How Much Help Do State and Local Governments Need? Updated Estimates of Revenue Losses from Pandemic Mitigation

I estimate that state and local governments have lost $141 billion of revenue from all sources in fiscal year 2020 (FY20) due to the COVID-19 mitigation shutdowns. Under three scenarios of increasing severity, I estimate that state and local governments will need to cut expenditures by between $59 billion and $350 billion in fiscal year 2021 (FY21) to offset impending losses of revenue. Some of the revenue losses can be offset by the rainy day funds that state and local governments have set aside during the expansion, but jurisdictions that lack a fiscal buffer may face painfully deep service cuts. This data brief updates and improves upon the estimates in the District Data Brief published on May 13, 2020. In these new estimates, I include all revenue sources that are likely to decline; the original brief covered only income and sales taxes.

To create these estimates, I took the following steps:

1. I calculate 2020 baseline expected revenue levels by increasing every revenue value reported in the 2017 Census of Governments (COG).¹ For income, sales, corporate, gaming, and other sources of tax revenues, I use the percentage growth reported in the National Association of State Budget Officers’ Fiscal Survey.², ³ For non-tax revenues, such as airport fees and public hospital charges, I increased them by the rate of inflation between 2017 and 2020.⁴ I assume the baseline expected revenue levels are the same for FY21 as they are for FY20 because no measure of growth is available for FY21 yet. With my approach, if the FY21 baseline were 1 percent higher, the estimated loses would also be 1 percent larger.
2. I pair a measure of the tax base to each type of revenue. For example, I used the Current Employment Statistics (CES) data with income taxes, the Personal Consumption Expenditures (PCE) data with sales taxes, and oil prices with severance taxes (a complete list of the pairings appears in table 4). With the tax base data, I estimate a decline from normal levels of activity to activity under a pandemic-mitigation shutdown.
3. I create three scenarios that specify how quickly the tax base declines are reversed and the base returns to normal. All the scenarios assume tax receipts came in at expected levels through 2020:Q1 and the tax base is at the shutdown level for 2020:Q2. In the V-shaped scenario, the declines are half-way reversed in 2020:Q3, and they completely reverse by 2020:Q4. In the middle scenario, based on the Great Recession, the declines are reversed over four quarters. The most severe scenario assumes the declines are halfway reversed in 2020:Q3, but a second wave of COVID-19 cases requires a complete shutdown in 2020:Q4. The second shutdown is followed by a slow recovery that takes all of 2021.
4. The COG reports the date when each government ends its fiscal year, and I use the scenarios to weight the combinations of shutdown, partially recovered, and normal quarters that fall within each government’s FY20 and FY21. Together, the quarterly levels imply a decline in the tax base below normal activity for the fiscal year.
5. Multiplying the projected fiscal year tax base percentage declines times the baseline revenue amounts gives a value of the revenue lost for each revenue type collected by each individual government. The results tables present these losses aggregated to national totals (table 1), state totals (tables 2 and 3) or totals by type of revenue (table 4).

The estimating equations and further details can be found in the Appendix of the District Data Brief published on May 13, 2020.

There are three important improvements in these estimates relative to the version published on May 13, 2020.

1. This version covers all revenue streams, including corporate income taxes, hospital revenue, airport revenue, and dozens of smaller sources of income. The previous brief covered only the two largest sources of revenue: individual income taxes and sales taxes.
2. Most of the revenue types are paired with line items in the NIPA PCE Table 2.4.5U.⁵ In this brief, I use the contrast between the February 2020 and April 2020 values in the monthly version of this table to estimate how much lower the tax base transactions are under the stay-at-home orders. In the previous brief, I used the decline from 2019:Q4 to 2020:Q1 to estimate the decline. Using the monthly data gives a more accurate estimate of the declines because any spending growth in January or February is not obscuring the drop in March or April.
3. In the previous estimate, I assumed all revenue sources grew at the rate of inflation to scale up from FY17 to FY20 revenue values. That suggested baseline values 7 percent higher than the 2017 values. Using the revenue growth values reported by NASBO, we can see that state revenues grew faster than inflation. On average, the baseline values are now 11 percent higher than the 2017 values. Because the increases are applied by state and by type of revenue, they more accurately reflect each state’s exposure to declines in different revenue streams.

The new national aggregate estimates appear in table 1. State governments may lose $85 billion of revenue in FY20. If the post-shutdown economic recovery is V shaped, I estimate that states will lose $45 billion in FY21. If the recovery is slow and similar to the recovery following the Great Recession, states could lose $122 billion. In a second-wave scenario, where shutdowns are re-imposed this fall, states would lose $238 billion. I estimate that local governments will lose $56 billion in revenue in FY20, and between $14 billion and $112 billion in FY21. These local estimates cover all revenue types, including income and sales taxes, and they use all of the additional data sources introduced in this Data Brief. The total estimates in this Data Brief are much higher than those in the previous Data Brief mainly because they cover additional major revenue sources such as corporate income taxes, hospital revenue, and airport revenue.

In table 1, I also present the losses that would remain if states split their rainy day funds between FY20 and FY21. This reduces the losses to $40 billion in FY20 and between $11 billion to $183 billion in FY21. Local governments also maintain rainy day funds, but there is no data set with a consistent measure of these funds, so the local revenue loss estimates cannot be adjusted.

Table 2 presents total loses for state governments and the losses as a share of own-revenue, that is, revenue the state collects rather than receives as a federal-to-state intergovernmental transfer. The losses in FY20 are substantial, ranging from 3.3 percent in Vermont to more than 10 percent in North Dakota. The FY20 losses could be particularly difficult to deal with because most states must offset them with cuts in the fourth quarter of FY20 to meet their balanced budget requirements.⁶ A second-wave shutdown, followed by a slow recovery would take 10 to 20 percent off of most FY21 state budgets and force major spending and job cuts. In table 3, the estimated losses of all of the individual local governments in the states are aggregated to state totals. The shares of own-revenue are means calculated with each government weighted by its total own-revenue. Even in the second-wave scenario, the average revenue loss is less than 10 percent in most states because local governments collect much of their revenue through property taxes. Property taxes usually do not decline until two years after the trough of a recession because the property tax assessment process operates with a multi-year lag.