Skip to main content
Jobseekers stand in line to attend the Dr. Martin Luther King Jr. career fair held by the New York State department of Labor in New York April 12, 2012. A report on Friday showed the economy created only 120,000 jobs last month, the fewest since October. The unemployment rate fell to a three-year low of 8.2 percent, but largely as people gave up the search for work.     REUTERS/Lucas Jackson (UNITED STATES - Tags: BUSINESS EMPLOYMENT) - GM1E84D0FT401

Using automatic stabilization programs to fight recessions and speed recoveries

Editor's Note:

This brief is part of the Brookings Blueprints for American Renewal & Prosperity project.

Blueprints for American Renewal & Prosperity



Every time there’s a big economic downturn, we scramble to agree on adjustments to Medicaid and the Unemployment Insurance Program, and we disagree about the size, scope, and timing of relief packages for states and localities. We should not be arguing about roof repairs during thunderstorms. Instead, we should preauthorize economic stabilization measures, the size and timing of which would we determined by quantitative triggers. When a state’s unemployment rate increases by a legislatively defined amount, the federal government’s matching payments for Medicaid and the Children’s Health Insurance Program (CHIP) would automatically increase, and it would qualify for extended and enhanced unemployment insurance as well. Direct assistance to states would be released on a state-by-state basis, so that states experiencing disproportionately large economic hits would get more per capita. One-third of this assistance would go directly to localities in a formula that reflects population, additional expenses to cope with the consequences of the downturn, and demonstrated revenue losses. As the downturn abates, federal aid in these three categories would be scaled back and would go to zero when conditions have returned to normal (as defined by legislation).

Back to top


As the United States began its long recovery from the Great Recession in 2010, Raymond C. Scheppach, former executive director of the National Governors Association, predicted that the steepest economic downturn since the 1930s would result in what he called a “Lost Decade” for states, which would not fully recover until “late in the next decade.” Sadly, his prediction turned out to be correct.

It was not until 2013 that state revenues, corrected for inflation, regained the level they had reached five years earlier. But the recovery was uneven. As late as 2018, nearly half the states were spending below their 2008 levels: More than half the states were spending less per capita for K-12 education, state funding for higher education was slashed by 13%, and state investment in infrastructure as a share of GDP hit its lowest level in more than half a century. Because state contributions to pension funds did not keep pace with their obligations, underfunding of state pension systems soared to $1.35 trillion by 2016 and remained elevated thereafter. State aid to localities shrank, as did the number of state employees.1

These changes in revenues and spending levels throughout our federal system had consequences with which the country continues to grapple. For example, reductions in state support for higher education forced public colleges and universities to raise tuition and fees. This increased the burden on students and their families and contributed to the student loan crisis that slows the economic progress of young adults today and delays long-term commitments such as homeownership. Spending for state public health departments fell by 16% and for local departments by 18%, while 38,000 state and local public health positions—about 12% of the total—were eliminated. As a result, when the COVID-19 pandemic hit, states were in a weaker position to respond effectively than they would have been a decade earlier.

“Not only did fiscal stringency make it harder for states to meet the needs of their citizens, it also slowed the national recovery from the Great Recession.”

Not only did fiscal stringency make it harder for states to meet the needs of their citizens, it also slowed the national recovery from the Great Recession. In a recent analysis, Ben Bernanke, former chair of the Federal Reserve during the Great Recession, argued that the $800 billion stimulus package enacted by Congress in 2009 was partly offset by cuts in spending and employment necessitated by steep state and local revenue declines and balanced budget requirements. “State and local budget cuts meaningfully slowed the recovery,” Bernanke concluded.

Two recent studies support his argument. In a contribution to the Brookings Papers on Economic Activity, authors Alan Auerbach, William Gale, Byron Lutz, and Louise Sheiner showed that between 2009 and 2018, states were a drag on economic recovery and growth. Not until 2019 did the states make a modest positive contribution to economic growth, a performance that is unlikely to be repeated in 2020. Another study found that, as meaningful federal aid to states and localities ground to a halt in the years after the Great Recession, declines in state and local government activities reduced GDP by 0.7% in the third quarter of 2012.

Back to top

Limits of historic and existing policies

Because balanced budget requirements limit the capacity of states to respond to economic downturns, only the federal government can take up the slack. Federal transfers to states and localities provide efficient and effective fiscal stimulus. But during the Great Recession, the national effort fell short. While states were still struggling, the 2009 stimulus bill expired in 2012 and nothing was done to replace it. Instead, concern about rising federal deficits and debt triggered restraints on discretionary spending that shifted federal policy toward austerity while unemployment remained elevated. This exacerbated the slowdown at the state and local level.

Now, says Bernanke, we are in danger of repeating this mistake. The failure of Congress to agree on the next phase of COVID-19 relief will force not only hard-hit sectors of the economy, such as airlines, but also state and local governments to slash spending, investment and jobs. The recovery will slow, the struggles of businesses and workers will continue, and people in need of assistance will suffer.

The problem we face is not trivial. Every percentage point increase in unemployment over the pre-recessionary baseline reduces revenues to states and localities by about 3.7%. Between 2008 and 2012, states and localities lost $283 billion in tax revenues, and the aftermath of the 2020 downturn will be worse. An increase of 5 percentage points in the unemployment rate will likely produce revenue losses of between 15% and 20%. State and local revenues are estimated to fall short of the pre-pandemic baseline by $155 billion this year, $161 billion in 2021, and $145 billion in 2022, and the shortfall is likely to persist for several years afterward.

“Although separated by a decade, these two deep recessions illustrate a common theme: The political will to stimulate the economy and maintain essential programs through increased federal spending wanes well before the need for it ends.”

Although separated by a decade, these two deep recessions illustrate a common theme: The political will to stimulate the economy and maintain essential programs through increased federal spending wanes well before the need for it ends. It is not hard to understand why. Not only do the major political parties disagree about the efficacy of this strategy, but also it is psychologically difficult to do what is necessary for as long as it takes. During recessions, the federal budget deficit expands even if the government appropriates no funds beyond what current legislation requires. In these circumstances, deciding to spend more and further increase the deficit goes against the grain. Because we would not act this way with our own money, we instinctively respond, “Why should we do so in our public policy?” This is all to say that, although the strategy of increasing government spending to counter downturns is supported by a mountain of economic research and practical experience, it challenges common sense. After the initial crisis gives way to a sustained downturn and a recovery that often is measured in years rather than months, the motivation to continue emergency assistance fades.

Back to top

Policy recommendations

There is a solution for this problem: We should do the equivalent of fixing the roof while the sun is shining. We should adopt policies that trigger different categories of federal spending—including assistance to states and localities—based on specific metrics such as rising unemployment and declining revenues. These policies, known as “automatic stabilizers,” would obviate the need for new legislation during economic downturns, and they would expire when the need for them ends rather than when the political will to extend federal assistance disappears.

We already use this strategy to some extent. For example, when higher unemployment shrinks family incomes, eligibility for some federal programs such as food stamps expands, and unemployment insurance payments rise in these circumstances as well. What follows is a sketch of one possible strategy that expands the use of this mechanism to buffer states against the consequences of economic downturns and to accelerate economic recoveries.

The overall objective of the proposed plan would be to replace 90% of the revenue losses each state experiences during a recession. States would be responsible for funding the remaining 10% by drawing on their rainy-day funds and lowering administrative costs.

Medicaid and the Children’s Health Insurance Program

A key focus of the plan is two large health-care programs—Medicaid and the Children’s Health Insurance Program (CHIP)—that are jointly funded by states and the federal government. Current legislation requires the federal government to match every dollar of state spending on these programs by at least one dollar. States with high per-capita incomes get the minimum, while poor states get more—in some cases, substantially more. For example, while the federal government pays for 50% of Connecticut’s Medicaid program, it pays for more than three-quarters of Mississippi’s.

The proposed reform would determine some non-recessionary baseline of each state’s unemployment rate and increase a state’s Medicaid and CHIP matching rates when its unemployment rate rises by a certain amount over its baseline. The cost of this program would be determined by four policy variables—the formula by which baseline unemployment is determined, the amount that unemployment must rise before the increase in matching rates is triggered, the percent point increase in the federal match for each percent point rise in unemployment, and the level to which unemployment must fall before the match terminates altogether. Based on current models, these parameters would be set to replace about one-third of states’ revenue losses. This would amount to about $50 billion for each of the next three years.

The Unemployment Insurance Program

A second key focus would be the Unemployment Insurance (UI) Program, created during the New Deal as a federal-state partnership. This system both mitigates the consequences of unemployment for workers and their families and acts as a macroeconomic stabilizer by compensating for a portion of the reduced consumption that joblessness produces.

The system is funded through a 6% tax on payroll (via The Federal Unemployment Tax Act) up to the first $7,000 of earned income. Although employers are responsible for paying the tax and remitting it to the federal government, they are eligible for a tax credit of 5.4%, leaving their maximum payment at $42 per employee. According to the Tax Policy Center, the federal government uses the funds to “cover administrative expenses, make loans to states that deplete their own reserves, and cover half of extended unemployment benefits made available when states experience prolonged periods of high unemployment. (States cover the other half of these extended benefits.)”

Over the past decade, the share of unemployed individuals receiving benefits has declined. A recent study has identified three partial explanations: Eight states have reduced the duration of benefits, many have stiffened administrative procedures—increasing the rate at which applications for benefits are denied—and because the number of workers experiencing extended periods of unemployment remained elevated for many years after the Great Recession, many unemployed individuals exhausted their eligibility for benefits well before they were able to find new jobs. Regrettably, as huge numbers of jobs are wiped out in the restaurant, hotel, and entertainment sectors, the rate of extended unemployment is likely to rise again in the aftermath of the COVID-19 pandemic.

There is a case to be made for scrapping the New Deal-era federal/state collaboration and making unemployment insurance a fully federal program. But to ensure wide support for needed changes, it is more practical to pursue reforms within the current structure. Among the key steps include:

  • To remain members of the system in good standing, each state should meet uniform federal requirements for eligibility, benefits, and information infrastructure, and they should adopt administrative procedures that encourage eligible workers to apply for benefits and process their claims efficiently. To deal with questionable claims, states should be required to establish review procedures to resolve disputes quickly and fairly.
  • The unemployment threshold that triggers an extension of UI benefits beyond the normal period should be simplified and aligned with the triggers for Medicaid and CHIP discussed above. When a state’s unemployment reaches this trigger-point, the federal government should pay 100% of the extended benefits, rather than the current 50%. The justification is straightforward: The higher a state’s unemployment, the more its revenues will decline, decreasing its ability to fund a benefits extension without depleting its UI reserves and borrowing from the federal government, as the current program permits. When states are forced to borrow in this manner, it typically takes them many years to repay the federal government, forcing cuts in other areas and slowing their economic recovery.
  • During severe downturns, such as the Great Recession and the COVID-19 recession, the federal government usually establishes emergency programs that increase benefits beyond what each state provides. These programs often start too late to mitigate the severity of the decline and end too early to help sustain the recovery. To correct these defects, a state’s entrance into the extended benefit program should also trigger federally funded emergency benefit supplements that would remain in place as along as the increased rate of unemployment persists. The level of supplemental benefits should be set so that the total benefits package would replace about 100% of low-income workers’ salaries and 75% of salaries for median-income workers. A recent study finds that, in states with average incomes, a supplemental payment of $200 per week would achieve these replacement levels.
  • The cost of the proposed UI reform is difficult to estimate, because everything depends on the depth and duration of the current downturn (and of future downturns as well). As a point of reference, increased federal outlays for extended benefits and emergency supplemental benefits averaged $50 billion per year between 2009 and 2013, when the country was struggling to recover from the Great Recession.2

Direct federal payments to states

The final component of the proposed plan would be a new program of direct payments to states, designed to replace about 30% of revenue losses relative to the pre-recession policy baseline; in the current downturn, this would average about $45 billion per year over the next three years.3 Triggers for entry into this program would include either a specified rise in a state’s unemployment rate or a specified loss in its revenues. Because unemployment statistics will typically be available before reliable data on revenue losses, most states would likely begin with the former. But the proposed program would allow them to switch if demonstrated revenue losses (for which unemployment is a proxy) turn out to be higher than the unemployment-based method assumes.

Unlike Medicaid, CHIP, and unemployment insurance programs that affect state finances, localities as well as states experience revenue losses. The current recession has revealed tensions between governors who want to retain control of federal emergency funds to address state budget deficits and localities that need a share of these funds to maintain essential services. This proposal would require states to distribute one-third of federal direct payment to their localities based on a formula that takes localities’ population, increased unemployment, and lost revenue into account.

Back to top


Taken together, these three components of the proposed plan would meet the objective of closing about 90% of state revenue losses during recessions. This would not solve all the problems in such periods, of course. For example, the receipts of transit systems often decline as fewer riders head to work each day. During the current pandemic, ridership has collapsed, plunging these systems into crisis.

Not only do revenues fall, but outlays also increase. For an undetermined period, states and localities must now refit schools, government offices, and other building to reduce the spread of the COVID-19 virus, and they must fund extensive testing as well. It remains to be seen whether the federal government will pay for the transportation and administration of a vaccine—assuming at least one meets rigorous standards of safety and efficacy—or leaves some of these expenses to the states.

Across the country, businesses have learned the hard way that their business interruption insurance policies are not designed to cover widespread health emergencies. The hastily designed Paycheck Protection Program, which has lapsed, helped many small businesses, but many others considered its ambiguous requirements a mixed blessing at best. As businesses fail in record numbers, more than 700,000 workers continue to file for unemployment benefits each week, and states and localities struggle to cope with foregone tax revenues.

“In the aftermath of the pandemic-induced recession, governments at every face will need to reinvent themselves to cope with the economic consequences of future disasters.”

In the aftermath of the pandemic-induced recession, governments at every face will need to reinvent themselves to cope with the economic consequences of future disasters. Using automatic stabilization programs to offset lost state and local revenues would be an important step, but it would address only part of a much broader systemic challenge.

Back to top


  1. Statistics in this paragraph via “‘Lost Decade’ Casts a Post-Recession Shadow on State Finances,” Pew Charitable Trusts, June 4, 2019.
  2. Calculation based on data from “Unemployment Compensation: The Fundamentals of the Federal Unemployment Tax (FUTA),” Congressional Research Service:
  3. Calculations based on “Fiscal Effects of COVID-19,” The Brookings Institution, September 24, 2020.


Get daily updates from Brookings