The unprecedented scale of income support provided to displaced workers due to the COVID-19 pandemic has made the economic downturn less severe than it might have been for millions of U.S. households. Although there has been considerable focus on the July 31, 2020 expiration of the federal government’s initial tranche of more generous unemployment insurance (UI) benefits, less attention has been paid to a looming crisis: the millions of workers who are at risk of losing UI when extended benefits lapse.
Under consideration in Congress are new automatic stabilizers for critical countercyclical programs, similar to what was highlighted in The Hamilton Project and Washington Center for Equitable Growth’s book Recession Ready. We show in this blog post that there is room for Congress to improve the triggers under current law that turn on and maintain the Unemployment Insurance Extended Benefits program to support the long-term unemployed when the labor market is weak. There is also room for states to take full advantage of provisions under current law by opting into more generous benefit extensions. In tandem, these federal and state actions would make it more likely that the U.S. unemployment insurance system doesn’t hinder the COVID-19 economic recovery.
Unemployment insurance payments, including the $600 extra payments and expanded eligibility provided by the CARES Act, totaled $361 billion from March through July 2020, providing households with a substantial boost to income and increasing demand for goods and services. Under current state laws, unemployment insurance benefits are typically time-limited to between 12 and 28 weeks, with the majority of states offering 26 weeks of benefits. Access to regular UI benefits may be dramatically curtailed as early as spring 2021, long before the labor market is expected to have recovered.
For prolonged economic downturns, the UI system includes a program called UI Extended Benefits, or EB, which automatically extends benefits for 6–20 additional weeks using automatic triggers based on particular economic indicators (e.g. state-wide unemployment rates or the share of people collecting UI benefits). Due to the swift and severe nature of the economic shock from the COVID-19 pandemic, these triggers have automatically turned on EB in nearly every state. This prompt EB triggering is unusual; in the past the lack of responsiveness of the automatic triggers spurred Congress to enact emergency UI extensions. Although the CARES Act similarly extended UI benefits, that action was preemptive rather than a reaction to a lack of responsiveness in the triggers: states started automatically triggering on for extended benefits less than one month after the passage of the CARES Act.
The foreseeable, looming crisis, documented below, relates to the potential for EB to lapse prematurely. While triggers turn on EB automatically, triggers also turn off EB automatically, potentially long before the labor market recovers. We expect the labor market recovery to be protracted, leading to more and more workers seeking extended benefits. However, in our estimation existing EB triggers would turn the program off too quickly. We expect that existing EB triggers would turn off UI extended benefits even as the overall unemployment rate remains elevated, jobs are difficult to find, and unemployment spells grow longer. We have seen this before with the EB triggers as currently specified. During the Great Recession, Congress had to step in twice to deal with this problem.
Unemployment insurance is funded by both federal and state taxes, and states maintain a lot of flexibility in administering the program. Some states are not taking full advantage of the EB program as it currently exists; state inaction limits access to benefits to workers that need them and dampens recovery efforts. Under the EB program, state governments are allowed to opt into using additional triggers that qualify them for EB with different economic indicators and allow for up to 20 weeks of benefits through the program. Nevertheless, only 15 states at the moment have opted into all of the optional triggers.
There are two rates that can be used to determine a state’s eligibility for extended UI benefits: a state’s Insured Unemployment Rate (IUR), which is the share of people receiving UI relative to the number of workers covered by the UI system; or, the state’s Total Unemployment Rate(TUR), which is the share of a state’s labor force that is unemployed.
There are three ways in which a state can become eligible for extended benefits. The first trigger is the default: EB is triggered on when a state’s 13-week average Insured Unemployment Rate (IUR) is above 5 percent and rising (specifically, it must be 20 percent above its average in the same period of the previous 2 years). There are two optional triggers that states may additionally put into place so that EB triggers on if any one of the triggers in place are satisfied. First, states can opt into an optional IUR trigger, which allows that when the 13-week average IUR is above 6 percent, extended benefits trigger on (without any requirement that the IUR be rising). As of August 2, 2020, 38 states and the District of Columbia had the optional IUR trigger in place. Second, states can opt into using a TUR trigger: when a state’s three-month TUR exceeds 6.5 percent and is 10 percent higher than the same time period in either of the previous two years, extended benefits trigger on. As of August 2, 2020, 19 states have the optional TUR trigger in place. Additionally, if states are using the TUR trigger, they receive an additional 7 weeks of benefits if the TUR is above 8 percent and rising.
EB triggers will turn off when none of the conditions of the triggers a state has opted into are met. In some circumstances this can happen because IUR or TUR rates are no longer elevated. However, EB can also be triggered off when the IUR and TUR are still high but not rising relative to the previous two years. For example, in the aftermath of the Great Recession, the recovery was protracted and sluggish. States’ IURs and TURs remained elevated long past the end of the recession, stretching into 2012 and 2013. States faced a cliff: Even though they continued to have elevated IURs and TURs, those rates had not risen relative to the peak during, or just after, the Great Recession. Congress stepped in twice, in 2010 and again in 2013, to keep EB in place by extending the lookback period for the IUR and TUR triggers, from two to three years.
Figure 1 shows the 13-week average IURs for states, the basis for the default EB trigger and one of the optional triggers. Because the default IUR trigger is based on both the level as well as the growth of IURs relative to the previous two years, crossing the 5 percent threshold (the dashed red line in figure 1) does not automatically mean that a state has triggered on by the IUR trigger. However, because the onset of the COVID-19 crisis was so rapid, all states that crossed the 5 percent threshold in 2020 triggered extended benefits.
Figure 1 shows that IURs rose rapidly starting in April 2020, triggering EB on as early as April 26, 2020 for Connecticut, Michigan, and Rhode Island. By the end of June 2020 every state and the District of Columbia had exceeded the universal IUR threshold, save South Dakota. Florida, highlighted in orange, which has had relatively low IURs over the past five years (a combination of the strong labor market and limited UI eligibility) was relatively slow to turn on EB in this crisis, but still passed the 5 percent threshold on June 7, 2020. South Dakota (the yellow line) has historically had low IURs, and today is the only state that has not triggered on for EB. Alaska (the purple line) has one of the most cyclical IURs due to the seasonal nature of its economy.
The TUR-based trigger has historically been a more responsive trigger than either of the IUR triggers, at least in part because workers not eligible for states’ regular UI benefits are more likely to be in the pool of workers laid off early in a downturn (figure 2). For example, in late 2009, 46 states and DC were eligible for EB based on the TUR trigger, compared to just 21 states using the IUR trigger. The 2009 Recovery Act increased states’ incentives to use the TUR trigger—and thus make EB more widely available—by fully funding extended benefits. In response, at the peak in 2011, 38 states and DC opted in, up from just 11 states early in the Great Recession.
In contrast to the experience of the Great Recession, the sudden nature of the current economic downturn has meant that states triggered on more quickly using the IUR trigger than the TUR trigger. That is because the IUR is a timelier snapshot of local economic conditions. Unemployment insurance data are released on a weekly basis with only a two-week lag, whereas TURs are monthly and often on a several-week lag. By the end of May 2020, two and a half months into the crisis, 42 states and DC had triggered on through IUR, while only 33 had TURs high enough to activate the trigger (regardless of whether they opted into using it—only 11 states had the TUR option in place in February before the crisis started). As of August 2, 2020, all states and DC met the conditions of the TUR trigger (even if they hadn’t opted into it), and all states except for South Dakota met the 5 percent IUR trigger.
Finally, the optional 6 percent IUR trigger seldom triggers on because historically IURs rarely exceed 6 percent. In 2009, only 7 states would have qualified for EB by it even though 38 states and DC had opted to add it as an additional trigger. Because the onset of the COVID-19 recession was so rapid and extreme, IURs have reached historic levels. As of August 2, 2020, 46 states and DC have IURs above 6 percent, but only 38 states and DC have opted into using the trigger.
That EB triggered on so quickly in this crisis is indicative of the severity of this downturn; but, the nature of Extended Benefits is that they are usually economically important after emergency measures end, when the labor market has been weak for some time and long-term unemployment (those unemployed for 27 weeks or more) has become elevated. In a typical state, an individual will need to exhaust their 26 weeks of regular UI benefits and their additional 13 weeks of emergency benefits (under the CARES Act)—a total of 39 weeks—before starting their extended benefits.
The default IUR trigger, which is typically used by states in the absence of federal funding for EB, is prone to trigger EB off too early when a labor market recovery is slow. As shown in figure 3, long-term unemployment in the aftermath of the Great Recession didn’t peak until May 2010, nearly a year after the recession ended, highlighting that the labor market was still weak and had many more years before it would recover. Because the default trigger requires that the IUR be high and rising, many states would have lost EB coverage in the wake of the Great Recession without intervention from Congress, as IURs remained high, but were not rising relative to the previous two years.
These federal interventions expired at the end of 2014, which meant that the extended benefits all but disappeared even while long-term unemployment was still well above pre-recession levels. As figure 3 shows, the gap between the number of long-term unemployed (purple line) and the number of people receiving emergency UI or EB increased when EB started to trigger off in most states in 2012, when the unemployment rate and long-term unemployment rate remained elevated.
The requirement that the IUR be rising is likely to be especially problematic in the current crisis. The unemployment rate peaked at a record level of 14.7 percent in April; despite expected declines over the next several years, unemployment over that period is likely to remain high by historical standards. Since EB typically triggers off when the unemployment rate is falling, even if it is high, the program could cut off at the worst possible time for unemployed workers. Without Congressional intervention (as during the Great Recession), we are likely to face a similar situation: EB will turn off prematurely, hurting the unemployed and pumping the brakes on the economic recovery.
There are looming cliffs facing the UI EB program. Full federal support of EB, as prescribed in the CARES Act, ends on December 31, 2020. In the absence of significant action by the federal government, states will likely face enormous fiscal pressures. With the abrupt ending of federal support for the UI system and UI EB in particular, states are likely to remove the optional triggers that would have helped EB to continue well into next year. Moreover, states are likely to curtail access to regular UI benefits more broadly, putting downward pressure on state IURs. As states move to using only the default IUR trigger, the continuation of EB will depend on IURs being well above their average over the preceding two years—which is unlikely to be the case.
We worry that by spring 2021, even if the national unemployment rate remains above 8.5 percent, as projected by CBO, EB could start to lapse in states. To avoid that outcome would likely require federal intervention, just as we saw in the years following the Great Recession.
In Recession Ready, Gabriel Chodorow-Reich and John Coglianese offer recommendations for improving the countercyclicality of UI, in part by reforming how extended benefits function. To this end, they propose to make EB fully federally funded, removing fiscal pressures on states that lead them to only use the default IUR trigger. They propose fixes to the triggers themselves, including removing lookback periods from the triggers and creating two new permanent EB triggers: extending eligibility to 60 total weeks when a state’s total unemployment rate crosses 9 percent and to 73 total weeks when a state’s total unemployment rate crosses 10 percent.
For automatic stabilizers to serve their role, they need to trigger on in a timely manner at the onset of a downturn, but also need to continue to support—and not put the brakes on—the recovery. Federal support should flow to states, current triggers should be reformed so as to not slow a recovery, and new triggers should be introduced to ensure benefit extensions in support of the long-term unemployed.
The authors thank Kristen Broady, John Coglianese, and Jay Shambaugh for their valuable feedback. Eliana Buckner, Jennifer Umanzor, and Sarah Wheaton provided excellent research assistance.