Since the beginning of the COVID-19 pandemic, the scope and amount of unemployment insurance (UI) in the United States has been expanded substantially through federal legislation. The legislative acts that created this change were passed when many businesses were shut down and many ordinary patterns of consumption were suppressed. Those unusual conditions have now largely ended, and much of the reasoning that advocates used to argue for the expansion no longer applies. It is a mistake to tie an explicit anti-work eligibility requirement to large amounts of cash income.

Background

Even prior to the pandemic, UI was a complex joint state-federal venture. Some funding comes from federal sources and some from state sources. States have some autonomy, but not full autonomy, over their policy details like formulas, administration, and eligibility. However, since the beginning of the pandemic, the federal role has increased, both in terms of the quantity of federal spending and in terms of rules for eligibility.

The recent expansion began with the Coronavirus Aid, Relief, and Economic Security (CARES) Act, which modified the UI system on several dimensions. First, it extended the duration of time an individual could collect UI to the end of 2020. Follow-up legislation, most recently the American Rescue Plan, has extended these benefits to September of 2021. This is a longer duration than normal (UI was typically available for only 26 weeks prior to the pandemic) though not as long as the 99-week benefits from the prior recession.

Second, the CARES Act expanded eligibility for UI to some classes of people who were previously ineligible, such as self-employed individuals, gig-workers, and those previously unemployed who had exhausted their benefits.

Finally, CARES introduced a federal unemployment supplement of $600 a week on top of state benefits. This combination resulted in 69% of the UI-eligible unemployed being paid more from unemployment than they did in total compensation prior to the pandemic. In subsequent legislation, it was reduced to $300 a week in December of 2020. As of the most recent legislation, the American Rescue Plan, the supplement was extended to September of 2021.

Anti-Work Requirements Are Relatively Unique

One theme of many policy debates is how government assistance changes incentives to work. There are two main points of view: the first is that the assistance should apply universally to working and non-working people alike. The second is that the assistance should come with work requirements—for example, to encourage self-reliance or a good-faith contribution towards that end.

One example of this debate in action is over the Child Tax Credit, which was also expanded in the American Rescue Plan. The benefit normally phases in with earned income; it effectively contained a work requirement. However, the American Rescue Plan recently expanded the credit to become a universal policy. Senator Marco Rubio recently published a critique; he would prefer a different proposal for expansion that he made alongside Senator Mike Lee, which would preserve the connection to work.

UI is a rare policy that takes neither the universal approach nor the work-requirement approach. Instead, it applies only to nonworking people. Simply put, it comes with an anti-work requirement.

Political rhetoric can often elide the distinction between universal policies and the anti-work requirement of UI. It is true that both are received by people who do not work, and both can make work less necessary by providing people with non-work sources of income. This effect, called an “income effect” by economists, is present in both policies; in both cases, they may work less because they have become richer.

However, compared to a universal policy, UI creates a second (and potentially much stronger) effect known as a “substitution effect.” A substitution effect is the result of policy that changes the effective exchange rate between two activities. People then respond by doing less of the activity that was made comparatively more expensive, and doing more of (“substituting”) the activity that was made comparatively less expensive.

UI creates a substitution effect in that it lowers the amount of post-tax-and-transfer money earned from a job. When a worker accepts a job, the worker’s income does not improve by the stated wage—rather, it improves only by the wage minus any unemployment benefits foregone. This makes a job relatively less attractive.

UI is therefore different from most policies, or sources of private-sector income, which lack a direct anti-work substitution effect. This effect is a substantial drawback that should not be taken lightly.

Magnitude and Circumstances Matter in UI Policy

The strongest arguments for UI do not justify UI policy in its expanded form under today’s economic circumstances.

For example, advocates have argued that UI is valuable in an insurance role: instability in income is undesirable, and UI can mitigate temporary income loss. However, insurance is typically carefully designed to avoid a situation where taking a loss and making a claim is more desirable than not taking the loss in the first place. For example, one typically cannot insure a piece of property for more than it is worth. In fact, to the extent that insurance exceeds replacement value, it creates income instability rather than removing it.

Determining replacement value is more difficult for jobs than for property insurance; a direct comparison of wage to benefits is not a complete analysis; for example, in addition to a wage, a job might be beneficial in terms of skillbuilding, but also costly in terms of time and effort. Seeking a job may require costly effort in itself. It is not possible to replicate the cost-benefit analysis of each worker’s decisions and precisely quantify the role of policy. However, UI benefits under current law are still large enough to exceed wages for many workers even after being scaled back to $300 a week. This suggests that—even in the best case—many workers are overinsured.

Advocates have also argued that the unusual circumstances of the peak of the COVID-19 pandemic made undesirable properties of high UI benefits (such as disincentivizing job search) less undesirable than usual. For example, in 2020 UI specialists at The Brookings Institution wrote that the pandemic “introduced some special circumstances unrelated to benefit levels that legitimately interfere with a worker’s ability to work.” They argued that this changed the calculus: if a worker’s job was shut down by the state, there was no immediate anti-work substitution effect from UI policy.

Those arguments do not hold up to present day realities: vaccinations for COVID-19 are readily available, case counts are low, employers are open again, and job openings have reached an all-time high. Instead of allowing workers to make the choices that are best for them, an expanded UI instead threatens to take away money from them—and make their new jobs less valuable—if they choose to take advantage of the hot job market.

The precise impact of the substitution effect may be difficult to measure; the policy is concurrent with other things that may be keeping workers away from work, like lingering pandemic restrictions or inconveniences. However, the evidence points in the direction of a policy-induced suppression of the recovery.

A recent analysis from Jason Furman and Wilson Powell shows that a high level of job openings typically results in a high percentage of unemployed people taking jobs. They posit that given the high level of job openings, we would ordinarily expect 34 percent of those unemployed in April 2021 to become employed as of May 2021. However, this did not come to pass: only 24 percent found new jobs.

Furthermore, labor force participation remains low—just 61.6 percent of Americans are either employed or looking for a job, down from a peak of 63.4 percent prior to the pandemic. While health and job openings metrics suggest that the economy can return to pre-pandemic levels of employment, workers are not returning. All of this suggests that the expansion of UI under the American Rescue Plan may be substantially disincentivizing job search.

State Policy Has Been Forced to Countermand Federal Policy

During the peak of the pandemic, some jobs were shut down by state order, or because there was a need to re-configure the workplace to make it safer, or for lack of consumer demand. Many states eliminated the requirement that an individual actively search for work while collecting unemployment insurance.

However, as states now want to return to business, federal policy is increasingly at odds with these goals, and states are searching for ways to mitigate the anti-work substitution effect introduced by federal policy. Some states have reinstated this work search eligibility requirement. 26 states have either withdrawn or will withdraw early from the federal unemployment supplement program. Most of the 26 states are also withdrawing from federal programs that expand benefits to workers who are normally ineligible for unemployment.

Some states are not only adding work search requirements and ending supplemental benefits, but also adopting programs to encourage unemployed individuals to re-enter the work force. Montana was the first state to declare that they would end federal benefits early, but the state is also offering a one-time bonus of $1,200 to unemployed persons who re-enter the work force. Arizona, New Hampshire and Oklahoma soon followed Montana’s example by adopting similar policies. Maine, Connecticut, and Colorado are not withdrawing early from federal benefits programs but are instead offering a similar return to work bonus for unemployed persons re-entering the work force.

The existence of state policies like this is an important marker; it suggests that states are attempting to countermand federal policy that is not working. That is, the “back-to-work” bonuses have a work requirement, and they offset the anti-work requirement imposed by UI. Importantly, this is happening not merely in states whose congressional delegations opposed the American Rescue Plan, but even those whose delegations supported it.

When governments at different levels attempt two offsetting policies like this, it is often a signal that the offsetting components should be removed and the policy simplified.

Workers on the Sidelines Contributes to Overheating

One persistent theme in debate over the American Rescue Plan is whether it would lead to economic overheating, a thesis espoused most boisterously by former Treasury Secretary Lawrence Summers. This thesis is that recent legislation has released too much income, in nominal terms, to Americans, which will result in too many dollars chasing relatively few goods, resulting in inflation.

Over the very short run, that inflation has materialized; many consumers are ready to return to normalcy but some businesses and supply chains are not yet able to keep up with demand, and have raised prices. To the extent that this is simply a one-time scramble by firms to return to previous production levels, that inflation could be transitory, as the Federal Reserve has argued. To the extent that policy inhibits people from working, though, inflation will continue in the longer run.

To help understand this debate more clearly, it is helpful to rely on a simple accounting identity, an economic relationship that is true by definition: growth in nominal gross domestic product (NGDP) is equal to growth in real gross domestic product (RGDP) plus inflation. Inflation, therefore, must come either from too-high NGDP or too-low RGDP.

NGDP is not too high; nominal household incomes were sustained throughout the pandemic, and households are spending approximately the amount, in dollars, that one might have expected them to spend if the pandemic had not happened. NGDP is now only slightly larger than it was in Q4 of 2019.

Instead, the problem is that RGDP is too low. There are several causes for low RGDP, and not all of them can be addressed by policy. Returning to normal after a big economic shock is inherently difficult under any circumstances. However, one likely cause is that there are fewer people working to create RGDP: just 145 million as of May, down from 153 million just prior to the pandemic.

Disincentivising Americans from returning to work suppresses real productive capacity; if spending returns to normal but productive capacity does not, then by definition only inflation can fill the void. 

Conclusion

Unemployment insurance contains a substantial substitution effect against working; it reduces the returns to work at the margin in a way that a universal policy, or a policy with a work requirement, would not. Although advocates argue that UI expansion sustained people’s incomes when their jobs were shut down by circumstances out of their control, those circumstances have now abated, and UI simply constrains their choices with a heavy-handed, counterproductive requirement against taking jobs. States looking to return to normalcy have been offsetting this substitution effect created by the American Rescue Plan, often using the funds given to them by the very same act. The current policy status quo now includes multiple layers of offsetting work requirements and anti-work requirements.

This inefficient state of affairs should be reversed as soon as possible. The federal unemployment supplement and expanded eligibility should be allowed to expire.

Alan Cole
Senior Economist