavirus, Aid, Relief, and Economic Security (CARES) Act (P.L. 116-136). It would be simple to do. More importantly, it would greatly simplify compliance and administration. And there would still be plenty of incentive for employers to maintain employment.
Paying the Rent
First, let’s look at how payroll costs for loan forgiveness are calculated. In the Small Business Administration’s interim guidance (and not in the statutory language), non-payroll costs — that is, interest, rent, and utility costs incurred by the borrower — cannot exceed 25 percent of loan forgiveness. (Because loan forgiveness equals payroll costs plus non-payroll costs, another way of expressing this rule is non-payroll costs cannot exceed one-third of payroll costs.)
Looking at this provision in isolation, for an employer that expects loan forgiveness to be less than the loan amount, and for whom the no-more-than-25-percent constraint is binding, any $1,000 increase in payroll will increase loan forgiveness by $1,333.
Example 1. At Gary's Garage, all 10 employees had an annual salary of $39,000 (with no benefits) in 2019. Thus, payroll costs for determining the loan amount are $81,250 (equal to 10 times $39,000 times 2.5 divided by 12). Gary's Paycheck Protection Program loan disbursement occurs on May 1. Gary had maintained payroll until then, but because of slow business and generous unemployment benefits, he lays off two employees May 1.
Gary should reconsider. During the eight-week forgiveness (“covered”) period, Gary would have had payroll costs of $60,000 (equal to 10 times $39,000 times 8 divided by 52). Also, during that eight-week period he expects to have rent, interest, and utility payments of $22,000. That means non-payroll expenses exceed 25 percent of loan forgiveness. Without the two layoffs, Gary's loan forgiveness would be $60,000 (of payroll costs) plus $20,000 ($22,000 of non-payroll costs limited to 25 percent of loan forgiveness). With the layoffs, Gary's loan forgiveness is $48,000 (of payroll costs, equal to 8 times $6,000) plus $16,000 ($22,000 of non-payroll costs limited to 25 percent of loan forgiveness). Because of the layoffs, Gary reduces his loan forgiveness by $16,000, from $80,000 to $64,000, while saving only $12,000 in salary costs — a net loss of $4,000.
For reasons that are not apparent, in addition to the loan forgiveness reduction that comes from any reduction in payroll costs, the CARES Act further reduces forgiveness for reductions in full-time employees (FTEs). In many cases this provides an inordinately large incentive to maintain employees. (We did not include the effect of this provision in Example 1.)
Now we will consider the impact of the FTE provision without the 25-percent-of-forgiveness constraint (in Example 2 below) and then with the constraint (in Example 3).
Example 2. As in Example 1, during the eight-week forgiveness period, Gary would have had payroll costs of $60,000 without layoffs. Unlike before, during that eight-week period he expects to have rent, interest, and utility payments of only $10,000, so the 25 percent of loan forgiveness does not come into play. Without the two layoffs, Gary's loan forgiveness would be $70,000 ($60,000 of payroll costs plus $10,000 of non-payroll costs). With the layoff of two employees, Gary's loan forgiveness is $58,000 ($48,000 of payroll cost plus $10,000). Because of the layoffs, Gary reduces his loan forgiveness by $12,000 (equal to $60,000 minus $48,000) and his salary costs by $12,000. Assuming away many details, it would have cost him nothing to retain the two employees. That's a pretty good employment incentive.
It turns out this $58,000 in forgiveness is only a tentative amount. Under the CARES Act’s section 1106(d), there is an additional reduction in forgiveness proportionate to reductions in FTEs. By reducing his employee count from 10 to eight, Gary has a loan forgiveness quotient of 0.8 (8 FTEs divided by 10 FTEs), so loan forgiveness turns out to be $46,400 (equal to 0.8 times $58,000). Thus, by his reduction in employment, Gary reduced his loan forgiveness by $23,600, from $70,000 to $46,400. Because he saved only $12,000 with layoffs, layoffs created a net loss of $11,600. Put differently, the incentive effect of the loan forgiveness provisions of the CARES Act in these circumstances is equal to a 197 percent wage credit (197 percent = $23,600 divided by $12,000).
It will not be uncommon for an employer that has reduced payroll to have the 25 percent constraint binding (because it is likely that interest, rent, and utility costs in total will not decline as much as payroll). So let’s combine the effect of the provisions in the two prior examples and consider an employer who has a binding 25 percent constraint and is considering a reduction in employment.
Example 3. Let’s begin where Example 1 left off. Gary would have had loan forgiveness of $80,000 without layoffs. By laying off two employees, his loan forgiveness was reduced to $64,000, without taking into account the additional reduction for his cut in FTEs. Because his full-time employment is down to 80 percent of prior levels, his loan forgiveness amount is reduced by 20 percent to $51,200 (equal to 0.8 times $64,000). The reduction in loan forgiveness resulting from the layoffs is $28,800 (equal to $80,000 minus $51,200). Thus, to save $12,000 of salary cost, Gary would give up $28,800 of loan forgiveness. To put it differently, in this case the incentive effect of the loan forgiveness provisions of the CARES Act is equivalent to a 240 percent wage credit.
Discussion
For simplicity of exposition, these examples have been designed to avoid some of the complicated quirks in the statute, particularly in the timing of periods for calculating averages before loan origination. But we hope readers agree that the situation described in the examples seems common. And we believe that in the examples, despite avoiding some complications, the basic thrust and apparent intent of the act’s provisions are not misrepresented.
Note that the incentive effects described above would even be larger if the tax-free treatment of loan forgiveness is coupled with the non-disallowance of deductions of expenses used to compute loan forgiveness. The taxpayer-friendly argument for non-disallowance is here. [Virginia Blanton, Michael Q. Cannon, and Jennifer A. Fitzgerald, “Double Tax Benefits in the CARES Act.”] The case for disallowance is here. [Eric Yauch, “Congress Asked to Clarify Tax Deductibility of PPP Loan Proceeds.”]
There are three reasons why Congress should consider the outright repeal of the limit on the amount of loan forgiveness (section 1106(d)). First, as demonstrated by these examples, the incentive effects they create are more than generous. The incentive effects without these provisions already provide at least full reimbursement, and can be greater than full reimbursement if the 25-percent-of-payroll restraint is binding or if deductions are not disallowed for expenses that generate tax-free forgiveness.
Second, these inordinately large incentive effects invite unproductive or perhaps abusive behavior. After Gary lays off two of his real employees, what is to prevent him from putting his nonmechanical nephew and brother-in-law on the payroll?
Third, in these examples we have deftly avoided the complex additional provisions in section 1106(d), including reductions in forgiveness for reductions in payroll costs and, working in the other direction, exemptions from these reductions for rehires.
The complexity and ambiguity of these provisions are extraordinary. Considering that they apply to hundreds of thousands of financially unsophisticated small businesses, this is a ridiculous situation. How bad is it? Anthony J. Nitti of RubinBrown LLP, one of the most prominent commentators on small business taxation, is one of the few who has been brave enough to try to puzzle out what these provisions mean. Exasperated, he concludes: “I give up. You should too. We need guidance.” (“Ten Things We Need to Know About Paycheck Protection Program Loan Forgiveness,” Forbes, Apr. 15, 2020).
In summary, it’s hard to see why Congress does not simply strike the provisions limiting loan forgiveness (section 1106(d)) from the CARES Act. They are extremely complex and serve no good purpose.