The Hamilton Project, Recession Remedies chapter, "Lessons Learned from Support for the State and Local Sector during COVID-19"
Summary
Chapter 6 of Recession Remedies: Lessons Learned from the U.S. Economic Policy Response to COVID-19, the Hamilton Project volume edited by Wendy Edelberg, Louise Sheiner and David Wessel, written by Louise Sheiner on support for the state and local sector. It notes that state and local governments account for about 13 percent of U.S. employment and tax revenues of about 9 percent of GDP, that spring 2020 projections put revenue losses as high as $900 billion, and that the CARES Act provided roughly $350 billion, with later legislation and the American Rescue Plan adding $640 billion. Revenues fell far less than feared and federal aid more than offset losses in every state, yet state and local employment in February 2022 remained three percent below its January 2020 level.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“The Hamilton Project, Recession Remedies chapter, ‘Lessons Learned from Support for the State and Local Sector during COVID-19.’”
Original at www.hamiltonproject.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/the-hamilton-project-recession-remedies-chapter-lessons-learned-from-s-455d39b982/
, SHA-256 ca294a43760659e084ba20a31c2937267458b19271375ea0c94e9c817332aa02.Full text
RECESSION
REMEDIES
Lessons Learned from the
U.S. Economic Policy Response to
COVID-19
Edited by
Wendy Edelberg, Louise Sheiner,
and David Wessel
Recession Remedies
Lessons Learned from the
U.S. Economic Policy Response
to COVID-19
Edited by
Wendy Edelberg, Louise Sheiner, and
David Wessel
Chapter 6
Lessons Learned from Support for
the State and Local Sector during
COVID-19
Louise Sheiner1
Introduction
State and local governments are significant players in the U.S. economy. Employ-
ment by state and local governments represents about 13 percent of total U.S.
employment, a larger share than the federal government. State and local tax
revenues represent about 9 percent of GDP, approximately half the share of
federal tax revenues.
Unlike the federal government, virtually all state and local governments
have to balance their operating budgets; they cannot borrow to finance large
deficits. Revenue losses experienced during recessions have to be financed by
savings or offset by spending cuts or tax increases. Governments typically make
most of these adjustments on the spending side, likely because tax increases
during economic downturns are particularly unpopular. Reductions in spending
deprive residents of valuable services and weaken the macroeconomy. Tight
spending at the state and local level in the aftermath of the Great Recession was
a factor behind the slow recovery. Tight budgets are particularly problematic
during a pandemic because much of the nation’s public health infrastructure
is at the state and local level.
In the spring of 2020, many analysts were projecting considerable revenue
losses in the state and local sector—with some estimates suggesting losses of
up to $900 billion over two years. In addition, state and local governments
were facing new demands on spending arising from the need to address
1. This chapter could not have been completed without the excellent research assistance of Sophia
Campbell, Lorena Hernandez Barcena, and Nasiha Salwati. I also thank Don Boyd, Wendy
Edelberg, Byron Lutz, Justin Marlowe, Tracy Gordon, and David Wessel for helpful comments.
215
216 | Recession Remedies
pandemic-related public health issues. Congress acted swiftly to provide aid.
The Coronavirus Aid, Relief, and Economic Security (CARES) Act, enacted in
March 2020, provided significant aid to state and local governments—roughly
$350 billion—and legislation passed in December 2020 and the American
Rescue Plan (ARP) enacted in March 2021 provided an additional $640 bil-
lion for a total of close to $1 trillion. This was far more than the roughly $275
billion provided to state and local governments during the Great Recession
(Congressional Research Service 2019). In addition, the Federal Reserve (the
Fed) launched the Municipal Liquidity Facility (MLF) to ensure that state and
local governments had access to credit.
So, what happened to state and local government revenues, employment,
and spending during the first two years of the pandemic? Revenues did not
decline nearly as much as had been first feared and federal aid was more than
sufficient to offset any revenue losses in every state. Nevertheless, state and
local government employment declined sharply, and the decline has been quite
persistent: employment by state and local governments in February 2022 was
three percent below the January 2020 level. Looked at another way, in Febru-
ary 2022, the state and local sector accounted for 23 percent of the shortfall in
U.S. employment from its pre-pandemic trend.2 Total nominal state and local
spending, as measured by the National Income and Product Accounts (NIPA),
(Bureau of Economic Analysis 2021b).
Thus, despite a large and rapid federal response, the state and local sector
once again appears to be lagging most other sectors of the economy. Of course,
unlike in the aftermath of the Great Recession, policymakers are now more
worried about excess demand than insufficient demand, so the weakness in
the state and local sector is much less of a concern from a macroeconomic
perspective. Still, it is helpful to understand what happened and why, and what
that might tell us about responses to future downturns.
This chapter addresses five questions:
1. Why were the revenue projections at the beginning of the pandemic
so inaccurate?
2. How much aid did state and local governments receive, and was it
sufficient to address revenue losses and increased costs related to the
pandemic spending?
3. How well did the Fed’s MLF work?
4. Why did employment decline so much? How much was specific to the
pandemic, and how much was related to budget concerns?
2. For this calculation, I assume that, absent the pandemic, both total and state and local
employment would have increased 0.7 percent per year, the annual rate of increase CBO
projected for total employment growth between 2019Q4 and 2021Q1 in its January 2020
economic projection (CBO 2020a).
State and Local Sector | 217
5. What are state and local governments doing with all the federal aid
they have received?
I conclude with a discussion of the lessons that can be learned from the
experiences during the pandemic, which I summarize here:
• Policies that provide fiscal support to households and businesses indi-
rectly support state and local revenues and should be considered in
determining the amount of direct aid to state and local governments.
• In order to prevent layoffs, aid to state and local governments should
be automatic or provided early in a recession.
• Aid should go directly to states and localities instead of only to state
governments where possible, and should have few conditions on its use.
• State and local governments are reticent about using one-time federal
aid to finance ongoing expenditures, which might preclude aid from
being used for the most effective purposes.
• The ability of state and local governments to borrow from the Fed in a
time of crisis can serve as an important backstop that can help stabilize
municipal bond markets.
• Timelier data on state and local government revenues and expenditures
are needed to assess ongoing economic conditions and to evaluate policy.
• The lesson of the Great Recession (i.e., that inadequate aid to state and
local governments can hamper an economic recovery) should not be dis-
carded because of the recent experience; the pandemic created unusual
economic conditions that are not likely to recur in future recessions.
Projected and Realized Revenue Losses
As shown in Table 6.1, virtually all analysts and policymakers projected large
and prolonged revenue losses. For example, Bartik (2020) projected losses
of $899 billion in fiscal years 2020 and 2021 while Bivens and Walker (2020)
projected losses of $345 billion. Others projected losses for only a subset of the
state and local revenues (e.g., just state taxes or just income and sales taxes).
These ranged from $130 billion (White, Crane, and Seitz 2020) to $395 billion
(McNichol, Leachman, and Marshall 2020). Most of these projections relied on
historical relationships between state and local revenues and the unemploy-
ment rate or the growth rate of personal income. Auerbach et al. (2020) took a
different approach, using a bottom-up method to project revenues by state for
each type of revenue. All revenue projections, regardless of approach, relied
on economic forecasts.
In fact, although tax revenues dipped at the onset of the pandemic, they
quickly recovered and have been quite healthy since. Table 6.2 compares tax
revenues from the U.S. Census Bureau’s Quarterly Summary of State and Local
Tax Revenue to a baseline where revenues increased 4 percent per year from
218 | Recession Remedies
Table 6.1
Projections of Revenue Losses From COVID in
the State and Local Sector
Economic
Revenue Losses Forecast
Revenue or FY2020 + FY2021 Underlying
Authors Spending (billions) Estimate Methodology
Bivens and Walker State and local $345 Goldman Sachs Historical relationship
(April 2020) Economic taxes (April 2020) augmented for assumed
Policy Institute local revenue effect.
A 1% increase in
unemployment associated
with a $60 billion decline in
state and local revenues.
McNichol, Leachman, State taxes $395 CBO (April 2020) Historical relationship: 1%
and Marshall (April increase in unemployment
2020) rate lowers revenues 3.7%.
White, Crane and State revenues $130 baseline; Baseline: Max 10% Proprietary model that
Seitz (April 2020) general funds $203 more severe decline real GDP, includes state-by-state
Moody’s Analytics scenario gradual recovery. regressions of state
More severe: revenues on economic
Max 14% decline revenues.
real GDP, gradual
recovery.
Bartik (May 2020) State and local $899 CBO (April 2020) Historical relationship
Upjohn Institute taxes augmented for assumed
local revenue effect.1%
increase in unemployment
lower state and local
revenues by $60 billion.
Clemens and Veuger State income and $148 CBO (April 2020) Historical relationship: 1%
(June 2020) sales tax decline in personal income
lowers revenues by 1.6%.
Whitaker (June 2020) All state and local $200–$490 Best: Recovery For income taxes: estimate
revenue (including complete by 2020 wage declines and assume
fees, charges, etc.) Q4. tax revenues decline
Worst: Second wave proportionally.
shutdown 2020 Q4. For sales taxes, use national
Economy recovered changes in portions of PCE
by Q4 2021. likely subject to sales tax.
Dadayan (July 2020) State taxes $200 States forecasts Estimated for all 50 states
Urban Institute Tax based on forecast data from
Policy Center 27 states.
Auerbach, Gale, State and local $270 (calendar CBO (July 2020) Detailed projections of tax
Lutz, and Sheiner taxes and fees years not fiscal bases and tax schedules.
(September 2020) years)
State and Local Sector | 219
Table 6.2
State and Local Government Revenues During
the Pandemic
Revenue Losses Relative to Counterfactual 4% Growth
(in billions)
FY2020 FY2021
Total current tax receipts -71 145
Personal Income taxes -38 70
Sales taxes -20 6
Property taxes -4 44
Taxes on corporate income -9 25
Source: Census Bureau 2021; author’s calculations.
Note: This table reports the difference between revenues collected
in FY2020 and FY2021 and revenues that would have been collected
had they increased at a 4 percent annual rate from 2020 Q1 on.
2020Q1 on, a bit above what state budget officials expected for state fiscal year
(FY) 2021 right before the onset of the pandemic (NASBO 2020). State and
local taxes were $71 billion lower in state FY 2020 but were $145 billion higher
in state FY 2021 (U.S. Census Bureau 2021).3,4 Looking at the components,
revenues were below baseline in FY 2020 but above baseline by FY 2021. This
is in contrast to the Great Recession, during which revenues fell sharply and
remained depressed for many years.
Why Were the Projections So Off?
It is important to understand why the historical relationships between state
and local revenues and projected economic conditions performed so poorly in
predicting actual revenue collections, particularly because many economists
have argued that federal aid to state and local governments should be triggered
automatically when economic conditions reach a certain threshold. For example,
Fiedler, Furman, and Powell (2019) argued that the federal share of Medicaid
should be increased by 3.8 percentage points for each percentage point by
which a state’s unemployment rate exceeds a threshold. Using the historical
3. Part of the decline in FY 2020 and recovery in FY 2021 reflects a shift in tax collections from
the delay in the tax filing deadline for individual and corporate income taxes.
4. Throughout this chapter, I define the fiscal year as beginning on July 1 and ending on July
30—that is, fiscal year 2020 is from July 1, 2019, to June 30, 2020. This timing is used by 46
states. The fiscal year begins on April 1 for New York, September 1 for Texas, and October 1
for Alabama, Michigan, and the District of Columbia.
220 | Recession Remedies
relationship between unemployment and state revenues, they estimated that
this increase in the Medicaid matching rate would be sufficient to offset two-
thirds of a state’s revenue losses. A key question, then, is whether these historic
relationships broke down during the pandemic or the economic forecasts on
which the revenue losses were based were too pessimistic.
Auerbach et al. (2020) examined the historical relationships between state
and local revenues and economic conditions (reproduced in Table 6.3). They
argued for excluding 2009, a particularly unusual year, and including a measure
of the stock market performance when estimating the historical relationships.
The inclusion of stock market performance was particularly important during
the pandemic because, rather than declining as it does during most downturns,
the stock market soared, boosting household wealth and taxable income. With
these adjustments, they showed that the predicted losses were smaller than
many others had projected.
Auerbach et al. (2020) also argued that this recession was sufficiently dif-
ferent from previous episodes that historical relationships might be misleading.
First, social distancing and remote work meant that sources of revenue that are
not typically cyclical (e.g., gas taxes, airport fees, motor vehicle licensing fees,
etc.) plummeted. Second, low-wage workers suffered disproportionately from
this recession, meaning that any given change in the unemployment rate had
a smaller effect on consumption and personal income and thus on state and
local revenues than is usually the case. Finally, generous federal aid to house-
holds strengthened household finances, thus supporting both sales taxes and
property taxes, and some of the federal aid directly boosted taxable income.
Many states tax unemployment benefits and Paycheck Protection Program
(PPP) loans boosted taxable profits.5
Another reason that the revenue forecasts were so off, however, is that the
analysts relied on economic forecasts that proved far too pessimistic. In July 2020,
for example, the Congressional Budget Office (CBO) projected that, as a result
of the pandemic, real GDP would be 8 percent below its pre-pandemic forecast
by the end of 2020 and 5 percent below it by the end of 2021 (Congressional
Budget Office 2020a and 2020b). Instead, real GDP was just 5 percent below the
pre-pandemic projection in 2020Q4 and 1 percent below in 2021Q4. Similarly,
CBO projected that the unemployment rates in Q4 of 2020 and 2021 would be
10.5 percent and 7.6 percent, respectively. Instead, they were 6.8 percent and 4.2
percent, respectively. Private sector forecasts were similarly overly pessimistic.
5. During normal times, 35 states tax UI benefits. However, 25 of these states adopted the federal
exemption included in the American Rescue Plan (enacted in early 2021) on the first $10,200
of benefits for most taxpayers or waived taxes on UI all together (Mengle 2022). Similarly,
although PPP loans were not taxable, it was originally thought that companies would not
be able to deduct the costs paid out of the loans (i.e., they could not double dip). However,
in the legislation enacted in December 2021, the Consolidated Appropriations Act of 2021,
Congress explicitly said that companies could do this. Many states automatically conform
with the federal tax law, so this reduced taxes for states as well.
State and Local Sector | 221
Table 6.3
Predicted Revenue Losses Given Actual Economic
Outcomes, Total FY2020 and FY2021 (Billions)
A. State and Local Tax Revenues and the Business Cycle (1985–2019)
Excluding Including Including Stocks
All 2009 Stocks and Excluding 2009
Dependent Variable: Log Change in Real per Capita State and Local Income Taxes
Change in UR Coefficient -4.90 -3.50 -3.30 -2.70
Adjusted R2 0.56 0.24 0.65 0.35
Log change real per Coefficient 2.00 1.40 1.40 1.10
capita personal income Adjusted R2 0.58 0.28 0.65 0.36
Log change real Coefficient 1.50 1.10 1.10 0.89
per capita “taxable” Adjusted R2 0.61 0.33 0.67 0.40
personal income
Dependent Variable: Log Change in Real per Capita State and Local Sales Taxes
Change in UR Coefficient -3.00 -2.30 -2.40 -2.00
Adjusted R2 0.72 0.44 0.75 0.49
Log change real per Coefficient 1.10 0.70 0.80 0.60
capita personal income Adjusted R2 0.60 0.27 0.64 0.31
B. Predicted Revenue Losses Given Actual Economic Outcomes, Total FY2020 and FY2021, Billions
Including
Excluding Including Stocks and
All 2009 Stocks Excluding 2009 Actual Change
Income Tax Revenues
Change in UR -$116 -$83 -$57 -$46 $48
Log change real per $5 $3 $23 $19 $48
capita personal income
(excluding EIPs )
Log change real -$44 -$32 -$14 -$11 $48
per capita “taxable”
personal income
(excluding UI)
Sales Tax Revenues
Change in UR -$70 -$54 -$49 -$41 -$14
Log change real per $23 $15 $25 $19 -$14
capita personal income
Source: Bureau of Economic Analysis 2021b; Bureau of Labor Statis-
tics 2022e; Wilshire Associates 2022.
Note: Change in personal income relative to a counterfactual in
which real income grows 2% per year.
222 | Recession Remedies
Panel B of Table 6.3 shows the decline in revenues over the FY 2020 and FY
2021 period that would be predicted from the regressions using actual economic
outcomes. Using the simple unemployment rate regression in column 1, the
actual increase in the unemployment rate would suggest income tax losses of
$116 billion, compared to the actual gain of $48 billion, and sales tax losses of
$70 billion, relative to an actual loss of $14 billion. These predicted losses are
much smaller than many of the losses predicted in the spring of 2020, but they
are still substantial. Even when excluding FY 2009 and including changes in
stock market performance, the losses using the unemployment rate regression
are predicted to be $87 billion for both sources of revenues combined, rather
than the $34 billion gain that actually occurred.
The predictions using personal income align more closely with actual
tax collections. Using the regression of income tax changes on the change
in total personal income (excluding the Economic Impact Payments [EIP])
suggests income tax gains of between $5 and $22 billion, depending on the
specification, which is closer to the actual gain of $38 billion. But using what
is a better-specified regression that relates revenues to a measure more closely
approximating taxable income (i.e., a measure that only includes sources of
income subject to tax) does less well. It predicts income tax losses of between
$11 billion and $44 billion.6 Turning to the sales tax predictions, the huge
increase in personal income would have predicted somewhat stronger sales
tax collections than actually collected.
What does this exercise suggest about the usefulness of these types of
regressions for predicting revenue losses in future recessions? First, to a large
extent, the overly pessimistic revenue forecasts were the result of overly pessi-
mistic economic forecasts. Policies that automatically provide aid to state and
local governments when economic conditions warrant it would automatically
adjust if economic projections turned out better (or worse) than expected. In
other words, the very large forecast errors during the pandemic do not imply
that these types of automatic stabilizer policies are misguided.
Second, changes in the unemployment rate or personal income are not
great predictors of revenue losses, even at the aggregate level. Regressions
that do not include FY 2009 had a relatively poor fit even before the pandemic
(R 2 ranging from .24 to .44, depending on the regression).7 This finding may
6. Personal income includes a lot of items that are not taxable and that should not affect tax
receipts, including employer-provided health and pension benefits, government transfers
like Medicare, Medicaid, and the Supplemental Nutrition Assistance Program, imputed
rental income on owner-occupied homes, and income earned by the nonprofit sector. This
regression excludes those components of personal income. Note that the regressions also
use lagged stock market performance, so the big gains in the stock market in 2021 are not
affecting predicted revenues.
7. This fact is perhaps not surprising. Recessions differ from each other, and economies and tax
structures differ across states. For example, New York is highly dependent on tourism, and
its lower-income individuals were deeply affected by the pandemic, which is reflected in the
State and Local Sector | 223
weaken the case for automatic aid to state and local governments, although
such policies would still be helpful even if the amount of aid provided does not
match revenue losses particularly well. For example, policies that are geared
toward replacing two-thirds of the lost revenues on average would help sup-
port an economic recovery—even if that aid sometimes proves too large and
sometimes too small. While Congress acted swiftly and forcefully this time to
support state and local governments, that may not be the case the next time.
Third, general fiscal support to households and businesses can indirectly
support state and local governments. When contemplating discretionary aid
to state and local governments during future downturns, it is important to
account for other policies that might have economically significant effects.
The generous fiscal support enacted during the pandemic—which not only
expanded Unemployment Insurance (UI) but also the EIPs and PPP—meant
that (a) the economic recovery was stronger than it otherwise would have
been, and (b) even holding economic conditions constant, high unemployment
did not translate into tight household budgets, thus supporting sales, excise,
and property tax collections for state and local governments. This is likely an
important reason why relying on increases in the unemployment rate to predict
state and local revenues losses yielded less accurate predictions than relying on
changes in personal income, even though both measures of the economy have
been similarly effective for predicting revenues historically.8
How Much Variation Is There in State and Local
Revenues Across the States?
When thinking about the effects of tight budgets on the state and local sector,
it is important to consider variation across the states: budget surpluses in one
state will not compensate for budget deficits in another in terms of the services
provided to the public. And if states are more likely to cut spending when
their budgets are in deficit than they are to raise spending when they are in
surplus, as suggested by Sorenson and Yosha (2001), cross-state variation in
revenue losses can also have implications for the level of employment and the
macroeconomy more generally.
The data available now only allow us to examine variation in tax collections
by state governments; we have no information on local government revenue
unemployment rate and even in employment growth. However, its economic and revenue
structure—a highly unequal economy, a financial-sector-driven economy, and a tax structure
that focuses on these features—meant that revenue growth was strong. Other tourism-focused
economies were not as fortunate. There is likely no single variable or single combination of
variables that can plausibly foretell fiscal stress across different kinds of recessions.
8. In addition, as discussed above, unemployment during this recession was unusually con-
centrated among low-wage workers, meaning that a given increase in unemployment had a
smaller effect on total wages.
224 | Recession Remedies
collections.9 Online Appendix Figure 6.1 reports the change in state revenues
averaged over FY 2020 and FY 2021 relative to revenues in FY 2019. There is a
great deal of variation across the states. States that had the largest revenue losses
or the smallest revenue gains were mostly energy producers or states heavily
dependent on tourism.10 One exception is Oregon, which experienced large
revenue losses in FY 2020 for a reason unrelated to the pandemic: the state’s
“kicker” law refunds tax collections if they come in two percent or more above
projections over the previous two years, which was the case in 2020 because
of strong revenues before the pandemic.
Still, aggregating both FY 2020 and FY 2021, revenues were only lower in
10 states than they would have been under a pre-pandemic baseline of 4 percent
annual growth. Revenues in Alaska, where 90 percent of general fund revenues
come from oil, were the weakest: roughly 35 percent below a pre-pandemic
baseline, reflecting weak oil demand and low oil prices (Understanding Alas-
ka’s Budget n.d.).
Regression analysis of revenue losses, shown in Online Appendix Table
6.1, find that oil and tourism states had lower revenues on average over the two
years, but other state characteristics—including the Biden share of the vote (a
measure of political leanings and attitudes toward the pandemic); the change
in the unemployment rate; personal income growth; the share of tax receipts
coming from sales, income, and property taxes; and even states’ own predictions
for revenue declines early in the pandemic—had no predictive power for total
state revenues over the two years.
Federal Support to State and Local
Governments
Federal Aid
In response to the large projected revenue losses and concerns about increased
demands on state and local budgets, Congress increased aid to state and local
governments by about $1 trillion—far more than the roughly $275 billion pro-
vided during the Great Recession. Table 6.4 details the sources of aid. About
$250 billion was provided through legislation enacted in the spring of 2020 (i.e.,
9. The U.S. Census Bureau will eventually publish data on local government revenues by state;
however, these data, which come from the Annual Survey of State and Local Government
Finances, come out with a long lag: data for FY 2020 (ending for most states on July 1, 2020)
will not be available until around June 2022.
10. The states that experienced the greatest losses or slowest revenue growth over FY 2020
and FY 2021 were, in order, Alaska, North Dakota, Wyoming, Texas, Hawaii, Oklahoma,
West Virginia, New Mexico, Oregon, Nevada, Florida, New Hampshire, and the District of
Columbia. Of these, Hawaii, Nevada, Florida, New Hampshire, and the District of Columbia
are tourism states while, other than Oregon, the rest are energy producers.
State and Local Sector | 225
Table 6.4
Total Enacted Aid to State and Local
Governments (Billions)
Families First Consolidated
Coronavirus Appropriations American
Response Act CARES Act Act, 2021 Rescue Plan
(Mar. 18, 2020) (Mar. 27, 2020) (Dec. 27, 2020) (Mar. 11, 2021) Total
Total $105 $250 $99 $542 $996
General Aid $100 $150 $350 $600
Coronavirus Relief Fund $150 $150
Coronavirus State and $350 $350
Local Fiscal Recovery
Funds
Enhanced Medicaid $100 $100
matching ratea
Targeted Aid $5 $100 $99 $192 $396
Aid for Unemployment $5 $12 $16
Insurance administrative
expenses
Aid to K–12 $17 $56 $123 $195
Aid to public institutions $12 $19 $35 $66
of higher educationb
Aid to health providersb $35 $1 $2 $37
Aid to transit agencies $25 $24 $33 $81
and transportation
infrastructure grants
Source: Committee for a Responsible Federal Budget 2020;
Ochieng et al. 2022; Department of Education 2021.
Note: (a) Author’s estimated value assuming public health emer-
gency ends on July 1, 2022. (b) Author’s estimate of share going to
public institutions.
the start of the pandemic). First, in mid-March, Congress increased the share
of Medicaid spending financed by the federal government by 6.2 percentage
points, retroactive to the start of 2020 and effective until the end of the public
health emergency. This enhanced match rate increased federal Medicaid grants
to state and local governments by about $40 billion per year. Presuming the
public health emergency is declared over by June 2022, this amounts to approx-
imately $100 billion in total. As part of that enhanced federal payment, states
were prohibited from terminating Medicaid coverage for existing beneficiaries
or to tighten eligibility criteria, leading to a surge in Medicaid enrollment.
Nonetheless, the net effect was to lower overall state Medicaid spending, thus
relieving pressure on state budgets (Auerbach et al. 2020).
226 | Recession Remedies
Second, as part of the CARES Act, Congress created the Coronavirus Relief
Fund (CRF), a $150 billion fund allocated to state and local governments for
the express purpose of addressing unanticipated expenses related to the pan-
demic. In addition, the CARES Act included provisions to help cover higher
UI administrative expenses and provided targeted aid to public educational
institutions, health providers (including public hospitals), and transit agencies.
Additional targeted aid was enacted in the Consolidated Appropriations Act
in December 2020 and the ARP in March 2021. The ARP also included an
additional $350 billion in direct aid to states.
How Was Aid Distributed Across the States, and
How Flexible Was It?
The CRF provided $142 billion in aid to state governments and some local gov-
ernments and $8 billion to tribal governments. Local governments of entities
(counties, cities, townships, etc.) with a population of at least 500,000 were
eligible to apply, with the amount paid to state government reduced by the
aggregate amount that was disbursed to eligible local governments within the
state. Aid was distributed based on population, but states received a minimum
of $1.25 billion, making it much more generous for smaller states. For example,
Vermont, South Dakota, and Montana each received aid exceeding 20 percent
of 2020 own-source revenues (i.e., revenues excluding federal grants), whereas
Iowa, Missouri, Mississippi, California, Connecticut, New York, and Washington
received aid of 6 percent of own-source revenues or less (Auerbach et al. 2020).
The ARP’s Coronavirus State and Local Fiscal Recovery Funds (CSFRF
and CLFRF) provided direct aid to many more entities than the CRF: While
only 171 local governments received direct funding through the CARES Act’s
CRF, tens of thousands of local governments received ARP’s direct aid, with
the total amounts provided to states, counties, cities, tribal government, ter-
ritories, and other local governments specified by Congress.11 States received a
total of $195.3 billion, with $25.5 billion allocated equally across the states and
the District of Columbia (again providing much more generous aid to small
states) and the remaining $168 billion distributed on the basis of the number
of unemployed individuals over the three-month period ending December
2020.12 Aid to local governments was distributed based on population in the
case of counties and economic need in the case of cities.
The funding provided by the relief funds—the CRF, the CSFRF, and the
CLFRF—was large enough to more than offset state revenue losses in every state.
11. For a description of both relief funds, including allocations methods, specified uses, and
actual allocations, see U.S. Department of the Treasury (n.d.).
12. Also, the District of Columbia got an additional $254.9 billion to compensate it for receiving
less than other states in CARES Act funding.
State and Local Sector | 227
And given their dependence on property taxes, local governments likely had
even more muted revenue losses than states, so it is almost certain that the aid
exceeded aggregate revenue losses by both state and local governments in each
state. That is not even counting the other sources of aid (e.g., higher Medicaid
match rate, aid targeted to K–12 education, public health providers, transit
agencies, etc.) that in aggregate was about equal to the size of the relief funds.
There were, however, restrictions on the purposes to which the federal
aid could be put. For example, the CARES Act’s $150 billion CRF was only
to be used to cover expenses incurred due to the public health emergency,
not to fund any items that were accounted for in the most recently approved
budget. That is, they were explicitly not intended to cover any revenue losses.
Of course, funding is fungible, so the restrictions on the use of particular funds
may not bind on state and local governments. Furthermore, by the summer of
2020, Treasury had issued guidelines that allowed fairly broad use of the funds
(Treasury 2020).13 However, it is possible that these restrictions slowed the use
of funds and possibly contributed to the declines in employment in the sector.
The funds made available by the ARP were far more flexible than the aid
provided by the CARES Act. In particular, recipients of ARP funds were per-
mitted to use the aid to replace lost public sector revenue, fund public health
efforts, and address the economic consequences of the pandemic (including
covering the costs of UI), provide premium pay to public sector workers, and
invest in water, sewers, and broadband infrastructure.14 However, the ARP
funds, which account for almost 60 percent of the total federal aid to state
and local governments enacted during the pandemic, came too late to prevent
employment losses in the sector, as I discuss below.
Federal Reserve Lending to State and Local Governments
In addition to direct fiscal support to state and local governments, the federal
government also took actions to reduce strains in the municipal bond market.
In March 2020, fearing that massive revenue losses would leave state and local
governments unable to service their debts, investors pulled a record $45 billion
from muni funds (mutual funds that hold the bonds of state and local govern-
ments). Spreads between the yield on muni bonds and Treasurys soared to levels
not seen since the Great Depression, and many governments had trouble bor-
rowing. The strains in the muni market were particularly problematic because
13. For example, any costs related to public health or safety were deemed an acceptable use of
CARES Act funding.
14. The funds were also explicitly prohibited from being used to finance tax cuts. However, the
Treasury’s implementation of that restriction leaves room for tax cuts. In particular, so long
as tax collections are above 2019 taxes adjusted for inflation, the Treasury will not claw back
any ARP money from states that cut taxes (Auxier 2021). The ARP funds were also explicitly
prohibited from being used to pay down unfunded pension liabilities.
228 | Recession Remedies
the traditional April 15 tax filing deadline was pushed to July 15, which meant
much lower cash flow than expected and greater need for short-term borrowing.
To support the flow of credit to state and local governments, the Fed
launched the Municipal Liquidity Facility (MLF) on April 9, 2020.15 The facil-
ity initially was designed to purchase up to $500 billion of short-term notes
directly from U.S. states, including the District of Columbia; U.S. counties
with a population of at least two million residents; and U.S. cities with a pop-
ulation of at least one million residents. The facilities were later expanded to
counties with a population of at least 500,000 and cities with a population of
at least 250,000. This was the first time that the Fed made direct loans to state
and local governments.
Under the MLF, the Fed would purchase newly issued state and local
government bonds at normal spreads over Treasury bonds (rather than the
elevated spreads prevailing in the muni market) plus a fee of 100 basis points,
later reduced to 50 basis points. The Treasury and the Fed jointly set the terms
to virtually guarantee that the program would not lose money in aggregate.
Nonetheless, any losses incurred by the Fed on these loans would be absorbed
by some of the $454 billion provided in the CARES Act to the Treasury to be
used to backstop Fed lending to businesses and state and local governments.
Take-up of MLF loans was very low: only the state of Illinois and the
New York Metropolitan Transportation Authority made use of the program,
borrowing $3.20 billion and $3.36 billion, respectively. Yet the MLF is widely
viewed as a successful intervention because it stabilized yields in the private
muni market. Bordo and Duca (2020), for example, estimate that muni yields
could have risen by as much as 8 percentage points more than they did in mid-
April had the Fed not launched the MLF. They argue that the MLF served as
an important backstop that eased investor fears: the availability of Fed loans
meant that state and local governments would be able to finance their debt
even if revenues plummeted. As might be expected, the benefits of the MLF
were particularly large for low-rated issuers. Comparing issuers just below
and above the population eligibility cutoff, Haughwout, Hyman, and Shachar
(2021) found that eligible low-rated issuers saw yields fall by about 72 basis
points relative to comparable ineligible issuers.
15. The Fed also supported the muni market prior to the launch of the MLF. In particular, on
March 23, 2020, the Fed announced that it would begin accepting variable rate muni demand
notes (long-term municipal bonds offered through money market funds) as collateral at
its new Money Market Mutual Fund Liquidity Facility. Haughwout, Hyman, and Shachar
(2021) note that municipal yields and other measures of market distress started falling the
day the Fed made that announcement.
State and Local Sector | 229
Employment of State and Local Workers
One motivation for the generous aid provided to state and local governments
was to allow them to finance pandemic-related expenses without laying off
workers. Yet, despite muted revenue losses in FY 2020, healthy revenue gains
in FY 2021, very generous federal aid, and ample borrowing capacity, state
and local employment fell sharply during the pandemic and has yet to fully
recover. State and local governments began laying off workers in March 2020,
and by May 2020 seasonally adjusted employment was 7 percent lower than it
had been in January 2020.
Analyses of state and local employment typically focus on four types of
workers: state workers in the education sector (about 2.5 million in 2019), state
workers outside of education (2.7 million), local education workers (8.0 million),
and local workers outside of education (6.6 million). As shown in Figure 6.1a,
in the first few months of the pandemic, employment fell sharply for state
government workers in the education sector (solid blue line) as enrollment in
institutions of higher education declined but fell only slightly for workers in
other sectors of state government. As shown in Figure 6.1b, in local government,
employment fell sharply in both the education and noneducation sectors.
Employment in education at both local and state governments began
to recover in January 2021; the recovery in local employment outside of the
education sector began in the fall of 2020.16 By January 2022 state education
employment was a bit above its January 2020 level while employment at local
governments remained 4 percent below. Of course, employment would have
been expected to increase somewhat over two years, suggesting somewhat larger
shortfalls in employment relative to a pre-pandemic baseline.17
These patterns are in sharp contrast to those in the Great Recession, shown
by the green lines in Figure 6.1. In particular, state education employment
increased in the Great Recession as enrollment in higher education increased,
and employment did not begin declining for other government workers until
close to two years after the start of the Great Recession. But state and local
government employment, other than in state education, fell consistently for
many years thereafter, not even beginning to recover until the end of 2012.
Understanding the patterns of employment declines during the pandemic
is critically important to understanding the efficacy of federal aid to state
and local governments. Is it the case that aid from the federal government is
16. The small increases in local and state education employment observed in the summer of 2020
likely reflect the unusual seasonal pattern in 2020. Many workers who typically would have
been laid off in the summer (e.g., cafeteria workers, bus drivers, maintenance workers, etc.)
were instead laid off in the spring. As a result, the typical summer layoffs were smaller than
normal, and this caused an increase in seasonally adjusted employment in the summer.
17. It is not clear what a good counterfactual baseline would be. Over the three years preceding
the pandemic, employment at state and local governments increased about 0.75 percent per
year, but the increase in 2019 was 1.1 percent.
230 | Recession Remedies
Figure 6.1
Employment Trends during the Great Recession
vs. the COVID-19 Pandemic
A. State Government B. Local Government
Employment Employment
104
102
Education
Percent change in employment
(Great Recession)
100
98
Excluding
96 education
(COVID-19
Recession)
Excluding
94
education
(Great
92 Recession)
Education
(COVID-19
90 Recession)
1 5 9 13 17 21 25 29 33 37 41 45 49 1 5 9 13 17 21 25 29 33 37 41 45 49
Months Since Peak Months Since Peak
Source: Bureau of Labor Statistics 2022a.
Note: Seasonally adjusted monthly employment relative to Decem-
ber 2007 for Great Recession and January 2020 for the COVID-19
pandemic recession.
generally ineffective at supporting employment, or was the pandemic just too
unusual to be informative? In particular, we need to understand the extent to
which the employment declines reflected actual or expected tight budgetary
conditions versus pandemic-specific conditions, like office and school closures
and lower supply of workers to the sector because of COVID-19 fears or vac-
cination mandates.
Figure 6.2 shows the changes relative to January 2020 in data on seasonally
adjusted job openings, hiring, and job separations from the Job Openings and
Labor Turnover Survey (U.S. Bureau of Labor Statistics 2022c). These data are
available for state and local education workers combined as well as state and
local noneducation workers combined. Examining the patterns over time, most
of the reductions in employment in the spring of 2020 came from the employer
side: the number of employees hired fell between 30 percent and 50 percent
while layoffs surged. In the education sector, quits and other separations, which
include retirements, also increased sharply in the spring and summer of 2020.
State and Local Sector | 231
Figure 6.2
State and Local Job Openings and Labor Turnover
A. Noneducation
150
Change in employment rate
Layoffs
120
90
60 Quits and other
separations
30 Openings
0
-30
Hires
-60
Q2 Q3 Q4 Q1 Q2 Q3 Q4
2020 2021
B. Education
150
Change in employment rate
120
90
60
30
0
-30
-60
Q2 Q3 Q4 Q1 Q2 Q3 Q4
2020 2021
Source: Bureau of Labor Statistics 2022c.
Note: Change in seasonally adjusted levels relative to January 2020.
The story is a bit different beginning in the second quarter of 2021, after
vaccines were rolled out. Then, layoffs actually fell below pre-pandemic levels,
accounting for most of the employment increases, while job openings rose.
Hiring rates, however, remained muted. The lackluster pace of hiring may
reflect the fact that wages in the public sector did not keep up with private
sector wages reducing employers’ abilities to attract workers (see Figure 6.3).
It seems clear that at least part of the surge in layoffs in the spring of 2020
reflects school and office closures. Cafeteria workers, bus drivers, classroom
232 | Recession Remedies
Figure 6.3
Change in Wages, Private Sector and State and
Local Sector
6
Quarterly change at an annual rate (percent)
5
Private
4
3
2
State and local
1
0
Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2019 2020 2021
Source: Bureau of Labor Statistics 2022b.
Note: Change from previous quarter in employment cost index for
total compensation, annual rate.
aides, and office workers were no longer necessary; furthermore, the availabil-
ity of generous federally financed UI meant that laying off workers instead of
keeping them on the payroll was beneficial to both the government employer
and the workers. Evidence that support staff were more likely to be laid off
initially comes from the distribution of job losses across wage quartiles. The
data in Figure 6.4 show the change in employment by occupation wage quartile
from the same quarter in 2019 using the Current Population Survey (CPS; also
referred to as the “Household Survey”).18 For all three sectors in which there
were significant layoffs in the spring of 2020—local education, local noneduca-
tion, and state education, the lowest wage workers suffered disproportionately;
this is especially true in education, where low wage workers suffered the brunt
of the early layoffs.
18. The data are sorted into quartiles based on the average wage by occupation in 2019. I use
this comparison because there are strong seasonal patterns in state and local employment
and the CPS data are not seasonally adjusted.
State and Local Sector | 233
Figure 6.4
Changes in State and Local Employment, by
Wage Quartile
A. Local Education B. State Education
5 4
4th
Percent change in employment
quartile
Percent change in employment
2
0
0
-5 -2
3rd -4
quartile
-1 -6
2nd 4th
quartile quartile
-8
-15 3rd
quartile
-10 1st 2nd
1st quartile quartile quartile
-20 -12
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2020 2021 2020 2021
A. Local Noneducation B. State Noneducation
8 1
6 9
Percent change in employment Percent change in employment
4 8
2 7
0 6
-2 5
-4 4
-6 3
-8 2
-10 1
-12 0
-14 -1
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2020 2021 2020 2021
Source: Census Bureau 2022; author’s calculations.
Note: Percent change in employment relative to same quarter in
2019 by occupation wage quartiles.
But employment declined even for those in the highest wage quartile,
and by the beginning of 2021, the employment declines were relatively evenly
distributed across the quartiles. This is consistent with the notion that declines
in employment in the spring of 2020 consisted of workers who were no longer
234 | Recession Remedies
needed or could not work remotely being laid off as well as broader layoffs that
might have reflected expected budgetary pressures from the widely predicted
revenue losses. Scarred by the large revenue losses of the Great Recession and
the painful spending cuts they required, state and local governments may have
acted more quickly this time.
Evidence from Cross-State Variation
There are three sources of data on state and local government employment
by state. The first, on which I rely, is the monthly Current Employment Sta-
tistics (CES) data—also known as the establishment survey (Bureau of Labor
Statistics 2022a). This is a large survey covering roughly 70 percent of state
and local employment. It provides data for four categories of state and local
workers—state education, state noneducation, local education, and local non-
education—on a not-seasonally adjusted basis.19 The literature that I discuss
below relies on different data, including the monthly CPS (Census Bureau
2022) and the Quarterly Census of Wages (QCEW; Bureau of Labor Statistics
2022). The CPS is a household survey that does not include as many state and
local workers as the CES. The QCEW has comprehensive administrative data
on employment—that is, it includes every state and local government—but is
only available after a six- to nine-month lag.20 As I discuss below, these data
sources can give somewhat different signals, reflecting the different definitions
of “employment,” the source of information (employer versus household), and,
importantly, the different sample sizes of the surveys.
Using the CES, I compare employment in 2020 and 2021 to employment in
the same month in 2019, which I call the “employment gap,” to roughly adjust
for seasonal patterns in employment. I compare four time periods: May 2020
(the lowest level of state and local employment in the pandemic), October 2020
(when many state and local economies had largely opened up), March 2021
(right before the vaccines became widely available), and December 2021 (the
most recent available data at the time of the analysis).
Table 6.5 summarizes the data. There is a lot of variation across the states,
with employment falling sharply in some states but rising or only falling a bit
in others. The correlations in employment gaps across time periods and types
(state/local, education/noneducation) are reported in Table 6.6. There are three
important findings:
1. The reductions in employment in the spring of 2020 are only loosely
correlated with the gaps in later time periods, as shown in panel A
19. The data for total state and local employment—that is, not broken down by sector—are also
available on a seasonally adjusted basis.
20. The QCEW only captures workers subject to the federal unemployment tax; this excludes
elected officials and students on work-study programs who are captured by the CES.
State and Local Sector | 235
Table 6.5
Employment Changes Relative to 2019
Standard
Mean Deviation Minimum Maximum
Local Education
May 2020 -9.4% 3.3% -18.7% -4.4%
Oct 2020 -7.0% 2.9% -15.7% -2.5%
Mar 2021 -6.2% 3.0% -13.7% -0.7%
Dec 2021 -3.6% 2.5% -9.3% 1.1%
Local Noneducation
May 2020 -7.3% 4.7% -28.2% -1.3%
Oct 2020 -3.6% 2.7% -11.2% 3.1%
Mar 2021 -3.6% 3.2% -14.1% 2.9%
Dec 2021 -3.9% 3.1% -11.2% 3.3%
State Education
May 2020 -7.7% 6.8% -25.3% 8.9%
Oct 2020 -8.3% 4.6% -18.3% 0.0%
Mar 2021 -8.1% 5.7% -17.7% 12.5%
Dec 2021 -6.3% 7.4% -22.9% 11.1%
State Noneducation
May 2020 -0.7% 2.4% -8.9% 3.4%
Oct 2020 -0.4% 2.6% -6.0% 7.0%
Mar 2021 0.1% 2.9% -6.7% 5.1%
Dec 2021 -2.2% 3.2% -11.8% 5.9%
Source: Bureau of Labor Statistics 2022a; author’s calculations.
Note: This table reports the percent change in employment by sec-
tor compared to the same month in 2019.
of Figure 6.5 for total state and local employment. For example, the
correlation coefficient between the local education employment gap in
May 2020 and the local education employment gap in October 2020 is
just 0.25. The correlation coefficients are somewhat higher for the other
sectors but still weak. In other words, state and local governments that
laid off a lot of workers in the spring of 2020 are not necessarily those
where employment remained low by the fall of 2020.
2. However, the rankings by state in employment gaps are fairly steady
after the first wave of layoffs and rehires (panel B of Figure 6.5), with
the correlation coefficients in most cases closer to 0.7 or 0.8. That is,
states where the level of employment in October 2020 was particularly
far below their pre-pandemic level are also states with the largest gaps
between pre-pandemic employment and employment in March 2021
and December 2021.
236 | Recession Remedies
Table 6.6
Correlations in Employment Changes across States
A. Correlation within a Sector across Time Periods
May October March
May 2020 May 2020 2020 and October 2020 and 2021 and
and October and March December 2020 and December December
2020 2021 2021 March 2021 2021 2021
State education 0.42 0.37 0.31 0.78 0.70 0.68
State noneducation 0.52 0.59 0.38 0.83 0.71 0.83
Local education 0.25 0.22 0.17 0.88 0.58 0.72
Local noneducation 0.67 0.64 0.55 0.89 0.77 0.83
B. Correlation within a Time Period across Sectors
May October March December
2020 2020 2021 2021
State education vs 0.21 0.06 0.10 0.15
State noneducation
State education vs 0.30 0.04 0.02 -0.04
Local education
State noneducation vs -0.06 0.01 0.00 -0.01
Local noneducation
Local education vs 0.15 0.21 0.27 0.01
Local noneducation
Source: Bureau of Labor Statistics 2022a; author’s calculations.
Note: The table reports the correlation coefficients for employment
declines relative to the same month in 2019 across time periods and
sectors.
3. There is little correlation across types of employment. That is, changes
in local education employment are not very correlated with changes in
local noneducation employment, nor are they correlated with changes
in state education employment.
The lack of correlation between employment gaps in the spring of 2020
and later suggests that the first wave of layoffs reflected different factors than
those that continue to keep employment levels down. The lack of correlation
across types of workers suggests that it will be difficult to find “the story” that
explains employment declines at state and local governments.
Understanding Cross-State Variation in Employment
Losses
As shown in Online Appendix Table 6.2, many variables that might be expected
to predict employment losses in the state and local sector during the pandemic
do not. For example, state and local governments that suffered larger revenue
State and Local Sector | 237
Figure 6.5
State and Local Employment Losses over Time
A. Spring 2020 vs. Fall 2020
0
Percent change in employment,
-1
SD WY
-2 MT TN
AR
OK TX
-3 AK AL DE
UT NC
-4 IA ID RI NY FL
ND NE
MS SC GA IN
-5 CT MO NV KS PA
October 2019–20
MN AZ
-6 ME OH
NJ LA
WI WV KY MD
-7 CO
VT IL
NH MA VA NM
-8 CA
MI
-9 HI OR
WA
-10
-20 -15 -10 -5 0
Percent change in employment, May 2019–20
B. Fall 2020 vs. Spring 2021
0 MT
Percent change in employment,
-1 AL
SC TX SD
-2 NE UT DE
ID WY
-3 ND FL AR TN
KS OK
-4 WV ME MO MS IA
CO AZ NV GA NY NC
-5 VA
March 2019–21
LA MD OH PA RI AK
-6 MA MN
WA IN
-7 OR IL KY WI NJ CT
MI NH CA
-8 NM
HI
-9
VT
-10
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Percent change in employment, October 2019–20
Source: Bureau of Labor Statistics 2022a.
losses during the Great Recession were not particularly likely to lay off workers
during the pandemic, nor were states that predicted large state revenue losses or
states that actually experienced larger revenue losses in 2020. Similarly, oil states
and tourism states, which did suffer larger revenue losses, did not on average lay
off workers disproportionately. States that had large budget balances before the
onset of the pandemic did not have smaller employment losses, nor did states
that received more federal aid as a share of own source revenues (Figure 6.6).
238 | Recession Remedies
Figure 6.6
Total State and Local Employment Losses and
Federal Aid
A. May 2020 Employment Losses and Federal Aid
0
Percent change in employment,
-2 TX IN AZ
NC FL
-4 SC AR GA TN
NY MD NM DE
-6 CO VAKS PAAL MA LA MS WY
NJ RI
CA UT OHNV OKKY ID MT
-8 MO NE
OR IL
May 2019–20
WV NH
-10 IA WAMN MI AK
WI ME
-12 CT ND VT
SD
-14
-16 HI
-18
0 10 20 30 40
Federal Aid enacted spring 2020 as percent of own source revenue
B. October 2020 Employment Losses and Federal Aid
0
-1
Percent change in employment,
-2 WY SD
TX TN MT
-3 OK AR
IA NY NC FLAL AK DE
-4 ID
GA NE RI
UT SC
-5 MO CT IN ND
MS
KS NV
October 2019–20
-6 MN PA AZ
OH NJ ME
WI MD KY
-7 CO MA LA WV
VA NM NH VT
-8 CA IL
-9 OR MI
WA HI
-10
0 10 20 30 40
Federal Aid enacted spring 2020 as percent of own source revenue
Source: Bureau of Labor Statistics 2022a; Auerbach et al. 2020.
State and Local Sector | 239
Two budget-related variables do predict employment gaps in the state and
local sector. First, the share of K–12 spending financed by state governments
predicts employment gaps in local education, perhaps because localities in these
states were more vulnerable to budget cuts coming from state revenue losses.21
Second, states that announced hiring freezes at the beginning of the pandemic
in response to expected revenue losses had lower levels in state employment
outside of education relative to 2019, particularly later in the pandemic.22,23
What about attitudes toward COVID-19? These attitudes affected official
decisions about whether offices, schools, and parks were closed or operating
at less-than-full capacity, and also affected workers’ willingness to work in
person and the level of demand for public services during the pandemic. It is
well-known that blue states have been much more concerned about the pan-
demic than red states. The share of voters who chose President Biden tightly
correlated with vaccination rates (Online Appendix Figure 6.2) and also with
measures such as the Oxford Stringency Index, which captures the degree to
which governments shut down economic activity (Hale et al. 2021).
As shown in Figure 6.7, the vaccination rate, measured here as the share of
population age five and older fully vaccinated in January 2022, is only loosely
related to employment gaps in the state and local sector in May 2020. However,
by the fall of 2020 it is strongly predictive of employment gaps: places that
ultimately will have high vaccination rates are also those where state and local
employment is depressed.24 The time pattern of the relationship makes sense,
as attitudes toward COVID-19 were much more similar across the states in
the spring of 2020 (e.g., nearly every school system went virtual) than in the
fall of 2020 and later (Ferren 2021).
Table 6.7 reports the coefficients from regressions of employment losses rel-
ative to 2019 on variables that seem to have some explanatory power. The effects
of state financing of education and hiring freezes are as described above. The
vaccination rate is strongly predictive of employment gaps in local education and
21. Data from the Annual Survey of State Finances (U.S. Census Bureau 2021) shows that aid
to local governments increased less in FY 2020 than in FY 2019, perhaps reflecting some
cutbacks in aid at the start of the pandemic. It is worth noting that state fiscal years typically
begin July 1, but in a number of areas, county and city fiscal years begin later (Gentry 2015).
22. Twenty states (AK, HI, IN, MD, ME, MI, MO, MN, NH, NJ, NM, NY, NV, OH, PA, TN, VA,
WA, WI, WY) announced statewide hiring freezes in the spring of 2020, and many were not
lifted until 2021 (with New York’s not lifted until September 2021). Most states that instituted
hiring freezes exempted positions necessary to protect health and safety while others exempted
“essential positions” more broadly. Information on hiring freezes was gathered from the National
Association of State Budget Officers and the National Association for Law Placement (2021).
23. These states did not, on average, experience larger revenue losses.
24. The Biden share of the vote and the Oxford Stringency Index (measured in the fall of 2020)
also predict state and local employment gaps; the Biden share is about equally good as a
predictor, while the fit of the Oxford index is a little weaker, perhaps because it does not
measure overall attitudes as well (Hale et al. 2021).
240 | Recession Remedies
Figure 6.7
Vaccination Rates and Changes in State and
Local Employment
A. May 2020
0
Percent change in employment,
-2 IN AZ
GA TX
-4 TN NC FL MD
SC DE NM NY
WY MS AR VA RI
-6 OK OH KS
AL LA PA CO NJ MA
ID KY NV UT CA
-8 MT
MO MI IL OR NH
NE
May 2019–20
-10 IA WA
WV MN ME
AK WI
ND
-12
SD CT VT
-14
-16 HI
-18
50 65 80 95
Percent of the population over age 5 fully vaccinated by January 2022
B. October 2020
0
Percent change in employment,
-1
WY SD
-2 TN MT
-3 TX
AL AR NC AK
ID OK IA DE
-4 SC UT FL NY RI
GA ND NE
-5 MS IN MO
October 2019–20
NV KS CT
MN
-6 WV OH AZ PA NJ ME
LA WI CO
-7 KY NM VA MD MA
CA VT
-8 IL NH
-9 MI
OR WA HI
-10
50 65 80 95
Percent of the population over age 5 fully vaccinated by January 2022
Source: Bureau of Labor Statistics 2022a; Center for Disease Con-
trol and Prevention 2021.
noneducation sectors in the fall of 2020 and the spring of 2021 and of employments
gaps in state education in the spring of 2021. But in later months, as vaccinations
were rolled out and the blue states opened up, the coefficient on the vaccination
rate became smaller and less significant. By December 2021 it no longer had a
State and Local Sector | 241
Table 6.7
Explaining the Cross-State Variation in
Employment Declines
Education Excluding Education
May Oct. Mar Dec. May Oct. Mar. Dec.
2020 2020 2021 2021 2020 2020 2021 2021
Local Employment
Vaccination rates, -0.04 -0.13*** -0.14*** -0.04 -0.13* -0.09** -0.10** -0.02
January 2022
State share K–12 -0.08* -0.06* -0.08** -0.06* 0.04 0.03 -0.00 0.02
State Employment
Vaccination rates, 0.03 -0.07 -0.17* -0.09 0.01 0.03 -0.02 0.02
January 2022
State hiring freeze 0.01 -0.01 -0.01 0.00 -0.01 -0.02** -0.02** -0.01
State and Local Employment
Vaccination rates, -0.08* -0.10*** -0.12*** -0.04
January 2022
State share K–12 -0.06* -0.03 -0.04** -0.03
State hiring freeze -0.01 -0.01 -0.01** -0.01**
Source: Bureau of Labor Statistics 2022a; Center for Disease Con-
trol and Prevention 2021; Congressional Research Service 2019.
Note: The change in employment is defined as the change in a par-
ticular month from the same month in 2019. Vaccination rate is the
share of the population 5 years and over who are fully vaccinated in
January 2022. Regressions use STATA robust command.
statistically significant effect on the employment gap. The insignificant effect of
the vaccination rate on state noneducation employment may reflect that states
that were very concerned about COVID-19 invested more in public health, which
could have boosted employment, offsetting the negative effects from office clo-
sures and the like. In addition, state government jobs may be more amenable to
remote work and thus less affected by attitudes toward COVID-19.
This analysis provides little evidence that weak employment has been
driven by tight budget conditions or that federal aid has been an import-
ant determinant of employment. To be sure, states that announced hiring
freezes did so in expectation of tight fiscal conditions, and these freezes, which
did not get lifted until 2021, clearly constrained hiring of state noneducation
workers. In addition, the fact that K–12 employment was cut more in states
where the state financed a larger share of education expenses also suggests that
fears of budget cuts affected employment. Still, the big determinants of fiscal
242 | Recession Remedies
conditions—revenue losses and federal aid—do not help explain the variation
in employment across states.
To some extent, the finding that budget conditions were not important
factors behind employment losses seems obviously correct; state and local gov-
ernments have received abundant aid and yet employment remains lower than
it was before the pandemic. But simply examining the relationship between the
extent of federal aid and the change in employment may not uncover the causal
effects of federal aid on employment. Federal aid was not randomly distributed:
small states got much more generous aid relative to their budgets, and aid to
K–12 was provided on the basis of need (as measured by Title I funding) so
that poorer states got more generous allotments. Each of these factors could
confound the analysis. For example, states that got a lot of federal aid were small
states, and if for some reason, small states are more likely to have large layoffs,
then it appears that generous federal aid is associated with worse outcomes. A
similar problem arises if states with few resources were more likely not only
to have layoffs but also to get generous K–12 funding.
Other Empirical Evidence
The finding using the CES that budget conditions—revenue declines and fed-
eral aid—had little effect on employment is contradicted by two papers in the
literature that attempt to carefully identify the causal effects of tight budgets.25
Relying on the CPS to measure state and local layoffs, Green and Louali-
che (2021) use two strategies to show that a large proportion of the decline in
employment in the state and local sector in the spring of 2020 was attributable
to tight budgets. First, they argue that states that depended on sales taxes as
an important source of financing expected larger revenue losses and show that
these states cut employment more. They calculate that sales tax exposure can
explain over 660,000 of the state and local jobs lost in April 2020, or about
two-thirds of the total decline. In addition, they exploit the kink in the CARES
Act formula for aid to show that states that got more CARES funding had
lower employment losses, finding that the CARES Act prevented the loss of
400,000 jobs. They also find that these effects are larger for states with smaller
rainy day funds.
In the online appendix, I reevaluate the findings in Green and Loualiche.
In particular, I compare the results using the CPS used by Green and Louali-
che to results using measures of employment changes from the establishment
survey (used above) and the QCEW. I show that the results in the Green and
Loualiche paper do not hold using the establishment data or the QCEW data,
25. In addition, Sheiner (2020) showed that states that got more federal aid and states that had
larger predicted revenue losses suffered larger declines in local education employment.
However, the CES data have been revised since then, and these variables no longer predict
the revised measures of employment losses by state.
State and Local Sector | 243
which are almost a complete census of state and local workers. Using these
alternative—and much more representative—data, there is no relationship
between state and local employment declines and the reliance on sales taxes
or the generosity of CARES Act funding (Online Appendix tables 6.3 and
6.4). The increases in layoffs and the declines in employment across states in
the CPS have little relationship to the declines in employment as measured by
the other two surveys, likely because the CPS sample sizes are too small for
reliable cross-state analysis.26
Haughwout, Hyman, and Shachar (2021) used the kink in CARES Act fund-
ing to counties—only counties with population greater than 500,000 were eligible
for direct aid—to explore the effects of federal aid on employment. Using the
QCEW to measure state and local employment, they found that direct CARES
Act funding led local governments to recall about 25 percent more education
workers in the first two months after the law was passed and that the effect per-
sisted into the fall for governments with good credit ratings. However, only $28
billion of the CARES Act funds went directly to counties, with the remaining
$132 billion going to states, territories, and tribal governments. Furthermore,
states where counties got money directly did not get more money overall: the
amount provided to counties was subtracted from the overall allocation, sug-
gesting that the direct targeting of counties mattered for employment.
Overall, it seems clear that the employment losses vary a lot by state in ways
that cannot fully be explained. Employment gaps—the differences between
employment during the pandemic and in 2019—were clearly affected by atti-
tudes toward COVID-19, and there is some evidence that fears of tight fiscal
conditions and direct federal aid to counties affected employment, but gen-
erous federal aid to states was clearly not sufficient to reverse or prevent all
the employment losses. One important question is, why not? What did state
and local governments do with the federal aid, and why didn’t they use it to
increase employment?
What Has Happened to Spending by State and
Local Governments?
Information on spending by state and local governments during the pandemic
is sparse. The Census Bureau’s Annual Survey of State Government Finances
for fiscal year 2020, which includes the first quarter of the pandemic for most
states, was released in December 2021. No data are yet available for either FY
2020 spending by local governments or for FY 2021 for either state or local gov-
ernments. The National Association of State Budget Officers (NASBO) publishes
26. The total number of state and local workers captured in the CPS in April 2020 was about
6,000, of whom 572 were unemployed. The median number of unemployed state and local
workers in a state was just nine.
244 | Recession Remedies
an annual report on state government expenditures, the latest of which includes
estimates for FY 2021. However, differences in accounting practices across the
states can make that report somewhat difficult to interpret, as I explain below.
Information for Fiscal Year 2020
According to the Census Bureau’s Annual Survey of State Finances, expendi-
tures by state governments increased 7.6 percent in 2020, up sharply from the
4.3 percent increase in 2019 (Online Appendix Table 6.5). However, most of that
increase reflects larger expenditures on insurance benefits, consisting mostly of
UI benefits. These expenditures are not subject to balanced budget requirements,
and states can adjust to the shock gradually over many years.27 Excluding such
expenditures, state expenditures increased 4.7 percent—roughly the same as
the 4.5 percent increase in 2019. The categories of spending did show marked
differences, however. Growth in spending on corrections, police protection,
health, hospitals, and public welfare picked up from 2019 while spending on
parks and recreation, natural resources, highways, governmental administra-
tion, and education increased at a slower pace than in 2019. The savings that
states realized from shutting down schools, offices, and parks likely allowed
for increased spending elsewhere.
Information for Fiscal Year 2021
Table 6.8 presents the data from NASBO for both FY 2020 and FY 2021. Overall
spending increased 8.7 percent in FY 2020 and 16.2 percent in FY 2021 according
to their data—the highest reading in the 35-year history of the NASBO report.
Spending financed by the federal government increased 21 percent in FY 2020
and 36 percent in FY 2021. One difficulty with the NASBO data is the incon-
sistent accounting for UI benefits across states: according to NASBO, some
states include only administrative costs associated with UI in their accounting,
while others include benefits costs as well, but they do not report which method
they use, nor is it clear whether they include all UI benefits or just the regular
benefits financed by states (Brian Sigritz, email communication, December 2,
2022).28 UI expenditures, to the extent they are included, are in the “all other”
category. Excluding this category of spending, spending increased 4.6 percent
in FY 2020—similar to the Census data—and 10 percent in FY 2021. Even
accounting for the higher inflation in FY 2021, this is a rapid pace of increase.
Furthermore, many other types of pandemic-related expenditures are included
27. Regular UI benefits are financed by states but paid out of a UI trust fund. When benefits
exceed available resources, states get automatic loans from the federal government. Fol-
lowing the Great Recession, states in aggregate eliminated their UI debt very slowly, only
extinguishing it in 2019 (Auerbach et al. 2020).
28. Brian Sigritz is the director of state fiscal studies at NASBO.
State and Local Sector | 245
Table 6.8
Increase in State Spending, by Source of Funds
and Category, FY2020 and FY2021
General Other State Federal Bond
Source of Financing Total Revenues Revenues Funds Issuance
Fiscal Year 2020
K–12 education 4.3% 4.9% 0.7% 7.8% 44.8%
Higher education 1.3% 3.4% 2.6% 13.8% 5.4%
Public assistance -8.0% 3.4% -17.5% 9.1%
Medicaid 6.5% 1.3% -1.3% 10.0%
Corrections 1.7% 0.3% 13.5% 261.6% 54.0%
Transportation 6.0% -6.9% 3.1% 6.4% -0.4%
All other (includes UI 18.9% 4.9% 5.1% 61.5% 0.5%
benefits in some states)
Total excluding other 4.6% 2.9% 1.7% 9.9% 7.8%
Total 8.7% 3.4% 3.1% 20.5% 4.7%
Fiscal Year 2021
K–12 education 28.5% 3.2% 19.0% 83.9% 46.4%
Higher education 4.0% 4.1% 0.5% 11.7% -3.2%
Public assistance 26.8% 15.6% 35.4% 28.5%
Medicaid 11.7% 0.1% 12.2% 15.8%
Corrections 0.2% -7.3% -1.5% 171.4% -42.7%
Transportation 8.5% 38.8% 4.2% 13.1% 17.4%
All other (includes UI 29.4% 8.5% 12.3% 73.0% 13.4%
benefits in some states)
Total excluding other 10.1% 2.3% 5.3% 21.4% 7.9%
Total 16.2% 4.1% 8.3% 35.7% 10.1%
Source: National Association of State Budget Officers (NASBO) 2021.
Note: The table reports the annual percentage increase in state
spending in fiscal years 2020 and 2021.
in “all other,” including spending on public health, housing assistance, eco-
nomic relief, aid to local governments, and broadband and other technology
upgrades, so excluding the category will understate the true increase in state
spending in response to COVID-19.
Thus, according to the NASBO (2021) data, state spending in 2021 was quite
robust, even though employment remained weak, and it appears that states
246 | Recession Remedies
are spending the federal aid at a relatively quick pace.29 Nonetheless, states
budget conditions remain very healthy, with total balances (rainy day funds
plus general budget surpluses) reaching a record 23.3 percent of expenditures
at the end of FY 2021. Furthermore, more than $100 billion in aid has yet to be
distributed (Parlapiano, Solomon, Ngo, and Cowley 2022).
State and local spending in the National Income and Product Accounts
(NIPA; Bureau of Economic Analysis 2021a and 2021b) looks quite different
from the state spending reported in the NASBO State Expenditure Report: in
the NIPA, nominal state and local expenditures rose 3.3 percent in FY 2020 and
just 3.7 percent in FY 2021.30 Nominal state and local purchases—the part of
state and local spending that enters directly into GDP—rose 3.2 percent in FY
2020 and 2.5 percent in FY 2021. But the data on which the purchase estimates
are based are quite sparse: They include data on employee compensation (which
comes from data on employment and wages), construction (which comes from
monthly Census surveys), and motor vehicle purchases (from R. L. Polk and
Company). Most other expenditure categories are estimated from historical
trends.31 Given the strength of the NASBO data, it seems likely that the BEA
has underestimated state and local purchases in FY 2021 and thus overstated
the drag of the state and local sector on the macroeconomy. Of course, not
all the increase in spending represents state and local purchases. As Table 6.8
shows, state spending on public aid increased sharply and this spending is not
included in NIPA purchases—so the eventual revisions to GDP, which will
occur during the annual revisions after BEA has the data from the Census,
are difficult to predict.
Why Haven’t State and Local Governments Used
Federal Aid to Increase Employment or at Least
Return It to Pre-Pandemic Levels?
It is hard to understand why employment in the state and local sector remains
depressed given the rollout of vaccines; the reopening of parks, schools, and
offices; and the healthy state budgets. In this section, I discuss several possible
29. Federally funded state expenditures were $473 billion higher than implied by a baseline
where federal grants increased four percent per year from 2019 on. Excluding the “all other”
category that contains at least some federally paid UI benefits, the increase in federally
financed expenditures is $163 billion. Of the roughly $1 trillion in funds, about $350 billion
went directly to states.
30. In the NIPA, all UI benefits are deemed federal spending, so they are not included in these
numbers. “FY 2020” is defined here as 2019Q3 through 2020Q2.
31. For total expenditures, as opposed to purchases, the Bureau of Economic Analysis also
has access to administrative data on Medicaid spending and some other social programs
(U.S. Bureau of Economic Analysis 2021).
State and Local Sector | 247
reasons. These are based on conversations with state officials and anecdotes
rather than on hard data.
One narrative that I have heard repeatedly is that “a rule of budgeting is
not to use one-time money for permanent expenses.” Thus, decisionmakers
are loathe to use federal aid on services for which there might be continued
demand after the federal money runs out. For example, one school superin-
tendent explained that he would not want to reduce class sizes using federal
aid—even if that might be particularly effective at remediating learning losses
from virtual schooling—because there would be continued demand for smaller
classes after the federal aid ran out. Instead, schools were looking for one-time
expenditures (e.g., new equipment or HVAC systems) to fund with the federal
aid. A counterargument is that many states are eyeing tax cuts, which similarly
hold down revenues and worsen budget conditions over the long run.
To the extent that layoffs in the spring of 2020 were caused by the fear of
tight budget conditions, the restrictions on the use of the CARES Act funds
(i.e., the prohibition on using them for expenses that were budgeted for before
the COVID-19 pandemic) may have limited their effectiveness at minimizing
layoffs. Similarly, the funneling of most of the aid money through the states—
rather than directing a significant portion to local governments—may have
meant the money was too slow to get to local governments.
When employers wanted to resume hiring, they faced an extremely tight
labor market. As shown in Figure 6.2, job openings have been elevated since
the second quarter of 2021, but hiring has not increased much, and as shown
in Figure 6.3, wages of state and local government workers have not kept pace
with raises in the private sector. This lack of pay parity likely exacerbated hiring
difficulties associated with the difficulties faced by many state and local workers
during the pandemic. School workers, for example, have had to deal with risks
of COVID-19 exposure, hybrid teaching, school closures, and staff shortages.
With the state and local workforce largely unionized, it is difficult to provide
higher pay to attract new employees without having to increase pay all around.
And again, state and local governments are wary of increasing pay because of
the implications for future years when federal money is no longer available.
Another possibility that several state officials mentioned is that the pan-
demic has improved the efficiency of the public sector. Whether it was laying
off workers whose services were not highly valued or the increased efficiency
from remote work and videoconferencing, government officials may not think
they need as many workers as they did before the pandemic. State noneduca-
tion employment fell during the Great Recession and never fully recovered,
suggesting that some of the declines in employment may be long lasting.
Finally, changes that occurred during the pandemic may have increased
the uncertainty about future spending needs and revenues. For example, given
the strains on the public health system during the pandemic, state and local
governments may want to increase investments going forward to prepare for
future health emergencies (Bosman and Fausset 2020). In addition, the transition
248 | Recession Remedies
to remote work leaves many cities unsure about whether economic activity
will return to pre-pandemic levels anytime soon, putting revenues from taxes
and fees at risk and also leading them to perhaps fundamentally rethink their
transportation and transit infrastructure. This type of uncertainty may have led
state and local governments to be very cautious in committing funds too rapidly.
Lessons Learned
Seven lessons can be learned from the experiences during the pandemic.
1. Policies that provide fiscal support to households and businesses indi-
rectly support state and local revenues. When contemplating the amount
of direct aid that might be necessary, these policies should be taken
into account.
2. To prevent layoffs and ensure adequate public service provision, aid to
state and local governments should be automatic or should be provided
early in a recession. While most of the employment declines in the
state and local sector over the past two years are likely related to pan-
demic-specific factors, there is evidence that some of the employment
losses reflected fear of tight budget conditions. At a minimum, the states
that imposed hiring freezes would likely not have taken that step had
they anticipated the substantial federal aid that would be forthcoming.
While not definitive, the experience during the pandemic suggests that
preventing initial layoffs is important.
3. Federal aid should go directly to states and localities instead of only
to state governments, where possible, and should have few conditions
placed on it. Although money is fungible, the way aid is distributed
does matter. The fact that states were explicitly prohibited from using
CARES Act funding to cover revenue losses and the targeting of aid
to states and only very large substate governments may have made it
less effective at preventing layoffs. The $350 billion in federal aid in the
ARP was much better on this front. It provided aid directly to local
governments, thus bypassing the possibly slow process by which states
redistribute aid to local governments, and allowed a much broader
array of purposes.
4. State and local governments are reticent about using one-time federal
aid to finance ongoing expenditures. That might preclude aid from
being used for the most effective purposes (e.g., increasing teaching
resources to address learning losses during the pandemic).
State and Local Sector | 249
5. The ability of state and local governments to borrow from the Fed can
serve as an important backstop that can help stabilize municipal bond
markets in times of crisis.
6. More timely data on state and local governments are needed. In 2020
the Bureau of Economic Analysis relied on the Urban Institute’s col-
lection of data from state agencies because these were available before
the Census Bureau’s Quarterly Summary of State and Local Taxes.
Similarly, NASBO published data for state spending in FY 2021 over a
year before the Census will release such data. Unfortunately, there are
no nongovernmental sources for data on spending by local governments,
and government data will not be released for FY 2020 (meaning through
2020Q2) until June 2022. The lack of timely official data made it difficult
to assess the fiscal conditions of state and local governments and to
know whether the enacted policies were successful. Lack of timely data
has also made it difficult to accurately assess the state of the economy
because the BEA has such limited information on purchases by state
and local governments.
7. This pandemic was different from most recessions. The fact that revenue
losses were modest and short-lived, and employment declined despite
healthy budgets, does not suggest that aid to state and local governments
is ineffective or unnecessary in economic downturns. The lesson of the
Great Recession—that inadequate aid to state and local governments can
hamper an economic recovery—should not be discarded because of the
recent experience; the pandemic created unusual economic conditions
that are not likely to recur in future recessions.
References
Auerbach, Alan J., William G. Gale, Byron Lutz, and Louise Sheiner. 2020. “Effects
of COVID-19 on Federal, State, and Local Government Budgets.” Brookings
Papers on Economic Activity (Fall). Brookings Institution, Washington, D.C.
Auxier, Richard C. 2021 “Treasury Will Allow States to Take ARP Funds and Cut
Taxes, With Some Guardrails.” TaxVox (blog). Urban-Brookings Tax Policy
Center, Washington, D.C. May 13, 2021.
Bartik, Timothy J. 2020. “An Updated Proposal for Timely, Responsive Federal Aid
to State and Local Governments during the Pandemic Recession.” W.E. Upjohn
Institute for Employment Research, Kalamazoo, Michigan.
Bivens, Josh, and Naomi Walker. 2020. “At Least $500 Billion More in Coronavirus
Aid Is Needed for State and Local Governments by the End of 2021.” Working
Economics (blog). Economic Policy Institute, Washington, D.C. April 9, 2020.
250 | Recession Remedies
Bordo, Michael D., and John V. Duca. 2020. “How New Fed Corporate Bond Pro-
grams Dampened the Financial Accelerator in the Covid-19 Recession.” Working
Paper 28097, National Bureau of Economic Research, Cambridge, MA.
Bosman, Julie, and Richard Fausset. 2020. “The Coronavirus Swamps Local Health
Departments, Already Crippled by Cuts,” New York Times, March 14, 2022.
Accessed March 20, 2022.
Centers for Disease Control and Prevention. 2021. “COVID-19 Vaccinations in the
United States, Jurisdiction.” Centers for Disease Control and Prevention, U.S.
Department of Health and Human Services, Washington, D.C. Last updated
March 13, 2022.
Center on Budget and Policy Priorities. 2020. “States Grappling with Hit to Tax
Collections,” Center on Budget and Policy Priorities, Washington, D.C. Accessed
March 16, 2022.
Clemens, Jeffrey, and Stan Veuger. 2020. “Implications of the COVID-19 Pandemic
for State Government Tax Revenues.” National Tax Journal 73(3): 619–44.
Congressional Budget Office. 2020a. “The Budget and Economic Outlook: 2020
to 2030.” Congressional Budget Office, Washington, D.C. January 28, 2020.
Congressional Budget Office. 2020b. “The Budget and Economic Outlook: 2020 to
2030.” Congressional Budget Office, Washington, D.C. July 2, 2020.
Congressional Research Service. 2019. “Federal Grants to State and Local Gov-
ernments: A Historical Perspective on Contemporary Issues.” Report R40638.
Congressional Research Service, Washington, D.C.
Dadayan, Lucy. 2020. “COVID-19 Pandemic Could Slash 2020–21 State Revenues
by $200 Billion.” TaxVox (blog), Urban-Brookings Tax Policy Center, Wash-
ington, D.C.
Federal Election Commission. 2020. “Official 2020 Presidential General Election
Results.” Federal Election Commission, Washington, D.C. Last updated Jan-
uary 28, 2021.
Ferren, Megan. 2021. “Remote Learning and School Reopenings: What Worked
and What Didn’t.” Center for American Progress, Washington, D.C.
Fiedler, Matthew, Jason Furman, and Wilson Powell III. 2019. “Increasing Federal
Support for State Medicaid and CHIP Programs in Response to Economic
Downturns.” Brookings Institution, Washington, D.C.
Gentry, Megan. 2015. “Staggered Fiscal Years Between States and Local Gov-
ernments.” Report No. 2015-R-0103. Connecticut General Assembly Office of
Legislative Research, Hartford, CT.
Green, Daniel, and Erik Loualiche. 2021. “State and Local Government Employment
in the COVID-19 Crisis.” Journal of Public Economics 193:104321.
Hale, Thomas, Noam Angrist, Rafael Goldszmidt, Beatriz Kira, Anna Pethernick,
Toby Phillips, Saumel Webster, et al. 2021. “A Global Panel Database of Pandemic
Policies (Oxford COVID-19 Government Response Tracker).” University of
Oxford, Oxford, United Kingdom. Accessed March 10, 2022.
State and Local Sector | 251
Haughwout, Andrew, Benjamin Hyman, and Or Shachar. 2021. “The Option Value
of Municipal Liquidity: Evidence from Federal Lending Cutoffs during COVID-
19.” Federal Reserve Bank of New York, New York, NY.
McNichol, Elizabeth, Michael Leachman, and Joshuah Marshall. 2020. “States
Need Significantly More Fiscal Relief to Slow the Emerging Deep Recession.”
Center on Budget and Policy Priorities, Washington, D.C.
Mengle, Rocky. 2022. “Taxes on Unemployment Benefits: A State-by-State Guide.”
Kiplinger, Washington, D.C.
National Association of State Budget Officers (NASBO). 2020. “The Fiscal Survey
of States: Spring 2020.” National Association of State Budget Officers, Wash-
ington, D.C.
National Association of State Budget Officers (NASBO). 2021. “State Expendi-
ture Report, Fiscal 2019-2021.” National Association of State Budget Officers,
Washington, D.C.
Ochieng, Nancy, Jeannie Fuglesten Biniek, MaryBeth Musumeci, and Tricia
Neuman. 2022. “Funding for Health Care Providers During the Pandemic: An
Update.” Kaiser Family Foundation, San Francisco, CA.
Parlapiano, Alicia, Deborah B. Solomon, Madeleine Ngo and Stacy Cowley. 2022.
“Where $5 Trillion in Pandemic Stimulus Money Went.” New York Times. March
11, 2022. Accessed March 16, 2022.
Sheiner, Louise. 2020. “Why Is State and Local Employment Falling Faster than
Revenues?” Up Front (blog), Brookings Institution, Washington, D.C.
Skinner, Rebecca R. 2019. “State and Local Financing of Public Schools.” Report
R45827. Congressional Research Service, Washington, DC.
Sorenson, Bent E., and Oved Yosha. 2001. “Is State Fiscal Policy Asymmetric Over
the Business Cycle?” Federal Reserve Bank of Kansas City, Kansas City, MO.
The National Association for Law Placement. 2021. “PSJD | Government Hiring
Freezes in the News (COVID-19).” Public Service Job Database. The National
Association for Law Placement. Accessed February 25, 2022.
U.S. Bureau of Economic Analysis (BEA). 2021a. National Income and Prod-
uct Accounts (NIPA). “Table 1.1.5. Gross Domestic Product.” U.S. Bureau of
Economic Analysis, U.S. Department of Commerce, Suitland, MD. Accessed
March 16, 2022.
U.S. Bureau of Economic Analysis (BEA). 2021b. National Income and Product
Accounts (NIPA). “Table 3.3. State and Local Government Current Receipts
and Expenditures.” U.S. Bureau of Economic Analysis, U.S. Department of
Commerce, Suitland, MD. Accessed March 16, 2022
U.S. Bureau of Economic Analysis (BEA). 2021c. “Nipa Handbook: Concepts
and Methods of the U.S. National Income and Product Accounts.” Bureau of
Economic Analysis, U.S. Department of Commerce, Suitland, MD. Accessed
December 2021.
U.S. Bureau of Labor Statistics (BLS). 2022a. “Current Employment Statistics—
CES.” U.S. Bureau of Labor Statistics, U.S. Department of Labor, Washington,
D.C. Last modified March 14, 2022.
252 | Recession Remedies
U.S. Bureau of Labor Statistics (BLS). 2022b. “Employment Cost Index.” U.S.
Bureau of Labor Statistics, U.S. Department of Labor, Washington, D.C. Last
modified January 28, 2022.
U.S. Bureau of Labor Statistics (BLS). 2022c. “Job Openings and Labor Turnover
Survey.” U.S. Bureau of Labor Statistics, U.S. Department of Labor, Washington,
D.C. Last modified March 9, 2022.
U.S. Bureau of Labor Statistics (BLS). 2022d. “QCEW Overview.” U.S. Bureau of
Labor Statistics, U.S. Department of Labor, Washington, D.C. Last modified
February 23, 2022.
U.S. Bureau of Labor Statistics (BLS). 2022e. “Local Area Unemployment Statistics.”
U.S. Bureau of Labor Statistics, U.S. Department of Labor, Washington, D.C.
Last modified March 14, 2022.
U.S. Census Bureau. 2021. “Annual Survey of State and Local Government Finances.”
U.S. Census Bureau, U.S. Department of Commerce, Suitland, MD. Last revised
October 8, 2021.
U.S. Census Bureau. 2022. “Current Population Survey (CPS) 2019–2021.” U.S.
Census Bureau, U.S. Department of Commerce, Suitland, MD. Last modified
January 14, 2022.
U.S. Department of Education (ED). 2021. “HEERF II Allocations for Public and
Nonprofit Institutions under CRRSAA section 314(a)(1).” Office of Postsecondary
Education, U.S. Department of Education, Washington, DC.
U.S. Department of the Treasury. n.d. “Coronavirus State and Local Fiscal Recovery
Funds.” U.S. Department of the Treasury, Washington, D.C.
U.S. Department of the Treasury. 2020. “Department of the Treasury Office of
Inspector General Coronavirus Relief Fund Frequently Asked Questions Related
to Reporting and Recordkeeping (Revised)” Office of Inspector General, Sep-
tember 2020.
Understanding Alaska’s Budget. n.d. “Understanding Alaska’s Revenue.” Accessed
March 10, 2022 at http://www.alaskabudget.com/revenue/.
Whitaker, Stephan D. 2020a. “Estimates of State and Local Government Reve-
nue Losses from Pandemic Mitigation.” Federal Reserve Bank of Cleveland,
Cleveland, Ohio.
Whitaker, Stephan D. 2020b. “How Much Help Do State and Local Governments
Need? Updated Estimates of Revenue Losses from Pandemic Mitigation.” Federal
Reserve Bank of Cleveland, Cleveland, Ohio.
White, Dan, Sarah Crane, and Colin Seitz. 2020. “Stress-Testing States: COVID-
19.” Moody’s Analytics, New York, NY.
Wilshire Associates. 2022. “Wilshire 5000 Total Market Full Cap Index.” Wilshire
Associates, Santa Monica, CA.
The COVID-19 pandemic posed an extraordinary threat to lives
and livelihoods. In the United States, the pandemic triggered a
sharp downturn. Yet, the ensuing economic recovery was faster
and stronger than nearly any forecaster anticipated due in part
to the swift, aggressive, sustained, and creative response of
U.S. fiscal and monetary policy. But when the next recession
arrives, it most likely won’t be triggered by a pandemic.
Recession Remedies examines and evaluates the breadth of
the economic-policy response to COVID-19. Chapters address
Unemployment Insurance, Economic Impact Payments, loans
and grants to businesses, assistance to renters and mortgage
holders, aid to state and local governments, policies that
targeted children, Federal Reserve policy, and the use of non-
traditional data to monitor the economy and guide policy.
These chapters provide evidence and lessons to apply to the
next recession.
Contributors
Anna Aizer • Robin Brooks • Tomaz Cajner
Gabriel Chodorow-Reich • Wendy Edelberg • Laura Feiveson
Jason Furman • Peter Ganong • Tim Geithner • Michael Gelman
Kristopher Gerardi • Fiona Greig • Laurie Goodman • Ben Iverson
Christopher Kurz • Lauren Lambie-Hanson • Pascal Noel
Claudia Persico • Jonathan Pingle • Louise Sheiner
Melvin Stephens, Jr. • Daniel Sullivan • Adi Sunderam
Stacey Tevlin • Joseph Vavra • Susan Wachter
David Wessel • Paul Willen
File and source
- File
- hamilton-2023-01-lessons-learned-support-state-local-sector-covid-19.pdf
- Size
- 1,628,625 bytes
- SHA-256
- ca294a43760659e084ba20a31c2937267458b19271375ea0c94e9c817332aa02
- Original
- www.hamiltonproject.org