One year after the onset of the coronavirus recession, the U.S. labor market is still a long way from its February 2020 employment levels, but saw important job gains last month. According to the latest Employment Situation Summary by the U.S. Bureau of Labor Statistics, the U.S. economy in February added 379,000 nonfarm payroll jobs—the greatest month-over-month gain since October of last year.
Today’s release also shows that between mid-January and mid-February the overall unemployment rate fell from 6.3 to 6.2 percent, with 50,000 workers re-entering the U.S. labor force. Across sectors, last month’s job gains were concentrated in leisure and hospitality, which added 355,000 jobs. Yet data on employment changes over the entire year show that the downturn remains especially hard for this sector and for service-providing industries in general. (See Figure 1.)
Figure 1
The economic pain brought on by the downturn continues to fall heaviest on some groups. The jobless rate stands at 9.9 percent for Black workers, at 8.5percent for Latinx workers, at 5.1 percent for Asian American workers, and at 5.6 percent for White workers. Disaggregating the data further shows that over the past 12 months, net job losses have been greatest for Black and Hispanic women and men—groups for whom employment declined by 9.7 percent and 8.6 percent, respectively. (See Figure 2.)
Figure 2
Hispanic men’s experience during the coronavirus recession
For Hispanic men, overall job losses are less severe than for women workers or Black men. Yet their experience during this recession also highlights important challenges they face in the labor market. For instance, Hispanic men are overrepresented in jobs that cannot be done from home. Despite accounting for about 8 percent of the U.S. workforce, these workers represent about a quarter of the construction workers and about 1 in 5 workers in the mining sector. In part as a result of their industry and occupational distribution, Hispanic men are facing risks associated with in-person work and have been more likely to experience joblessness than their White, non-Hispanic peers—a trend that risks entrenching longstanding inequities between the two groups of workers.
Consequently, as of the last quarter of 2020, Hispanic men were earning the lowest median weekly earnings of any other group of men, and just a bit higher than the weekly earnings than for Hispanic women.
Even though earnings disparities between Hispanic men and their White non-Hispanic peers are often attributed to differences in education, these pay disadvantages persist even among workers with the same level of education. A recent study shows that whereas 6 percent of White men with an advanced degree hold low-wage jobs, 13 percent of Hispanic men do. An analysis by the Economic Policy Institute shows that Hispanic men make about 15 percent less than White men who live in the same geographic region and have the same level of education and work experience. That gap, moreover, remains relatively unchanged since the early 1970s.
Researchers also find evidence that Hispanic men—and especially those who also are immigrants—are more likely to take low-wage and low-quality jobs, since they often lack the private networks and access to social insurance programs that would allow them to engage in longer job-search periods. Consistent with this evidence is that Latinx workers who are part of a labor union experience a particularly large pay boost. On average, workers covered by a union contract are paid 11 percent more than their nonunionized peers. Among Latinx workers, however, the wage premium associated with being represented by a union contract is more than 20 percent.
Yet research by Jake Rosenfeld of the University of Washington-St. Louis and Meredith Kleykamp at the University of Maryland, College Park also finds that Latinx immigrants are less likely to be part of a union than U.S.-born Latinx workers, suggesting that stronger networks and U.S. citizenship might protect workers against hostile responses to unionization efforts.
Conclusion
As the U.S. economy went into a pandemic-driven tailspin one year ago, almost 21 million workers lost their jobs between mid-March and mid-April alone. In February 2021, the U.S. labor market is short 9.5 million jobs relative to February 2020. These losses remain starkest for Black and Latina women, and other vulnerable groups of marginalized workers, highlighting the importance of policy in setting the groundwork for an equitable economic recovery.
Above all, policymakers should prioritize the enforcement of existing labor law. Even though this should be a priority for policymakers during booms as well as contractions, research shows that wage theft rises and falls with the business cycle—as recessions hit and the jobless rate rises, so does the share of workers who suffer a minimum wage violation.
Another way to tackle these U.S. labor market inequities is to lift the federal minimum wage, now frozen at $7.25 an hour for more than a decade. More than 40 percent of U.S. workers make less than $15 dollars per hour. In the food services industry, where Latinx workers make up more than a quarter of all workers, a whopping 78 percent of workers earn less than $15 dollars an hour. A large share of U.S. workers and an even larger share of women workers and workers of color would receive a much-needed pay boost should the federal minimum wage increase, helping drive a faster and more equitable economic recovery.
As federal policymakers debate President Joe Biden’s proposed $1.9 trillion American Rescue Plan, critics now argue that it will “overshoot,” meaning it would boost economic activity beyond the desired level. The economic track record of the past 10 years shows these concerns are misplaced. Undershooting the policy response would be a far more dangerous prospect and could lead to a repeat of the slow and inequitable economic growth that followed the previous U.S. recession.
After the Great Recession of 2007–2009, then-President Barack Obama and the U.S. Congress passed an insufficient stimulus, then pivoted too quickly to debt reduction. This was a crushing mistake that left many U.S. workers and their families stuck in the doldrums for years, facing stagnant, or even declining, incomes. The slow and uneven recovery was the direct result of these policies. This time, federal policymakers would be wise to err on the side of doing too much rather than too little.
A new data series from the U.S. Department of Commerce’s Bureau of Economic Analysis—its Distribution of Personal Income series—shows just how devastating this pattern was for most U.S. workers and their families. The agency’s new data series charts the distribution of aggregate growth in Personal Income and Disposable Personal Income, and reports how much growth accrues to the bottom, middle, and top of the income distribution. A new release in December 2020 updates the dataset through 2018, capturing most of the recovery from the Great Recession.
The lesson from the data is clear. The policy response to the Great Recession left millions of low- and middle-income families struggling while wealthy families saw significant additions to their incomes. Consider the cumulative growth in disposable personal income of high- and low- or moderate-income Americans. High-income households in the top 10 percent initially suffered a steep drop in income at the onset of the Great Recession due primarily to the collapse of income from assets such as stocks and bonds, which cratered early in the economic downturn, alongside business income losses. Low- to middle-income households in the bottom half of the income distribution did not initially suffer as dramatic a fall. (See Figure 1.)
Figure 1
One reason for the less dramatic shock to income seen in Figure 1 among households in the bottom 50 percent was because disposable personal income incorporates transfers from the federal government to households, so losses in this group were partially compensated for by rising Unemployment Insurance payments, Supplemental Nutrition Assistance Program benefits, and other government benefits. But this group then became mired in years of stagnant, or even declining, income as these benefits ended amid a still-tepid economic recovery, and did not experience substantial income gains until 2015.
By comparison, households in the top 10 percent of income recovered almost immediately after the end of the Great Recession and ended 2018 up 22 percent, compared to 2007. Importantly, the jump in top incomes in 2012 and steep decline in 2013 seen in Figure 1 does not represent a real decline, but rather is the result of households retiming their income to occur in 2012 so they could avoid rising top-income tax rates in 2013 as a result of the expiration of high-income tax cuts first enacted during the George W. Bush administration.
This issue brief breaks down the new BEA data series to discern why the U.S. economic recovery from the Great Recession was so weak and uneven, and what lessons can be learned from the policies of economic austerity enacted by Congress too swiftly after the end of the previous recession. Those lessons, briefly, are that policymakers should enact strong fiscal stimulus today and should invest in the development of distributional data so they can understand in real time how broadly shared the coming economic recovery is.
What happened to families’ income during the Great Recession?
The Bureau of Economic Analysis’ new data series provides a holistic look at how incomes changed for households in the bottom 50 percent of income during the Great Recession and the subsequent recovery. Personal Income consists of wages, transfers from the government, rent income, interest and dividends, and income from individually owned businesses and businesses organized as partnerships. Americans at the bottom and middle of the income distribution derive a significant amount of their income from the first two of these, wages and government transfers.
Government transfers initially blunted the impact of the 2007–2009 recession for low- and middle-income Americans. But once the initial stimulus wore off, transfers declined, and wages languished, providing positive but small contributions to income. (See Figure 2.)
Figure 2
Households were bolstered in 2014 by the implementation of the Affordable Care Act. But wages continued to stagnate until around 2016. It wasn’t until the tail end of the recovery from the Great Recession that the labor market started to really work for people in this group.
By contrast, households in the top 10 percent by income recovered quickly and were helped along by a mix of income sources, such as dividends from stocks, interest from bonds, rising business income, and much stronger wage growth, than what other Americans experienced. (See Figure 3.)
Figure 3
The result is that the top 10 percent of households routinely flirted with 6 percent income growth in the years after the Great Recession because they did not rely on any single source of income, but rather combined good wage growth with resurging business income and returns on their financial assets throughout the recovery period.
What’s more, this data series may understate the resurgence of top incomes after the Great Recession. The dataset-based definition of Personal Income notably excludes capital gains—the profits made by selling financial assets such as stocks. Income from capital gains is notably concentrated at the top of the income distribution.
The lesson: More economic stimulus is needed today to spur a more sustained economic recovery
What happened after the Great Recession that led to 5 years of mediocre economic progress for most Americans? Many factors may have contributed, but 2010 was famously the year that the United States—and most of the rest of the world—pivoted to austerity politics, slashing budgets in response to the supposed largesse of economic stimulus enacted in 2009. And that stimulus was comparatively small; lawmakers in 2020 spent more money over a much shorter span of time to prop up the economy.
Fast forward to today. The ongoing coronavirus recession is punishing for most families. Unemployment remains high, and it is concentrated among low- and middle-income earners. Yet, according to Opportunity Insights, high-wage employment is actually rising. And just like during the Great Recession, financial assets initially suffered but are now going up. The stocks that comprise the S&P 500 index are now up about 14 percent over their pre-pandemic high in February 2020, providing a significant boost to high-wealth households. More than half of all stocks are owned by the wealthiest 1 percent of U.S. households, and the top 10 percent own more than 80 percent of all stocks.
In recent months, aggregate job growth slowed considerably and has been especially weak for women and workers of color, strongly suggesting the initial signs of recovery amid the coronavirus recession may not last, absent more government intervention. And federal policymakers don’t yet know how low- and middle-income households will fare when eviction and foreclosure moratoriums are lifted—two economic cliffs that could impose sudden hardships on a large number of Americans. Nor is there a full picture of how this pandemic is impacting incomes yet.
The $1.9 trillion stimulus package that is currently working its way through Congress would provide significant support to incomes for low- and middle-income families. President Biden’s package includes enhanced Unemployment Insurance that will help sustain these families through the recovery, stimulus checks that will help them pay their bills now and will boost aggregate demand, and aid to states that will help them retain essential workers and open schools. These programs, and others included in the bill, will boost the incomes of workers outside the top 10 percent and help the economy avoid the pitfalls of the Great Recession recovery.
The BEA data on the previous economic recovery show the dangers of an inadequate policy response. Indeed, there are two specific lessons policymakers should take from the new BEA data.
First, Congress needs to pass a strong relief package and should continue to spend if economic indicators suggest that the U.S. labor market is not recovering quickly enough. After all, the overarching goal of federal policy is to get back to a tight labor market as quickly as possible and to avoid economic scarring that will permanently impact the economic prospects of the young and the economically vulnerable.
Second, these new distributional data provide actionable insights on how families fare during recessions and recoveries. To manage an effective recovery, Congress and the Biden administration need to understand how the economy is functioning in real time.
Right now, we are largely reliant on imperfect indicators that only hint at how families are faring. Policymakers know unemployment is up and aggregate Gross Domestic Product is down. But if the administration and Congress had access today to the kind of disaggregated statistics from the BEA running through 2018, then they could act with full knowledge of who was hurt most by the recession and who is being left out by the recovery.
Just imagine if federal policymakers could know in detail now just how incomplete the recovery from the current recession is today. If they did, then they could consider further economic stimulus with a clear target for how much stimulus is needed and how it should be targeted. Alas, policymakers are always seeing the economy imperfectly because data are often not available until long after the fact.
Piloting the U.S. economy through recessions and recoveries will be easier if policymakers know how the economy is faring for different kinds of families across levels of income, geography, race and ethnicity, and gender. Work by the BEA, the U.S. Department of Labor’s Bureau of Labor Statistics, and the U.S. Census Bureau to provide more granular economic data series should be considered an essential part of our economic crisis toolkit and funded accordingly.
Appendix
A note on data-analysis methods
The BEA data series provides total shares of Personal Income for each income decile of U.S. households. It reports, for example, that the bottom decile held 2.17 percent of all Personal Income in 2018. From this, it is easy to calculate growth rates for the various deciles, but there is no population adjustment. Because the dataset uses households as the unit of analysis (BEA starts with the Current Population Survey, which surveys households), I adjusted for the number of households in the economy to create Figure 1.
This “population” adjustment does not necessarily lead to a data series that closely tracks the per capita Personal Income series. Households sometimes grow faster or slower than the overall U.S. population does. Over the years analyzed in this issue brief, households generally grew faster than the population, leading to lower rates of growth across the board. Per capita Personal Income shows much stronger income growth in the aftermath of the Great Recession than my per household measure does, for example. In other years, the effect is reversed: The Great Recession was deeper in per capita Personal Income than in per household Personal Income.
To create Figures 3 and 4, I used Household Income rather than Disposable Personal Income, because it is easier to reconcile this measure with the distributional tables that the Bureau of Economic Analysis provides. At the recommendation of the agency, I did not adjust for household growth in these figures.
Notably, the BEA dataset is cross-sectional rather than panel. That is, when it refers to the top 10 percent, it does not mean the same group of people in each year but rather the 10 percent of people in the U.S. economy who happen to have the highest incomes in that particular year. In a given year, a large number of people may move up or down the distribution and find themselves in new groups.
To strengthen the federal government’s response to the coronavirus pandemic and the resulting economic downturn, President Joe Biden seeks to enact his American Rescue Plan to stop the spread of the coronavirus and COVID-19, the disease caused by the virus, provide relief to devastated families and businesses, especially those headed by Black and Latinx Americans who are bearing the brunt of the economic pain, and stimulate the U.S. economy as it struggles to recover from recession.
President Biden’s far-reaching proposal includes a number of critical enhancements to key programs that support families and buttress consumption. But there is one problem: The Biden plan includes arbitrary cut-off dates that halt emergency programs regardless of the state of the U.S. economy. In a few instances, the plan allows for the possibility of automatic extensions, but it contains no specifics. No matter how high unemployment remains when these programs expire, Congress would be forced to go through yet another challenging and unpredictable legislative negotiation to continue them.
As we saw this past July, after some the provisions in the Coronavirus Aid, Relief, and Economic Security, or CARES, Act expired, and then again in December, when the remainder of those provisions ended, premature expirations of aid leave suffering families uncertain about putting food on their tables, exacerbate preexisting challenges facing program administrators, and undermine economic recovery as consumers cut back spending in fear of what might come next. Unemployment Insurance beneficiaries, for example, suffered from significant delays and lapses over the course of 2020 due to the various congressional stalemates, as well as the UI system’s chronic problems.
There is a solution to this problem. The next coronavirus relief bill can be written to keep critical benefits in place until they are no longer needed. So-called off-triggers that depend on economic data can ensure that benefits phase out as the economy recovers, not on a random date. There are a number of possible mechanisms that can serve as off-triggers, winding down enhanced benefits based on the state of the economy in individual states and nationally.
Automatic stabilizers
Programs that expand in a recession, such as Unemployment Insurance, are known as automatic stabilizers, as they help to stabilize a weak economy automatically. But policymakers and economists alike have seen time and time again that in past recessions—and the current one—the current structure of these programs provides too little too late to either stop the U.S. economy from declining or revitalize it so that workers are reemployed, wages rise, and families can feel secure. So, Congress routinely steps in to boost benefits—especially for programs such as Unemployment Insurance and the Supplemental Nutrition Assistance Program, formerly known as food stamps—and lengthen their duration. But the partisan wrangling that often afflicts Congress usually delays assistance at the beginning of a recession, and the assistance falls short. Then, Congress struggles to extend benefits, often letting the programs lapse for spells or prematurely pulling the plug altogether.
This was certainly true during and after the Great Recession of 2007–2009. As Rep. Don Beyer (D-VA) recently noted, “In the Great Recession, they re-upped Unemployment Insurance 13 different times. That’s 13 political battles between the House and the Senate, needing 60 votes and the president.” The result of inadequate and unreliable support from the federal government was a long, drawn-out recovery in which the U.S. labor market took years to recover its vitality and the economy was characterized by lagging employment, wage growth that was sluggish at best, and continuing anxiety and financial challenges for U.S. workers and their families.
Recession Ready
What was lacking in the Great Recession, and what is needed now, is a mechanism for ensuring that benefits remain available as long as they are necessary, with no need to wait for Congress to act. There are a number of solid, evidence-based ideas for such a mechanism. In 2019, the Washington Center for Equitable Growth and The Brookings Institution’s Hamilton Project published Recession Ready, a book of essays with comprehensive plans to strengthen automatic stabilizers, including recommendations for automatic triggers to activate supplemental benefits to families and state governments, and automatic triggers to turn them off.
Before the next recession, it will be important to establish on-triggers that automatically implement supplemental benefits without the need for new legislation when economic data make it clear that a recession is imminent or already occurring. But at this moment, what is important is that off-trigger mechanisms be included in the COVID relief legislation now being debated in Congress. The House and Senate have already seen the introduction of several bills that contain such mechanisms. Most of these proposals focus on the level of unemployment or on health conditions, either nationally or by state, and vary in technical ways. Fundamentally, however, they are all intended to sustain spending until job markets have sufficiently recovered.
Options for off-triggers
There are a number of ideas for ensuring that unemployed workers do not prematurely lose any of the special Unemployment Insurance benefits enacted by Congress when, on March 14, they would expire under current law, including extended benefits, supplemental payments, or the inclusion of workers who are usually not eligible for the program. These same ideas could be considered for when President Biden’s American Rescue Plan would end them at the end of September. Essentially, these off-triggers retain extended benefits until the economy improves to a sufficient level. (See Table 1.)
Table 1
Sens. Michael Bennet (D-CO) and Jack Reed (D-RI) and Reps. Beyer and Derek Kilmer (D-WA) introduced legislation in 2020 to extend and increase unemployment benefits. The legislation divides the states into six tiers based on their rate of unemployment and reduces or increases various types of benefits based on the tier where each state, as well as the District of Columbia, Puerto Rico, and other U.S. territories, falls over time, with additional benefits phasing out entirely when the 3-month moving average of their individual unemployment rate drops below 5.5 percent and the 3-month moving average of the national unemployment rate is below 5.5 percent and has been declining for 2 consecutive months.
They also propose to mix in a trigger based on public health, by supplementing weekly unemployment benefits by $600 (as opposed to the current $300 weekly supplement and the $400 weekly supplement proposed by President Biden) until the end of the COVID-19 National Emergency Act declaration first issued by former President Donald Trump in March 2020. The benefit would then decline to $450 for 13 weeks, and then to $300 or $200 depending on the state’s tier, ending completely in a state when the state’s 3-month moving average unemployment rate sinks below 7.5 percent and the 3-month moving average of the national unemployment rate is below 5.5 percent and has been declining for 2 consecutive months. The bill allows for the resumption of the $600 payment if the emergency declaration is extended or a new COVID-19 emergency occurs.
Another measure tied to unemployment benefits is a Recession Readyproposal by Gabriel Chodorow-Reich of Harvard University and John Coglianese of the Federal Reserve Board. The authors add two triggers to the normal extended Unemployment Insurance triggers to lengthen benefits for longer periods: to a total of 60 weeks (including the original 26 weeks) when a state’s unemployment rate crosses 9 percent and to 73 weeks when a state’s unemployment rate crosses 10 percent. This trigger is based simply on state unemployment rates.
Unemployment rates are also the basis for off-triggers for other proposed automatic stabilizers. Another Recession Ready proposal put forward by former Washington Center for Equitable Growth Director of Macroeconomic Policy Claudia Sahm suggested a system of annual direct stimulus payments during recessions. They would be triggered by a sustained rise in unemployment and would continue annually if the national unemployment rate stayed 2 percentage points above the level at the time of the first payment. Using the $2,000 payments enacted early during the pandemic as an example, the off-trigger would be achieved when the unemployment rate falls below 5.5 percent, which is 2 points higher than the rate in February 2020.
Similarly, Reps. Madeleine Dean (D-PA) and Matt Cartwright (D-PA) recently introduced the Payments for the People Act, which would send most families quarterly rather than annual payments. The amount would be based on the national level of unemployment, and their off-trigger is simple and not set by a formula. Payments would just continue until the national rate of unemployment falls below 5.5 percent for 2 consecutive months.
And then, there is legislation introduced by Sen. Bob Casey (D-PA) based on the Recession Ready Medicaid and Child Health Insurance Program proposal of former Chair of the President’s Council of Economic Advisers and current Equitable Growth Steering Committee member Jason Furman, Matthew Fiedler of The Brookings Institution, and Wilson Powell III of Harvard University. Rep. Susie Lee (D-NV) introduced a similar measure in the House, and the Take Responsibility for Workers and Families Act—a coronavirus relief measure introduced by House Democratic leadership in 2020—also contained a similar provision.
Both the legislation and the program by Furman and his co-authors establish on- and off-triggers related to state unemployment rates. They would make a state eligible for increased support in any quarter in which its unemployment rate exceeded a threshold level, set by a formula designed to ensure that it is targeted to serious downturns rather than small fluctuations in unemployment. The Furman proposal would make a state eligible for relief in any quarter in which its unemployment rate exceeded the 25th percentile of the distribution of the state’s unemployment rates over the preceding 15 years, plus 1 percentage point. The percentile measure is intended to approximate the state’s unemployment rate at full employment. This additional percentage point allows for normal fluctuations and ensures that the trigger is targeted to serious economic downturns.
The Casey and Lee bills differ somewhat in that the trigger is unemployment exceeding the lower of either the 20th percentile of the distribution of the state’s quarterly unemployment rates for the most recent 60-quarter period, plus 1 percentage point, or the state’s average quarterly unemployment rate for the previous 12-quarter period, plus 1 percentage point. Under all of these formulas, the amount of additional federal support would depend on the extent to which a state’s unemployment rate exceeded the threshold.
The increased support would continue based upon quarterly assessments of each state’s unemployment rate and would phase out using the same triggers. Effectively, as a state’s unemployment rate declined, its level of additional support would also decline, until the state fell below the initial threshold, at which point additional federal spending would cease.
The labor market indicators used for off-triggers generally use the traditional unemployment rate known as “U3,” which measures the number of people actively looking for work. But policymakers could choose to use other measures, such as “U6,” which includes discouraged and involuntary part-time workers; the prime-age employment-to-population ratio, or EPOP, which is a measure of the labor market participation rate of those ages 25 to 55; or even disaggregated versions of these statistics that focus on the employment prospects of historically disadvantaged groups that tend to lag behind national averages. There are pros and cons to each of the possible alternatives, but all of them would be an improvement on no triggers at all.
A different kind of off-trigger is proposed by Hilary Hoynes of the University of California, Berkeley and Diane Whitmore Schanzenbach of Northwestern University, whose Recession Ready proposal is to increase the Supplemental Nutrition Assistance Program’s maximum benefits by 15 percent during periods of high national unemployment. They propose an off-trigger similar to one included in the 2009 American Recovery and Reinvestment Act based on the food assistance program’s typical inflation indexing. Under their plan, the new maximum benefit would remain in place until inflation in food prices raised the previous maximum by that same amount, presumably a period of some years. At that point, when the original maximum benefit caught up with the 15 percent increase, the supplement would effectively disappear and benefits would resume increasing with inflation.
Conclusion
There are a number of ideas for Congress and the Biden administration to consider, but whatever they choose, it is important to note that the enactment of triggers in place of cut-off dates does not preclude additional action by Congress. When enhanced benefits or payments are set to expire automatically because of an off-trigger, but policymakers believe that conditions have not recovered sufficiently, Congress can bypass the trigger and keep additional spending in place. Or they can boost support even further by delivering, for example, monthly direct payments. Of course, the opposite is also true: Congress can determine that the economy has recovered sufficiently before an off-trigger is reached, and reduce or shut down the additional payments. Triggers are a backstop, not a replacement, for legislative action.
While they may be only a backstop, off-triggers are far superior to artificial stop dates. Because they don’t include such a date, including them in the next coronavirus package might increase the bill’s sticker price, as calculated by the Congressional Budget Office. But the cost, in human suffering and economic uncertainty, of an arbitrary and premature expiration date would surely be higher.
If relief and aid programs are inadequate, and if they depend on Congress returning again and again to extend them, then these measures are fighting with one hand tied behind their backs. Establishing off-triggers maximizes the power of relief measures and assures the workers and families depending on this support that Congress will not allow benefits to lapse. The new Congress and the new administration should enact them now, and let these programs do their essential work.
Taxis line up outside of Union Station in Washington, D.C., June 2019.
The U.S. Bureau of Labor Statistics just released its Employment Situation Summary for the month of January, showing that the labor market’s path toward a recovery lost steam over the past few months. In December, the U.S. economy shed 227,000 jobs, breaking a seven-month streak of net job gains. In January, less than 50,000 jobs were added, making December and January the worst months for employment since April, when more than 20 million workers lost their jobs.
At 9.2 percent, the unemployment rate for Black workers fell 0.7 percentage points between mid-December and mid-January, but continues to stand above the jobless rate of any other major racial or ethnic group. The unemployment rate for Asian American workers was the one to increase between December and January, jumping from 5.9 percent to 6.6 percent. The jobless rate for Latinx workers stands at 8.6 percent and for White workers at 5.7 percent. (See Figure 1.)
Figure 1
The Jobs report also sheds light on workers’ experience by race, gender, and ethnicity. For workers 20-years-old and over, Latina’s and Black women’s employment shrunk the most between January of 2020 and January of 2021, and Latina workers experienced a massive 22 percent drop in employment between January and April of last year. After almost a quarter of a million Latino workers lost their jobs between November and December, last month they won back 116,000 jobs. (See Figure 2.)
Figure 2
The leisure and hospitality industry lost almost half a million jobs in December and the sector continued to shrink last month, losing 60,000 jobs in January. Employment in the sector is down 23 percent relative to its January 2020 levels. Within leisure and hospitality, the arts, entertainment, and recreation subsector experienced an even deeper 32 percent decline in employment. Amid the continuing coronavirus recession, workers in low-wage service-providing jobs are experiencing especially tough economic outcomes, raising questions about the labor market dynamics that have put these workers in a particularly fragile position to withstand this shock. (See Figure 3.)
Figure 3
In addition to being more likely to experience unemployment, income losses, and be at greater risk of getting sick, the prevalence of tipped work in low-wage service industries exposes workers to a two-tiered system where employers are required to pay a federal tipped minimum wage of only $2.13 per hour. The Fair Labor Standards Act establishes that if a workers’ tips and wages do not add up to at least $7.25 per hour—the federal wage floor—their employers must cover the difference. In practice, however, it is common for tipped workers to earn less than the minimum wage as well as to not get paid overtime for additional hours of work.
A 2010–2012 study by the U.S. Department of Labor investigating almost 9,000 restaurants found that almost 84 percent participated in at least one type of wage and hour violation. And, alarmingly, research on the Great Recession of 2007–2009 found that wage violations became more prevalent as unemployment increased, with the probability of experiencing wage theft being highest for non-U.S. citizens, Latinx workers, Black workers, and women workers.
Both as a result of labor law violations and because many tipped occupations are among the lowest paid in the United States, tipped workers—among them bartenders, hair stylists, taxi drivers, and food servers—are twice as likely as their non-tipped counterparts to experience poverty. In addition, because women and workers of color are overrepresented in many of the occupations with the largest number of workers earning wages at or below the federal minimum, tipped work also entrenches pay disparities.
Occupational and industrial sorting by race, gender, and ethnicity explains an important chunk of the disparate effects the coronavirus recession has had on different groups. For instance, workers holding jobs that require in-person interactions face a higher risk of unemployment. Yet, research also finds that job segregation does not tell the full story. Case in point, a recent study finds that even when accounting for factors such as industry and occupation effects, age, geographic region, and education, women workers and workers of color were more likely than White men to lose their jobs, with Black and Latina women facing the greatest risk of becoming unemployed as the coronavirus recession hammered the U.S. labor market between March and April.
In addition, disparities in job displacement are neither a new development nor a phenomenon that only occurs during downturns. Research found that at least since the early 1980s, Black workers have been more likely than their White counterparts to involuntarily lose their jobs—a disparity that has increased since the 1990s.
Almost a year since the onset of the coronavirus recession, the U.S. economy is down almost 10million jobs with respect to February 2020. As the crisis continues to exacerbate and shed light on long-standing inequities in the U.S. labor market, policymakers should ensure that workers in precarious industries, particularly workers from historically marginalized groups who have been sorted into low-wage occupations, can access the social safety net. This includes tipped workers being assured access to Unemployment Insurance benefits, ramping up labor law enforcement efforts to prevent wage theft, and making sure that minimum earnings requirements do not keep benefits from some of the workers who need them the most.
The coronavirus pandemic and the ensuing economic recession has changed the structure of the U.S. labor market. Public health measures require limiting in-person services, and as a result, many industries saw sharp declines in employment in the springtime. Some industries, such as leisure and hospitality, then experienced a second downturn as the pandemic surged in the winter. Other industries that faced structural changes over the long term have seen some of those processes sped up.
In this column, we examine how the retail industry has been affected by the coronavirus recession, how this has been influenced by preexisting trends in the industry, and what this means for worker well-being of those employed across different types of retail jobs.
The latest Employment Situation Summary, released on January 8 by the Bureau of Labor Statistics, also known as the Jobs Report, shows that between mid-November and mid-December of 2020, the U.S. economy lost 140,000 nonfarm payroll jobs. Among the major U.S. industries, employment declines were especially large for the leisure and hospitality sector, which lost nearly half a million jobs. Government lost 20,000 jobs, making December the fourth consecutive month in which public-sector employment declined. But other industries saw important gains. The professional and business services and retail trade sectors added 161,000 and 121,000 jobs, respectively. (See Figure 1.)
Figure 1
Despite these gains, in December 2020, the retail sector employed 410,000 fewer workers than in February of that year. Moreover, the coronavirus recession hit some parts of the industry much harder than others. For instance, the number of jobs in sports, hobby, book, and music stores shrunk more than 17 percent between February 2020 and December 2020. Almost 1 in 4 clothing store jobs have been lost since the onset of the crisis. At the same time, garden supply stores, nonstore retailers, and general merchandise stores—which include warehouse clubs and supercenters—have actually grown amid the economic downturn. (See Figure 2.)
Figure 2
This uneven effect on retail reflects two important features of the coronavirus recession. First, because the measures needed to contain the dual economic and health crises affected both demand for goods and services and the operations of many retailers, workers employed in nonessential businesses and holding jobs that require face-to-face interactions have generally been more exposed to layoffs. For the frontline retail workers who kept their jobs, going to work became much riskier.
Second, and as with other economic trends, the downturn could be accelerating dynamics that were reshaping the retail sector well before the onset of the recession. Over the past decade, the sector’s somewhat sluggish recovery from the Great Recession of 2007–2009 was marked by the growth of e-commerce and bankruptcies of well-known apparel and department stores that media reports called a “retail apocalypse.” Even though there is evidence that the degree to which the industry is declining is overstated, retail employment fell somewhat since reaching its peak in early 2017, and is not projected to grow in the next decade.
Additionally, the same parts of the industry that are being hardest hit by the current downturn—clothing, sporting goods, and personal care stores, for example—have shrunk relative to the size of the retail sector since at least 2007. (See Figure 3.)
Figure 3
What does this mean for U.S. retail workers? In the immediate term, continued job losses in the retail sector are hurting some of the most vulnerable workers in the U.S. economy. For instance, 2019 research shows that retail salespersons—a position in which women workers, Black workers, and Latinx workers make up a disproportionately large share of the workforce—represent more than 8 percent of all low-wage U.S. workers, a significantly higher share than any other single occupation. More broadly, the sector pays the second-lowest average wages of any major U.S. industry. Only a small fraction of workers has access to employer-provided benefits. In addition to poor compensation, features of bad-quality jobs, such as unpredictable schedules and high rates of turnover, are rife across the retail industry.
As for longer-term implications, research from the Urban Institute finds that retail jobs are often the first rung in workers’ career ladders, making good jobs in retail an important piece of career advancement and influencing lifetime earnings growth. And while many workers transition out of the retail sector when switching jobs, workers of color in general and Black women in particular are less likely than their White counterparts to exit service and retail occupations. As such, policymakers should prioritize measures that improve labor standards amid the cyclical downward pressures on job quality.
One way to do this is to raise the wage floor. Even though more than one-fifth of retail jobs earn the minimum wage, nearly 30 states increased minimum wages at the beginning of this month. Maintaining minimum wage increases is a critical part of boosting wage growth in the eventual economic recovery. Research from Kevin Rinz and John Voorheis of the U.S. Census Bureau estimates that some of the worst earning losses of the Great Recession would have been partially offset if the federal minimum wage was increased at the magnitude of the increase in Seattle between 2013 and 2016.
In addition to maintaining and increasing job standards, the overall decline in job growth in December and the persistent decline of employment in retail through the coronavirus recession points to the continued need for direct relief to these hard-hit workers. Unemployment insurance systems need to be shored up and more generous unemployment benefits and direct stimulus payments need to be expanded. Research by Ammar Farooq of Uber Technologies Inc. and Adriana Kugler and Umberto Muratori of Georgetown University shows that Unemployment Insurance benefits improve job quality as workers have the time and financial security they need to move into employment opportunities that better match their skills and interests.
Protecting the economic well-being of workers on the bottom rungs of the wages ladder would power a more robust U.S. economic recovery and improve the resiliency of workers in the long-run so that future economic growth would be more broadly shared.
The Washington Center for Equitable Growth is announcing a trio of research grants to scholars exploring the impact of paid family and medical leave and paid sick leave policies amid the coronavirus pandemic and ensuing recession. These studies, conducted by faculty members at leading U.S. colleges and universities, will answer important research questions on paid leave’s role in protecting public health and economic security during the ongoing crisis.
But first, some background on Equitable Growth’s initiative in these arenas in light of the coronavirus recession: Equitable Growth is committed to advancing academic knowledge on paid family and medical leave, an issue that has been pushed to the forefront of policy debates in light of today’s serious public health crisis. As noted in a June analysis by Chantel Boyens at the Urban Institute, several states with their own paid family and medical leave systems experienced a sizable increase in paid leave claims at the start of the pandemic.
Shortly thereafter, the U.S. Congress passed the Families First Coronavirus Response Act. The FFCRA provided emergency paid leave to workers affected by sickness or child care responsibilities related to the coronavirus pandemic and COVID-19, the disease caused by the virus. Though only temporarily available, this emergency paid leave is the first ever federal paid leave guarantee in the United States.
Understanding how state paid leave programs and this federal emergency leave addressed the needs of workers and affected the functioning of the U.S. economy during this crisis is crucial as policymakers consider options for permanent paid family and medical leave guarantees in the future. We are pleased to announce investments in three research projects poised to shed light on how governmental paid leave programs addressed the current crisis.
In their project, “COVID-19 and Paid Leave: Assessing the Impact of the FFRCA” Kristen Smith of Dartmouth College, Tanya Byker of Middlebury College, and Elena Patel of the University of Utah will investigate how leave-taking was influenced by the passage of the Families First Coronavirus Response Act. The researchers will analyze the impact of the FFCRA on several employment and leave-taking outcomes, including labor force participation, usual hours worked, and reasons for work absences. Using data from the Current Population Survey, the researchers will examine how workers trade off the alternatives to leave-taking, including working while sick or separating from the labor force. The authors plan to compare leave-taking in states that either do or do not have state-based paid leave programs. This study will join an emerging body of literature studying the efficacy of the FFCRA.
Also investigating how paid leave can impact families’ finances and well-being in times of crisis is Jane Waldfogel of Columbia University in her project, “Access to Paid Leave during the COVID-19 Pandemic: Evidence from New York City.” This study will explore access, use, and outcomes associated with paid leave during the pandemic in New York City, using data from the New York City Longitudinal Study of Health and Wellbeing, also known as the Poverty Tracker. This survey follows representative samples of New York City residents, interviewing them every 3 months for up to 4 years and collecting a wealth of data on poverty, material hardship, health and well-being, and asset and debt levels.
Waldfogel will re-interview a sample of respondents who have already participated in 4 years of pre-pandemic surveys. In their first interview since the onset of the pandemic, the sample will be asked about their use of employer-, state-, or FFCRA-provided paid leave during the pandemic, as well as their experiences with poverty, hardship, and health and well-being since the emergence of the coronavirus and COVID-19.
In particular, the research team is interested in examining whether people in states with guaranteed paid sick leave were better able to adopt social distancing measures. Additionally, the team will examine if there are differences across states with and without paid leave systems in reported measures of illness, leave-taking, and economic and psychological distress associated with the pandemic.
The COVID-19 pandemic will end someday, but the lessons learned during this period will be informative to policymakers for years to come. Should another public health crisis emerge, policymakers will be better informed and better prepared to implement effective policy strategies to protect our public health, our families, and the broader U.S. economy.
More broadly, policymakers can draw on these lessons to construct a robust and equitable paid leave system for all workers with medical and caregiving needs to ensure the recovery from the coronavirus recession is strong and broad-based. Equitable Growth is proud to support this research and looks forward to working with this impressive slate of grantees to disseminate their findings and to continuing to build a bridge between those who make paid leave policy and those who can provide the evidence needed to do so in a way that supports an economy that works for all.
The effects of the coronavirus recession extend beyond jobs and income to include the emotional well-being of parents and their children.
As job losses in the United States persist amid the coronavirus recession, the economic and mental health fallout becomes ever clearer. Historically, research on the effects of job losses and unemployment on parental mental health and family dynamics has emphasized its ill effects, highlighting, for example, the increase in parental stress and depression and disruptions in parent-child relationships in households suffering from job loss. Many of these adverse impacts can be traced to the financial hardships that can accompany sudden unemployment, particularly in economically vulnerable households.
During the pandemic, the expansion of the Unemployment Insurance system and other economic aid programs mitigated, at least temporarily, the most devastating consequences for those who are able to access those benefits. By the end of December, however, all of these expanded unemployment benefits will expire, while other fiscal measures to protect existing jobs from the ravages of the current third wave of the coronavirus pandemic and COVID 19, the disease caused by the virus, are stalled in the U.S. Congress.
What, then, has been the impact of job losses so far on family well-being during the pandemic? It turns out the answer is more nuanced than one might first assume. In our new working paper, I and my co-authors Susan Mayer and Rohen Shah at the University of Chicago find that there are, indeed, negative effects for mothers in low-income households, many of whom are experiencing increased levels of stress and depressive symptoms, compared to before the pandemic. Yet there also are positive effects for some mothers and their children. The difference is largely around whether mothers lose income due to pandemic circumstances.
These findings offer new insights into the effect that economic shocks, including the coronavirus pandemic, can have on the quantity and quality of time spent between mothers and children, as well as on the financial and emotional health of mothers. I’ll now walk through some of our specific findings.
Pandemic effects on family dynamics
Our survey research on the effects of the pandemic on family dynamics focuses on low-income families, where the virus has been more prevalent. These families, which often already face stresses associated with limited income, face additional pressures during the pandemic, including stress arising from stay-at-home orders and social distancing, separation from friends and other support resources, and disruptions to employment and daily routines. Children in low-income families tend, on average, to have lower academic and socio-emotional skills, compared to their peers in economically better-off households. The families in our study have preschool age children who are vulnerable to a slowdown, or even a reversal, in their skills development due to their preschools having closed in the spring.
On the one hand, given the stresses and pressures placed on low-income families during the pandemic, one would assume that mothers are facing more ill-effects than otherwise and that such negative effects are affecting relationships with their children. This conforms with much of the news coverage and other cultural assumptions prevalent during the pandemic.
On the other hand, many of those mothers received federal stimulus funds in late spring 2020 and, if they lost their jobs, received expanded Unemployment Insurance that, in some cases, amounted to a higher income during their unemployment. Some also may have partners who increased their work hours, and hence their income, due to the pandemic. In effect, these families experienced a type of paid family leave for at least one partner, albeit with an uncertain future looming directly ahead.
We also investigated the effects on parental mental health and stress, parent-child interactions, and children’s behavioral adjustment. We did this by comparing the relative importance of pandemic-induced economic hardships, such as job and household income losses and the inability to make ends meet, as well as pandemic-induced social conditions, including oneself or one’s family members becoming ill with COVID 19, and pandemic-induced increases in child care time.
The findings about the relative effects of job and income losses are especially relevant to discussions of economic support for low-income families. In particular, mothers who both lost their jobs and suffered substantial drops in household income were significantly more likely to report depressive symptoms, as well as increases in life stress, compared to those survey respondents who experienced neither of these events. These mothers also experienced less hope for the future. Importantly, however, losing a job with no substantial loss of household income had no impact on these mental health indicators.
Furthermore, parents who lost their jobs but did not experience a drop in their incomes reported having more positive interactions with their children and were more likely to report that their children enjoyed spending time with them. Parents struggling with losses of income, by contrast, report more often losing their temper and yelling at their children.
Conclusion
The coronavirus pandemic induced a severe economic recession that negatively affected millions of U.S. households. Much of this effect is captured in employment and income data, but for most of those families, the effects extend beyond jobs and income to include the emotional well-being of parents and their children. This is especially true as families experience shelter-in-place orders and children attend school from home.
In our new research, I and my co-authors show that the important factor in determining parents’ well-being is loss of income and not necessarily their employment status. For policymakers and others concerned about the well-being of U.S. households, a key lesson is to consider the extent of income losses, which shape parents’ mental health, how they relate with their children and, ultimately, children’s adjustment in a way that job losses by themselves do not. These results also highlight the importance of time for parenting. It is important for parents to have both income and time to spend with their children. If parents can have more time with no loss of income, then parent-child interactions can improve. This may suggest generous paid family leave is a promising way to improve child outcomes.
Unfortunately, these critical unemployment programs are set to expire at the end of the year, potentially leaving millions of Americans vulnerable to economic and material hardships. Leaders of both major parties have expressed support for renewing the programs in some form, but Congress has been unable to reach a deal to do so. Our findings underscore the urgency of these efforts.
An essential worker restocks the shelves at a Walmart in Salem, Ore., May 2020.
Today’s U.S. Bureau of Labor Statistics Jobs Report demonstrates a stalled jobs recovery following the deep and rapid economic decline brought about by the coronavirus pandemic. The unemployment rate has decreased slightly to 6.7 percent, but still remains higher than the 3.5 percent unemployment rate in February prior to the start of the coronavirus recession.
The workers who have been hit the hardest continue to be Black workers (10.3 percent), and Latinx workers (8.4 percent). Women workers saw a decrease in unemployment (6.1 percent) that was accompanied by a slight decline in labor force participation (57.1 percent). These workers are also disproportionately low-wage workers as the service sector industries in which most of these workers are employed are among the hardest hit. To put this in context, when the deepest effects of the pandemic hit the economy in April, the unemployment rate reached a post-World War era high of 14.7 percent and the prime-age employment-to-population ratio plummeted to 69.7 percent—a 10 percentage point drop from the previous month.
The same month, however, one important economic indicator pointed to another direction. Seemingly contradicting one of the main laws of economic theory, just as labor demand plummeted—when economists would expect wages to decline in tandem—average hourly earnings jumped 8 percent in April relative to the same month of 2019. This massive rate of growth is not even observed during booms: Using this same measure, average wage growth did not surpass 3.8 percent at any point during the 2009–2020 economic expansion. This month wage growth normalized somewhat, with an average worker in the private sector earning $29.58 $an hour, a 4.19 percent increase from the year before.
The Bureau of Labor Statistics provides a disclaimer in the Employment Situation Summary that “the large employment fluctuations over the past several months—especially in industries with lower-paid workers–complicate the analysis of recent trends in average hourly earnings.” While higher-wage workers are more likely to be able to telecommute during the current public health crisis, low-wage workers are more likely to lose their jobs. The coronavirus recession has led to a unique contraction of the service sector, changing the composition of employment in the overall U.S. labor market.
As research by Peter Ganong at the University of Chicago Harris School of Public Policy and Pascal Noel and Joe Vavra at the University of Chicago Booth School of Business shows, unemployment this year is greater for low-income workers. They estimate that workers’ whose pre-job loss weekly earnings put them in the bottom 20 percent of earnings distribution were experiencing a jobless rate of 16 percent between April and July of 2020. Those in the top 20 percent of the income ladder, by contrast, were experiencing an unemployment rate below 4 percent.
As lower-income workers disproportionately lost their jobs, higher-earning workers were more likely to remain employed, which taken together mechanically increases average earnings. Similarly, research by the Federal Reserve Bank of San Francisco shows that in these first months of the coronavirus recession, workers in the bottom quartile of the earnings distribution represented about 1 in 2 of all transitions out of employment.
When looking at other metrics of wage growth, another picture emerges. Accounting for changes in the composition of occupations by measuring wage and salary growth for a fixed set of jobs, the Employment Compensation Index of the National Compensation Survey shows that rather than accelerating, wage growth has slowed down. (See Figure 1.)
Figure 1
At the onset of the pandemic there was some expectations that the earnings of essential and front-line workers would rise. According to the hypothetical premise of “compensating wage differentials” in labor economics, employers need to make more attractive offers in order to convince workers to do risky or undesirable jobs. In the early months of the health and economic crises, there was some evidence that this would be the case. Companies such as Amazon.com Inc., Costco Wholesale Corp, and Walmart Inc. offered front-line and essential workers—cashiers, warehouse, and fast-food workers, for example—hazard pay, bonuses, and sick paid leave.
That is largely no longer the case. Many workers on the front lines and concerned about getting sick are no longer receiving extra compensation. According to a poll commissioned by the Economic Policy Institute, in mid-May only about 30 percent of respondents working outside from home were receiving some kind of hazard pay, even as more than 50 percent were concerned about contracting the coronavirus and falling ill from COVID-19, the disease caused by the virus. Throughout the late summer and fall, even as the health crisis continued to put workers at risk, many businesses cut those benefits.
As the public health crisis worsens, states are maintaining partial opening or strengthening so-called lockdowns to limit the surge of the pandemic. Economist Trevon Logan at The Ohio State University estimates that hour reductions during peak hours for bars after 10 p.m. as well as limited capacity requirements in restaurants and bars are translating into at least a 50 percent wage cut for many service workers.
It’s important for economists and policymakers alike to consider where the U.S. labor market is today compared to prior to the pandemic. Back then, wage growth had finally started to ramp-up for the first time in years. As many states and cities increased their minimum wage and the labor market tightened, wage growth accelerated the most for the lowest earners, yet significant income inequality and differences in unemployment rates by race remained persistent. Now, after the coronavirus recession caused an unprecedented number of jobs lost and widespread pay cuts, elevated unemployment rates will likely exert downward pressure on wages as workers find fewer jobs options and grow increasingly desperate as unemployment benefits expire.
This is why it’s essential to put forward policies that help maintain labor standards, including increasing earnings through minimum wage increases and ensuring workplace safety amid the pandemic so that an eventual recovery is robust and resilient, rather than further entrenching economic inequality. Recent research and today’s Job Report further emphasizes the need to center economic policies around the well-being of the hardest hit workers to ensure they won’t have to wait another 10 years for strong wage growth.
At the outset of the coronavirus pandemic earlier this year, millions of U.S. workers lost their jobs. In March, the U.S. Congress took notice and decided that the 26 weeks of Unemployment Insurance typically provided by states were not enough for an employment crisis that would likely extend beyond 6 months. Through the new Pandemic Emergency Unemployment Compensation program, it provided people who lost jobs through no fault of their own with an additional 13 weeks of benefits.
Today, coronavirus cases counts are skyrocketing, hospitals are filling, businesses are closing, and long-term unemployment is surging. But some workers have already run out of benefits, and the PEUC program is set to expire on December 26, leaving 8.1 million workers without any income support through this emergency 13-week benefit extension.
The expiration of the Pandemic Emergency Unemployment Compensation program means devastation for workers and their families this December. But research released earlier this year by Adriana Kugler and Umberto Muratori at Georgetown University and Ammar Farooq at Uber Technologies Inc. shows that the effects of the PEUC expiration are likely to persist far into the future for workers whose benefits expire, and ripple through the U.S. economy to affect firms and workers searching for well-matched employment relationships.
The basic idea behind this research is intuitive. When people have resources to meet their basic needs while out of work, they can take the time they need to find the right job for them, rather than taking the first work opportunity that comes along. Workers benefit because they find jobs that pay more and meet their needs more broadly, and firms benefit because they recruit workers with the right skills for the job. While the idea makes sense, it hasn’t been backed by much evidence in the U.S. context, in part because of data limitations.
This makes it particularly exciting that this research team was able to access data from the Longitudinal Employer Household Dynamics database, which matches administrative information about workers and their earnings with administrative information about their employers. Using these data, as well as information from the Current Population Survey and the natural experiment that occurred when different states offered different durations of unemployment benefits during the Great Recession of 2007–2009 and its aftermath, Farooq, Kugler, and Muratori take a close look at how job seekers and firms fare when the duration of unemployment benefits is extended. (See Figure 1.)
Figure 1
Farooq, Kugler, and Muratori find that the longer you have access to unemployment benefits, the higher paying the job that you eventually find is and the more your skills and training get put to use. (See Figures 2 and 3.) Access to the full extended emergency benefits during 2009, which increased benefits from 26 weeks to 79 weeks, increases re-employment wages by 2.6 percent and also increases the probability that a worker will be re-employed in a job that requires more education than their previous job by 11.7 percentage points.
Figure 2
Figure 3
The idea that workers match their talents with job openings that put their skills to use is not just good news for those workers. It matters for firms and the broader economy as well. Farooq, Kugler, and Muratori take advantage of their dataset, which allows them to follow individual workers over time and also to see the full workforce of the firms with which they eventually find re-employment. Using these features of the data, they find that higher-quality firms are better able to recruit workers that have the abilities they need. This creates a chain reaction that spreads through the U.S. economy. The authors describe it eloquently:
These results suggest that if a worker can receive UI benefits for a longer period, she will be able to find a job with an employer that is closer to her in terms of quality. This worker then is likely to leave another job open for someone else who is also likely to be better matched, and in turn that other worker can also leave vacant another job and relieve it to someone else, generating a chain reaction that makes many other workers, beyond the one receiving the UI extension, match better in the labor market.
In this manner, the benefits of extended unemployment benefits ripple past a worker’s current situation to affect their future job prospects, the productivity of firms, and the experiences of workers across the economy.
Particularly striking is the fact that the positive effects of extended unemployment benefits are larger for workers who are members of groups that lack funds for a rainy day. Because of labor market discrimination, unequal access to education and training, caregiving obligations, and stagnating wage levels, members of demographic groups, including women, people of color, and less-educated workers, typically lack the private savings held by their counterparts who are members of more advantaged groups.
Indeed, this research finds that during the Great Recession and its aftermath, women, people of color, and low-educated workers improve their job matches when they have access to unemployment benefits even more than their counterparts who are members of wealthier demographic groups. The research team finds that a 53-week increase in UI benefits improves the job-match quality by 0.9 percent for White workers, but improved match quality even more for workers of color: The effect size for workers of color is 1.2 percent.
Similarly, Farooq, Kugler, and Muratori find that workers of color (along with women, less-educated workers, and young workers) see larger returns to increased weeks of benefits when it comes to wage levels. This supports the idea that the improved fit between workers and the jobs they find stems from workers’ ability to spend more time searching for a job that is a good match, without sacrificing their basic needs in the short term.
Their findings also suggest that insufficient durations of unemployment benefits during moments of macroeconomic contraction exacerbate inequality in the U.S. labor market. This concern is magnified during today’s coronavirus recession because the sectors of the U.S. economy that have been hardest hit are the same ones where women, people of color, and low-educated workers are overrepresented.
There is another reason that policymakers should be deeply concerned about the relationship between job matching and extended unemployment benefits during this crisis in particular. While this research speaks to job match in terms of workers’ education levels and wages, we can guess that the extended time provided for job search also allows workers to find jobs that are better matches for them along other dimensions—schedule flexibility, location, and working conditions. During the current public health crisis, working conditions are of outsize importance. The ability to be choosy about the health and safety conditions of one’s workplace is more likely to be the difference between life and death for a worker or a member of the worker’s family than is typical outside of the context of a public health crisis.
So, what to do about the upcoming expiration of the PEUC program? There is the obvious short-term fix—extend the duration of PEUC benefits so that all workers who are unemployed through no fault of their own can receive unemployment benefits for the amount of time it takes to find a job that is a good match for their skills and their health.
And then, there is the bigger structural issue. It’s clear from a macroeconomic perspective that the duration of unemployment benefits should extend automatically during moments of economic contraction. Indeed, a permanent program called the Extended Benefits program is designed to do just this. But the formulae we rely on (known as “triggers”) to turn on Extended Benefits are broken, and we are repeatedly left in the situation we find ourselves in today: waiting on political horse trades to enact common sense, economically necessary policy decisions through one-off emergency benefit extensions.
So, to make unemployment compensation work for workers and for the economy as a whole in both the short- and long-term, policymakers should extend PEUC today, but they shouldn’t stop there. They should also redesign Unemployment Insurance permanently with improved automatic stabilizers so that payment extensions trigger automatically when economic conditions warrant.
Remdesivir may be one possible drug to combat COVID-19.
Overview
The extraordinary ongoing effort to find treatments to cure, ameliorate symptoms of, and vaccinate against the novel coronavirus and COVID-19, the disease caused by the virus, is a reminder of how much the U.S. healthcare system, the U.S. economy, and we as individuals and families rely on these products, as well as the public research that goes into discovering them and the pharmaceutical companies that develop and manufacture them. This issue brief discusses the ownership and costs of prescription drugs, as well as potential policy initiatives in this area, through the lens of one prescription medication, the antiviral drug remdesivir.
Remdesivir is important because it became the first signal of hope amid the coronavirus pandemic, when research published in April 2020 suggested that it reduced the duration of symptoms for patients with severe COVID-19 infections by about 4 days on average, although it did not have a significant effect on mortality. Those results led the U.S. Food and Drug Administration to permit its use under an emergency use authorization, and for months, it was the only drug available with at least some proven real clinical benefits for COVID-19. On October 22, the agency formally approved remdesivir for the treatment of COVID-19 requiring hospitalization.
Approval by the Food and Drug Administration, however, means that a drug has some benefit but it does not guarantee any particular level of benefit. Indeed, in a recent study, we found that only about one-third of new drugs approved by the agency over the past decade were rated as having high therapeutic value. For remdesivir, recent data have shaken the foundations of its therapeutic potential. According to a pre-print of a major World Health Organization solidarity study from October 2020, remdesivir does not appear to affect patients’ mortality or their hospitalization duration, and it is not clear whether its effects on symptoms are additive to dexamethasone, another drug shown to work for severe COVID-19. Remdesivir still seems to display some anti-COVID-19 activity and may have a role in physicians’ therapeutic armamentarium, but what that role will be will continue to evolve as further data emerge.
Gilead Sciences Inc. owns all the relevant patents for remdesivir. After the drug was authorized for use against COVID-19, the company announced plans to donate 1.5 million doses. But that commitment will not meet demand, and the company announced in June a list price of more than $3,000 for a full treatment regimen. At that price, Gilead is reportedly on track to make more than $9 billion in revenue on remdesivir in 2020 and 2021 alone. With FDA approval in hand, the company will be able to set whatever price it wants for the drug for likely more than a decade.
Remdesivir, then, offers a good reminder of the need for policies that can achieve continuing prescription drug innovation, while also maintaining a fair price for U.S. patients and the health care system more broadly. This issue brief will provide a short history of the development of remdesivir, summarize the intellectual property laws governing prescription drugs in general and remdesivir in particular, and present policy options to affect the pricing of remdesivir and drugs like it going forward.
Briefly, the policy recommendations detailed below are divided among four basic stages of drug development, approval, and production:
Stage 1: the discovery process, leading up to approval by the U.S. Food and Drug Administration
Stage 2: the 12- to 14-year period of market exclusivity for brand-name drugs
Stage 3: the transition to a competitive market with the introduction of generic versions of the medication
Stage 4: the period when multiple generic versions of the drug can be on the market
Remdesivir has just completed stage 1. Other potential treatments for COVID-19, such as monoclonal antibodies, are at even earlier stages. While these policy recommendations represent an ambitious agenda with significant implications for a broad range of drugs, the crisis brought on by the coronavirus pandemic makes clear that such an agenda not only is necessary but can and should be considered in the near future.
In short, now is the best time to put many of these provisions in place, before remdesivir and other treatments and vaccines make further progress along the approval path.
Let’s turn first to the history of the development of remdesivir to demonstrate the clear role of public financing in the discovery and development of the drug, and its public policy implications.
The discovery and development of remdesivir
Prescription drugs can be very costly, which affects not only U.S. healthcare and the economy broadly but also individual health outcomes. High prices can keep important medications out of the hands of patients who need them to get well, or even to live. The price of brand-name prescription drugs has risen markedly faster than overall inflation over the past two decades. (See Figure 1.)
Figure 1
The federal government and not-for-profit organizations, especially U.S. universities—funded in good part by federal agencies—support or conduct most of the basic and translational research that lays the foundation for the medications we use. The role of public funding is more pronounced among the most transformative new drugs. The prescription drug industry also invests a great deal in bringing brand-name products to market—usually at the clinical testing, regulatory approval, and manufacturing phases of development—but the prices they charge are not related to the costs they incur in developing medications.
Yet U.S. patent laws give what is essentially monopoly pricing power to companies that hold the intellectual property related to valuable drugs, even if that intellectual property is derived from insights originating in the basic and translational research completed by publicly funded institutions and their scientists. And these firms sometimes use that power to charge prices far beyond the value that the drug provides. Given the federal government’s critical role, and the vital public interest in supporting public health, there are a number of concrete steps U.S. policymakers could take to better manage prescription drug costs at every stage of the approval process for these medications—without stifling innovation.
Remdesivir is a potent case in point.
The road that led to remdesivir’s development so that it could be effective against COVID-19 was paved by public funding and university research. Its initial development was the culmination of several years of collaboration between Gilead Sciences, the U.S. Army Medical Research Institute of Infectious Diseases, and the U.S. Centers for Disease Control and Prevention. The company and the two federal agencies were already collaborating in search of possible antiviral candidates in Gilead’s library of nucleoside analogues when the Ebola virus broke out in West Africa in 2014. The initiative identified the precursor to remdesivir, which company researchers and the government, led by the Army’s infectious diseases institute, further developed.
Researchers at Gilead Sciences and universities then turned to studying remdesivir as a potential treatment for various viruses, including coronaviruses, supported in part by another U.S. Department of Defense research agency. This joint public-private research involved groups such as the National Institutes of Health, that included the University of Alabama, the University of North Carolina, Chapel Hill, and Vanderbilt University. This research led to the discovery that remdesivir could be useful against Middle East Respiratory Syndrome, or MERS, and severe acute respiratory syndrome, or SARS.
Eventually, research conducted during the Ebola outbreak, sponsored by the government of the Democratic Republic of Congo and the African Coalition for Epidemic Research, Response, and Training, determined that remdesivir was not sufficiently effective against Ebola. But remdesivir was one of the earlier products tested against COVID-19, based on its usefulness against other coronaviruses causing MERS and SARS. Clinical trials were launched around the world beginning in early 2020 to evaluate the safety and efficacy of remdesivir in hospitalized patients with severe cases of the disease.
After several promising but insufficient trials, the National Institute of Allergies and Infectious Diseases, or NIAID, initiated the first double-blind randomized U.S. trial of remdesivir in February 2020. The trials showed that the drug could reduce recovery time by 4 days but did not lead to a significant reduction in deaths. It was not a gamechanger, but it was effective—and was the first drug identified to have a palliative effect on COVID-19.
Intellectual property and remdesivir
The results of remdesivir’s NIAID trial, along with other supporting data, led to its emergency use authorization and Gilead Sciences’ announcement in May that it would provide a set number of free doses of remdesivir, and then to its subsequent price announcement. Gilead set a price, however, that raised important concerns about the cost and availability of what was then the primary useful drug against COVID-19. Indeed, a number of state attorneys general urged federal action to address these issues, and state treasurers are calling on Gilead Sciences to lower the drug’s price. These calls have become more compelling with the recent data from the large multinational trial suggesting that the drug does not have benefits in terms of mortality or hospitalization.
Now that the drug is FDA-approved, Gilead remains in control of manufacturing, pricing, and distribution until its patents and other statutory exclusivities expire, and generic versions are approved by the Food and Drug Administration. There are several laws that provide exclusivity to manufacturers, but two that apply specifically to new medicines are especially relevant when discussing remdesivir. The Drug Price Competition and Patent Term Restoration Act of 1984—the Hatch-Waxman Act—gives drug companies a minimum of 5 years to 7 years of exclusivity for new drugs. Yet most drug manufacturers have patents lasting much longer than that. In fact, Gilead currently owns at least 12 patents on remdesivir. The last of them does not expire until 2039. But such patents can sometimes be challenged or, if necessary, designed around to enable generic drugs to enter the market long before patents expire. The Hatch-Waxman Act created a pathway for such challenges to occur, and it created a mechanism for generic manufacturers to seek FDA approval when the patents expire.
Another law offering exclusivity at the time of drug approval is the Orphan Drug Act, enacted in 1983, which provides 7 years of exclusivity and a valuable tax credit for drugs used in treating rare diseases (those affecting fewer than 200,000 people). The purpose of this law was to encourage pharmaceutical companies to develop drugs for rare diseases that might not be profitable, given the limited market for them. Early in the COVID-19 pandemic, when the number of infections was below 200,000, Gilead Sciences requested Orphan Drug Act designation for remdesivir for the treatment of COVID-19. The Food and Drug Administration approved this status in March, despite the rising numbers of cases—until public scrutiny led the company to withdraw this designation voluntarily.
Public financial support for development of remdesivir
As noted above, remdesivir, like most drugs, stands on the shoulders of federally funded research. In some respects, that is truer for this drug than for many others, given the specific partnerships formed between federal agencies and Gilead Sciences, culminating with the recognition that the drug might be useful for COVID-19. (See Table 1.)
Table 1
U.S. government research laboratories and university research funded by the federal government—mainly the National Institutes of Health—are the origin of many fundamental discoveries on which new drugs are based. Drug companies then spend a considerable amount developing and bringing a product to market. This is generally accomplished by the pharmaceutical company that comes to own the intellectual property for a given compound. Yet companies charge high prices to the very taxpayers who funded the research that made their product possible, and do not direct proceeds to the federal government to help fund new research. In the case of remdesivir amid the coronavirus recession, these considerations are now front and center for policymakers.
Policy options
There are numerous legislative and regulatory actions that policymakers could take to address drug prices that have risen far beyond inflation in the past three decades. All of them taken together would significantly lower spending on prescription drugs while ensuring continued funding for true innovation. Applying such measures to the case of remdesivir can be further justified by the federal funding that provides all or part of the foundation for the advances subsequently funded by Gilead. (For a more thorough description of these ideas, which address problems in four key periods of the development of pharmaceuticals, see my essay in Equitable Growth’s Vision 2020: Evidence for a stronger economy.)
Drug discovery period
When it supports research that could lead to the discovery or understanding of prescription drugs, the National Institutes of Health could require a reasonable pricing provision to be attached to the provision of funding. This provision could, for example, require that the ultimate price of the product be no greater than its value-based price—a price reflecting the drug’s potential ability to improve patient outcomes over comparable interventions—as determined by independent organizations.
In the case of remdesivir, the Institute for Clinical and Economic Review used cost-effectiveness analysis to judge that a fair price for remdesivir initially would be $4,460, assuming it had an effect on mortality, which was not statistically significant in the initial trial. Without an effect on mortality, as was confirmed by the more recent multinational trial, the institute estimated that a cost-effective price of remdesivir would be just $310 for a course of therapy.
Another option—not relevant in the case of remdesivir—would be for the National Institutes of Health to change its interpretation of the “march-in” provision of the Bayh-Dole Act of 1980, which established the basic rules for commercialization of technology arising from government funding. Under Bayh-Dole, the NIH retains a license on patents resulting from federally funded research and can “march in” to invalidate an exclusive commercialization license if the product is not made available on “reasonable terms.” The institute has never applied this provision to pricing matters, but this can be a tool for curbing high prices in a limited number of cases. This provision would not be relevant in the case of remdesivir because the government does not have a direct stake in the patents held by Gilead.
Brand-name-only period
Until the availability of generic substitutes for remdesivir, there will be no direct competition to bring down prices. Introduction of other brand-name products indicated for the same purpose have generally not been shown to lower prices to a substantial extent. Until generic competition, then, the best way to achieve fair prices for brand-name drugs is to empower the buyers to negotiate better terms with the manufacturers. There is no larger U.S. buyer than the 44 million Americans covered by Medicare.
Medicare, however, is barred from negotiating drug prices—unlike the U.S. Department of Veterans Affairs—a prohibition that does not apply to any other health care service it covers. So, the best solution for managing drug prices during this period is to provide the entire federal government, including Medicare, the authority to negotiate reasonable prescription drug prices that better reflect the value of treatments, as well as the government’s budget and the drugs’ origins.
To accomplish this, the United States could follow the model established by numerous other countries. U.S. policymakers could create a health technology assessment organization that would evaluate a newly approved drug’s clinical value and help determine a fair price based on how well it is expected to perform against other available treatments as well as other relevant factors. That assessment would provide the basis for negotiations with the manufacturer. Future prices could rise and decline based on inflation, how well the drug continues to perform in real-world use, and the introduction of related products. It would make sense to create such an organization within the U.S. Department of Health and Human Services.
Another way of addressing prices during this period is to provide better information to prescribing physicians and patients. Currently, the vast majority of information that both doctors and their patients receive is a result of the billions of dollars that companies spend promoting their products. What is needed is more objective, noncommercial information about drug benefits, risks, and cost effectiveness. Policymakers should support independent programs that generate unbiased information about evidence-based management of disease and disseminate this educational information to physicians. This can translate into more cost-effective prescribing.
Transition to a competitive market period
Because the only type of competition that consistently and substantially lowers drug prices comes from the introduction of interchangeable, FDA-approved generic drugs, brand-name manufacturers often employ strategies to extend market exclusivity periods. Companies exploit ambiguities in the Patent Act to obtain dozens of “secondary” or “tertiary” patents on peripheral aspects of their approved brand-name drugs. The subjects of these follow-on patents can include anything from the medicine’s coating or a change in its delivery from a pill to a capsule, to using a new device such as an injectable pen. Incremental changes are commonly covered by these patents, but they often do not provide advancements in drug efficacy, safety, or convenience that are commensurate to the higher prices being charged.
In addition, manufacturers have usedvarious other strategies to prevent the timely entry of generic drugs. These have included filing citizen petitions with the Food and Drug Administration raising frivolous concerns about the interchangeability of a potential generic, restricting supplies of their product for generic manufacturers to use in the bioequivalence studies needed to prove that a generic matches the original drug, and entering into settlements with generic manufacturers to drop patent challenges and delay their plans to market a competing generic product.
Policies to make those strategies more difficult or impossible to carry out can speed the introduction of competition and thus price reductions. A number of bills have been considered in Congress that address generic-delaying strategies in a piecemeal way. And in 2019, Congress passed and President Donald Trump signed into law bipartisan legislation intended to stop the practice by brand-name drugmakers of keeping supplies of medications out of the hands of generic manufacturers to prevent them from creating competing products.
Other measures Congress could take include requiring greater disclosure of a product’s patents (particularly for biologic drugs), preventing abuses of the Food and Drug Administration’s petition process, addressing remaining problematic settlement agreements with generic manufacturers, and giving the Food and Drug Administration greater authority to approve as interchangeable generics that are slightly different from the original drug if the differences are not clinically relevant. More frequent use of the Patent Trial and Appeals Board’s administrative patent review process—such as automatic review at the time any drug patent is FDA-listed—also could help weed out insufficiently innovative patents.
Another policy solution that Congress could enact is to restrict a brand-name drug’s market exclusivity period to a particular time period and not permit secondary or tertiary patents—or any of the other strategies—from being able to block FDA approval of a generic version. Manufacturers could be restricted to the single patent for which they seek and receive patent term restoration (a period of up to 5 years to account for time spent in clinical trials and FDA review), plus the 6-month patent extension manufacturers receive for testing their drugs on children. At the end of this period, generics would be permitted to enter, no matter what other patents have been obtained.
Multisource generic drug period
After a drug has lost exclusivity protection, prices may not fall if there are not enough generic manufacturers in the market, and other market conditions might also work against price declines. The Food and Drug Administration has deployed user fees, which began in 2012, to make substantial reductions in previous delays in the approval process for generic drugs. More resources must be invested at the agency to ensure that there are no unnecessary delays in generic drug approval, as well as in providing guidance on the types of studies generic manufacturers need to complete to receive FDA approval.
There are also some instances of older pharmaceuticals that no longer have exclusivity but for which not enough generic manufacturers have entered the market to reduce price substantially. In these cases, importation is a possible solution. This would involve setting up a regulatory system that would allow mutual recognition between the Food and Drug Administration and similar regulatory systems in Europe and in Canada, Australia, and Japan.
Another solution would be to authorize government manufacturing of generic drugs. Some private organizations and other nonprofit drug manufacturers have emerged in recent years, and a government-run manufacturing plant could ensure a continued supply of off-patent products that for-profit generic manufacturers have lost interest in producing. This idea has been proposed at both the national and state levels.
Conclusion
The pricing of remdesivir is emblematic of problems with the current process of developing and marketing prescription drugs in the United States. The policies described in this issue brief won’t discourage innovation, and they won’t reduce the availability of lifesaving and life-extending drugs—quite the opposite. They will turn the attention of successful companies away from efforts to create incrementally innovative products or to extend patent protection for their existing money-making products and toward the development of new ones. And they will relieve the financial burden that high prices place on consumers, businesses, and governments.
