Congress has provided multiple rounds of relief to state and local governments in legislation responding to the economic shock of the coronavirus, but much more will be needed by these governments in coming years. We estimate that roughly $500 billion more will be needed by the end of 2021 to keep state and local governments from becoming a significant drag on economic recovery after the public health crisis passes.

The recently passed CARES Act allocated $150 billion to help state and local governments grapple with the costs of responding to COVID-19. But this amount does not come close to what is needed to address the severity and likely duration of the public health and economic crises. As economic activity has collapsed, it has triggered a dramatic downturn in state and local revenues even apart from new spending demands imposed by the coronavirus. Unlike the federal government, most state governments are required by law or constitution to balance their budgets. As revenues decline because of lower incomes and reduced spending, state and local governments face serious fiscal constraints, often leading to budget cuts that further depress demand in the economy.

Already, states and localities are announcing austerity measures and severe budget shortfalls exactly when public spending is most critical—both for protecting workers and for priming the economy for a rapid bounceback when the shutdown ends. Ohio Governor Mike DeWine has proposed an across-the-board 20% budget cut; New York State Comptroller Thomas DiNapoli estimated tax revenue would be between $4 billion and $7 billion below projections for fiscal year 2020; Arkansas’s projected revenue decline is more than double the size of the state’s reserves; and California is projected to spend down its sizable cash reserves in mere months, despite previously being on track to build the largest cash reserve in its history of more than $20 billion.

Local governments are announcing severe revenue shortfalls, too: This week, Arlington County, Virginia, announced a shortfall of $56 million for FY 2021, and cities like Seattle and New Orleans are each projecting shortfalls of at least $100 million this year.

Clearly, the aid to state and local governments passed so far is not sufficient, and we estimate that at least $500 billion will be needed by the end of 2021. Here’s how we got to this number.

In a recent blog post for the Brookings Institution, Matthew Fiedler and Wilson Powell III highlight the need for more federal aid to state governments. They draw on previous work to show that every one-percentage-point increase in the national unemployment rate is associated with a $45 billion (or 3.7%) decline in state revenues, holding tax policy changes constant. Their recommendation is a triggers-based increase in the federal share of Medicaid payments (normally Medicaid is financed by state governments and the federal government jointly), with the triggers being state-specific unemployment rates that are abnormally high.

Their recommendation for a trigger-based increase in Federal Medical Assistance Percentage (FMAP) spending should certainly be adopted—it is well crafted to deliver federal aid as long as state economies remain impaired. However, these trigger-based FMAP increases need to be supplemented with other federal transfers to state and local governments, for a couple of reasons.

First, Fiedler and Powell only highlight the damage done to state revenue from an economic downturn, but local governments also see revenues crater in the face of downturns. Local governments spend as much as state governments do, so anything that puts downward pressure on this spending will be a large drag on economic recovery in coming months.

Second, their proposal for a triggered FMAP increase would neutralize roughly two-thirds of the overall negative fiscal shock to state governments over the course of a downturn. While this is a huge improvement over the status quo, we should aim to do even better in the face of the extraordinary economic crisis spurred by the coronavirus.

Third, recent forecasts highlighting the extraordinary coronavirus economic shock and the long recovery period already assume substantial increases in state and local aid from the federal government will be forthcoming in the next few weeks. If one looks, for example, at the most recent Goldman Sachs unemployment projections and uses these to infer the negative shock to state and local finances, this inference would be underestimated because even the baseline level of unemployment is premised upon more state and local aid coming forth than has so far been passed by Congress.

To get a rough estimate of how much state and local tax revenue will be damaged by excess unemployment in coming quarters, we apply the Fiedler and Powell estimates that each percentage-point increase in unemployment is associated with a $45 billion shortfall in state tax revenues. We also estimate what the total state and local shortfall would be if the same rule of thumb applied to local revenues. In 2018, local revenues were about two-thirds as large as state revenues (excluding intergovernmental transfers). If local revenues were just as responsive to rising unemployment as state revenues, this would imply that each percentage-point increase in unemployment was associated with a $75 billion reduction in state and local revenues.

From here, we can then compare the next seven quarters of Goldman Sachs’s unemployment forecast to the level of unemployment that prevailed in the first quarter of 2019. From this, we calculate the average amount of excess unemployment over the next seven quarters (5.7 percentage points) and multiply this by the $45 billion associated with declining state revenue by Fiedler and Powell, the $75 billion in state and local revenue losses that would result if local revenue was as sensitive to business cycle conditions as state revenue, and $60 billion, the average of the two.

This implies revenue shortfalls of $260 to $430 billion, with an average of $345 billion. However, even this path of excess unemployment assumes at least an additional $100 billion in aid to state and local governments that has not yet passed Congress. Therefore, the next relief bill should provide state and local governments with at least $500 billion in direct, unrestricted aid by the end of 2021, which is roughly equivalent to the federal government picking up all Medicaid costs for two full years. This new aid should include a requirement that funds are allocated to localities, including the localities with populations below 500,000 that were left out of the CARES Act, and the aid should flow to state and local governments immediately in order to avoid a drag on the economy.