The COVID-19 Pandemic and Challenges Facing State and Local Governments.

• Author: Clemens, Jeffrey

Economic crises bring questions about the design and implications of fiscal systems to the forefront. In the United States, state and local governments employ roughly one in seven workers and spend an amount equivalent to one-fifth of GDP. Because many of these entities operate with balanced budget requirements, downturns create pressure because declines in revenue coincide with a rise in demand for public services. These pressures come with some urgency, as state and local governments play roles in the administration and financing of safety net programs, the delivery of public health services, and the provision of public transit and education.

At the outset of the COVID-19 pandemic, concerns over the budgetary health and service performance of state and local governments were top of federal policymakers' minds. This was driven in part by the experience of the Great Recession, after which the state and local public sectors were widely perceived as a drag on the broader economy. In an effort to avoid a repeat of this, federal policymakers legislated close to $1 trillion in fiscal assistance to state and local governments, substantially exceeding the roughly $225 billion in fiscal assistance appropriated during the Great Recession through the American Recovery and Reinvestment Act (ARRA).

Three distinct sets of questions relate to the design of federal fiscal assistance. One involves the design of formulas through which the assistance is delivered. Another addresses the macroeconomic impacts of federal fiscal assistance, an issue on which research blossomed following the Great Recession. A third set relates to the core functions of state and local governments: how was fiscal assistance deployed and what impacts did it have on outcomes under the purview of public health officials, safety net program administrators, school districts, and other government agencies?

The Stabilization Problem

At the pandemic's outset, Stan Veuger and I projected the potential effects of the pandemic on the revenues of state and local governments, as did a number of independent research teams. (1) An objective of our work was to inform policymakers regarding the amount of aid that might be justified on revenue stabilization grounds. We illustrated how the Congressional Budget Office's (CBO's) early-pandemic forecasts for personal income and personal consumption expenditures could be used as forecasts of the evolution of state and local tax bases. Multiplied by historical estimates of the elasticity of revenues with respect to fluctuations in tax bases, CBO's forecasts of declines in economic activity could be translated into forecasts of revenue shortfalls.

As Veuger and I explained later, two lessons emerged from our analysis. (2) First, in a predictive sense, revenue forecasts tended to perform better when they relied on close rather than distant proxies for state and local governments' tax bases. At the COVID-19 pandemic's onset, forecasts that relied on the historical relationship between revenues and states' unemployment rates produced relatively inaccurate predictions. This is illustrated in Figure 1, which shows one set of projections, by Timothy Bartik of the Upjohn Institute, (3) that relied on forecasts of the unemployment rate, and another, by Veuger and me, that was based on projections of aggregate income and consumption. Because realized revenues would ultimately--and, to be clear, surprisingly--exceed prepandemic forecasts, larger shortfall forecasts were less accurate than smaller shortfall forecasts. (4) Forecasts that relied on disaggregated consumption and income data performed even better. (5) The shift in consumption towards goods and away from services led sales tax revenues to be more robust than most analysts expected. Predictions based on forecasts of disaggregated consumption data thus performed better than predictions based on forecasts of aggregate data.

Second, revenue forecasts suffered from a reliance on forecasts of economic activity that, in CBO's tradition, reflected "current law." Consequently, the associated forecasts for the evolution of states' tax bases did not account for the effects of not-yet-passed pandemic-related aid for households and businesses. As a result, the forecasts of revenue shortfalls were based on a conceptual error of viewing revenue shortfalls and household and business financial stress as separate rather than interconnected phenomena.

The pandemic experience raises interesting questions about the tradeoffs...