Program on Race, Ethnicity and the Economy (PREE)

Section 5. Benefit levels: Increase UI benefits to levels working families can survive on


Key proposals

  • Wage replacement rates: Increase replacement rates to levels that truly alleviate severe economic hardship, particularly for low-wage workers, with a progressive formula that replaces at least 85% of wages for the lowest earners and gradually decreases to replace 50% of wages for higher earners, 30% of wages for very high earners, and so on.
  • Dependent allowance: Help alleviate food and housing insecurity when parents or caregivers lose a job by providing a minimum dependent allowance of $35 (inflation adjusted) per dependent per week.
  • Subminimum wage earners and tipped workers: Treat subminimum wage earners fairly by calculating benefit amounts based on what they should have earned if they were paid the prevailing minimum wage or, for tipped workers, wages with tips, whichever is greater.
  • Part-time workers: Support rather than discourage part-time work as a bridge back to employment by implementing an earnings disregard that keeps part-timers’ UI benefits, combined with their part-time earnings, from falling short of their pre-layoff average weekly wage.
  • New entrants and reentrants: Support job seekers who are newly entering or reentering the labor market with an allowance of $200 per week or 20% of the state’s average weekly wage, whichever is greater.
  • Economic downturns: Make unemployment benefits a more powerful tool for recovery by increasing wage replacement rates, minimum and maximum benefit amounts, the dependent allowance, and the jobseeker’s allowance when extended benefits are triggered on:
    • 100% replacement for lowest-tier earnings and 65% of wages for highest-tier earnings
    • Minimum benefit the greater of $300 per week or 40% of the state’s average weekly wage (up from the greater of $250 or 30%)
  • Taxation of benefits: Increase UI take-up and get more UI benefits recirculating as spending in the economy by fully exempting households with less than $100,000 in adjusted gross income from federal taxation of UI benefits.


The problem

On average, unemployment insurance benefits replace just about 40% of a worker’s prelayoff wages. When accounting for the time-limited nature of UI income and the nonwage benefits that are part of the compensation package for many workers (benefits like employer-provided health insurance and retirement contributions), the true pay replacement rate is much lower.1 It is the states that set benefit levels, and states vary tremendously in the benefit levels they offer, due to different minimum and maximum benefits, as well as formulas for computing benefits based on wage history. There are currently no federal standards for UI benefits, and many states pay benefits that are so low they allow jobless workers to fall quickly into poverty during a spell of job loss (GAO 2020). In no state are UI benefits enough to cover a worker’s basic needs (Roberts and Schweitzer 2020).2

Black and Latino workers face greater hardship when they receive low levels of benefits because they are less likely than white workers to have private savings to use to supplement income from unemployment insurance (Gould-Werth 2021; Gould, Perez, and Wilson 2020; Gould and Wilson 2020). Women and workers of color are further disadvantaged because they are disproportionately clustered in low-paying jobs, and thus receive lower levels of benefits than their more highly compensated counterparts (Tucker and Vogtman 2020). In this way, the UI system entrenches and deepens existing inequities, particularly the systemic racism that has prevented workers of color from building family wealth (Bhutta et al. 2020; Kijakazi et al. 2019).

Workers with dependents, workers earning subminimum wages, tipped workers, and part-time workers all face particular difficulty making ends meet with the benefit levels to which they are entitled. In the current system people who are struggling as they seek to secure their first job or to reenter the labor market after an absence (for reasons that may extend from child-rearing to incarceration) receive no benefits at all. And despite the fact that unemployment spells are longer when economic conditions are worse, and that increasing unemployment benefits when the economy is in a downturn helps both individual workers and the macroeconomy by supporting spending, the regular UI program does not adjust benefit levels in response to economic trends (BLS 2021a; Bachas et al. 2020).

The solution

We can do better than a system that fails to alleviate severe economic hardship and deepens inequities. The policy proposals laid out below would increase the level of UI benefits, including making the benefit formula more progressive so that low-wage workers are better supported. The proposals also address the special difficulties that some groups of workers, particularly women and workers of color, face when trying to meet their need under the current system of determining unemployment benefit levels. Benefit levels should meet a guaranteed minimum in all states.


Benefits offered by state-run UI systems should meet federally guaranteed minimums. To aid states in the determination of benefit levels, the federal Department of Labor should establish a system that computes benefit amounts from wage records that are collected in national data sets such as the Quarterly Census of Employment and Wages (QCEW). Given that the recent work and earnings history of potential UI recipients is fully contained in QCEW data, it would be a straightforward task to compute each potential recipient’s UI benefit in a system where benefit levels were harmonized across the country. This would be a huge administrative efficiency improvement when compared with today’s patchwork system of benefit formulas that vary significantly by state.

Determining replacement rates

Each state sets its own formula for determining weekly benefit amounts—so the share of weekly wages that is “replaced” by unemployment insurance benefit payments vary widely by state. While the wage replacement rate thus varies by state, the U.S. Department of Labor’s Employment and Training Administration tracks replacement rates by state and region and computes a national average. Except for a brief period in the UI program’s early years, the wage replacement rate has remained below 40% (O’Leary and Rubin 1997; U.S. DOL-ETA 2021).

Context: How current low and uneven wage replacement rates fall short of what workers need

While UI benefits are higher for low-wage workers, they rarely exceed 50% of a worker’s previous wages. Social scientists agree that current wage replacement rates in the U.S. are too low (von Wachter 2019).

Because UI benefits replace only wages, and not benefits such as employer-paid health insurance premiums and employer-paid retirement contributions, the replacement rate of total compensation is even lower. Typically about 30% of what employers spend on employee compensation goes toward benefits (BLS 2021b), but UI payments replace only a portion of wages and ignore the much higher true monetary benefits that workers receive from work. Changes in the law in the 1970s and 1980s that made UI benefits taxable also reduce the value of benefits relative to workers’ prelayoff earnings.

Black and Latino workers are particularly harmed by low benefit amounts. For one, while many states have made it harder to access benefits in recent years, the states with the deepest cuts to their UI programs are mostly Southern states and have higher shares of Black residents (Kofman and Fresques 2020). Further, Black and Latino households have far lower private wealth on average than white households, and lower levels of wealth mean there is less ability to absorb the shock of income loss stemming from a spell of joblessness (Bhutta et al. 2020; Gould and Wilson 2020; Gould, Perez, and Wilson 2020).

In addition, there is a real problem in this country of workers being underpaid for their work, whether it is because they are deprived of overtime pay or paid a subminimum wage (Cooper and Kroger 2017). This reduces their benefits amounts and makes it difficult for them to cover basic necessities.

Policy proposal: Increase replacement rates using a progressive structure

To better assist all workers in meeting their basic needs while they search for work, UI benefit levels should be harmonized across all states by the U.S. Department of Labor and set progressively to provide low-wage workers with a higher replacement rate of pre-termination wages than their high-wage counterparts. Under the proposed progressive system for calculating UI benefits, different benefit rates apply to different levels of earnings for each worker, with rates falling as earnings rise (kind of like how marginal tax rates work, but in reverse). Table 5.1 shows the proposed replacement rates for earnings at different wage levels. The calculation used to determine the level of wage replacement at each level of earnings begins with dividing total earnings across the base period by the number of weeks in the base period to determine a worker’s average weekly wage. This number should be compared with the state’s average weekly wage (AWW) to determine the level of wage replacement. As shown in the table, under this progressive replacement rate structure, UI replaces 85% of wages a worker earns up to 50% of the state AWW, 70% of wages a worker earns that are between 51% and 100% of the state AWW, and 50% of wages a worker earns that are greater than 100% of the state AWW.

Table 5.1

Progressive replacement rates for unemployment benefits by earnings level and under a maximum benefit cap

Earnings level

For that portion of a worker’s average weekly wage (AWW) that is

Replacement rate

The share of lost wages replaced by UI benefits is

Between 0% and 50% of state AWW 85%
Between 51% and 100% of state AWW 70%
Over 100% of state AWW 50%
Minimum benefit amount The greater of 30% of state AWW or $250 per week
Maximum benefit amount 150% of state’s AWW

Note: The earnings level is a worker’s average weekly wage as a share of state average weekly wage during the base period. Under the proposed progressive system for calculating UI benefits, different benefit rates apply to different levels of earnings for each worker, with rates falling as earnings rise (kind of like how marginal tax rates work, but in reverse). See Table 5.2 for how this translates to effective tax rates for workers at different earnings levels.

Copy the code below to embed this chart on your website.

Policy proposal: Set minimum and maximum benefit levels that ensure adequate benefit levels

Though formulas that set benefit levels as a percentage of weekly wages work well for workers in the middle of the earnings distribution, they are less effective for workers at the top and bottom of the earnings distribution: Low earners may end up with benefit levels that are too low to help them handle living expenses when their work income is reduced. State UI systems have recognized this issue and set minimum and maximum benefit levels. However, minimum and maximum benefit levels vary greatly across the states, and there is no federal standard states must meet.

Scant weekly benefits don’t pay bills for out-of-work restaurant worker

Ruby De La Rosa, Massachusetts

When the pandemic hit, I was given UI almost immediately, as my restaurant had closed for quarantine. At first, it was so easy and helpful. [But] once the additional $600 was removed from UI, I could no longer afford my rent, car payment, and other bills, [as I was] only approved for $174 a week. This is an unlivable wage. When I was working prior to the pandemic, I was making $600+ per week as a waitress.

We recommend setting standards for minimum and maximum benefit amounts as follows:

  • minimum benefit amount of at least 30% of the state’s average weekly wage or $250 per week,3 whichever is greater
  • maximum benefit amount of no more than 150% of the state’s average weekly wage

The choice to cap benefits for any individual worker at an amount equal to 150% of the state’s AWW is consistent with a sound social insurance principle—followed, for example, by the Social Security program—that benefits should rise in line with income so long as tax contributions rise as well.4 Table 5.2 shows how the rates and caps outlined in Table 5.1, when applied to earnings, produce progressive effective replacement rates for workers at a variety of points along the earnings distribution.

Table 5.2

Examples of effective replacement rates for unemployment benefits by worker earnings level, accounting for minimum and maximum benefit amounts

Earnings level

If a worker’s average weekly wage is

Effective replacement rate

The worker’s unemployment benefits equal

50% of state’s average weekly wage At least 85% of preseparation wages*
100% of state’s average weekly wage 78% of preseparation wages
125% of state’s average weekly wage 72% of preseparation wages
150% of state’s average weekly wage 68% of preseparation wages
200% of state’s average weekly wage 64% of preseparation wages
300% of state’s average weekly wage 50% of preseparation wages
500% of state’s average weekly wage 30% of preseparation wages

Note: The effective replacement rate is how much of a worker’s preseparation wages the worker gets in UI benefits. It is the net effect when the different replacement rates in Table 4.1 are applied to different portions of preseparation wages and added together and any minimums or maximums accounted for. The "At least" qualifier in row one accounts for the $250 floor on minimum benefits. If $250 > 85%*50% of the state’s average weekly wage, then the worker would get more than 85% of their preseparation wage.

Copy the code below to embed this chart on your website.

Appendix Table 5a at the end of this section provides a detailed look at the calculations illustrating the progressive wage replacement structure proposed here, using examples that assume that the state’s average weekly wage is $1,000.

Ensuring that benefit levels are adequate for all groups of workers

Policy proposal: Implement a dependent allowance

Households with children are much more likely to face food and housing insecurity when a job is lost (Keith-Jennings, Nchako, and Llobrera 2021). Systemic racism contributes to this vulnerability for households headed by women of color, as women of color are overrepresented in low-wage work (Matthews and Wilson 2018). Women of color also are especially vulnerable as they make up a disproportionate share of single-parent households. Nondiscretionary expenses are higher for workers caring for dependents, such as young children, adult children with disabilities, aging parents, and full-time college students. Many of these expenses, such as child care, tuition, diapers, groceries, medical prescriptions, and care giving, can’t be reduced without great harm when households are faced with the income constraints associated with job loss.

Ten states currently offer dependent allowances, though amounts and eligible dependents vary by state (Whittaker and Isaacs 2019; U.S. DOL-ETA 2020). A national minimum standard is called for: In all states unemployment benefits should include additional supports for workers caring for dependents. We propose adding a dependent allowance of at least $35 per week per dependent (indexed to inflation) to workers’ benefits.

The definition of eligible dependents should be broad and include at least:

  • children ages 18 and younger, including foster children, stepchildren, and children for whom the worker has at least 50% custody in shared-custody arrangements
  • full-time students, up to the age of 26
  • nonworking adults in the household ages 60 and older, such as workers’ parents
  • adults with disabilities in the household

Table 5.3 shows how this change, combined with implementation of the progressive replacement rates outlines above, would affect benefit levels for hypothetical workers in Mississippi and Massachusetts.

Table 5.3

How progressive replacement rates and a dependent allowance would affect unemployment benefits for low-wage earners in Mississippi and Massachusetts

Worker profile State average weekly wage Preseparation earnings Current weekly benefit amount and dependent allowance Proposed weekly benefit amount and dependent allowance Total proposed vs. current benefit amount
Mississippi single mother supporting two children and disabled parent $800 $400 per week $200 UI benefits

$0 dependent allowance

$340 ($400 x .85) wage replacement

$105 dependent allowance ($35 per dependent)

$445 per week vs. $200
Massachusetts single mother supporting two children and disabled parent $1,200 $400 per week $250 UI benefits

$0 dependent allowance

$340 ($400 x .85) wage replacement

$105 dependent allowance ($35 per dependent)

$445 per week vs. $250


Copy the code below to embed this chart on your website.

Policy proposal: Calculate benefit amounts for subminimum wage earners and tipped workers based on the prevailing minimum wage or the worker’s wages with tips, whichever is greater, so that subminimum wage workers aren’t penalized twice by their low earnings

There are 3 million tipped workers—disproportionately women and people of color—and over 100,000 workers with disabilities who are paid a subminimum wage (Allegretto and Cooper 2014; West 2019). Further, employers of some specific classes of workers, such as certain workers with disabilities, full-time or young workers in some sectors, and a few categories of agricultural workers carved out from the minimum wage provisions of the Fair Labor Standards Act, may be allowed to pay less than the prevailing minimum wage. Many workers are also paid less than the amount they are legally entitled to under the law. This phenomenon, known as wage theft, includes not paying promised pay, paying workers less than the legal minimum wage, failing to provide legally required overtime pay, confiscating tips, taking illegal deductions from wages, asking employees to work off the clock, and misclassifying employees as independent contractors to pay a wage lower than the legal minimum. Wage theft disproportionately hurts low-wage workers and has a very high cost for workers, as shown in a study on just one form of wage theft, which found that in the 10 most populous states, 2.4 million workers are paid below minimum wage, with their losses averaging about $3,300 per year per worker (Cooper and Kroeger 2017).5

Whether workers receive subminimum wages legally or illegally, their low earnings depress their UI benefit levels. In cases where workers qualify for unemployment insurance benefits but their reported wages are below the prevailing minimum wage,6 we propose calculating their benefits as though they had earned the prevailing minimum wage. Tipped workers who earn a subminimum wage should be able to self-report their tips and then have their UI benefits calculated on the basis of their earnings with tips or the prevailing minimum wage, whichever is greater. Finally, it is worth highlighting that the minimum benefit formula outlined above should go a long way in improving UI benefits for workers earning subminimum wages.

Policy proposal: Mandate that administrative data collected from UI programs across the country include hours worked by employee, so that programs can calculate more adequate benefit levels for subminimum wage workers

Note that there are data collection implications for the subminimum wage earner benefits proposal. Calculating what workers would have earned had they earned the minimum wage requires multiplying their hours worked by the minimum wage. Currently, only a small number of states collect data on hours worked by individuals. As part of their data collection and reporting to the federal Quarterly Census of Employment and Wages or QCEW (the federal database of administrative data from UI programs across the country), states would have to collect data on the hours worked by individuals. But this data capacity would already be expanding as we pursue the recommendations made in other sections of this report. Specifically, in Section 2, on finance, we recommend that states switch to an experience rating method for determining an employer’s UI tax rate that measures the hours worked of each employer’s workforce (instead of basing the tax rate on unemployment insurance claims filed by the firm’s workers). If states are already collecting this data, there is only a small incremental effort needed to record these hours by worker. Given that states such as Minnesota and Washington already maintain these records, hours collection is well within the capacity of other state systems. Collecting hours-worked in the QCEW provides important benchmarks for other federal data surveys, and thus would also offer significant informational benefits for researchers and policymakers.

Policy proposal: Pay fair benefits to part-time workers

State UI systems badly mistreat underemployed workers, imposing harsh and economically irrational rules on workers who return to the workforce on a part-time basis. In theory, most states allow a worker to continue to claim UI even if they collect some part-time wages, but existing state rules make little sense. Workers can earn up to a set amount, known as the “disregard,” but then every dollar earned above that amount reduces UI benefits dollar for dollar. In effect, this is a remarkably inefficient tax system: it imposes a zero tax rate up until a cliff, then a 100% rate after the cliff. Further, in many states, the earnings disregard is so low—$30 or $40—that there is no real point to working at all.

In a UI program with minimum federal standards, the earnings disregard should be set at a level that enables part-time workers to access unemployment insurance while they search for full-time work, and encourages unemployed workers to work part time as they search for full-time work. We propose implementing an earnings disregard in which all earnings from part-time work are disregarded until the sum of a worker’s part-time wages and UI benefit exceeds 110% of prelayoff wages.7 Workers’ combined benefits plus part-time earnings could modestly exceed their previous weekly wage because even at equal earnings, part-time work typically is not a full replacement, either in terms of benefits or security. As an alternative, federal law could require a disregard of no less than 50% of the worker’s average weekly benefit, as several states now permit.

Arts worker pays high penalty for working part time

Nate Claus, New York

As a freelance union stage manager for theater, dance, opera, and other live events, when my show closes, it almost always means that the entire company shuts down. That puts us all out of work simultaneously. Because of this cycle, folks in my business rely on a combination of “day jobs” and unemployment insurance to help us to survive as we search for the next gig.

Unfortunately, day jobs, by nature, are often on call or part time. I have had to accept part-time work for earned wages that reduced my unemployment benefits to the degree that my income those weeks was less than it was for weeks where I didn’t work at all. If our UI system were set up in a way that put unemployed workers first, it would make such a critical difference in my life.


Policy proposal: Support job seekers newly entering or reentering the labor force with an allowance of $200 per week or 20% of the state’s average weekly wage

The unemployment insurance system should be reimagined to support all job seekers in the labor force, not just those with a narrowly defined qualifying loss of a job under a traditional employer-employee relationship. In addition to supporting them with financial UI benefits, the prospect of financial benefits may encourage UI applications from job seekers and bring them into other employment assistance and workforce development programs. As described in the eligibility chapter of this report, we propose establishing a jobseeker’s allowance to workers newly entering the labor force and searching for work, and workers who are reentering the labor force after taking time away. We propose that the benefits for eligible job seekers should be set at the greater of 20% of the state’s average weekly wage or $200 per week (indexed to median wage growth). Workers eligible for the jobseeker’s allowance would also be eligible for dependent allowances.

Setting benefit levels during economic downturns

Policy proposal: Make weekly benefit amounts responsive to periods of economic downturn

Benefits should be increased when labor demand is depressed, not only to better support workers, but also to provide more effective macroeconomic stabilization. When the economy is depressed and vacancies fall below the number of jobless workers, it is low economywide spending that is holding back employment growth, not workers who aren’t looking hard enough for jobs. More-generous UI benefits can help boost this spending. Yet until 2020, the only past effort to increase UI benefits to respond to the business cycle was a relatively tiny $25 weekly addition during the Great Recession.

These past failures to boost the low baseline levels of UI benefit generosity during recessions hampered the system’s ability to serve as an effective macroeconomic stabilizer. Indeed, evidence from increased benefit levels during the COVID-19 crisis indicates that increasing benefit levels during moments of macroeconomic contraction results in large boosts to aggregate demand and strong spending.8

As described in Section 4, on benefit duration, our proposed extended benefits triggers are reliable indicators of macroeconomic contraction. In periods when extended benefits trigger on, we propose adjusting the replacement rate formulae for all benefits as follows:

  • 100% for wages up to 50% of the state’s average weekly wage, up from 85% in non-EB periods
  • 85% for wages between 51% and 100% of the state’s average weekly wage, up from 70% in non-EB periods
  • 65% for wages greater than 100% of the state’s average weekly wage, up from 50% in non-EB periods

In addition, we propose:

  • increasing the minimum benefit to the greater of $300 per week or 40% of the state’s average weekly wage
  • increasing the weekly allowance per dependent to $50 (indexed to inflation)

Finally, we propose increasing jobseeker’s allowance benefits to the greater of $250 or 30% of the state’s average weekly wage per week (with the $250 threshold indexed to median wage growth).

Reforming taxation of UI benefits

The real value of UI benefits to claimants is reduced through taxation. UI benefits were fully tax exempt for the first 40 years of the U.S. system, but then taxes were phased in between 1978 and 1986. The historic rationale for imposing UI taxes was to reduce the desirability of receiving UI relative to the desirability of working.9 Yet as described in Section 4, contemporary research shows that there is little empirical basis for concerns about the economic effects of increasing the take-home value of UI benefits.10

Indeed, taxing benefits adds additional complexity and compliance costs to the system and reduces take-up of UI benefits (Anderson and Meyer 1997), preventing the program from effectively serving its dual mandate of smoothing consumption for workers who lose jobs through no fault of their own and stabilizing the macroeconomy. There are, however, some reasons to tax UI benefits. In the absence of UI benefit taxation, well-compensated workers who draw benefits part of the year, or high-income dual-earner households with one separated worker, will receive a higher after-tax replacement rate than low earners.

Policy proposal: Exempt UI income from federal taxes for low-income households

We recommend exempting UI income received by low-income households from federal income tax.11 This will increase take-up and increase effective benefit levels for households with the highest marginal propensity to consume—to spend their benefits and thus help boost aggregate demand—thereby improving UI’s ability to function as a macroeconomic stabilizer (Bivens 2017). At a minimum, to reduce the extent to which federal taxes pull fiscal resources out of states with more-generous UI systems, federal income tax on UI benefits could be redeposited in state trust funds.

While we strongly recommend exemption, were low-earner benefits to be taxed, we would recommend allowing them to be treated as “earnings” for Earned Income Tax Credit (EITC) purposes.

Policy proposal: Phase out exemption so that UI benefits received by households whose  adjusted gross income exceeds $150,000 are fully taxed

To increase equity when generating federal revenue, we recommend phasing out exemption for high-income households. The American Rescue Plan fully taxed UI benefits for households with adjusted gross income greater than $150,000, and we recommend following that benchmark, although with a more gradual phase-out beginning at $100,000 AGI.

Appendix Table 5A

Sample weekly benefit amount calculations in a state where average weekly wage (AWW) is $1,000

Wage replacement in wage category
Worker’s AWW during base period 0%–50% of state AWW 51%–100% of state AWW >100% of state AWW Sum of replaced wages, no min or max Weekly benefit amount Effective replacement rate Explanation
$200 0.85 *$200 = $170 0.7*0 = 0 0.5*0 = 0 $170 $300 150% Sum of replaced benefits lower than minimum benefit ($250 or 30% of state AWW, or $300 in this case), so minimum benefit received
$350 0.85*$350 = $297.50 0.7*0 = 0 0.5*0 = 0 $298 $300 86% Sum of replaced benefits lower than minimum benefit ($250 or 30% of state AWW, or $300 in this case), so minimum benefit received
$750 0.85*$500 = $425 0.7*250 = $175 0.5*0 = 0 $600 $600 80% Sum of replaced wages is greater than minimum and less than maximum—follows formulas exactly
$1,500 0.85*$500 = $425 0.7*500 = $350 0.5*$500 = $250 $1,025 $1,025 68% Sum of replaced wages is greater than minimum and less than maximum—follows formulas exactly
$3,000 0.85*$500 = $425 0.7*500 = $350 0.5*$2,000 = $1,000 $1,775 $1,500 50% Sum of replaced wages is greater than maximum—maximum benefit applies
$5,000 0.85*$500 = $425 0.7*500 = $350 0.5*$4,000 = $2,000 $2,775 $1,500 30% Sum of replaced wages is greater than maximum—maximum benefit applies

Copy the code below to embed this chart on your website.


1. For example, see Table 1 in the federal government periodic summary of what employers spend on employee compensation for the shares attributable to wages versus benefits (BLS 2021b).

2. According to Roberts and Schweitzer (2020), “there isn’t a single county, parish, borough, census area, or city in the United States where state unemployment benefits alone can cover a set of very modest monthly expenses (for example, the 40th percentile of area rent) for a typical one-adult, one-child family.”

3. The $250 weekly benefit amount would change over time, as it would be indexed to median wage growth.

4. A key proposal in section 2 of  this report would increase the taxable maximum for the base of the Federal Unemployment Tax Act (FUTA) tax to be equal to the taxable maximum for the base of the Social Security (Federal Insurance Contributions Act or FICA) tax, which was just under $143,000 in 2021. A worker earning $143,000 annually makes roughly $2,750 each week, and under the formulas proposed in this section would qualify for a UI benefit of roughly $1,775. This rounds up roughly to 150% of the average weekly wage in 2020, which was just under $1,200 nationally.

5. See Cooper and Kroeger 2017 for an overview of wage theft in the U.S. economy and an estimate of its reach.

6. At the local, state, or federal level.

7. If macroeconomic conditions warrant the triggering of extended benefits, it is beneficial to the macroeconomy to raise benefit levels, so during extended benefit periods, we propose that all earnings from part-time work be disregarded unless the sum of a worker’s part-time wages and UI benefit exceeds 125% of their prelayoff wages.

8. See Ganong et al. 2021 for evidence on the large macroeconomic boost provided by pandemic-related UI expansions, including increases in benefit generosity.

9. It is sometimes argued that UI benefits should be taxed because they are replacements for wages, but this is also true in a general sense for benefits from Social Security, Supplemental Nutrition Assistance Program (SNAP), Temporary Assistance for Needy Families (TANF), and other safety-net programs. These programs are generally untaxed or only partly taxed because it is simpler to adjust benefit levels directly, rather than issuing benefits and then taxing back part of the value of that benefit. Further, federal taxation of state-funded UI benefits is effectively a federal surtax on state UI tax collections (Galle and Pancotti 2021). For any given target replacement rate a state might choose, federal taxation of benefits requires the state to pay more, and therefore to impose more in state taxes, to achieve the identical replacement rate.

10. The literature on the employment effects of potential benefit duration discussed in Section 4 is applicable here, as longer durations increase the net take-home value.

11. The arguments for exempting state UI benefits from federal tax are stronger than the case for states exempting their own benefits. We do not recommend requiring states to exempt UI from their own income tax, although federal exemption would automatically result in state exemption in about half of states.


Allegretto, Sylvia, and David Cooper. 2014. Twenty-Three Years and Still Waiting for Change: Why It’s Time to Give Tipped Workers the Regular Minimum Wage. Economic Policy Institute, July 2014.

Anderson, Patricia M., and Bruce Meyer D. 1997. “Unemployment Insurance Takeup Rates and the After-Tax Value of Benefits.” Quarterly Journal of Economics 112, no. 3 (August): 913–937.

Bachas, Natalie, Peter Ganong, Pascal J. Noel, Joseph S. Vavra, Arlene Wong, Diana Farrell, and Fiona E. Greig. 2020. “Initial Impacts of the Pandemic on Consumer Behavior: Evidence from Linked Income, Spending, and Savings Data.” National Bureau of Economic Research working paper no. 27617, July 2020. DOI 10.3386/w27617

Bhutta, Neil, Andrew Chang, Lisa Detting, and Joanne Hsu. 2020. “Disparities in Wealth by Race and Ethnicity in the 2019 Survey of Consumer Finances.” Federal Reserve Board FEDS Notes, September 28, 2020.

Bivens, Josh. 2017. Inequality Is Slowing U.S. Economic Growth: Faster Wage Growth for Low- and Middle-Wage Workers Is the Solution. Economic Policy Institute, December 2017.

Bureau of Labor Statistics (BLS). 2021a. “Duration of Unemployment” (data graphic related to Employment Situation Summary April 2021).

Bureau of Labor Statistics (BLS). 2021b. “Employer Costs for Employee Compensation –December 2020” (news release). March 18, 2021.

Cooper, David, and Teresa Kroeger. 2017. Employers Steal Billions from Workers’ Paychecks Each Year. Economic Policy Institute, May 2017.

Galle, Brian, and Elizabeth Pancotti. 2021. The Case for Forgiving Taxes on Pandemic Unemployment Aid. The Century Foundation, February 2021.

Ganong, Peter, Fiona Greig, Max Liebeskind, Pascal Noel, Daniel M. Sullivan, and Joseph Vavra. 2021. “Spending and Job Search Impacts of Expanded Unemployment Benefits: Evidence from Administrative Microdata.” Becker-Friedman Institute Working Paper No. 2021-19, February 2021.

Gould, Elise, and Valerie Wilson. 2020. Black Workers Face Two of the Most Lethal Preexisting Conditions for Coronavirus—Racism and Economic Inequality. Economic Policy Institute, June 2020.

Gould, Elise, Daniel Perez, and Valerie Wilson. 2020. Latinx Workers—Particularly Women—Face Devastating Job Losses in the COVID-19 Recession. Economic Policy Institute, August 2020.

Gould-Werth, Alix. 2021. Research on Race and Consumption Shows Why Racial Disparities in Unemployment Insurance Receipt are Detrimental to Economic Recovery in the United States. Washington Center for Equitable Growth, March 2021.

Government Accountability Office (GAO). 2020. COVID-19: Urgent Actions Needed to Better Ensure an Effective Federal Response. Report to Congressional Committees. November 2020.

Keith-Jennings, Brynne, Catlin Nchako, and Joseph Llobrera. 2021. Number of Families Struggling to Afford Food Rose Steeply in Pandemic and Remains High, Especially Among Children and Households of Color. Center on Budget and Policy Priorities. April 2020.

Kijakazi, Kilolo, K. Steven Brown, Donnie Charleston, and Charmaine Runes. 2019. What Would It Take to Overcome the Damaging Effects of Structural Racism and Ensure a More Equitable Future? Urban Institute. May 2019.

Kofman, Ava, and Hannah Fresques. 2020. “Black Workers Are More Likely to Be Unemployed but Less Likely to Get Unemployment Benefits.” ProPublica. August 24, 2020.

Matthews, Madison, and Valerie Wilson. 2018. “Separate is Still Unequal: How Patterns of Occupational Segregation Impact Pay for Black Women.” Working Economics Blog (Economic Policy Institute), August 6, 2018.

O’Leary, Christopher, and Murray Rubin. 1997. “Adequacy of the Weekly Benefit Amount.” In Unemployment Insurance in the United States: Analysis of Policy Issues, edited by Christopher O’Leary and Stephen Wandner, 163–210. Kalamazoo, Mich.: W.E. Upjohn Institute for Employment Research.

Roberts, Lily, and Justin Schweitzer. 2020. “You Can’t Afford to Live Anywhere in the United States Solely on Unemployment Insurance.” Center for American Progress. September 2020.

Tucker, Jasmine, and Julie Vogtman. 2020. When Hard Work is Not Enough: Women in Low-Paid Jobs. National Women’s Law Center, April 2020.

U.S. Department of Labor Employment and Training Administration (U.S. DOL-ETA). 2020. “Monetary Entitlement” in Comparison of State Unemployment Laws 2020,

U.S. Department of Labor Employment and Training Administration (U.S. DOL-ETA). 2021. UI Replacement Rates Report (online database). Accessed April 2021.

von Wachter, Till. 2019. “Unemployment Insurance Reform.” Annals of the American Academy of Political and Social Science 686, no. 1 (November): 121–146.

West, Rachel. 2019. “Everyone is Overlooking a Key Part of the New $15 Minimum Wage Bill.” TalkPoverty, June 19, 2019.

Whittaker, Julie M., and Katelin P. Isaacs. 2019. Unemployment Insurance: Programs and Benefits. Congressional Research Service (CRS) Report RL33362, October 2019.