Above photo: People wait in line as SF-Marin Food Bank hands out 1600 food bags in San Francisco on April 20, 2020. Work furloughs and layoffs are driving thousands to seek food assistance. San Francisco Chronicle | Hearst Newspapers via Getty Images.
5.3 million workers will likely lose their jobs by the end of 2021.
Last week, EPI hosted a bipartisan panel of economists who called upon policymakers to pass significant federal aid for state and local governments in coming months. This panel’s judgement was unanimous that federal aid for subnational governments is crucial for helping the economy mount a rapid recovery from the current crisis. In this post, we highlight that:
- If policymakers do nothing at the federal level to address these shortfalls, the United States could end 2021 with 5.3 million fewer jobs, with losses in every state.
- Further, if Congress passes some level of aid that is insufficient—less than $1 trillion—they will needlessly guarantee a significant job gap by the end of 2021.
- If they pass $500 billion of aid over that time, the jobs gap will likely be roughly 2.6 million. If they pass $300 billion of aid, the jobs gap will likely be roughly 3.7 million.
- While empirical estimates of the shortfall should guide policymakers’ thinking, they can (and actually should) avoid putting a firm sticker price on state and local aid by tying this aid to economic conditions. If the economy recovers faster than the forecasts driving the $1 trillion estimated shortfall indicate will happen, then less aid would be needed. If instead recovery lagged, more would be needed.
- Finally, filling in the estimated shortfalls would merely return state and local governments to their pre-crisis fiscal status quo. But the unique features of the current economic shock will put greater demands on public services than existed before the crisis. To go beyond macroeconomic stabilization and promote the general welfare, even more federal aid to these governments is likely needed.
Because a weakening economy undercuts state and local tax revenues, and because states operate under balanced budget constraints, the coming months will see intense downward pressure on state and local spending. Reductions in this spending will in turn significantly slow recovery from the current economic crisis. This is not an abstract concern—the historically slow recovery in state and local spending following the Great Recession by itself delayed a recovery in unemployment to pre-crisis levels by four full years.
Recent estimates indicate that state and local governments will face a shortfall approaching $1 trillion between now and the end of 2021. The methodology behind this estimate is straightforward: High-quality research shows that each one-percentage-point rise in the unemployment rate leads to a budget shortfall for state governments of $45 billion. Given that local government revenues are nearly two-thirds as large as state revenues, a conservative adjustment would imply that each percentage-point increase in the unemployment rate increases state and local budgets shortfalls by $70 billion. With these estimates in hand, researchers have compared the forecasted path of unemployment rates over the next seven quarters and multiplied the excess of forecasted unemployment over the first quarter unemployment rate (3.8%) by this $70 billion.1
If this $1 trillion shortfall is not filled in by the end of 2021, then state and local government spending would be roughly $430 billion lower at the end of 2020 and $570 billion lower at the end of 2021. Each dollar in state and local spending cuts triggers a multiplier effect as governments end contracts with local businesses and public-sector employees see income drops and, in turn, pull back on their consumption spending. After accounting for these ripple effects, the shortfall in public spending will lead to losses in overall gross domestic product (GDP) of just under $800 billion by the end of 2021. This $800 billion represents about 3.7% of forecasted GDP by the end of 2021.2
If this percent reduction in GDP translates one-for-one into employment reductions, this would imply 5.3 million job losses, based on current estimates of the level of payroll employment by the end of 2021. All told, without aid to state and local governments, 5.3 million workers will likely lose their jobs by the end of 2021.
This job shortfall scales with the level of the overall state and local fiscal shortfall. If Congress passes only $500 billion in relief, then the shortfall will be smaller, but will still exceed 2.6 million jobs by the end of 2021. If Congress passes only $300 billion, then the shortfall will be 3.7 million. We should note that a job shortfall of 3.7 million would exceed the entire employment losses seen in the recessions of the early 1990s and early 2000s. In short, $300 billion in aid to state and local governments would not even move the economy’s health to the level it sat at during recent recessions—and this is 18 months from now.
Of course, we don’t know for sure how large the fiscal shortfall will turn out to be. It could be smaller if recovery is faster than forecasted. It could also be larger, depending both on the pace of recovery and on the unique demands this particular crisis is putting on state and local governments. These governments are on the front lines of health and education investments, and both health and education have been profoundly affected by the coronavirus crisis.
Luckily, we don’t need to guess how much aid will be needed—we could make this aid contingent on economic conditions rather than arbitrary calendar timelines. But, if Congress unwisely follows past practice and insists on arbitrary calendar timelines and set amounts of aid, they should use the right estimates. And these right estimates argue that anything less than $1 trillion of aid will threaten to drag significantly on growth over the next 18 months.
Below we allocate these national losses across states. We allocate the national estimates to states based on an average of three state-specific weights: the state’s share of private employment, the share of state and local employment, and the share of overall state and local spending. These three components roughly capture the distribution of how state and local spending cutbacks ripple through the economy. Previous research indicates that more than half of the effect of declining state and local spending is felt in private-sector employment reductions. Some of this effect on the private sector is due to multiplier effects, and some is due to the fact that private firms often supply inputs into public-sector activities (think textbooks for schools and automobiles for police forces).
Further, while some state and local spending cutbacks hit public-sector workers directly, some of these cutbacks are in the form of reduced income transfers, and this can hit both public and private sectors.
Finally, the current economic crisis is unique in some important ways—besides just its size and suddenness—that will require even more aid. First, whereas recessions tend to start elsewhere and then sweep up low-wage workers in the spillover damage, the current crisis is hitting low-wage workers first. Second, it’s a health shock—both to spending and in terms of the intense demands it is putting on the narrow sector of the health care complex that must deal with COVID-19. Third, it is directly affecting children and education through school shutdowns.
All of these factors will put historically large strains on state and local governments even above and beyond the strains that result from high unemployment and depressed economic activity. For example, state and local governments often are the ones dealing with the crush of safety net applicants for programs like unemployment insurance (UI). Crucially, these governments are also on the front lines of the public health response to the coronavirus crisis. In order to ensure a safe return to more normal economic activity, they will need to ramp up public health spending significantly on resources like testing and tracing.
Perhaps the single biggest challenge to returning to more normal life with the virus is how to open schools safely. However this happens, there is no doubt at all that it will be more expensive. Kids and teachers will need testing and personal protective equipment (PPE), at a minimum. Further, there may be needs to rotate kids in and out of school and have some taught online while others are simultaneously at school physically on given days. This requires more, not fewer, school personnel to manage and maintain educational quality.
Pushing the economy back to macroeconomic stabilization is a vital goal for policymakers, and if they fall short of this goal, millions will suffer. But the unique features of the coronavirus economic shock also mean that a larger and better functioning public sector will be needed for at least a few years to meet the health and educational challenges the disease has put in front of those sectors. All of this argues for going big, not cheap, when Congress decides how to help state and local governments in coming months.
1. As of late May, Tim Bartik of the Upjohn Institute calculated that this method implied state and local government shortfalls of just under $1 trillion by the end of 2021. Recently, he used a more updated projection from the Congressional Budget Office (CBO) that sees lower unemployment over the next 18 months, and the estimated shortfall has fallen to closer to $900 billion. However, there are reasons to think this improved unemployment rate forecast by the CBO should be taken with a large grain of salt. For example, CBO estimates that the unemployment rate will average 15.1% in the second quarter of 2020. In April and May, the unemployment rate averaged 14%, so this might seem like a pessimistic forecast by CBO. However, a very large number of people were likely misclassified in these months as “employed, not at work due to illness” instead of unemployed. Further, there has been a very large reduction in labor force participation since February, and these people very likely want a job and just couldn’t actively search due to economic shutdowns. Accounting for these groups, the more accurate unemployment rate likely averaged 21% in those two months. Even if misclassification and the issue of labor force participation goes away entirely just in one quarter, this six-percentage-point difference in unemployment measures in the second quarter alone would push estimated budget shortfalls back over $1 trillion, even with the rest of the newer CBO unemployment forecast adopted.
2. To get an estimate for nominal GDP at the end of 2021, we take estimates from the Congressional Budget Office (CBO) for nominal GDP at the end of 2020 and multiply by 1.045, assuming a 4.5% growth in nominal GDP between the fourth quarter of 2020 and the fourth quarter of 2021. The CBO forecasts a 2.8% growth in real GDP in this time; a 4.5% nominal rate hence assumes a 1.7% growth in the GDP price deflator, on par with its growth in recent years. For payroll employment in the fourth quarter of 2021, we take the average of four forecasts of this measure, from the Congressional Budget Office, Moody’s Analytics Economy.com, J.P. Morgan, and Goldman Sachs. Our estimate for the multiplier of state and local spending is taken from the various estimates surveyed in Bivens 2011.