The expansion of unemployment insurance (UI) benefits in 2020 and 2021 through the COVID-19 pandemic has come in fits and starts, with benefits increasing at times to levels well in excess of what workers were earning at their jobs and at other times dropping drastically without regard to changing labor market conditions. We need a better approach.
With unemployment insurance, timing is everything
The goal of unemployment insurance is to provide assistance to workers when they need it. Unemployment insurance (UI) provides a lifeline to workers when they lose their jobs. It is a state-run system with federal oversight. During recessions the federal government has frequently stepped in to enhance what states provide to better enable workers who lost jobs manage their income loss.
Ideally, benefit levels should increase for only a brief period when the labor market is at its nadir—when jobs just are not available and there is very little disincentive effect of UI—but supporting consumption and maintaining aggregate demand is vital.
Ideally, expansion of UI should be well-timed to the economic contraction, starting quickly when a recession hits to help those workers and to prevent further deterioration of the economy. At the same time, any enhanced level of aid should wind down as the economy recovers to maintain work incentives that will allow that economy to grow as rapidly as possible.
These benefits, though, should not linger through the end of the recession. As the economy is regaining steam and jobs begin to come available (as limited as they may be), incentives to look for work need to be restored.
The size of UI benefits has fluctuated in the current recession and the timing of the fluctuations was determined to a large extent by the political process. The CARES Act, enacted in late March 2020 included several provisions related to the unemployment insurance system augmenting benefit levels, expanding the duration of benefits, and making UI available to workers who are not typically covered.
In particular, it temporarily boosted state UI benefit amounts through the Federal Pandemic Unemployment Compensation (FPUC) program, which made an additional $600 in weekly benefits available to all individuals eligible for state UI benefits beyond the amount for which they would normally receive. These benefits were authorized through July 31, 2020. Just after those benefits expired, President Donald Trump signed an executive order to pay Lost Wage Assistance of $300 out of FEMA funds until that ran out in September.
Congress made subsequent ad-hoc adjustments, providing additional benefits at some points, but not others, and at various levels (see chart). Although the initial expansion of benefits was well timed with respect to the start of the recession, the subsequent policy changes are largely disconnected from economic conditions.
Benefit levels should be proportional to lost wages. UI benefits are typically lower than the wages that they are replacing. However, at the start of the coronavirus pandemic, the federal government sought to fully replace workers’ lost wages since workers were stopping work due to the shutdowns motivated by public health. But antiquated state technology infrastructure required that the enhanced benefits be provided in constant dollar amounts to all recipients rather than individually tailoring beneficiaries’ “replacement rate” (weekly benefit/usual weekly wage).
As a result, many workers have been better off on UI than at their previous jobs. For example, for a full-time, full-year worker in California earning $20 per hour pre-pandemic, their replacement rate jumped from 50% to 125% to 87.5% to 50% and then settled back at 87.5%. For lower-wage workers, the replacement rates were even higher during the periods in which the additional benefits were available. Providing workers with benefits greater than their lost wages is a disincentive to work, particularly when those higher benefits extend into the early period of economic recovery.
A federal increase of UI benefit amounts over the state benefit in response to a recession is new. Traditionally, the main federal policy lever used to enhance UI benefits during a recession is to increase the length of time workers can receive benefits—but the timing of this type of enhancement has also faced issues in the past. The permanent, joint state-federal extended benefit (EB) program automatically extends the number of weeks of benefits available when a state’s unemployment rate crosses a statutory threshold.
Although the EB program has a formal system to trigger the increase that does not depend on the political process, these automatic triggers have been criticized for years for being ineffective, often failing to turn on even when economic conditions warrant it. This led the federal government to routinely implement ad-hoc adjustments to the maximum duration of benefits whenever a recession occurs.
Importantly, extended benefits does not help individuals or the economy until regular UI benefits have been exhausted (in the majority of states, this is after 26 weeks). Increasing the level of UI benefits is most useful as soon as the labor market crashes.
The role of the UI system during a recession is to provide income support to those who lose their jobs and to help stabilize the economy by increasing spending. A two-tiered system of enhanced UI benefits during a recession could accomplish both goals better than our historical approach of relying exclusively on extended benefits.
Ideally, benefit levels should increase for only a brief period when the labor market is at its nadir—when jobs just are not available and there is very little disincentive effect of UI—but supporting consumption and maintaining aggregate demand is vital. In the early period of COVID, the benefit increase did just that.
These benefits, though, should not linger through the end of the recession. As the economy is regaining steam and jobs begin to come available (as limited as they may be), incentives to look for work need to be restored. Maintaining the traditional extended benefit system for longer, though, will continue to lessen the blow for workers who still need some form of longer-term support.
Using two tiers of federal assistance—enhancing state benefits and extending the period that benefits are available—would require using two different triggers.
Implementing a two-tiered system requires adopting two sets of automatic triggers to turn on and off each type of benefit. Increasing benefits early on means triggering off increased benefits as soon as there are clear signs of a strengthening labor market. A delayed trigger makes sense for extended benefits to provide additional long-term support for unemployed workers who need it.
Relying on legislative or executive actions to turn each set of benefits on and off makes it unlikely that the timing is most appropriate to the economic conditions. These ad hoc program changes have not worked well with extended benefits; the timing is unlikely to improve after adding an additional layer of complexity for a second tier of benefits.
That said, trying to identify rules that fit the needs of both a pandemic-induced labor market crash and a more prolonged downturn, as in the Great Recession, is difficult. The devil is in the details in arriving at appropriate triggers for this two-tiered system.
It is beyond the scope of this analysis to propose detailed alternatives. We note, though, that analysts have proposed, and some states have adopted, alternative triggers for the extended benefit system that are designed to address some of the past deficiencies. Those proposals could be modified to fit the needs of a two-tiered system.
If it were possible to get an accurate and timely measure of vacancies, such information also could be useful in setting the triggers to turn the increased benefits on or off. Recent reports of labor shortages in many industries indicate that such a measure would likely have triggered off increased benefits by now.
What this Means:
Now that we have implemented increased benefits once, it seems likely to us that they will be proposed again in the next recession (which will hopefully be a long time from now). Given this likelihood, it makes sense to think hard now about how to modify the system to do a better job of timing the benefit increase. Ideally, there will be time (and money) to upgrade state computer systems before the next recession, providing the ability to better tailor increased benefits to the individual worker’s situation.
While designing the appropriate automatic triggers will need further research, no matter the specific system of triggers that are implemented, it should be an improvement over our reliance on ad-hoc measures, which result in inadequate policies.
Patricia M. Anderson is a professor of economics at Dartmouth College. Her research interests include food insecurity, child obesity, and social insurance programs. Phillip Levine is a professor of economics at Wellesley College. His research focuses on social issues such as abortion and teen childbearing, and on evaluating policies designed to improve the well-being of disadvantaged youth.