Controversies and Confusions about Pandemic Wage Subsidies

In the pandemic’s first year, governments in the US, Canada, and numerous other countries enacted large subsidies to help businesses survive lockdowns and retain workers. The US Paycheck Protection Program (PPP) extended USD 800 billion of business loans that, subsequently, were mostly forgiven. The Canada Emergency Wage Subsidy (CEWS), delivering cash grants to businesses totaling CAD 83.5 billion, had a proportionately similar fiscal cost.  At the time, the subsidies enjoyed strong support across political parties. But policy specialists worried about their potentially inadequate targeting, and not surprisingly, the now-expired subsidies are generating lingering controversies. The White House, for example, recently defended President Biden’s controversial student loan relief plan by attacking Republican politicians who benefitted from the PPP as business owners. 

One recent study casts the PPP in a particularly negative light. A group of economists led by David Autor estimated that only 23-34% of PPP dollars ‘went directly to workers who would otherwise have lost jobs’, with the balance accruing to the benefit of business owners. The program was ‘highly regressive’ in its economic impact, with 3/4 of PPP funds going to the top quintile of US households. Although the authors consider Canada’s CEWS better targeted, this and other studies are likely to confirm misgivings about large-scale pandemic wage subsidies in Canada (and elsewhere) as well. In fact, many believe that the most charitable take on the PPP, CEWS, etc. is that they achieved exceptional speed in benefit delivery—at the cost of (massive) mis-targeting.

It is easy to say: were we to do it all over again, we would surely choose differently.  But in truth, I believe not only do we not yet have the benefit of full hindsight, critiques of pandemic wage subsidies also suffer from mis-guided comparisons and implausible benchmarks.

Consider first some misguided comparisons. Many commentators (Autor et al included) believe that short-term work (STW) programs—used most widely in Europe and known as work sharing or short-time compensation in North America—are a particularly well-targeted type of wage subsidy. In an STW program, employers agree with the government to reduce the work hours of designated employees. Such employees get reduced pay from employers, but receive compensation for the work hours lost from the government, set at unemployment insurance (UI) replacement rates (ie percentages of regular hourly wages).

STW programs thus both provide income support to workers and preserve jobs. This combination may at first appear to be exactly the goal of wage subsidies like PPP and CEWS. Are STW programs thus an alternative to PPP and CEWS? Canada’s Finance Minister Chrystia Freeland, who spoke admiringly of Germany’s STW scheme as superior to CEWS in 2020, is not alone in thinking so.

But STW programs are not really comparable to wage subsidies. STW’s main purpose is to neutralize dis-employment effects of the UI system. Suppose that UI offers a 60% wage replacement for laid-off workers, but a business operating during the lockdown is able to justify maintaining work hours only at 40% of normal levels. In this scenario, the employee would be better off being laid off (receiving 60% of regular income while not working) than continuing to work but receiving only 40% of regular income. UI’s availability, therefore, may in itself destroy jobs. STW neutralizes this effect by ensuring that the worker remains better off being employed. However, under STW the employer receives no direct subsidy from the government: by assumption, the employer has already cut labor costs by reducing work hours.

Wage subsidies, by contrast, assume that many employers cannot easily reduce work hours or wages. By getting the government to bear a part of employers’ fixed labor costs, wage subsidies help financially constrained firms. (Because wage subsidies reduce hourly labor costs, they also incentivize hiring, whereas STW does not.) In summary, while STW programs may be an important complement to UI programs, they are not a substitute for wage subsidies, unless one wishes away the problem of downward wage rigidity.  

Consider next problematic benchmarks implied by PPP/CEWS critics. It is easy to misread Autor et al.’s conclusion that only 23-34% of PPP dollars ‘went directly to workers’ to imply that, ideally, 100% of PPP dollars should go to preventing layoffs. This would, however, directly contradict the declared intent of PPP, CEWS and other similar programs, which considered helping employers cover all fixed costs. Job retention was never PPP’s sole purpose.

But there is a more important point. Even businesses subsidies specifically intended to promote the hiring or retention of workers have ‘leakage’—they end up delivering benefits to employers. For example, when France enacted a large payroll tax cut to promote employment of low- and middle-wage workers (during ‘regular’, non-emergency times), about 50% of the tax cut’s benefit went to business owners. Similarly, a Swedish payroll tax cut aimed at encouraging the hiring of young workers also shifted substantial benefits for unintended beneficiaries. Even the Earned Income Tax Credit, paid directly to low-income workers, may partially benefit employers. 

The key question, then, is what is the baseline of tolerable ‘leakage’ for wage subsidies. If the ‘regular’ leakage of wage subsidies of 50% is acceptable, then a 70% leakage for a rapidly implemented subsidy that were also meant to benefit business owners may not look so terrible after all. 

Crucially, one cannot evaluate policies only on the basis of distributional impact. As Autor et al.’s study itself shows, stimulus paychecks to households may be distributionally more desirable than UI (which makes large payments to high-income workers who lose their jobs). But that surely is no reason to abandon UI. Nor do many people think that the U.S. government should have spent more on stimulus paychecks (USD 800 billion) or UI payments (USD 680 billion) during 2020-21: they arguably already showered too much largesse. Like these other programs, PPP may be faulted for its excessive generosity. But it did possess its own distinctive purpose not achievable through the other programs: namely, providing businesses with liquidity and sustaining employment in circumstances of fixed labor costs.

A decisive argument against pandemic wage subsidies, in fact, could be made if the pandemic would have destroyed few valuable job matches: UI itself would be sufficient if most job separations took the form of temporary layoffs after which workers returned to the same jobs. While few predicted that this would happen, there is now some surprising evidence for it.

We can thus certainly expect to become wiser with more hindsight. But critiques of PPP, CEWS and similar programs so far have been based more on mistaken preconceptions than new insights.

Wei Cui is Professor of Law at the Peter A. Allard School of Law, University of British Columbia.