Last summer, NELP published a brief asking, “Are State Unemployment Systems Still Able to Counter Recessions?” In it, we examined how states erected administrative hurdles to ensure that fewer workers qualified for diminished benefits. This perfect storm of benefit cuts and access restrictions so destabilized the foundation of the unemployment insurance (UI) program—a program that workers in the United States have historically relied upon for countercyclical support—that we are now witnessing its dysfunction in real time. In 2017, NELP published an analysis describing the policy choices that dramatically reduced both recipiency and access to benefits. In extraordinarily poorly performing states, such as Florida and North Carolina, we have worked to sound the alarm.
Now, we have reached the point where it is too late to prepare states for the massive tsunami of unemployment that is rocking the economy and state UI systems. Fortunately, the federal government stepped in to encourage states to lift some of the more onerous restrictions and add enough money to weekly benefits for four months to ensure that workers maintain some basic buying power and have at least a semblance of economic stability. For many workers and their families, this has helped avert some of the most catastrophic consequences—for now.
However, broken UI infrastructures—particularly in Florida, which created an application process designed to discourage workers from accessing their earned benefits—are wreaking havoc with the claims filing process. As a result, workers who need benefits are unable to access them. While even some states with better computer systems and smarter recession planning are also struggling, states that have designed programs to fail may collapse under the weight of the surge of new claims.
Furthermore, proof of unequal access to UI and the reemployment services attached to it is demonstrated by unemployment rates that are far higher in communities of color—many of which have never recovered from the last recession. Brookings Institution fellow Andre M. Perry released a comprehensive analysis regarding unemployment and Black workers. While the unemployment rate for Black workers hovers around double the rate for white workers, and the rate for Latinx workers also well exceeds that for white workers, it is important to specifically look at communities that are left behind when the economy expands. Perry found that the average unemployment rate in majority-Black cities with populations over 65,000 was 7.1 percent, and Black unemployment in those communities was 12.11 percent. The unemployment rate in Detroit is 17.4 percent, yet policymakers in Michigan cut the maximum duration of benefits to 20 weeks. Communities like Detroit need greater investments but instead receive fewer weeks of UI and therefore less access to reemployment services and connections to job training.
To be sure, even states that worked to establish the best UI systems are having trouble in the face of the historic demand for benefits. With an increase in claims to 3.3 million the week ending March 21, and to 6.6 and 6.8 million in the next two weeks, states are facing unprecedented demand. At the same time, state administrative funding is based on last year’s record-low unemployment levels. Every state is struggling to meet demand.
But the states that we previously examined may well have more difficulty processing claims and paying benefits, while other states that also implemented harmful benefit cuts and restrictions have, at least temporarily, restored benefit durations and are working diligently to get payments to unemployed workers. This brief outlines where we identified challenges in the past, where those challenges persist, and where we find hope for the future.
The most obvious way that states cut benefits was by reducing the maximum duration of benefits to below 26 weeks. Ten states cut duration of benefits. The most recent was Alabama: last June, it cut maximum duration to just 14 weeks. Three states cut maximums from 26 to 20 weeks: Michigan, Missouri, and South Carolina. Arkansas cut maximum benefit duration to 16 weeks. Florida, Georgia, North Carolina, Kansas, and Idaho adopted sliding scales tied to state unemployment rates.
These benefit duration cuts have a disparate impact on communities of color. As of the third quarter of 2019—during a period of record-low unemployment—the average duration of an unemployment spell was 21 weeks. For white workers, a 20-week cutoff would not affect the average worker who remained unemployed for 19 weeks. However, Black or African American workers averaged 25.9 weeks in their unemployment period. A 26-week benefit period would completely cover average unemployment duration for Black workers; any duration reduction therefore statistically harms Black workers more. Since the COVID-19 pandemic began, Michigan, Kansas, and Georgia have all acted to restore 26 weeks of benefits. Other states should do the same.
States also increased administrative requirements for workers, such as by increasing the number of work searches and the documentation required within systems. Some states couched their cuts as an increased focus on program integrity—generally meaning overpayments. In Michigan, for example, the legislature passed measures in 2011 that steeply increased penalties for overpayments. Soon after, a new information technology system went live that flagged between 20,000 and 40,000 Michiganders for fraud with an inaccuracy rate of 93 percent. Massive penalties created serious financial challenges for falsely accused claimants, drove workers into bankruptcy, and some even to suicide. It is important to note here that flawed algorithms also tend to have a racial bias, as has been widely reported by technology writers including Virginia Eubanks and Cathy O’Neill.
As NELP’s policy brief “Ain’t No Sunshine” reported in 2015, Florida’s disastrous legislative changes in 2011 and 2012, combined with its unreliable new CONNECT information technology system, created the perfect storm to reduce benefit access, thwart timely payments, and depress benefit recipiency. Among those changes: cutting duration of benefits from 26 weeks to a first-of-its-kind sliding scale of 12 to 23 weeks (currently 12); adding a (now repealed) 45 question “initial skills assessment”; adding onerous new work search requirements; and requiring all claims to be filed online. Together, these harmful measures have steeply reduced recipiency—that is, the share of unemployed workers receiving a benefit. As a result, as of April 18, the state had only processed 5 percent of the claims submitted since the beginning of the pandemic, with no count of how many workers were unable to file at all.
In 2013, NELP filed a complaint with the U.S. Department of Labor (DOL) challenging the Florida program’s compliance with federal law. In a separate proceeding, DOL’s Civil Rights Division found the program to be uniquely discriminatory and recommended corrections, yet the state continues to lag. As NELP previously reported, just 12 percent of unemployed workers in Florida received any UI benefits in 2015. That number fell to 8.3 percent in the first quarter of 2019, the lowest in the nation, and averaged only 11 percent for all of last year. The processing of claims in Florida failed DOL timeliness performance standards. Approximately half of those lucky enough to receive benefits run out of weeks of eligibility before they find work—the second worst rate of benefit exhaustion in the nation. The state’s average weekly benefit of $252.87 replaces only 38 percent of pre-unemployment average wages. If the state’s 339,150 insured unemployed workers had been paid just the national average benefit of $370.82, that would have meant $40,002,742 in additional benefits paid to claimants in Florida just for the week ending April 4, 2020. Because that most likely represents a low number of people compared to the number actually qualified to receive benefits, we will be updating these findings when it becomes clearer how many workers are falling through the cracks.
North Carolina is one of 10 states that reduced the maximum duration of benefits from what was previously the standard of 26 weeks. The UI legislation passed in 2013 reduced maximum benefit duration from 26 weeks down to a range of 12 to 20 weeks (currently 12), depending on the unemployment rate. The unemployment rate required to hit the 20-week maximum is an astoundingly high 9 percent; an unemployment rate below 5.5 percent triggers a 12-week maximum duration.
Several other changes enacted in 2013 also contribute to North Carolina’s low recipiency rate. For example, the new law increased the waiting period in certain circumstances to receive a benefit. Most states have an initial waiting week, but no other state has more than one waiting week in the case where a worker files more than one claim in a year. The law also removed qualifying good-cause reasons for leaving a job, including family caregiving or leaving because a closure is pending, or to follow a spouse who has relocated for work. It is also harder now for workers whose hours are reduced to qualify for benefits.
The 2013 legislation also dramatically reduced the average benefit paid to North Carolina jobless workers. Before the 2013 changes, the maximum benefit was $525 and was indexed to increases in the average weekly wage; the new law cut the maximum to a flat $350, where it remains today. The 2013 legislation also changed the method for calculating benefit amounts, causing a decline in the share of prior earnings being replaced for claimants, as well as a correlated decline in the average weekly benefit paid to claimants. In North Carolina, the average weekly benefit was $264.70 in the third quarter of 2019. If the 359,151 insured unemployed workers in the state for the week ending April 4 were paid the national average of $370.82, then $38,113,104 more in benefits would have been paid to workers in the state. Of course, that doesn’t take into account the number of workers that could not access benefits. We will provide an update when it becomes clearer how many workers are unable to access benefits.
The percent of jobless workers receiving benefits has plummeted and is now the lowest in the United States. It was 9.1 percent for 2019, dropping as low as 8.3 percent in the third quarter. At 47.1 percent, the share of recipients who run out of benefits in North Carolina before finding work is fourth highest in the nation. Its timeliness in paying benefits is one of the worst in the nation as well, with only 73.9 percent of benefits paid in a timely manner.
Other causes of declines in recipiency relate to new online computer systems that are more difficult for workers to navigate or that may increase improper denials or inaccurately flag workers for fraud. Out of necessity, the state shifted to a new system from an antiquated one. But pre-rollout testing may have been less than rigorous. Many claimants reportedly had great difficulty accessing benefits due to system errors soon after the agency rolled out the new platform.
Indiana passed sweeping UI legislation in 2011. The state reduced benefits by shifting to four quarter averaging for benefit calculations rather than basing benefit amounts on the highest quarter of earnings. This kind of averaging is particularly harsh on people who are marginally attached to the workforce and trying to move toward full-time work, such as workers who are involuntarily part time, or people who accept temporary work assignments as they can find them. Here is a chart detailing the timing of the benefits decline:
Indiana also limited access to UI for several categories of workers, including those whose workplace was temporarily shut down but expect to return to work, workers receiving severance or unused vacation pay, Head Start workers, and workers getting a buy-out in exchange for their separation. As a result of those limitations, as well as several other potential causes, Indiana’s recipiency level has fallen dramatically, even compared to neighboring states. Today, only 18 percent of Indiana’s unemployed workers receive UI benefits, and the $300 average weekly benefit only replaces about 36 percent of previous wages. If the state paid the national average in weekly benefits, the 152,609 insured unemployed workers for the week ending April 4 would have been paid $10,807,769 more in benefits in that week alone. Again, this does not take into account workers who are unable to access benefits; we will provide updates once the data is available.
Meanwhile, the average tax rate among all employers subject to UI taxes in Indiana is 1.77 percent as of the second quarter in 2019. This is below the national average of 2 percent, and well below the 2.6 percent rate in Illinois, as well as Kentucky’s 2.1 percent, Ohio’s 2.27 percent, and Michigan’s 3.19 percent. For Indiana, that translates to $204 on average per worker per year, while in Illinois it is $364, Kentucky $209, Michigan $352, and $247 in Ohio. There was plenty of room within the taxation structure to increase tax levels to raise benefits before this public health crisis.
Arizona passed several measures that, taken together, reduced access to benefits. In 2012, the state dramatically raised the amount of money a worker must earn to qualify for benefits and extended the amount of time in which eligibility determinations can be made. In 2016, it added a disqualification for failing a drug test, added that workers have to search for work at least once per day on four separate days rather than three per week, and said that workers are independent contractors if they ever sign a declaration saying that they are. (For years, NELP has led work to shine a light on employers’ requiring workers to sign statements against their interests in order to work.) As a result, only 11 of every 100 unemployed workers in Arizona are able to receive a UI benefit. Only 83.8 percent of claims are processed in a timely manner, according to the DOL, which considers less than 87 percent to be unacceptable. The average weekly benefit of $233.78 only replaces 36 percent of prior wages, on average. If the 116,782 insured unemployed workers in the state in the week ending April 4 had received just the national average benefit, that would have put an additional $16,003,805 in the pockets of Arizonans that week alone. Of course, that number does not include workers who either couldn’t file or didn’t receive their benefits that week; we will continue to monitor the data and provide updates as it becomes available.
Wisconsin is another state in which lawmakers chose to erect barriers to UI benefits. The state increased work search requirements from two per week to four per week, and five for workers who are unemployed for seven weeks or more (if the agency makes a determination that the claimant “is placing unreasonable limitations as to salary, hours, or conditions of work in accepting new work or is not engaging in work search efforts as would a prudent person who is out of work and seeking work”). It also established a voluntary system for employers to report failed drug tests to the state agency in order to more easily disqualify workers. It reduced wage levels that qualify as suitable work, increased penalties for fraud determinations, and eliminated good cause for failure to appear weekly to report information about benefits and recipients’ demographic characteristics.
While then-Governor Scott Walker touted these reforms and the effect they had on UI trust fund solvency, Black workers were criminalized by the system while the state’s Black/white unemployment gap is among the worst in the nation. While 29 percent of unemployed workers receive benefits—a little above average—it marks a steep decline from the 50 percent recipiency rate in 2007. The average weekly benefit of $323.90 replaces 43 percent of previous income.
In 2011, South Carolina passed legislation to reduce maximum duration of benefits from 26 weeks to just 20 weeks. In 2015, the state added a provision disqualifying workers who fail an employer-provided drug test. In 2016, it increased policing of work search claims and increased the disqualification of benefits following a negative finding from 8 weeks to 19 weeks. And, in 2017, it began requiring online documentation of work searches.
South Carolina routinely falls to the bottom in several key measures of program effectiveness. The recipiency rate in the state is just 16 percent. More than 38 percent of workers run out of benefits before they find a job—the 12th highest exhaustion rate in the United States as of the last quarter of 2019. South Carolina’s average weekly benefit amount of $273.74 replaces just 40 percent of pre-unemployment income. If that benefit were just raised to the national average weekly benefit, the 125,376 insured unemployed workers in the state in the week ending April 4 would have received an additional $12,171,502 in benefits that week alone. This does not take into account the workers unable to access the system; we will monitor and provide updates as the data becomes available.
Louisiana has historically ranked near the bottom for several of our metrics, and that was the case even before the last recession. It is one of the states that did not enact major legislative changes to UI in the wake of that recession. With just a 12 percent recipiency rate last year, and a paltry average benefit of $215.99—which only replaces about 33 percent of average wages—Louisiana’s is a system that has long failed workers. If the 213,338 insured unemployed workers in Louisiana in the week ending March 28 had been paid the national average benefit, they collectively would have received an additional $33,031,122 in benefits that week alone. These figures do not account for those workers left out of the system; we will monitor that data and provide updates as information becomes available.
Tennessee is one of the states that did not reduce core benefits or duration, but instead began making the process to apply for UI far more difficult. The state added language about failing drug tests, eliminated the alternative wage base (which had made it easier for workers with uneven work histories to qualify), eliminated dependent allowances, and reduced how much a job must pay to be considered an offer of “suitable work” that a worker would need to accept or lose benefits. The state’s recipiency rate is only 13 percent, and the average weekly benefit of $240.13 only replaces 37 percent of average wages. If the state’s 206,622 insured unemployed workers had received the national average benefit in the week ending April 4, workers in Tennessee would have received an additional $27,003,429 in benefits that week alone. These data do not include workers who are falling through the cracks and unable to access the system; we will provide updates once these numbers are available.
Georgia is a good example of a state that took an axe to benefits in the wake of the Great Recession, but now appears to be recognizing its mistake and has begun restoring benefits. In 2012, the state reduced maximum benefits and then in 2014, cut duration to 14 weeks (with a sliding scale of 14 to 20 weeks). It also cut workers out of the program and increased penalties for overpayments. As of 2019, recipiency was down to just 14 percent. Now, however, in the midst of the COVID-19 pandemic and the associated economic contraction, Georgia Governor Brian Kemp signed executive orders restoring duration to 26 weeks and exempting the first $300 of weekly earnings from counting against UI eligibility. These encouraging moves show that it is possible to quickly improve state UI systems on an emergency basis.
Michigan not only cut benefit duration from 26 to 20 weeks under the previous administration, but it also adopted sweeping new fraud rules allowing for additional penalties. This is another example of how important it is to have thoughtful, claimant-informed modernizing of IT systems. The new benefit system adopted by the state relied on a faulty algorithm to flag claimants for fraud, falsely charging between 20,000 and 40,000 workers for fraud and wrongfully “recouping” massive amounts of money from them. The problem was the program was 93 percent inaccurate. Instead of receiving benefits they had earned, people who were truly unemployed and needed benefits were charged money they didn’t have—driving some into bankruptcy and even, reportedly, to suicide.
The new administration’s approach to UI is a 180-degree turnaround. Michigan Governor Gretchen Whitmer signed an executive order restoring benefit duration to 26 weeks. She also appointed a long-time claimant advocate, Steve Grey, to head the agency. By all accounts, his commitment to pay benefits fairly to all workers losing their livelihoods in this pandemic is to be commended.
NELP remains concerned about any state that has seen a sharp decline in UI benefit recipiency since before the last recession. These declines generally represent lack of access, and every state benefit cut or limitation to access is a story about a state making decisions to limit access to an earned benefit to save money. This is a story of why we can’t ignore systems for decades and then throw money at them only when there is an emergency. We cannot lose this lesson about what happens when we let important systems fall into disrepair during economic expansions.