Economic Impact Payments and Household Spending During the Pandemic — NBER Working Paper 30596
Summary
NBER Working Paper 30596, dated October 2022, is by Jonathan A. Parker, Jake Schild, Laura Erhard and David Johnson and was prepared for the Fall 2022 Brookings Papers on Economic Activity (BPEA) conference. The abstract says households spent only a small fraction of their 2020 Economic Impact Payment within a couple of months of arrival. It says households with lower liquid wealth entering the pandemic and those less able to earn while working from home each raised consumption more following receipt of their payment, and that estimates suggest even less spending on average in response to the second and third rounds. It says the point estimates imply less short-term spending on average than in response to economic stimulus payments in 2001 or 2008.
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
- Issuer
- Source documents
- Document type
- Economic Impact Payments And Household Spending During The Pandemic
- Case
- Economic Impact Payments And Household Spending During The Pandemic
Cite this document
“Economic Impact Payments and Household Spending During the Pandemic — NBER Working Paper 30596.”
Original at www.nber.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/economic-impact-payments-and-household-spending-during-the-pandemic/
, SHA-256 5e1b395c7c654aa9cebbd7d245ca5aacc8c343ba703d73dddd7fcd9b5b69217c.Full text
NBER WORKING PAPER SERIES
ECONOMIC IMPACT PAYMENTS AND HOUSEHOLD SPENDING
DURING THE PANDEMIC
Jonathan A. Parker
Jake Schild
Laura Erhard
David Johnson
Working Paper 30596
http://www.nber.org/papers/w30596
NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
October 2022
Prepared for the Fall 2022 Brookings Papers on Economic Activity (BPEA) conference.
Jianmeng Lyu and Ian Sapollnik provided excellent research assistance. For useful discussions,
we thank Peter Ganong, Thesia Garner, Jianmeng Lyu, Claudia Sahm, Gianluca Violante, and in
particular Karen Dynan, Jan Eberly, and Matt Rognlie as well as participants at the 2022 Society
of Government Economists Annual Conference, the Conference on Direct Stimulus Payments to
Individuals in the Covid-19 Pandemic at the Athens University of Economics and Business and
University of Exeter Business School (online), a Bundesbank "Friendlyfaces" Workshop, and the
Fall 2022 BPEA Conference. The initial phases of this work occurred while David was at the
University of Michigan. The views expressed herein are those of the authors and do not
necessarily reflect the views of the National Bureau of Economic Research.
NBER working papers are circulated for discussion and comment purposes. They have not been
peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies
official NBER publications.
© 2022 by Jonathan A. Parker, Jake Schild, Laura Erhard, and David Johnson. All rights
reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit
permission provided that full credit, including © notice, is given to the source.
Economic Impact Payments and Household Spending During the Pandemic
Jonathan A. Parker, Jake Schild, Laura Erhard, and David Johnson
NBER Working Paper No. 30596
October 2022
JEL No. D14,D15,E21,E62,G5,H31
ABSTRACT
Households spent only a small fraction of their 2020 Economic Impact Payment (EIPs) within a
couple of months of arrival, consistent with i) pandemic constraints on spending, ii) other
pandemic programs and social insurance, and iii) the broader disbursement of the EIPs
compared to the economic losses during the early stages of the pandemic. While these EIPs did
not fill an urgent economic need for most households, the first round of EIPs did provide
timely pandemic insurance to some households who were more exposed to the economic
losses from the pandemic. Households with lower liquid wealth entering the pandemic and
those less able to earn while working from home each raised consumption more following
receipt of their EIP. While our measurement for later EIPs is not as reliable, our estimates
suggest even less spending on average to the second and third rounds of EIPs. Our point
estimates imply less short-term spending on average than in response to economic stimulus
payments in 2001 or 2008. While our analysis lacks the power to measure longer-term spending
effects, the lack of short-term spending contributed to strong household balance sheets as the
direct economic effects of the pandemic on households waned.
Jonathan A. Parker Laura Erhard
MIT Sloan School of Management US Bureau of Labor Statistics
100 Main Street, E62-642 Division of Consumer Expenditure Surveys
Cambridge, MA 02142-1347 2 Massachusetts Avenue NE
and NBER Washington, DC 20212-0001
JAParker@MIT.edu Erhard.Laura@bls.gov
Jake Schild David Johnson
Bureau of Labor Statistics National Academies of Sciences,
Division of Price Index and Engineering and Medicine
Number Research 500 Fifth Street, NW
2 Massachusetts Avenue NE Washington, DC 20001
Washington, DC 20212-0001 djohnson@nas.edu
Schild.Jake@bls.gov
A data appendix is available at http://www.nber.org/data-appendix/w30596
In response to the economic consequences of the pandemic, the United States govern-
ment distributed three waves of Economic Impact Payments to American households. In
March of 2020, following the declaration of a national emergency, Congress passed the
Coronavirus Aid, Relief, and Economic Security (CARES) Act. The Act authorized more
than $2 trillion of spending on programs that included the disbursement of $300 billion
in Economic Impact Payments (EIPs) to the vast majority of Americans. In December
2020 with the pandemic continuing, the Coronavirus Response and Relief Supplemental
Appropriations (CRRSA) Act authorized a second wave of roughly $150 billion in EIPs,
and in March 2021, the American Rescue Plan (ARP) Act authorized a third round of just
under $400 billion in EIPs.
While these payment programs were modelled on stimulus payment programs that
the government had implemented at the beginning of recessions in both 2001 and 2008,
the economic situation in the pandemic was entirely different. The pandemic caused a
large collapse in production as well as demand, as people — partly at the behest of the
government — cut back on both producing and consuming goods and services which
risked exposure to COVID-19. Thus, the EIPs were not intended to stimulate demand for
consumption but rather to provide pandemic insurance, ensuring that people who had
unexpectedly lost their livelihoods could continue to cover their consumption needs and
financial obligations. The EIPs were not targeted to those who had lost their incomes, but
were widely distributed, presumably for reasons of feasibility and expediency, as well as to
get aid to people who were impacted but not eligible for aid through other programs.
In this paper, we study the responses of consumer spending to the arrival of the EIPs
and evaluate the extent to which the EIPs provided widespread, urgently-needed pandemic
insurance. Focusing first on the spending response to the first round of EIPs, we estimate
that the spending of the average household rose only a small amount over the couple of
months following the arrival of their EIP, when compared to households that received later
EIPs or did not receive EIPs at all, suggesting that the typical recipient was not in dire
need of the EIP. We do however find larger spending responses both for those households
with low levels of ex ante liquid wealth and for those more reliant on earnings from jobs
with tasks that could not be done from home. While our data do not measure the arrival
of the second and third rounds of EIPs as well as they do the first round, our estimates
suggests even lower average, short-term spending responses to these final two rounds.
Finally, we find some evidence of spending over the three months following our initial
short-term spending estimates, but lack the statistical power to measure the spending
effects of any round of EIPs over a longer period; we can only conclude that the lack of
1
short-term spending contributed to strong household balance sheets as the economic effects
of the pandemic waned following the three rounds of EIPs.
Our results are based on analysis of the Consumer Expenditure (CE) Interview Survey.
We measure the average response of consumer spending to the receipt of an EIP using
variation across households in receipt, in amount conditional on receipt, and in when they
received a payment. As a baseline, we compare our estimates of spending to those reported
in Johnson et al. (2006) and Parker et al. (2013) for the 2001 and 2008 tax payments using
exactly the methodology employed in these papers. But there are substantial differences not
only between program goals, but also between the structure of these payment programs
and the structure of the EIP programs. The EIPs were disbursed more widely, more
rapidly (and so less drawn out over time), more by direct deposit, and rounds one and
three were larger than the payments in 2001 and 2008. Most importantly, the EIPs were
disbursed without any randomization. Thus, while we compare our estimates to the
spending responses estimated in the earlier literature, our main analysis uses an estimator
that is both more robust to non-random differences in spending responses over time and
better-suited for the variation across households in the EIP programs. In terms of more
robust, our main analysis employs a method that is unbiased in the presence of significant
difference in spending responses over time (for the same round of EIPs), a concern of a
recently literature on treatment effects (see e.g. Borusyak et al., 2022; Orchard et al., 2022).
In terms of better-suited for the variation across households in the EIP programs, each
round of EIPs was distributed mostly during one month and without any random variation
across months. For example, the first round of EIPs had the most variation in timing and
almost half of these EIPs were disbursed by direct deposit during the week of April 10
and almost 90% of 2020 EIPs disbursed within the first five weeks.1 As a result, our main
analysis leans heavily on comparing the spending of similar households that do and do
not receive EIP and that receive EIPs of different amounts relative to their typical spending
amounts. Receipt status is primarily driven by whether the IRS had the information to
disburse the payment and whether the household was ineligible due to too high income or
citizenship status.2 Section III present our method including how we further modify the
canonical method for the extreme volatility in expenditures during the pandemic.
Our first main finding is that the CE data show only small short-term spending increases
1 We do not study the spending responses to EIPs that were received as part of income tax refunds or
implicitly as lower tax payments.
2 For the first round of EIPs for example, 3.8% of eligible households did not receive an EIP in 2020 because
the IRS did not have the necessary information to disburse their EIP, and 16% of tax units were not eligible
for an EIP because their incomes were too high or they did not meet the citizenship requirements (e.g., a
couple with one non-citizen spouse that filed jointly; see Sections I and II).
2
on non-durable goods and services in response the receipt of an EIP. For the first round
of payments in 2020, ninety-five percent confidence intervals imply that people increased
their spending on non-durable goods and services as measured (roughly 44% of total
expenditures measured in the CE) by between 4.6 and 15.8 percent of their EIP during
the three-month CE reference period during which the EIP arrived.3 We find a similar
average propensity to increase consumer spending (MPC) for the second, smaller round
of EIPs, disbursed mainly in January 2021 when the economy was somewhat more open.
For the third round of EIPs in the spring of 2021, our estimates imply almost no spending
response. An important caveat to these second two results is that receipt of these EIPs
appear to be under-reported in the CE survey, and therefore these spending responses
may be underestimated. Nonetheless, all three estimated spending responses on the broad
measure of non-durable goods and services in the CE Survey are small, and suggest that
most EIP dollars were not providing urgently-needed pandemic assistance.
These relatively low spending responses are consistent with the fact that the EIPs were
disbursed far more broadly than the income losses caused by the pandemic, with the
presence of pandemic constraints on spending, and with the large increase in household
account balances during the pandemic. Roughly 145 million EIPs were disbursed by
mid-2020 while employment dropped by 22 million during the pandemic recession.4
Particularly during the first wave of EIPs, many types of consumption were constrained
by the prevalence of the disease and/or by government restrictions which, together with
diminishing marginal utility on unaffected goods and services, could have held back the
overall expenditure response to the payments. Indeed, Guerrieri et al. (2020) make this
assumption to study the macroeconomic consequences of the pandemic, and our results
show some evidence of additional spending on durable goods for the first two EIP rounds,
consistent with the shift in aggregate retail spending from services and towards durable
goods during the pandemic.5
Particularly for the second and third round of EIPs, these low spending responses are
also consistent with households on average already having plenty of liquid funds. As the
constraints on spending relaxed, the pandemic reduction in spending coupled with other
3 This propensity to increase consumer spending within a few weeks of the arrival of the first round of
EIPs are somewhat lower than found in previous studies using aggregated data or information on select
populations, issues we discuss below.
4 Cajner et al. (2020) and Cox et al. (2020) document the large diversity in outcomes in the pandemic
recession.
5 In total, we estimate that about 24% of EIPs were on average spent in the three-month period in which
they arrived on all CE expenditures. The spending responses to the EIPs were on average more tilted to
durable goods than the spending responses to the 2001 tax rebates, but not that dissimilar from those to the
2008 economic stimulus payments.
3
government support (e.g. the paycheck protection program and expanded unemployment
insurance benefits) including earlier EIPs may have raised average liquidity and lowered
the need for households to spend during the second and third waves of EIPs. Finally, the
third round of EIPs was large relative to all other payments and larger transitory increases
in income in theory raise liquidity themselves and lead to smaller shares of the increase
being spent in the short-run.
Did the EIPs cause later spending? We find some evidence of continued higher spend-
ing in the months following the three-month period of receipt, although these are fairly
statistically uncertain. We estimate that the roughly 45 percent (round 1) and 60 percent
(round 2) of people’s EIPs were spent after the concurrent and subsequent three month
period. We measure essentially no spending increase in response to the third-round of
EIPs at any horizon. Our analysis has no power to estimate spending responses at longer
horizons. However, following the disbursement of the EIPs, credit card balances decreased,
liquid account balances increased, and stock prices for “meme”’ retail stocks increased (see
Grieg et al., 2022; Greenwood et al., 2022). Strong household balance sheets typically raise
expenditures, and so may have contributed to higher demand as the pandemic waned.
Figure 1 summarizes these findings by showing that the direct, short-run spending
responses to the EIPs were relatively small. The Figure plots observed real personal con-
sumption expenditures (PCE) and the same series subtracting off the increase in spending
implied by our estimates assuming that the contemporaneous spending response occurs
evenly over the month of receipt and the first following month and that the lagged spend-
ing response occurs evenly over the following three months. The lines without different
EIPs in Figure 1 are thus not true counter-factuals, but are simply PCE without the partial-
equilibrium effect of the EIPs on consumer spending based on this simple accounting
exercise. Figure 1 not only shows the relatively small increase in direct spending implied
by our estimates, but also highlights the extremely strong rebound in consumer demand
for non-durable goods and services to which the EIPs may have contributed with delay
through temporary decreases in debt or increases in saving.6
Our second main finding is that, while the average spending response to the EIPs are
modest, we find significantly higher short-term spending responses for households that are
more exposed to the economic losses from the pandemic, consistent with these households
using the EIPs to fund spending that they could not easily do otherwise. Our first measure
of exposure is low ex ante liquid wealth. For the first round of EIPs, households in the
6 Note that for non-durable PCE, we use MPC estimates from a CE measure that includes some services
and semi-durable spending so likely over-estimates the spending effects of the EIPs.
4
Figure 1: Implied change in Real Personal Consumption Expenditures directly due to disbursement of EIPs
(a) Personal Consumption Expenditures (b) Non-durable Personal Consumption Expenditures
1200 260
1150
Without EIP3
240 Without EIP3
Billions of 2012 dollars Billions of 2012 dollars
1100
1050 Without EIP2 Without EIP2
220
1000 Without EIP1
200
950 Without EIP1
900
180
Jul 19 Jan 20 Jul 20 Jan 21 Jul 21 Jan 22 Jul 22 Jul 19 Jan 20 Jul 20 Jan 21 Jul 21 Jan 22 Jul 22
Personal consumption expenditure Non-durable personal consumption expenditure
Trend, 2012-2019 Trend, 2012-2019
Notes: Monthly personal consumption expenditures in billions of 2012 dollars (August 17, 2022). The
trend line is the average monthly growth rate of real PCE from January 2012 to December 2019 applied
to the real value of PCE from July 2019. Without EIP series are constructed by subtracting from PCE the
spending implied by the MPC estimates from Table C.11 and the monthly EIP payments from the EIP
Dashboard, Bureau of the Fiscal Service as of December 15, 2021. We assume that the contemporaneous
spending occurs evenly in the month of receipt and the subsequent month, and that lagged spending
occurs evenly over the following three months. We assume negative estimated spending is actually zero.
bottom third of the distribution of liquid wealth – those with less than $2,000 available ex
ante – spent at roughly two and a half times the rate of those in the middle third, while
those in the top third of the distribution of liquid wealth (above $12,500) had roughly no
spending response. Later differences in liquidity across households are less important for
the second two rounds.7 Our second measure is based on whether a households earns a
significant share of its income from work that is unlikely to be able to be done from home
or remotely. Households with lower ability to work from home spent more out of their
first-round EIPs when they arrived. We find no such evidence for later rounds of EIPs.
In sum, while on average the EIPs appear to have gone to many households with
incomes that were unharmed by the pandemic, some of the EIPs, mainly in the first round,
did support short-term spending for some households, primarily those with low ex ante
liquid wealth and those reliant on income that could not be earned by working from home.
In terms of future policy, both this paper and the research on consumption responses to tax
payments more generally suggest that greater targeting of households with little liquid
7 For the second round, we find essentially no spending response in the top third of the distribution of
liquid wealth, but similar spending responses between the bottom two thirds. Finally, the only evidence for
spending in response to receipt of the third round of EIPs is for the middle third of the distribution of liquid
wealth.
5
wealth and low debt capacity would be more efficient in the sense of generating more rapid
increases in demand for purposes of stimulus programs or getting more of the payment
money to those households most vulnerable to income losses for pandemic insurance.8
However, there are also potential moral hazard costs of targeting economic need or low
liquidity more directly. One approach to minimizing these costs would be to base payments
on household characteristics that are less responsive, for example not sending pandemic
insurance payments to people who were not previously employed and therefore not at risk
of losing their jobs (e.g. people who were retired in 2019 did not lose their jobs in 2020 and
on average had increases in wealth). Alternatively, either stimulus or pandemic insurance
could be delivered through increasing temporarily the generosity or eligibility of existing
government programs that are based on direct targeting, such as unemployment assistance,
Temporary Assistance for Needy Families, etc. where the disincentives of these programs
are better understood and potentially better minimized (e.g. see Ganong et al., 2022).9
Most studies of the spending response to previous tax payments have estimated the
response to payments using variation in spending between recipients and non-recipients
(e.g. Bodkin, 1959; Agarwal and Qian, 2014; Kueng, 2018), over time (e.g. Souleles, 1999;
Parker, 1999; Stephens, 2003; Farrell et al., 2019; Baugh et al., 2020), and using randomization
in policy in either dimension (Agarwal et al., 2007; Broda and Parker, 2014; Parker, 2017;
Lewis et al., 2021, in addition to those already cited).10 The disbursement of the EIPs
was not randomized in any way across households or time. Because of this, the present
study as well as existing studies of the spending response to the EIPs focus on comparing
spending before receipt to spending after receipt, comparing spending between recipients
to non-recipients, and comparing households receiving different sized EIPs.11
The first rapid analysis of the spending changes caused by the EIPs, Meyer and Zhou
(2020), used Bank of America transactions data and reports large increases in aggregated
card spending on the day of and the day following receipt of an EIP associated with bank
accounts that received EIPs on April 15 (when over 40% of EIPs were disbursed) relative to
8 Past payments sent out either as pandemic insurance or stimulus programs have increasingly targeted
these populations to some extent by excluding households with high previous-year incomes.
9 For pandemic insurance, Romer and Romer (2022) also suggest a role for policy in providing hazard pay.
For the purposes of economic stimulus, it is also worth noting that government spending generates immediate
spending by definition, and so in this sense is equivalent to an MPC of 100% out of a payment program. That
is, rapid government spending raises aggregate demand by more than equivalent-cost payment programs,
although obviously the goods and services purchased will differ, as will the distributional effects of the
policies.
10 Most closely related, Fagereng et al. (2021) measures the spending response of (random) lottery winners.
11 Kubota et al. (2020), Feldman and Heffetz (2020), and Kim et al. (2020) measure the spending responses
to tax payments disbursed in response to the pandemic in Japan, Israel, and South Korea respectively.
6
those that did not. Daily spending increased by an average of 50% year over year between
April 15 and 16 for households with incomes below $50,000 and by only 3% for households
with incomes above $125,000. Also using aggregated data, Chetty et al. (2021) finds that
over this same couple of days, card spending in zip codes in the bottom quarter of the
distribution of average household income rose by 25% while those in the top quarter of the
distribution rose by only 8%. Finally, also using zip code level data and using incidental
differences in timing in EIP disbursements across zip codes, Misra et al. (2021) infers an
MPC of 50% in the few days after an EIP arrives.
Our evidence shows lower spending responses than measured in existing studies,
all of which use account-level data on financial transactions to measure the spending.
Karger and Rajan (2021), Baker et al. (forthcoming), and Cooper and Olivei (2021)) find
that people’s out-of-account spending rises cumulatively by 46%, 25-35%, and 66% of
their first-round EIPs, respectively, within a few weeks of receipt.12 One likely reason for
these larger spending responses than found in the CE Survey data is that these account-
level studies cover populations that are likely to have larger spending responses than
average.13 There are other possible reason also, such as the different ways in which the
studies measure consumer expenditures. Account-level data on transactions may mis-
characterize debt payments or saving as consumption (e.g. paying debt on un-linked
credit cards, payments of overdue bills from past consumption, or transfers to investment
accounts).14 Alternatively, respondents in the CE Survey could forget to report EIP-induced
purchases. Finally, the differences could arise in part from statistical issues, both the
statistical uncertainty inherent in any estimator and the statistical methods that we use.15
12 Karger and Rajan (2021) also estimate a 39% MPC for the second round of EIPs.
13 The accounts used in Karger and Rajan (2021) are skewed towards lower income households (average
annual income of $20,880), the households Baker et al. (forthcoming) are those that have opted to use a
financial app designed to help them save (and have average incomes of $36,000), and Cooper and Olivei
(2021) uses Facteus data covering lower-income households many of whom are un-banked.
14 Baker et al. (forthcoming) include car loans and mortgage payments as consumption-related, whereas
this paper includes interest payments on mortgage loans as part of consumption-related spending, but not
the principal.
15 The CE is a small dataset, with a similar number of recipients to that in Baker et al. (forthcoming),
and standard errors are a substantial share of the differences among the estimates across the papers. The
randomness of the estimator may also explain the difference between our estimated spending propensities
and those estimated in the CE during previous tax rebate episodes.
7
I The Economic Impact Payments
In response to the economic fallout from the pandemic, the Federal government passed
three pieces of legislation each of which authorized the disbursement of a round of what
came to be called Economic Impact Payments (EIPs): the CARES Act of March 2020, the
Coronavirus Response and Relief Supplemental Appropriations (CRRSA) Act of 2021,
and the American Rescue Plan (ARP) Act of March 2021.16 We organize our description
of the EIP programs around the three ways in which EIPs differed across households:
differences in dollar amount conditional on receipt, differences in the time of receipt of the
EIP, and whether a household did or did not receive an EIP at all. Unlike when payments
were disbursed in 2001 and 2008, none of these three sources of variation are completely
unrelated to household characteristics.
In terms of amount, the first round of EIPs (which we call EIP1s) consisted of a base
payment of $1,200 for an individual, $2,400 for a couple filing jointly, and additional
payments of $500 for each qualifying dependent under age 17. The CARES Act set upper
income thresholds for receiving the full payment of $75,000 for an individual, $112,500 for
a head of household, and $150,000 for couples filing jointly, where income was based on
2019 adjusted gross income (AGI) if the taxpayer had already filed their 2019 tax return in
2020, otherwise income was based on 2018 AGI as reported in 2019 tax filings.17 For every
$100 of adjusted gross income over the threshold the stimulus payment was reduced by
$5.18
Second round EIPs, EIP2s, were smaller, consisting of a base payment of $600 for an
individual or $1,200 for a couple filing jointly, and additional payments of $600 for each
qualifying dependent under age 17. The upper income thresholds and phase-out rate for
this second round of EIPs were the same as for the first round.19
The third round of EIPs, EIP3s, were substantially larger than EIP1s or EIP2s. They
consisted of a base payment of $1,400 for an individual, $2,800 for a couple filing jointly, and
16 The Coronavirus Response and Relief Supplemental Appropriations Act was include as a part of the
Consolidate Appropriations Act of 2021, which was signed into law on December 27, 2020.
17 In December 2020, the phase-out threshold for a qualifying widow(er) increased from $75,000 to $150,000,
according to the IRS. This change does not affect our analysis.
18 In an article released by The Hill (Bolton, 2020), Republican senators are referenced saying they want to
model the recovery rebate on the stimulus checks former President George W. Bush sent out during the 2008
financial crisis. The 2008 rebate had income thresholds of $75,000 for individuals and $150,000 for couples
filing jointly, and were phased out at a rate of $5 for every $100 of income over the threshold.
19 For the second round of EIP, income is defined as the tax filer’s 2019 AGI reported on their 2020 tax
filings. If a tax return had not been filed by the time the payments were distributed, the tax filer did not
receive an advanced payment and had to claim the Recovery Rebate when filing their 2020 tax return in 2021.
8
Figure 2: Economic Impact Payment Amounts as a Function of AGI and Family Structure
(a) CARES Act (b) CRRSA Act (c) ARP Act
additional payments of $1,400 for each qualifying dependent. They were also distributed
slightly more broadly along several small dimensions, including that the definition of
“qualifying dependent” was expanded to include dependents over the age of 17. The upper
income thresholds were the same as the first and second rounds; however, the phase-out
rule was more aggressive so that the larger amounts did not lead to EIPs being received
higher up the income distribution. Specifically, rather that a constant phase out rate, income
thresholds were set such that tax filers with 2020 AGI above $80,000 for an individual,
$120,000 for a head of household, and $160,000 for a couple filing jointly, regardless of the
number of qualifying dependents, did not receive an EIP3.20 For example, an individual
with no dependents, base payment of $1,400, had a phase out rate of $28 for every $100 of
AGI over $75,000, whereas an individual with one qualifying dependent, base payment
of $2,800, had a phase out rate of $56 for every $100 of AGI over $75,000. Figures 2a, 2b,
and 2c display the EIP amounts as a function of income for various family structures for
the first, second, and third round of EIPs, respectively.
In addition to households receiving different amounts of EIPs, households also received
them at different times. In each round, most taxpayers who had included their bank
information when filing a recent tax return (e.g., for a refund) received their EIP during the
first week of disbursement. For EIP1, bank information came from a 2018 or 2019 tax return,
and for EIP2 and EIP3, bank information came from a 2019 or 2020 tax return. The IRS also
20 If a 2020 tax return had not yet been filed, then 2019 AGI from the 2019 tax return filed in 2020 was used
instead.
9
launched a web page where households could enter their information for the IRS if they
either had omitted bank information from their returns or were eligible but had not filed
2018 or 2019 returns.21 For EIP1, this constituted roughly 35 million households. The IRS
also collected information on eligible households from the Social Security Administration
and the Veterans Administration (and the Railroad Retirement Board).
The IRS began depositing EIP1s into bank accounts on April 10, 2020, and using the
information that the IRS was able to gather and process in time, roughly 105 million or
about 63% of all EIPs were disbursed in April 2020. For eligible households without the
necessary bank information, the EIPs arrived starting two weeks after April 10 by mailing
a paper check or pre-paid “EIP” card. The disbursement of checks occurred with a greater
delay. By the end of April only about 7 million checks (4% of EIPs) were sent out. Most
of the checks were sent out in May, about 27 million or 16% of EIPs, and all of the “EIP”
cards were sent out in May, about 4 million or 2% of EIPs. About 95% of all first round
EIPs were delivered in the first two months of disbursement. The remainder of the EIPs
continued to trickle over the rest of 2020. Figure 3a shows the minimal variation in timing
of the distribution of CARES Act EIPs.
In contrast, the disbursement of the second round of EIPs has almost no variation across
months. For EIP2, almost all of the second round EIPs were distributed in January 2021
(see Bureau of the Fiscal Service (2022)). Daily Treasury Statements show some EIP2s also
being disbursed in February, which is due to reissuing payments that were initially unable
to be delivered.
The disbursement of the third round of EIPs was slightly more drawn out over time
than that of the EIP2s, but still more concentrated over time than the first round of EIPs.
A full 74% of all EIP3s were distributed in March 2021 (62% by direct deposit; 8.5% by
check; and 2.7% by EIP cards). By the end of April about 92% of all third round EIPs had
been distributed, with the remaining 8% distributed over the remainder of 2021. Although
the IRS distributed a smaller percentage of EIP3 in the first two months of disbursement
compared to EIP1, about 5 million more EIPs were distributed during March and April of
2021 than compared to April and May of 2020. Additionally, about 20 million (7%) more
EIPs were distributed by direct deposit. Figure 3 displays the variation in the timing of
disbursement of EIP1s and EIP3s.
Finally, there is a set of households that either did not receive EIPs at all or who received
their EIPs after filing their taxes as part of their income tax refunds or implicitly as reduced
21 IRS web page “Get My Payment;” https://www.irs.gov/coronavirus/get-my-payment (downloaded
October 2021)
10
Figure 3: The disbursement of EIP payments over time and by mode of distribution
(a) EIP1: CARES Act, 2020 (b) EIP3: ARP Act, 2021
Source: Data acquired from the Bureau of the Fiscal Service (2022). Months are the disbursement
months.
income tax payments. There are three main reasons why a household did not receive an
EIP during each primary disbursement period. First, an individual was ineligible for an
EIP if they did not have a Social Security Number (SSN) valid for employment. The CARES
Act was worded such that families were ineligible if they had filed jointly and one of the
spouses was not a US citizen, a situation affecting an estimated 14.4 million people (Gelatt
et al., January 15 2021). The CRRSA Act changed this requirement. A married couple
filing a joint return became eligible for a partial Recovery Rebate credit when only one
spouse has a SSN. This change resulted in 2.9 million people becoming eligible.22 The ARP
Act further expanded the eligibility criteria to anyone with a SSN, which resulted in an
additional 2.2 million eligible individuals.23
Second, eligible households did not receive an EIP disbursement if they had changed
accounts and/or addresses during the relevant previous year, if they had not given their
information to the IRS, or if the IRS did not otherwise have their information (e.g., from
22 Of these 2.9 million people, 1.4 million were US citizens or legal immigrants and spouses of an unau-
thorized immigrant, and 1.5 million were children with one unauthorized immigrant parent. The change in
eligibility criteria was applied retroactively, which means not only did these individuals now qualify for the
second EIP, but they were also able to claim the first EIP through the Recovery Rebate tax credit on their 2020
tax filing.
23 These 2.2 million individuals are children whose parents (or parent) are unauthorized immigrants. Since
no parent had a SSN, they were ineligible for the first and second EIP, which means their children were also
ineligible.
11
the Social Security Administration). For example, four months after the CARES Act (by
the end of July), 10 percent of EIPs had not been disbursed, and 5 percent or nine million
eligible households had not received an EIP by the end of September (Murphy, 2021). For
EIP2 or EIP3, people who re-located even temporarily during the pandemic and formally
changed their addresses or banks accounts became ineligible for EIP disbursement.
Finally, the third reason that households would not receive an EIP is that EIP amounts
declined to zero as income increases. As shown in Figure 2 high-income households were
not eligible and a significant number of higher-income households that received EIPs in
the first two rounds were not eligible for an EIP3.
Taxpayers that fell into either of the first two categories and so were ineligible for a
disbursed EIP but were eligible for an EIP, could receive their EIPs as tax credits when they
filed their 2020 taxes in 2021 for EIP1 and EIP2, and when they filed their 2021 taxes in
2022 for EIP3. More generally, taxpayers were also eligible to receive a tax credit for any
amount by which the EIP they were due based on their final tax information exceeded the
amount they had been disbursed. These “true ups” amounted to roughly $44 billion in tax
years 2020 and $18 billion in tax year 2021 (Splinter, 2022). There was no corresponding
payment required however if a disbursed EIP exceeded the amount that should have been
disbursed based on the later tax information.24
In aggregate $271 billion was disbursed during the first EIP round, $141 billion during
the second EIP round, and $390 billion during the third EIP round (Internal Revenue
Service, June 15 2022). Alone, any one of these rounds is much larger than the previous
2008 program which disbursed $120 billion in 2020 dollars, which in turn was close to
double the total of the 2001 rebate program. Combined, the three rounds of EIP disbursed
more than six times the amount disbursed with the 2008 program. About $260 billion
worth of EIPs were disbursed in the second quarter of 2020, which corresponds to about
5.3 percent of GDP or 8.0 percent of PCE in that quarter (Figure 3 and Internal Revenue
Service, May 22 2020). The first quarter of 2021 saw $473 billion of EIPs disbursed, from
both the second and third waves. This represents 2.1 percent of GDP and 3.2 percent of
PCE. The third EIP wave additionally disbursed $67 billion in the second quarter of 2021,
corresponding to 0.29% of GDP and 0.42% of PCE. The next section describes the EIPs as
recorded in our CE dataset.
24 These safe harbor amounts were roughly $21 billion in tax years 2020 and $43 billion in tax year 2021
(Splinter, 2022).
12
II The Consumer Expenditure Survey
Data for this study are from the Consumer Expenditure (CE) Interview Survey, a
household survey run by the Bureau of Labor Statistics. The CE data set contains spending,
demographics, and other financial information on households living in the U.S. The Bureau
of Labor Statistics (BLS) structures the CE so that a consumer unit (CU) at a given address,
which we will refer to a household, is interviewed up to four times at three month intervals
about their spending over the previous three months (”reference period”). New CUs are
added to the survey every month, and while a significant dollar share of spending data is
reported at the monthly level, a little over half of spending is only reported for the entire
three-month reference period. Thus, we use the data at the (overlapping) three-month
frequency.25 Appendix A.2 contains more details about CE files and variables we use in
this study.
Following the passage of the CARES Act, the BLS added a module of questions about
the EIPs to the CE survey starting with the June 2020 interviews and continuing until the
October 2021 interviews, with the exception the questions were not fielded in January
2021.26 These questions were worded similarly to questions that the BLS added to the CE
about stimulus payments in 2008. The questions measure the date of receipt, the number
of EIPs received, the amount received, which member or members of the household the
payment was for, and the mode of receipt (by check, direct deposit, or debit card).27 The
questions were phrased to be consistent with the style of other CE questions and the
questions on previous CE surveys about the 2001 and 2008 tax rebates. Although the
wording did not follow exactly previous CE surveys, the module of questions also asked
whether the EIP was used mostly to add to savings, mostly to pay for expenses, or mostly
to pay off debt. Appendix A.1 contains the language of the CE survey instruments.
The fact that the EIP questions were not included in the May 2020 interview ques-
25 ”Overlapping” means for CUs interviewed within two months of each other the three-month reference
period for reporting spending will include some of the same months. For example, a CU who is interviewed
in June has a three-month reference period of March, April, and May, and a CU interviewed in July has a
three-month reference period of April, May, and June. Both reference periods include April and May; thus,
we consider them overlapping.
26 The module was developed by the BLS partly based on the similar questions from 2008 and in consultation
with others in the federal statistical system, particularly those working with the Household Pulse Survey (in
which EIP questions had already been asked), and outside researchers, two of whom are co-authors of this
paper.
27 Starting with the July interview the mode of receipt question was expanded to include via tax rebate.
Any instances of receipt via tax rebate were dropped, which resulted in 5 relevant rebates being excluded.
Prior to the July interview, CUs who received an EIP via their tax rebate were asked to not include it when
reporting EIP receipt.
13
tionnaire means that, even for EIP1 where the distribution of EIPs was somewhat drawn
out over time, we have very little power to identify the impact of the arrival of EIP1s on
spending using only variation in the timing of receipt across households. The vast majority
of EIP1s were disbursed in April and May. And while April and May are in different
three-month expenditure recall periods for households on the May interview cycle, they
are not for households on the June or July interview cycle. Thus, we cannot compare how
spending differs between April and May depending on whether an EIP1 is received in
April or May. Since EIP2 and EIP3 have very little variation in the timing of receipt, and
since only about 10% of EIP1s arrive after May 2020, we focus primarily on analysis that
leans heavily on other sources of variation, like amount and recipient status.28
A second implication of the lack of EIP questions on the May 2020 survey is that we
have no way to tell whether households interviewed in May received the EIP1 or not during
the previous three months. The reference period for the May 2020 interview includes April
when over half of all EIPs were disbursed. Thus, we drop all households on this interview
cycle because we cannot compare the spending of those receiving different EIPs at different
times (or not at all) since we do not have the EIP information. More precisely, we restrict
our sample to households that had an interview during June or July of 2020 when the
EIP questions were asked and the three-month recall periods include April and May 2020.
This restriction drops roughly one third of households – those in the interview cycle that
includes May 2020, as well as any other households that are missing interviews in June or
July 2020 interviews. To be clear, we use all available interviews for the households that
have interviews in June or July 2020 (provided the observation has the other necessary data
and a consecutive interview also with valid data). However, the loss of the observations on
the May interview cycle reduces statistical power.
We face a similar, but less significant challenge for households interviewed in January
2021. In this case, we assume no EIPs were received in the references period (October,
November, and December) for households interviewed in January 2021 (when the EIP
questions were not asked).29
We construct two main samples of CE households for each EIP round. For each round,
we limit the sample to households with interviews during the main period of disbursement:
June and July 2020 for EIP1, February, March, or April 2021 for EIP2, and April, May, or
28 We investigated measuring the spending response to the EIP1 using the data at the monthly frequency
and only the CE expenditure categories that are collected by month, but found weak statistical power
(consistent with the conclusions of prior work with the CE).
29 Less than 2% of EIP1s were distributed over October, November, and December. EIP2s began being
distributed by direct deposit during the last few days of December, but did not clear until January 4th, the
official payment date according to the IRS. Checks for EIP2 did not begin being distributed until January.
14
June for EIP3. For each, we construct first a broad sample we refer to as all households that
makes minimal additional drops and follows exactly earlier analyses of tax rebates in the
CE. Details are provided in Appendix A.5.3. Second, motivated both by the unprecedented
nature of the pandemic and programmatic differences between the EIPs and previous tax
rebates, we construct our final sample by adjusting the way in which older households and
households with very low levels of reported expenditures are dropped and dropping high
income households who are mostly ineligible for EIPs (see details in Appendix A.5.3 and
Table C.5 to Table C.7). We discuss these choices in detail in the next section.
Tables 1 show that the monthly distribution of EIPs reported in the CE line up reasonably
well with other official statistics. The first two columns of Table 1 show statistics for our
final sample (which drops high-income households as described subsequently); the second
two columns show statistics for the CE data including all (available) interview months. For
EIP1, April data for the raw CE sample is adjusted up by fifty percent to account for our
dropping one third of recipients, those interviewed in May when the EIP questions were not
asked. The CE data have slightly fewer EIP1s reported during the peak month of April and
more in the following months than the US Treasury reports. This difference is consistent
with some time delay between disbursement and receipt for mailed payments and with
some households taking time to notice EIPs deposited into their accounts (and with the
possibility that some households report a later date of receipt than actually occurred).30
For later rounds of EIP, the monthly distribution lines up well with what we know from
other sources also.
Columns 3 and 4 of Panel B in Table 1 show that 24% of households do not receive an
EIP1 according to the CE data compared to 20% in reality (3.2% of households were eligible
tax units who were non-recipients in 2020, and 16% of households were not eligible for
EIPs). In our final CE dataset, about 17% of households do not receive an EIP1 because we
drop households with high incomes (as noted on page 21). As shown in panels D and F,
these numbers are larger for EIP2 and EIP3, and while EIP3 was phased out more rapidly
with income, so that fewer households received the payments, these numbers suggest that
the CE data is missing some EIPs.
In terms of dollar amounts, the average value of EIP1s received in a reference period,
conditional on a positive value, is $2,098, slightly higher than the average individual EIP of
$1,676 reported by the IRS (Internal Revenue Service, June 15 2022).31 The average EIP2
30 In the final sample, about 10% of households that get EIPs report multiple EIPs. About 50% of these
report EIPs in more than one month of which about 60% report receiving EIPs in different reference periods.
31 When using all CE data, and without aggregating to the three-month reference period level, the average
15
Table 1: Percent of EIPs by month and percent of households not receiving EIPs
Census Bureau’s
Unweighted Weighted Unweighted Weighted Household Pulse
CE final CE final CE CE Survey and
sample sample (adjusted) (adjusted) U.S. Treasury
Panel A: The distribution of EIP1s across months, in percent
April 2020 53.8 54.6 53.1 54.1 66.4
May 2020 36.3 35.4 35.3 34.3 25.7
June 2020 7.5 7.7 8.9 9.0 1.1
Jul to Nov 2020 2.4 2.3 2.7 2.6 6.8
Panel B: Percent of households or tax units not receiving an EIP1
Total (households) 17.0 17.0 24.7 24.6
Ineligible (tax units) 16.2
Eligible (tax units) 3.2
Panel C: The distribution of EIP2s across months, in percent
December 2020 24.3 24.2 19.6 19.4 0
January 2021 68.6 68.5 64.2 63.7 100
February 2021 7.1 7.3 16.2 16.9 0
Panel D: Percent of households or tax units not receiving an EIP2
Total (households) 50.8 51.9 52.2 53.0
Panel E: The distribution of EIP3s across months, in percent
March 2021 68.2 68.8 65.8 66.2 73.8
April 2021 23.7 23.3 25.9 25.8 18.8
May 2021 3.5 3.2 3.6 3.4 2.3
Jun to Dec 2021 4.6 4.7 4.6 4.6 5.2
Panel F: Percent of households or tax units not receiving an EIP3
Total (households) 29.4 29.0 40.5 40.3
Notes: Weighted data using the average of FINLWT21 across all interviews. All samples use available CE
data, so interviews through and including September 2021. See Appendix A.5.3 for CE sample construc-
tion and adjustments for months in which EIP questions were not asked. ‘Unweighted CE’ includes all
households with interviews in these months. In Panels A, C, and E, months are recipient months in the first
four columns but are disbursement months in the last column. In the final column of Panels B ineligible
households is as self-reported in the Census Pulse Survey from Garner et al. (2020) and eligible households
not receiving payments are counted through October 2020 as reported in Murphy (2021). For Panels C and
E, the disbursement data comes from the Bureau of the Fiscal Service, Department of the U.S. Treasury.
16
amount is $1,301, and the average EIP3 amount is more than double this amount, $2,814.
Appendix tables C.1, C.2, and C.2, shows the distribution of total EIP amounts received
across household-reference-periods in our CE final sample (unweighted, unadjusted) and
shows households (correctly) report most amounts at the standard EIP amounts disbursed
in each round. For example, consistent with the payments specified by CARES, most
reported EIP1s are at the base amounts or in multiples of $500 above them: about 55% of
households report payments of $1,200 (the basic payment for a single filer) or $2,400 (a
couple filing separately or getting the basic payment as joint filers or a single filer with two
children).
According to the IRS, there were 162 million first-round EIPs disbursed in 2020 totaling
$271 billion, 147 million second-round EIPs totaling $141 billion (as of early Februray 2021),
and 167 million third-round EIPs disbursed in 2021 totaling $390 billion (Internal Revenue
Service, June 15 2022). In the weighted CE data, and scaling up for the interviews missing
for first-round EIPs, we find 138 million first-round EIPs totaling $261 billion, 79 million
second-round EIPs totaling $106 billion, and 111 million third round EIPs totaling $254
billion.32 Households that receive EIP1 and EIP2 by direct deposit on average have slightly
higher expenditures, are slightly younger, have higher incomes, lower liquidity, and have
larger EIPs, than households that receive the payments by mail, but for EIP3, households
that receive the payments by direct deposit are slightly older, and have lower incomes.
The fractions of EIPs reported by households as received by direct deposit, by paper
check, and by debit card match very closely the fractions reported by the Treasury as
disbursed by these methods. Panel A of Table 2 shows that 75% of EIP1s in the CE were
reported as being received by direct deposit, 23% by paper check, and 2% by debit card.
The Treasury reports that 76% of EIP1s were disbursed by electronic deposit, 22% by paper
check, and 2% by debit card during 2020.33 Though there were no explicit instructions,
CE respondents likely reported EIPs that were deposited onto federal benefit cards (Direct
Express Cards) as received by debit card, and while directly comparable numbers from the
Treasury are not available, through June 2020, 3% of EIP1s had been distributed by debit
card and an additional 1% by deposit onto benefit cards (Murphy, 2021). Consistent with
the increase in direct deposit across waves, the CE shows the share of households receiving
their EIP by direct deposit increasing in each subsequent wave.
(unweighted, unadjusted) EIP is $1,837.
32 The lower number in the CE for first-round EIPs is in small part a result of not including information
from CE interviews after December 2020, and similar for third-round EIPs, since interviews after September
2021 is not yet published.
33 https://www.irs.gov/statistics/soi-tax-stats-coronavirus-aid-relief-and-economic-security-act-cares-
act-statisticsEIP1 (Downloaded Oct 28, 2021).
17
Table 2: The share of EIPs by method of disbursement and reported main use
EIP1 EIP2 EIP3
Panel A: Distribution of payment methods, in percent
By direct deposit 74.5 77.7 84.6
By check 23.4 15.8 11.7
By debit card 2.1 6.5 3.7
Panel B: Distribution of reported main use, in percent
Mostly for expenses 56.4 54.5 51.9
Mostly paid off debts 17.8 19.8 19.1
Mostly added to savings 25.9 25.7 29.0
Notes: Statistics based on ‘CE final sample’ include only CE households with certain interviews (June or July
2020 for EIPI, February, March, or April 2021 for EIPII, and April, May, or June 2021 for EIPIII), with income
that does not exceed a certain threshold determined by marital status and family structure, and cleaning
described in Appendix A.5.3. Weights applied are the average of CU weights across all interviews for that CU.
The BLS also asked households to report on the CE Survey whether they spent or
saved their EIPs (the reported preference methodology of Shapiro and Slemrod, 1995). The
responses suggest greater spending than our analysis of expenditures does. Panel B of
Table 2 shows that 56% of households report using their EIP1s mostly for expenses, and
this fraction declines slightly across EIP waves. There is also a significant increase in the
share of households reporting mostly saving their EIPs in round three relative to earlier
EIPs. In 2008, the BLS added different questions to the CE survey that were more similar
to those in Shapiro and Slemrod (1995, 2009) and found that 32% of households would
“mostly spend” their tax payments and 51% would “mostly pay down debt.”
More comparable over time, Sahm et al. (2010) and Sahm et al. (2020) ask the same
questions in both 2008 and 2020 (not in the CE Survey) and the changes in answers suggest
only very slightly lower spending responses in 2020 than in 2008. In response to the EIPs,
18% of respondents report that their EIPs will cause them to “mostly increase spending,”
only one percent lower than in 2008, which suggests little difference in rate of spending
between the EIPs and earlier stimulus payments.34
Following previous research on spending responses using the CE, we construct four
measures of consumer expenditures at the three-month frequency: 1) food, which includes
34 Schild and Garner (2020), Garner et al. (2020), and Boutros (2020) provide in depth analysis of the U.S.
Census Bureau’s Household Pulse Survey (HPS) in which 59% of respondents state that they “will mostly
pay for expenses” with their EIPs. More similar to the Sahm et al. (2020) shares, Coibion et al. (2021) shows
that only 15 percent of households in the Nielsen Consumer Panel report that they mostly spent or expect to
spend their EIPs. Among these households, the average spending rate is 40%. Armantier et al. (2020) reports
a slightly larger number in the New York Fed Survey of Consumer Expectations survey in which households
on average say that they consumed 29% of their EIPs.
18
food consumed away from home, food consumed at home, and purchases of alcoholic
beverages; 2) strictly nondurable expenditures, which includes some services and adds
expenditures such as household operations, gas, and personal care following Lusardi
(1996); 3) non-durable expenditures on goods and services, which adds semi-durable
categories like apparel, reading materials, and health care (only out-of-pocket spending
by the household) following previous research using the CE survey; 4) total expenditures,
which adds durable expenditures such as home furnishings, entertainment equipment,
and auto purchases.35
Relative to the administrative data used in the studies of the EIPs discussed in the
introduction, there are three main advantages of using the CE interview survey as well
as three weaknesses. The first advantage is that the CE contains detailed measures of
consumer expenditures rather than just the transaction counterpart, or, for some trans-
actions like checks or cash, just the amount.36 Second, the CE tracks spending and EIP
receipt by individual consumer units, rather than by accounts (and linked credit or debit
cards). Finally, the CE is a stratified random sample of U.S. households constructed by
the U.S. Census and so when weighted is representative of the U.S. population (along the
dimensions of the census-based strata and conditional on participation in the survey). The
main weakness relative to existing studies are the relatively small sample size, sampling
(e.g., non-response) error, and the presence of measurement error in expenditures and EIP
receipt.
The next section discusses how and why our estimation methodology differs from
previous approaches, as well as presenting the results of applying the previous methodol-
ogy exactly to estimate the average spending response to the EIPs. The following section
presents our baseline estimates of spending rates based on an approach that account for
the differences both between previous tax rebates and the 2020 EIPs, and between previous
recessions and the pandemic recession.
III Estimation method
In this section, we first briefly present the way Johnson et al. (2006) and Parker et al.
(2013) estimate the consumer expenditure responses to the tax rebates disbursed in 2001 and
2008. We then refine this methodology and adopt identifying assumptions that are better
suited to estimating the spending effects of these EIPs given programmatic differences, the
35 The exact definitions are given in Appendix A.3.
36 E.g., terms like Amazon or Starbucks or Sammy White’s. Payments to un-linked credit cards and transfers
to other accounts are also difficult to categorize as spending for consumption, debt payment, or saving.
19
pandemic situation, and the possibility of cross-cohort differences in spending propensities
within each EIP round.
Using samples analogous to our sample of all CE households, the previous papers
estimate an equation analogous to the following equation for household i with consumer
expenditures, Ci,t , observed for (overlapping) three-month period t:
S
∆Ci,t = ∑ β s EIPni,t−s + X i,t γ + τt + ϵi,t (1)
s =0
The key regressor is EIPni,t−s , the total dollar amount of economic impact payments from
round n ∈ {1, 2, 3} received by household i in three-month period t − s.37 The variable τt
is a complete set of time effects for every period in the sample that control for the seasonal
variation in consumer expenditures as well as the average effect of all other concurrent
aggregate factors. The control variables X i,t contain age (agei,t ) and change in family
size (∆FamSizei,t ) which control for the life-cycle pattern of spending and for changes
in consumption needs. Finally, ϵ captures movements in consumer expenditures due to
individual-level factors such as changes in income, expectations, and consumption needs,
as well as measurement and recall error in expenditures.
Provided ϵ is uncorrelated with the other right-hand-side regressors (and for now
maintaining the assumption that β (or its distribution over i) does vary with EIP arrival
date), the key coefficient β s measures the average partial-equilibrium response of household
consumer expenditures to the arrival of the EIP during the three-month period s periods
after the EIP arrives. In our main analysis, in which EIPni,t−s is regressed on ∆C, β s
measures the share of the EIP spent, or the marginal propensity to increase consumer
expenditures (MPC).38 These estimated MPCs are based on three sources of variation:
whether a household receives an EIP or not, variation in the (overlapping) three-month
period in which the EIP is received, and variation in the amount of the EIP.
As we show at the end of Section IV, estimates of the spending responses based on this
exact methodology — while having the advantage of being most comparable to earlier
work — are small, statistically weak, and unstable compared to these earlier analyses. The
first finding may simply reflect reality, but the second two may be indicative of problems
37 In Table C.8 and in additional results in the appendix, we sometimes replace this regressor with
1[ EIPni,t−s > 0], an indicator variable for whether an EIP from round n is received (in the period t − s) at all.
In the appendix, we present some results that use change in log consumption as the dependent variable.
38 When 1[ EIPn
i,t−s > 0] is regressed on ∆C, β s measures the dollars spent. And when 1[ EIPni,t−s > 0] is
regressed on ∆lnC, 100 ∗ β s measures the percent increase in spending.
20
with the methodology, driven by: i) differences between previous tax rebate programs and
this one, ii) differences between previous recessions and the pandemic recession, and iii)
concerns raised recently about consistent estimation if MPCs vary across households such
that the distribution of β s,i changes over time.
Our first main concern, is differences between previous tax rebates and these EIPs.
Relative to the earlier studies, the timing of the disbursement of the EIPs was not random-
ized in any way and was far more limited, both in reality (as described in Section I) and
observed in our data (for the reasons described in Section II). Therefore our estimation
necessarily relies more on differences in spending patterns across households with different
EIP amounts, including those that do not receive EIPs (at least only as part of lower tax
payments or higher refunds in the following year).
Our solution is to make the sample of non-recipients more similar to recipients by
excluding households with high incomes from our analysis. Motivated by the phase-out
of the EIPs described in Section I, for each EIP round, we first posit an income cutoff at
the nearest $25,000 above the income level (rounded to the nearest $25,000) at which a
household would no longer receive an EIP. Different cutoffs apply to households with
different family structures – whether the household contains children and whether it has
one single adult, a married individual or couple, or multiple adults. In addition, note that
recipient status is not a clean function of CE income because EIPs are disbursed based
on adjusted gross income rather than the pre-tax income we observe in the CE, because
reported income has some error, and because the IRS uses calendar year income for either
2018 or 2019 and neither year nor filing status is collected as part of the CE Survey.39
Thus, we adjust each income cutoff up in increments of $25,000 until more than 80% of the
observations with incomes in the $25,000 range just above the cutoff are from non-recipients.
Additionally, we set the cutoff for households with kids to be no lower than the cutoff for
households that are otherwise the same but without kids (i.e., married without kids and
married with kids), if the former has a lower cutoff after increments.40 This process omits
a few recipients. However, more importantly, it leaves some households in our analysis
who are non-recipients due to having too much income but who still have incomes similar
to our recipients and who therefore are potentially a good comparison group for those
households who do receive EIPs. We refer to the three resulting samples — one for each
39 Information on income is collected as part of the CE Survey, but these questions ask about income earned
in the past twelve months, which may not correspond to a calendar year. Additionally, tax filing status is not
asked about in the survey, but imputed values are provided in the data. Imputations of filing status and tax
liabilities are done using the National Bureau of Economic Research’s TAXSIM program.
40 Appendix Tables C.5, C.6, and C.7 show the the selection of resulting cutoffs and the number of recipients
in the $25,000 income ranges above and below each cutoff.
21
EIP round — as our final samples and it is these samples that are tabulated in Section II.
Another differences between previous tax rebates and these EIPs is that there are three
waves of EIPs in reasonably rapid succession, and in equation (1), the estimated spending
responses to one EIPs may be biased by responses to other EIPs. In response, in our main
analysis of the spending responses to EIP2 and EIP3, we include in X as control variables
the same distributed lags of the two other EIPs when observed as we do for the main EIP
of interest. This control is imperfect since we do not observe all earlier EIPs received and
since there is cross-household correlation between recipients status and potentially even
amount for EIP2 and EIP3. Thus we also check (and find similar results) when we estimate
our responses without these controls.
Our second main concern is related to the fact the pandemic was a time of unprece-
dented consumption volatility during which people with different levels of consumer
expenditures had vastly different dollar changes over time. During the early stages of
pandemic in particular, households with higher incomes have much larger changes in
dollar spending on average.41 These differences across households suggest that the time
dummies in equation (1) do a poor job of capturing the average dollar change in spending
for households with different incomes. Since income and average expenditure are also re-
lated to recipient status and EIP amount, these differences may well create bias in estimates
of MPCs. For example, if there are large changes in dollar spending in April 2020, when
most EIP1s were disbursed, that are not caused by EIP receipt or amount conditional on
receipt and yet correlated with receipt or amount, then estimates from equation (1) would
be inconsistent.42
However, groups of people with different incomes — and so with different average
levels of consumption spending — experienced roughly similar percent changes in con-
sumer spending over time (e.g., see Cox et al., 2020). We find for example that, for a given
time period t, differences in ∆ ln C across terciles of the income distribution are lower than
differences in ∆C (see Appendix Figures C.1.b and C.1.c).
Our solution therefore is to scale all the variables in our regression by C̄i , the average
consumer expenditure (of each type) for family i and also to allow a different regression
intercept for households that never receive a given EIP. Letting X̃i,t = Xi,t /C̄i for any
variable X and R(i ) be an indicator variable that equals one for households that receive at
41 Appendix Figures C.1.a and C.1.c show this across terciles of the income distribution.
42 Previous recessions analyzed in earlier work had less variation in average change in dollar spending
by income. And previous analyses found similar MPCs across different specifications, most importantly
between results using log change in consumer spending and those using dollar change.
22
least one EIPn, we infer MPCs from the equation:
S
ei,t = ∑ β s EIPn
∆C ] i,t−s + X
e i,t γ + τt + α R(i) + ϵi,t (2)
s =0
where X contains (scaled) age, change in family size, and possible previous EIPs. The
main coefficient of interest, β s , still measures the propensity to spend out of an EIP, but by
scaling all variables we have transformed the τ from absorbing the average change in dollar
spending across households in that period to absorbing the average percent change in
consumer expenditures across households in that period. Similarly, α R(i) allows a different
average growth rate of expenditure between recipients and non-recipients, and the residual
is in terms of a percent deviation of consumer expenditure rather than dollar deviation. In
the CE survey, the average percent change in spending measured in this way is significantly
more similar for households across terciles of standards of living as measured by their
average level of income (compare Appendix Figures C.2.a to C.2.b and C.2.c to C.2.d).
Our third and final main concern is related to the developing literature addressing
potential bias in difference-in-differences type estimation with both different groups treated
at different times and heterogeneity in average treatment effect across groups (e.g. Borusyak
and Jaravel, 2018; de Chaisemartin and D’Haultfœuille, 2020; Goodman-Bacon, 2021; Sun
and Abraham, 2021; Callaway and Sant’Anna, 2021; Wooldridge, 2021; Borusyak et al.,
2022). In our context, estimation of equation (1) would be biased if there is variation in
average MPC, or β s , across households receiving the EIP in question in different months.
The bias would arise from (implicitly) comparing the expenditure responses of households
receiving EIPs at different times to infer the evolution of expenditure after EIP receipt.
Equation (1) assumes that each household’s expenditure response is given by β s instead of
β s,t .43 To be clear, any variation in the tendency to spend out of EIPs in different waves (1,
2, and 3) would not create any bias.
On the one hand, variation in β s across households receiving the EIP at different times
could be significant because when each household received its EIP is non-random (unlike
in previous payment programs). Later recipients tended to be households for which the
IRS did not have their bank information or physical address, and so have slightly lower
incomes and expenditures on average. In addition, the pandemic period was a period of
unprecedented economic volatility, and variation in β s over time could arise from variation
43 In a dynamic specification where leads and lags are added, there is also the additional problem of
“contamination,” see Sun and Abraham (2021) for details.
23
in the economy or the pandemic situation.44 On the other hand, most of our variation comes
from comparing recipients to non-recipients (always a valid comparison) and comparing
people receiving different amounts of EIPs. Further, Parker et al. (2021) shows through
simulation that there is minimal bias for quite substantial variation in average treatment
effect over time for the first round of EIPs, where the variation in timing of receipt is the
greatest of the three.
Our solution is to follow Borusyak et al. (2022), which allows differences in MPC or β s
over time and is unbiased under generalized parallel trends (and no treatment anticipation)
assumptions.45 The estimation method can be clearly described as a three-step procedure.
Denoting the set of never-treated and not-yet-treated observations as Ω0 , in the first step
we estimate the time dummies and coefficients on controls using only Ω0 :46
∆C
ei,t = X
e i,t γ + τt + α R(i) + ηi,t ∀{i, t} ∈ Ω0 (3)
In the second step, for treated observations only, we compute the difference between
observed scaled change in expenditure and the scaled change in expenditure predicted by
controls and time, denoted by ∆Či,t .
∆Či,t = ∆C
ei,t − X
e i,t γ̂ − τ̂t − α̂ R(i) / Ω0
∀{i, t} ∈ (4)
Thus, ∆Či,t is an estimate of the household-level spending response to the EIPs. In the
third step, we run WLS regression of the new dependent variable on the EIP variable(s) of
interest:
S
∆Či,t = ∑ β s EIP1
] i,t−s + ϵ̌i,t (5)
s =0
Our method solved the issue created by “forbidden comparison,” but note that the third
step deviates from Borusyak et al. (2022) – we rely on regressions to compute average MPC
instead of aggregating individual effects using proposed weights. This change allows us to
exploit the differences in treatment intensity and to compare different specifications. To the
44 Also, the CE interview structure could lead to heterogeneity. Even for households that received the
payment on the same day and had the same spending response in reality, if they were interviewed in different
months and hence had different reference periods, the measured spending response would differ.
45 The estimator is also efficient under homoskedasticity and is “asymptotically conservative” when
standard errors are clustered.
46 As noted, for EIP2 analysis, ]
EIP1i,t−s and ]
EIP3i,t−s are added as controls. Similarly, for EIP3 analysis,
EIP1i,t−s and ]
] EIP2i,t−s are added as controls.
24
best of our knowledge, those features cannot yet be achieved for our specific setting by any
of the new estimators to date. The disadvantage is that the weights used in the regressions
are not as explicit, and could be hard to interpret.47
To better approximate the average response, we also use the average CE weight across
all interviews for each household. In practice whether one weights or not (or whether one
uses replication weights) makes very little difference to the estimates.48
IV The average MPC in response to the arrival of each EIP
This section presents the results of our analysis of the spending responses to all three
rounds of the EIPs using the same survey data source, the CE Survey, as was used in
studying the 2001 and in 2008 tax payments. We show that the estimated, short-term
spending responses out of EIPs are small whether we use the new and improved estimation
method just described or the exact same method as used in the studies of the 2001 and
2008 payments. The estimated spending responses are small both relative to the responses
estimated for the past tax payments and relative to other estimates of spending responses
to these EIPs that are based on other populations and datasets.
Table C.8 displays the main spending responses to all three rounds of EIPs, both the
average fraction of the EIP that is spent shortly after arrival (first four columns) and the
average dollar amount that is spent (last four columns). These results come from our main
estimation method of equation 2 (the three-step, unbiased procedure) with S = 1.
The first four columns of the first row of Panel A show that the first round of EIPs
was not spent rapidly after receipt and so on average was not providing urgently-needed
pandemic insurance. During the three-month reference period in which a payment was
received, a household on average increased its spending on non-durable goods and services
by 10.2% of EIP1, and on all CE-measured goods and services by 23.4% of EIP1. Taking the
perspective of classical statistics, the 95% confidence intervals of cumulative spending rule
out spending in excess of 16% of the EIP on non-durable goods and services and 35% on
47 However, some early evidence shows that after addressing “forbidden comparison”, the weighting issue
is unlikely to lead to significant bias since the estimate will be a convex weighted average, see Baker et al.
(2022) and Roth et al. (2022) for the stacked regression method, for example.
48 We make three other choices that differ slightly from previous analyses. As in previous papers, we drop
the bottom 1% of the distribution in broad non-durable consumer expenditures after adjusting for family size,
but instead of estimating the bottom one percent using a quantile regression on a linear trend, we drop the
bottom 1% in each interview to account for the volatility across time during our sample due to the pandemic.
Second, we do not drop households older than 85. Finally, we choose to follow Panel A of Table 3 in Parker et
al. (2013) rather than Table 2, which means allowing a different average growth rate of expenditure between
recipients and non-recipients. Our estimates are largely insensitive to these three choices.
25
Table 3: The contemporaneous response of consumer expenditures to EIP receipt
Food Strictly Nondurable All CE Food Strictly Nondurable All CE
and Nondurables goods and goods and and Nondurables goods and goods and
alcohol services services alcohol services services
MPC Dollars spent
Panel A. EIP1
EIP1
] 0.011 0.075 0.102 0.234
(0.016) (0.020) (0.028) (0.059)
1[ EIP1
^> 0] 6.5 96.4 80.8 336.5
(25.3) (36.6) (46.4) (96.6)
Panel B. EIP2
EIP2
] 0.034 0.103 0.083 0.247
(0.021) (0.031) (0.039) (0.090)
1[ EIP2
^> 0] 18.8 80.8 65.6 156.7
(23.6) (44.0) (52.2) (114.4)
Panel C. EIP3
EIP3
] 0.036 0.030 0.009 0.015
(0.017) (0.016) (0.018) (0.043)
1[ EIP3
^> 0] 99.5 86.8 55.1 -36.0
(33.8) (40.8) (42.2) (102.4)
Average quarterly household spending across three waves
$2,292 $4,516 $5,996 $14,401 $2,292 $4,516 $5,996 $14,401
Notes: Table reports estimation of equations 3 to 5 with S = 1, with scaled dollar change in consumption
as the dependent variable and using weighted least squares using average weights. Each pair of rows
uses the final sample for that EIP round. Standard errors included in parentheses are adjusted for arbi-
trary within-household correlations and heteroskedasticity. Besides separate intercepts, regressions also
include interview month dummies, scaled age and change in the size of the CU, and controls for the other
EIPs for EIP2 and EIP3. For EIP1, the four columns have 3,541, 3,543, 3,543, and 3,544 treated observa-
tions, and 2,264 never-treated or not-yet-treated observations except for the first column that has 2,261.
For EIP2, the columns have 3,171, 3,171, 3,175, and 3,175 treated observations, and 5,002, 5,004, 5,004, and
5,005 never-treated or not-yet-treated observations. For EIP3, the columns have 3,566, 3,566, 3,568, and
3,567 treated observations, and 3,465, 3,474, 3,477, and 3,474 never-treated or not-yet-treated observations.
all CE goods and service.
First four rows of Panels B show similar low spending responses for the second round
of EIPs. The third and fourth columns show that 8% and 25% of the EIP2s were used for
expenditures on non-durable goods and services and total CE-measured expenditures
respectively within the three-month period of receipt. These first two panels are consistent
with the hypothesis that, because households tilted spending towards durable goods
during the pandemic, the spending response to the EIPs was similarly tilted towards
durable. Compared to past stimulus programs, the share of spending going to durable
26
goods does appear higher than in 2001, but it is not higher than in 2008 and the statistical
strength of both comparisons is weak.
Finally, the first four rows of Panel C show even lower spending responses for the third
round of EIPs than for the first and second round EIPs. Spending in response to EIP3
receipt was economically (and statistically) close to zero. As noted, because it is possible
that some households that received EIP2 or EIP3 payments failed to report them, one
should be maintain some skepticism that the actual spending response were quite this low,
particularly for the third round of the EIPs. However, the lower spending response are
consistent both with the rise in liquid balances throughout the pandemic (see Grieg et al.,
2022) and with the large dollar size of the third round of the EIPs.
How might our estimated spending response to EIP2 and EIP3 be lowered by under-
reporting of EIP receipt in the CE? Under-reporting implies that some households in our
control group were actually treated, and so reduces the difference we measure between
groups. To shed light on this possibility, we calculate EIP receipt and amount from the
rules of each round of EIP and the TAXSIM imputations contained in the CE as described
in Appendix A.6. We create alternative measures of our EIP variable for each round of EIP
by assuming that any EIP arrived in the first two months of that round. We then conduct
our main analysis using these imputed EIPs and dropping any CE household with a recall
period that does not contain both of the critical two months. Appendix Tables ?? to C.10
show the results of our analysis. Using imputed EIPs, the estimated MPCs for EIP2 are
smaller than those in our main analysis suggesting we are not overestimating the spending
response in the second round. For the third round however, analysis of these alternative
measures suggest that our estimated MPCs out of EIP3 are indeed underestimated but this
alternative analysis still finds them to be relatively small.
The last four columns of Table C.8 show the dollar spending response to receipt of an EIP
(rather than the MPC) and imply smaller spending responses. These columns are based on
our main estimation but replacing our measure of EIP amount with an indicator variable
of EIP1 receipt, 1[ EIPni,t−s > 0], so that these estimates do not identify the spending
effect using any information about EIP amounts across recipients. The estimated dollar
spending responses to the arrival of EIP1 are $81 or 3% of the average EIP1 on non-durable
goods and services (statistically insignificant, column 7) and $337 or 16% of the average
EIP1 on all measure CE expenditures (statistically significant, column 8). For EIP2 the
spending responses of $66 and $157 respectively (statistically insignificant), are 5% and
12% of the average EIP2 and so imply even less spending than the specification in the first
four columns. Finally, the last four columns also continue to show very small spending
27
Table 4: The longer-term response of consumer expenditures to EIP receipt
Dependent variable: scaled dollar change in spending on
Panel A: EIP1 Panel B: EIP2 Panel C: EIP3
Strictly Nondurables All CE Strictly Nondurables All CE Strictly Nondurables All CE
non- goods and non- goods and non- goods and
durables services durables services durables services
EIPn
]t 0.075 0.102 0.234 0.103 0.083 0.247 0.030 0.009 0.015
(0.020) (0.028) (0.059) (0.031) (0.039) (0.090) (0.016) (0.018) (0.043)
EIPn
] t −1 -0.011 -0.080 -0.017 0.030 -0.013 0.107 0.000 -0.049 -0.150
(0.020) (0.028) (0.070) (0.038) (0.045) (0.124) (0.010) (0.019) (0.049)
Implied cumulative fraction of EIP spent over two three-month periods
0.139 0.124 0.452 0.235 0.153 0.601 0.059 -0.030 -0.119
(0.051) (0.068) (0.158) (0.086) (0.104) (0.257) (0.036) (0.047) (0.112)
Notes: Table reports β 0 and β 1 from estimation of equations 3 to 5 with S = 1. Regressions also include inter-
view month dummies, a separate intercept for non-recipients, scaled age, and change in the size of the CU.
Panels B and C additionally control for the other EIP waves. The sample is the final sample which includes
only CE households with income that does not exceed a certain threshold determined by marital status and
family structure. Regressions are conducted using weighted least squares, where the weights applied are aver-
age weights. Standard errors included in parentheses are adjusted for arbitrary within-household correlations
and heteroskedasticity. For Panel A, observations are those with an interview in June or July 2020; the columns
have 2,264 never-treated or not-yet-treated observations and 3,543 treated observations. For Panel B, observa-
tions are those with an interview in February, March or April 2021; the columns have 4,815, 4,817, 4,818 never-
treated or not-yet-treated observations and 3,171, 3,175, and 3,175 treated observations, respectively. For Panel
C, observations are those with an interview in April, May or June 2021; the columns have 3,474, 3,477, 3,474
never-treated or not-yet-treated observations, and 3,566, 3,568, and 3,568 treated observations, respectively.
responses to the third round of the EIPs, particularly because the average EIP3 is 1/3 bigger
than the average EIP1.
We have measured EIP-driven spending in the short term to evaluate whether the
EIPs provided urgently-needed pandemic insurance, and we now turn to evaluating
subsequent spending, which is informative both about pandemic insurance but over a
three-month longer period and, for longer horizons, about the contribution of EIPs to the
rapid pandemic recovery and potentially inflation. In terms of pandemic insurance, we
find some evidence of continued higher spending for EIP1 and EIP2, but no evidence of
any continued spending for EIP3. In terms of increases in demand over any longer periods,
we lack the statistical power to add any evidence on the potential contribution of EIPs to
strong demand or inflation during the second half of 2021 or beyond.
Table C.11 shows the longer-run response of spending to the receipt of an EIP. The
coefficient β 1 on EIP
g i,t−1 measures the decline in spending during the three-months follow-
ing receipt, so that β 0 + β 1 measures the increase in spending in the second three months
28
Table 5: Estimated MPCs on CE-measured non-durable goods and some services
Full Sample, Recipients Only, Full Sample
Three-months Three-months Three months of receipt
of receipt of receipt and subsequent three months
2001 Economic Rebates 0.386 0.247 0.691*
(0.135) (0.213) (0.260)
2008 Stimulus Payments 0.121 0.308 0.347
(0.055) (0.112) (0.155)
2020 EIP 1 0.102 -0.062 0.124
(0.028) (0.072) (0.068)
2020 EIP 2 0.083 0.153
(0.039) (0.104)
2021 EIP 3 0.009 -0.030
(0.018) (0.047)
Source: Johnson et al. (2006)), Parker et al. (2013), and Parker et al. (2021) and current paper. The ∗ denotes a
large MPC driven in part by one outlier in spending on food.
relative to the previous three months. The bottom row of the table reports β 0 + ( β 0 + β 1 ),
the sum of the contemporaneous spending and this additional spending, which is then
the total spending during both the three-month period of receipt and the subsequent
three-month period (as a percent of the EIP).
For EIP1, the cumulative MPC on strictly nondurable and broad non-durable goods and
services are both roughly 13% and on all CE goods and services is 45% (with a standard
error of 15.8%). For EIP2, the MPCs are slightly higher, consistent with the more open
economy and the smaller size of the payments. Finally, for EIP3, we find no evidence that
EIP3s were spent during the three months of receipt or during the subsequent three month
period. Appendix Tables ?? to C.13 shows that using our imputed EIP measures described
above does not change the conclusion of small spending effects.
Table 5 summarizes our finding of low spending response to these EIPs, and compares
the spending responses to those of earlier stimulus payment programs. The MPCs out of
the EIPs are substantively lower than MPCs out of tax payments disbursed in 2001 and in
2008 according to studies using the same survey data.
Are these relatively low spending response due to our differences (improvements) in
methodology? No. To show this, we apply the methodology of Johnson et al. (2006) and
Parker et al. (2013) exactly and estimate spending responses to each round of EIPs on
the sample of all CE households. The estimated spending responses are unstable across
29
Table 6: The response of consumer expenditure to EIP arrival estimated on recipients and
non-recipients using the methodology previously applied to tax rebates
Food Strictly Nondurable All CE Food Strictly Nondurable All CE
and Nondurables goods and goods and and Nondurables goods and goods and
alcohol services services alcohol services services
MPC Dollars spent
Panel A. EIP1
EIP1 0.043 0.071 0.077 0.280
(0.032) (0.044) (0.059) (0.217)
1[ EIP1 > 0] 157.3 296.4 375.0 1278.8
(89.9) (130.2) (167.8) (647.5)
Panel B. EIP2
EIP2 0.011 0.037 0.030 0.008
(0.029) (0.044) (0.055) (0.325)
1[ EIP2 > 0] -57.1 -11.1 -10.1 -498.7
(51.7) (79.3) (99.5) (749.8)
Panel C. EIP3
EIP3 0.001 0.001 0.005 0.222
(0.013) (0.017) (0.023) (0.149)
1[ EIP3 > 0] 14.2 -6.3 22.7 702.1
(45.1) (70.3) (91.4) (648.7)
Notes: Table reports β 0 from estimation of equation 1 with S = 0 with dollar change in consumption as the
dependent variable and using weighted least squares using average weights. Standard errors included in
parentheses are adjusted for arbitrary within-household correlations and heteroskedasticity. Regressions also
include interview month dummies, age, and change in the size of the CU. The samples are constructed as in
previous research papers (see Appendix). Panel A has 5,634 observations and includes the sample of all CE
households with an interview in June or July 2020. Panel B has 8,302 observations, includes the sample of all
CE households with an interview in February, March, or April 2021, and additionally includes controls for
EIP1 and EIP3. Panel C has 7,335 observations, includes the sample of all CE households with an interview
in April, May or June 2021, and additionally includes controls for EIP1 and EIP2.
specifications and columns, but on average are not inconsistent with Table C.8 above for
EIP1 and EIP3 (results for EIP2 suggest even smaller spending responses).
More precisely, we estimate equation (1) on samples that are constructed exactly as in
these earlier papers, and replicate Table 2 in both of these papers, for all three rounds of
EIPs. As shown in the first four columns of Table 6, for EIP1, point estimates suggest MPCs
of 4.3% on food, 7.1% on strictly nondurables, 7.7% on the broad measure of non-durable
goods and services, and 28.0% on all goods and services. While all these estimates are
statistically insignificant, these point estimates are consistent with those in Table C.8. But
this methodology leads to wildly different conclusions for other specifications, unlike found
in the analysis of the 2001 and 2008 tax payments, and consistent with the arguments for
our preferred specification in Section III. The last four columns of Panel A show estimates
30
using an indicator variable for receipt in place of EIP1 amount and implies that, in the three
months in which the EIP1 arrives, spending increases by $157 on food, $296 on strictly
nondurables, $375 on non-durables, and $1,279 on all goods and services, with all but
the first being statistically significant. For the average EIP1, these estimates would imply
MPCs of 7%, 14%, 18%, and 61% respectively, roughly double those from estimating the
MPC directly (the average of EIPi,t conditional on receipt is $2,098). Appendix Table C.4
shows the results of estimation for the two other specifications used in previous research
and these estimated spending responses are all statistically insignificant and, again, imply
quite different MPCs than Table 6. 49
V EIPs as Pandemic Insurance
While we find low average spending responses to the EIPs, the EIPs may nonetheless
have filled urgent economic needs for some subset of households, presumably those who
were the most impacted economically by the pandemic. In this section, we construct observ-
able measures of economic vulnerability to the economic consequences of the pandemic
and evaluate whether households that were more exposed spent more of their EIPs to
maintain or increase their consumer spending in the short run. We focus both on house-
holds with little ex ante liquid wealth and on households with labor income exposed to the
pandemic as measured from their ability to work from home. While the average spending
response to the EIPs are low, consistent with payments not being required to fill short-term
spending needs for most households, we find two pieces of evidence that the EIPs did
raise spending and so provided potentially important assistance to some households. First,
we show households that entered the pandemic period with little ex ante liquid wealth
spent a larger share of their EIP1s. For EIP2 and EIP3, there is little to no evidence that
households with low liquid wealth had higher MPCs. Second, we show that households
whose incomes were more exposed to the pandemic — those with lower ability to work
from home – spent more out of their first-round EIPs when they arrived. For the second
round of EIPs we find no such pattern of MPC related to the ability to work from home.
49 Johnson et al. (2006) and Parker et al. (2013) both report estimates of MPCs (in Table 3) that rely only
on variation in time of receipt by dropping all households that never receive stimulus payments. In these
earlier episodes this variation was closer to purely random. Given the lack of variation in timing in the
EIP programs, estimates of the MPC in analogous samples that drop households that never receive EIPs
have very large standard errors. For EIP1, the program with the largest variation in timing of disbursement,
Appendix Table C.1 in Parker et al. (2021) shows that the standard errors are typically 50% to 100% larger
than in Table 6 and C.4, as expected given the lack of variation. Additionally, the estimates are more variable
and many are negative; so, we learn little from this exercise.
31
For the third round, there is some evidence of a small effect.
We estimate different MPCs for different groups of recipients by interacting the EIP
variables in equation 2 with a group-membership indicator variable, denoted g(i ) so that
the spending response of interest varies by group as well as horizon. We use the equation:
S
ei,t = ∑ β g(i),s EIPn
∆C ] i,t−s + X
e i,t γ + α g(i) + τt + ϵi,t (6)
s =0
which also allows the intercept or average growth rate of spending to differ by group
(α g(i) ). For studying the MPC of EIP2 and EIP3, we also interact the controls for other EIPs
(in X)
e with the indicator for group membership. To be clear, consider the MPC for EIP2.
We estimate equation 6 using our imputation estimator and the procedure described in
equations equations 3-5.
First, we split the sample of households by their ex ante liquid wealth and find that
households that entered the pandemic with low liquidity had strong spending responses
to the first round of EIPs in the CARES Act. We measure liquid wealth as the sum of
balances in checking accounts, saving accounts, money market account, and certificates
of deposits at the start of the households first interview (reported in the last interview).50
Panel A of Table 7 shows that households in the bottom third of the distribution of liquidity
– those with less than $2,000 available, which is still a substantial amount – have statistically
significant MPCs of 6%, 22%, and 48% on food, non-durable goods and services, and all
CE goods and services respectively. While the difference between each of these MPCs and
the corresponding MPC of either of the other third of the distribution is not statistically
significant, they are economically large, and we can reject the equality of MPCs across these
three groups for spending on both non-durable goods and services and all CE goods and
services.
Previous research on tax rebates that uses the CE Survey has not consistently found a
statistically-significant decreasing relationship between spending responses and liquidity.
However, analyses with better measures of liquidity have generally found a larger MPC
50 Even the low liquidity group has substantial reported wealth, and in particular the distribution of
reported liquid wealth is much higher in this 2020 data than it was in 2008. In Parker et al. (2013) the 33rd
percentile in the distribution of liquid wealth was only $500. One possibility is changes in the distribution
of respondents, although this appears unlikely as we discuss in Appendix A.4. More likely, this difference
reflects changes in the CE Survey and the financial accounts that it covers. While in 2008 the CE asked
about balances in checking and saving accounts separately, in 2013 the CE survey switched to asking a
single question about total liquidity across a larger set of types of accounts, and starting in 2017 the survey
introduced an initial question asking whether there was a zero balance in these accounts. The latter change
was associated with a reduction in the number of households reporting zero balances.
32
Table 7: The contemporaneous response of consumer expenditures to EIP by liquidity
Dependent variable: scaled dollar change in spending on
Panel A: EIP1 Panel B: EIP2 Panel C: EIP3
Food Nondurables All CE Food Nondurables All CE Food Nondurables All CE
and goods and and goods and and goods and
alcohol services alcohol services alcohol services
Bottom third: ≤ 2, 000 Bottom third: ≤ 2, 000 Bottom third: ≤ 2, 000
Top third: ≥ 12, 667 Top third: ≥ 12, 000 Top third: ≥ 10, 000
EIPn
]t 0.039 0.087 0.178 -0.032 0.112 0.220 0.078 0.132 0.081
(0.033) (0.064) (0.155) (0.071) (0.111) (0.326) (0.035) (0.062) (0.197)
EIPnt × Bottom third
] 0.016 0.130 0.301 0.050 0.009 0.191 -0.018 -0.099 -0.065
(0.051) (0.095) (0.210) (0.084) (0.157) (0.403) (0.048) (0.087) (0.219)
EIPnt × Top third
] 0.013 -0.188 -0.275 -0.090 -0.255 -0.272 0.048 -0.129 -0.057
(0.046) (0.102) (0.243) (0.085) (0.139) (0.449) (0.101) (0.099) (0.267)
p-value for test of 0.942 0.011 0.044 0.107 0.077 0.492 0.784 0.355 0.957
equality of responses
Implied propensity to spend by group
Least liquid 0.055 0.217 0.479 0.018 0.121 0.411 0.060 0.033 0.016
third (0.039) (0.070) (0.142) (0.046) (0.112) (0.237) (0.034) (0.062) (0.095)
Most liquid 0.052 -0.101 -0.097 -0.122 -0.143 -0.052 0.126 0.003 0.024
third (0.032) (0.079) (0.188) (0.048) (0.083) (0.309) (0.094) (0.077) (0.180)
Notes: All regressions use our imputation estimator to estimate equation (6). Also included are interview
month dummies, scaled age and change in the size of the CU, separate intercepts by thirds of the liquidity dis-
tribution interacted with other EIPs. The sample is the final sample and all results are from WLS regressions
using average weights. Standard errors included in parentheses are adjusted for arbitrary within-household
correlations and heteroskedasticity. The tests of equal responses are joint test for H0 : β 0,Bottom third = 0 and
β 0,Top third = 0. For Panel A, observations are those with an interview in June or July 2020; the columns
have 1,065, 1,066, and 1,066 never-treated or not-yet-treated observations, and 1,608, 1,609, 1,609 treated
observations, respectively. For Panel B, observations are those with an interview in February, March or
April 2021; the columns have 1,795 never-treated or not-yet-treated observations and 1,211 treated obser-
vations. For Panel C, observations are those with an interview in April, May or June 2021; the columns
have 1,387, 1,390, and 1,387 never-treated or not-yet-treated observations, and 892 treated observations.
for households with lower liquidity (e.g. Parker, 2017; Olafsson and Pagel, 2018; Ganong et
al., 2020; Baugh et al., 2020; Fagereng et al., 2021).
For the second round of EIPs, the spending responses are higher for households in the
bottom two thirds of the liquidity distribution and we can no longer reject equality of the
MPCs across the thirds of the distribution of liquid wealth. No spending responses are
statistically significant, but point estimates suggest the least liquid households spent 12%
of their EIPs on non-durable goods and services, the middle third in terms of liquidity
spent 11%, while the most liquid households are estimated to spend a negative amount.
The MPCs on total expenditures are more related to liquidity: 41%, 22% and -5% as we
move from the lowest to highest third of the distribution of liquid wealth but again with
33
no estimate being statistically significant. These findings are not inconsistent with Schild
and Garner (2020) which shows that in the Household Pulse data, households reporting
higher levels of financial difficulty are more likely to use their EIP2s mostly for spending.
Finally, for the third round of EIP — the largest in dollar terms and the latest in the
pandemic and the most likely understated due to data issues — the middle of the liquidity
distribution is the only group estimated to have a statistically significant spending response
to the arrival of their EIPs: 13% (6%) on non-durable goods and services, compared to
3% (6%) and 0.3% (8%) for the bottom and top thirds of the distribution of liquid wealth
respectively. Again, we cannot reject the null hypothesis of no differential response.
These patterns suggest that the first round of EIPs did meet important liquidity needs
for households with little liquid wealth in the early stages of the pandemic, when the
economy was most shut down. But later EIP rounds appear less beneficial on this front (or
their benefits were less related to liquid wealth). The second round payments were broadly
spent at the same average rate as EIP1, consistent with the tendency for households to
spend out of small, transitory increases in liquidity, and also with similar constraints on
consumer spending from the pandemic as EIP1. And the low spending of the final round of
payments, particularly among households with little liquidity is consistent with the large
size of the payment, although again our caveat about the low rate of EIP receipt reported
in the CE survey applies.
Analysis of our second measure of whether the EIPs provided effective pandemic
insurance — based on households ability to work from home – paints a similar picture:
the first round of EIPs appear to fill a pandemic insurance need for households but later
rounds do not.
We measure the exposure of income to the inability to work from home for EIP1 by the
share of pre-pandemic household income that cannot be earned from home. Specifically, for
the reference person and any secondary earner, we calculate the share of tasks associated
with their job based on their industry and education level following a mapping into the
Mongey et al. (2021) and Dingel and Neiman (2020) classifications by occupation and
education. For individual’s with no earned income (valid missing earnings), like retirees
or people not in the labor force, the measure is zero. We then multiply this share by
each person’s wage and salary income, sum to the household level, and divide by family
income. Because we require pre-pandemic income, we only use this measure to analyze
EIP1. Appendix B.3 contains complete details.
Table 8 shows that households most reliant on labor income from jobs that cannot be
done at home account for most of the spending response to the first round of EIPs. The
34
Table 8: The response of consumer expenditures to EIP1 receipt by the exposure of income
to inability to work from home in 2020
Dependent variable: scaled dollar change in spending on
Food and alcohol Nondurables All CE goods and services
Fraction of EIP1 spent over contemporaneous three-month period
EIP1t
] 0.021 0.052 -0.049
(0.022) (0.055) (0.119)
EIP1t × Middle third
] 0.030 0.176 0.258
(0.043) (0.089) (0.232)
EIP1t ×Least able third
] 0.036 0.064 0.367
(0.038) (0.083) (0.188)
p-value for test of 0.731 0.225 0.210
equality of responses
Cumulative fraction of EIP1 spent over contemporaneous and next three-month period
Most able third -0.007 -0.135 -0.435
(0.057) (0.159) (0.349)
Middle third 0.126 0.365 0.181
(0.100) (0.190) (0.622)
Least able third 0.117 0.285 0.842
(0.080) (0.156) (0.448)
Notes: All regressions use equation 6. Also included are interview month dummies, scaled age and change
in the size of the CU, and separate intercepts by thirds of the distribution. The sample is the final sample
which includes only CE households with an interview in June or July 2020, with income that does not exceed
a certain threshold determined by marital status and family structure. The work-from-home measure used
is the income-based measure. All results are from WLS regressions. Weights applied are average weights.
Standard errors included in parentheses are adjusted for arbitrary within-household correlations and het-
eroskedasticity. The tests of equal responses are joint test for H0 : β 0,Least able third = 0 and β 0,Middle third = 0.
third of households with little to no income exposure have point estimates that imply
EIP1 lowered their spending. The third of households with income that was moderately
exposed, had an average MPC of 36% (19%) on non-durable goods and services while
the most exposed third had a similar average MPC of 29% (16%) (and an MPC on total
expenditures of 84% (45%)) during the three-month period of receipt and the subsequent
period.
For later EIPs, given the rotating panel structure of the CE, we cannot measure pre-
pandemic incomes, and earnings after the onset of the pandemic may already reflect losses
35
Table 9: The contemporaneous response of consumer expenditures to EIP receipt by the
ability to work from home
Dependent variable: scaled dollar change in spending on
Food Nondurables All CE Food Nondurables All CE Food Nondurables All CE
and goods and and goods and and goods and
alcohol services alcohol services alcohol services
Panel A: EIP1 Panel B: EIP2 Panel C: EIP3
Bottom third: ≤ 89.4% Bottom third: ≤ 87.0% Bottom third: ≤ 84.1%
Top third: ≥ 99.1% Top third: ≥ 98.8% Top third: ≥ 97.7%
EIP
gt -0.014 0.048 0.189 0.061 0.131 0.265 0.026 -0.021 0.015
(0.027) (0.041) (0.086) (0.035) (0.071) (0.144) (0.036) (0.038) (0.080)
EIPt × Middle third
g 0.031 0.090 0.158 -0.042 -0.131 -0.211 0.024 0.033 0.219
(0.038) (0.062) (0.149) (0.049) (0.095) (0.230) (0.039) (0.050) (0.122)
EIPt × Least able third
g 0.090 0.109 0.239 -0.004 -0.024 -0.072 0.016 0.066 0.143
(0.037) (0.069) (0.144) (0.053) (0.100) (0.216) (0.039) (0.048) (0.106)
p-value for test of 0.046 0.188 0.222 0.648 0.326 0.656 0.817 0.376 0.174
equality of responses
Implied spending by group
Middle third 0.016 0.138 0.346 0.019 0.000 0.054 0.049 0.012 0.233
(0.026) (0.047) (0.121) (0.034) (0.063) (0.179) (0.015) (0.033) (0.093)
Least able third 0.076 0.157 0.428 0.057 0.110 0.193 0.042 0.045 0.158
(0.026) (0.055) (0.116) (0.040) (0.070) (0.160) (0.016) (0.029) (0.070)
Notes: All regressions use equation 6. Also included are interview month dummies, scaled age and
change in the size of the CU, and separate intercepts by thirds of the distribution. The sample is the
final sample which includes only CE households with an interview in June or July 2020, with income
that does not exceed a certain threshold determined by marital status and family structure. The work-
from-home measure used is the non-income measure. All results are from WLS regressions, and the
weights applied are average weights. Standard errors included in parentheses are adjusted for arbi-
trary within-household correlations and heteroskedasticity. The tests of equal responses are joint test for
H0 : β 0,Bottom third = 0 and β 0,Middle third = 0. For Panel A, all regressions have 3,470 observations For
Panel B, all regressions have 3,099 observations. For Panel C, all regressions have 3,463 observations.
incurred by an inability to work from home. Therefore, in order to investigate differences in
consumption responses across ability to work from home for EIP2 and EIP3, we construct a
work-from-home measure that does not rely on observing pre-pandemic wage and salary
earnings. We construct a second measure based of the share of wage and salary (potential)
earnings that cannot be done from home and the assumption that earners within a family
have equal earnings. This measure requires only information on the industry and education
of (potential) earners, whether currently working or not (see Appendix B.3 for details).
Using this second measure, Panel A of Table 9 shows findings for EIP1 that align well
with our first measure of the ability to work from home based on pre-pandemic income.
That is, we find all spending is done by the two thirds of households with the highest level
of income exposure during the pandemic, as we did in Table 8. Panels B and C show no
significant differences in spending propensities related to the ability to work from home
36
for either of the second two rounds of EIPs, consistent with the waning of the economic
impact of the pandemic. If anything, EIP2 spending responses are concentrated among
households with no income exposure to the pandemic. For EIP3, only those with incomes
that are the most exposed to the pandemic have statistically significant spending response
on non-durable goods and services.
In sum, while on average the EIPs appear to have gone to many households with
incomes that were unharmed by the pandemic (e.g. retirees, those employed and able to
work from home, etc.), some of the EIPs, mainly in the first round, did support short-term
spending for some households, those with low ex ante liquid wealth and those reliant on
income that could not be earned by working from home.
VI Concluding remarks
The pandemic limited the types of goods and services that people could purchase
and many households reduced spending. There were also policy responses besides the
EIPs, including extended and expanded unemployment insurance, and the Paycheck
Protection Program that transferred money to small and medium sized businesses with
some incentives to maintain payroll, both of which were intended to help offset any lost
income. Finally, the depth and duration of the pandemic were uncertain, particularly when
the first round of EIPs were being disbursed. These factors appear to have led to less
spending on non-durable goods and services (CE-measured) in response to the arrival
of the first round or EIPs than out of the tax rebates in 2001 and 2008, and to have tilted
what spending response there was towards durable goods. We observe similar spending
responses to the second-round of EIPs, but very little short-run spending in response to
the third, consistent with high pre-existing high levels of financial resources; although, the
response is not as cleanly measured as the previous EIPs.
Were the EIPs effective? The goal of previous tax rebates programs was to increase
demand and so their efficacy is largely related to the speed and size of the spending
responses. In contrast, the policy goal of the EIPs was insurance, that is, to provide money
to those who lost or would lose employment and who would not be covered by government
aid programs. For these individuals, the EIPs could be initially saved and then used to
cover a later loss. We find significant spending responses for households with low levels of
ex ante liquidity in response to the first round of EIPs during the national emergency at
the onset of the pandemic. The smaller amount of spending following the arrival of the
December 2020 payments was due to a spending response by those outside the top third of
37
the liquidity distribution. Finally, we find substantially higher spending responses by those
reliant on earnings from jobs with tasks that could not be done from home in response to
the first-round EIPs (and little evidence on this issue for later EIP rounds).
The small, short-term spending response and its pattern suggest that the EIPs went to
many people who did not need the additional funds as urgent pandemic insurance (e.g.
Sahm, 2021, debates these issues). However, despite the lack of much immediate spending,
the EIPs could have filled the role of pandemic insurance for some households beyond the
time horizon accurately measured by this (and other) studies. On the other hand, from a
demand management perspective, the unspent EIPs have contributed to strong households
balance sheets over the past year, a period of strong demand and rising inflation.
38
References
Agarwal, Sumit and Wenlan Qian, “Consumption and Debt Response to Unanticipated
Income Shocks: Evidence from a Natural Experiment in Singapore,” American Economic
Review, 2014, 104 (12), 4205–4230.
, Chunlin Liu, and Nicholas S. Souleles, “The Reaction of Consumer Spending and
Debt to Tax Rebates: Evidence from Consumer Credit Data,” Journal of Political Economy,
2007, 115 (6), 986–1019.
Armantier, Olivier, Leo Goldman, Gizem Koşar, Jessica Lu, Rachel Pomerantz, and
Wilbert van der Klaauw, “How Have Households Used Their Stimulus Payments and
How Would They Spend the Next?,” Technical Report, Liberty Street Economics. October
13 2020.
Baker, Andrew C., David F. Larcker, and Charles C.Y. Wang, “How much should we trust
staggered difference-in-differences estimates?,” Journal of Financial Economics, 2022, 144,
370–395.
Baker, Scott R., R. A. Farrokhnia, Steffen Meyer, Michaela Pagel, and Constantine
Yannelis, “Income, Liquidity, and the Consumption Response to the 2020 Economic
Stimulus Payments,” Review of Finance, forthcoming.
Baugh, Brian, Itzhak Ben-David, Hoonsuk Park, and Jonathan A. Parker, “Asymmetric
Consumption Smoothing,” American Economic Review, 2020, 111, 192–230.
Bodkin, Ronald, “Windfall Income and Consumption,” American Economic Review, 1959,
49 (4), 602–614.
Borusyak, Kirill and Xavier Jaravel, “Revisiting Event Study Designs,” SSRN, 2018.
, , and Jann Spiess, “Revisiting Event Study Designs: Robust and Efficient Estimation,”
SSRN, 2022.
Boutros, Michael, “Evaluating the Impact of Economic Impact Payments,” Manuscript,
Duke University, December, 2020.
Broda, Christian and Jonathan A. Parker, “The Economic Stimulus Payments of 2008 and
the Aggregate Demand for Consumption,” Journal of Monetary Economics, Fall 2014, 68,
S20–36.
39
Bureau of the Fiscal Service, “EIP Dashboard,” 2022. data acquired from the Bureau of the
Fiscal Service, U.S. Department of the Treasury.
Cajner, Tomaz, Leland Crane, Ryan Decker, John Grigsby, Adrian Hamins-Puertolas,
Erik Hurst, Christopher Kurz, and Ahu Yildermaz, “The U.S. Labor Market During the
Beginning of the Pandemic Recession,” Brooking Papers on Economic Activity, Summer
2020.
Callaway, Brantly and Pedro H. C. Sant’Anna, “Difference-in-Differences with multiple
time periods,” Journal of Econometrics, December, forthcoming 2021, 225 (2), 200–230.
Chetty, Raj, John Friedman, Nathaniel Hendren, Michael Stepner, and The Opportunity
Insights Team, “The Economic Impacts of Covid-19: Evidence from a New Public
Database Built Using Private Sector Data,” NBER Working Paper 27431, June 2021.
Coibion, Olivier, Yuriy Gorodnichenko, and Michael Weber, “How Did U.S. Consumers
Use Their Stimulus Payments?,” NBER Working Paper 27693, 2021.
Cooper, Daniel and Giovanni Olivei, “High-Frequency Spending Responses to Govern-
ment Transfer Payments,” 2021.
Cox, Natalie, Peter Ganong, Pascal Noel, Joseph Vavra, Arlene Wong, Diana Farrell, and
Fiona Greig, “Initial Impacts of the Pandemic on Consumer Behavior: Evidence from
Linked Income, Spending, and Savings Data,” Brookings Papers on Economic Activity, June
2020.
de Chaisemartin, Clément and Xavier D’Haultfœuille, “Two-Way Fixed Effects Estimators
with Heterogeneous Treatment Effects,” American Economic Review, September 2020, 110
(9), 2964–2996.
Dingel, Jonathan I. and Brent Neiman, “How many jobs can be done at home?,” September
2020, 189, 104235.
Fagereng, Andreas, Martin B. Holm, and Gisle J. Natvik, “MPC Heterogeneity and
Household Balance Sheets,” American Economic Journal: Macroeconomics, October 2021, 13
(4).
Farrell, Diana, Fiona Greig, and Amar Hamoudi, “Tax Time: How Families Manage Tax
Refunds and Payments,” Technical Report, JPMorgan Chase Institute March 2019.
40
Feldman, Naomi E. and Ori Heffetz, “A Grant to Every Citizen: Survey Evidence of the
Impact of a Direct Government Payment in Israel,” NBER Working paper 28312, December
2020.
Ganong, Peter, Damon Jones, Pascal J. Noel, Fiona E. Greig, Diana Farrell, and Chris
Wheat, “Wealth, Race, and Consumption Smoothing of Typical Income Shocks,” NBER
Working Paper 27552, 2020.
, Fiona Greig, Pascal Noel, Daniel M. Sullivan, and Joseph Vavra, “Spending and Job-
Finding Impacts of Expanded Unemployment Benefits: Evidence from Administrative
Micro Data,” Draft Manuscript, August 2022.
Garner, Thesia I., Adam Safir, and Jake Schild, “Receipt and use of stimulus payments in
the time of the Covid-19 pandemic,” Beyond the Numbers: Prices and Spending, August
2020, 9 (10).
Gelatt, Julia, Randy Capps, and Michael Fi, “Nearly 3 Million U.S. Citizens and Legal
Immigrants Initially Excluded under the CARES Act Are Covered under the December
2020 COVID-19 Stimulus,” https://www.migrationpolicy.org/news/cares-act-excluded-
citizens-immigrants-now-covered, Migration Policy Institute January 15 January 15 2021.
Goodman-Bacon, Andrew, “Difference-in-differences with variation in treatment timing,”
Journal of Econometrics, December, forthcoming 2021, 225 (2), 254–277.
Greenwood, Robin, Toomas Laarits, and Jeffrey Wurgler, “Stock Market Stimulus,” NBER
Working Paper Series, No. 29827, March March 2022.
Grieg, Fiona, Erica Deadman, and Tanya Sonthalia, “Household Finances
Pulse: Cash Balances during COVID-19,” JPMorgan Chase Institute,
https://www.jpmorganchase.com/institute/research/household-income-
spending/household-finances-pulse-cash-balances-during-COVID-19, downloaded
August 2022 September 2022.
Guerrieri, Veronica, Guido Lorenzoni, Ludwig Straub, and Ivan Werning, “Macroeco-
nomic Implications of COVID-19: Can Negative Supply Shocks Cause Demand Short-
ages?,” NBER Working Paper 26918, 2020.
Internal Revenue Service, “SOI Tax Stats - Coronavirus Aid, Relief, and Economic Security
Act (CARES Act) Statistics,” https://www.irs.gov/statistics/soi-tax-stats-coronavirus-
aid-relief-and-economic-security-act-cares-act-statistics June 15 June 15 2022.
41
, “Treasury, IRS release latest state-by-state Economic Impact Payment figures for May
22, 2020,” https://www.irs.gov/newsroom/treasury-irs-release-latest-state-by-state-
economic-impact-payment-figures-for-may-22-2020 May 22 May 22 2020.
Johnson, David S., Jonathan A. Parker, and Nicholas S. Souleles, “Household Expendi-
ture and the Income Tax Rebates of 2001,” American Economic Review, December 2006, 96
(5), 1589–1610.
Karger, Ezra and Aastha Rajan, “Heterogeneity in the Marginal Propensity to Consume:
Evidence from Covid-19 Stimulus Payments,” Federal Reserve Bank of Chicago Working
Paper, Working Paper, No. 2020-15, February 2021.
Kim, Seonghoon, Kanghyock Koh, and Wonjun Lyou, “Spending impact of COVID-19
stimulus payments: Evidence from card transaction data in South Korea,” SSRN, 2020.
Kubota, So, Koichiro Onishi, and Yuta Toyama, “Consumption responses to COVID-19
payments: Evidence from a natural experiment and bank account data,” Covid Economics,
December 2020, 62 (18), 11–23.
Kueng, Lorenz, “Excess Sensitivity of High-Income Consumers,” Quarterly Journal of
Economics, 2018, 133 (4), 1693–1751.
Lewis, Daniel, Davide Melcangi, and Laura Pilossoph, “Latent Heterogeneity in the
Marginal Propensity to Consume,” Federal Reserve Bank of NY Staff Report, June 2021.
Lusardi, Annamaria, “Permanent Income, Current Income, and Consumption: Evidence
from Two Panel Data Sets,” Journal of Business and Economic Statistics, January 1996, 14
(1), 81–90.
Meyer, Michelle and Anna Zhou, “COVID-19 and the consumer: data through April 16,”
Bank of America Global Research Data Analytics, 22 April 2020.
Misra, Kanishka, Vishal Singh, and Qianyun Poppy Zhang, “Impact of Stay-at-home-
orders and Cost-of-living on Stimulus Response: Evidence from the CARES Act,” NYU
Stern School of Business Working Paper, 2021.
Mongey, Simon, Laura Pilossoph, and Alexander Weinberg, “Which workers bear the
burden of social distancing?,” The Journal of Economic Inequality, 2021, 19 (3), 509–526.
Murphy, Dan, “Economic Impact Payments: Uses, payment methods, and costs to recipi-
ents,” Economic Studies at Brookings, February 2021.
42
Olafsson, Arna and Michaela Pagel, “The Liquid Hand-to-Mouth: Evidence from Personal
Finance Management Software,” Review of Financial Studies, 2018, 31 (11), 4398–4446.
Orchard, Jacob, Valerie A. Ramey, and Johannes F. Wieland, “Micro MPCs and Macro
Counterfactuals: The Case of the 2008 Rebates,” 2022.
Parker, Jonathan A, “The Reaction of Household Consumption to Predictable Changes in
Social Security Taxes,” American Economic Review, 1999, 89 (4), 959–973.
, “Why Don’t Households Smooth Consumption? Evidence from a 25 Million Dollar
Experiment,” American Economic Journal: Macroeconomics, October 2017, 9 (4), 153–183.
Parker, Jonathan A., Jake Schild, Laura Erhard, and David S. Johnson, “Household
Spending Responses to the Economic Impact Payments of 2020: Evidence from the
Consumer Expenditure Survey,” BLS Working paper WP-544, 2021.
, Nicholas S. Souleles, David S. Johnson, and Robert McClelland, “Consumer Spending
and the Economic Stimulus Payments of 2008,” American Economic Review, October 2013,
103 (6), 2530–53.
Romer, Christina D. and David H. Romer, “A Social Insurance Perspective on Pandemic
Fiscal Policy: Implications for Unemployment Insurance and Hazard Pay,” 2022, 36 (2),
3–28.
Roth, Jonathan, Pedro H. C. Sant’Anna, Alyssa Bilinski, and John Poe, “What’s Trend-
ing in Difference-in-Differences? A Synthesis of the Recent Econometrics Literature,”
arXiv:2201.01194 [econ.EM], 2022.
Sahm, Claudia R., “Robust evidence for $1400 checks: Relief for families and recovery for
all,” Jain Family Institute Policy Brief, February 17, February 17 2021.
, Matthew D. Shapiro, and Joel Slemrod, “Household Response to the 2008 Tax Rebate:
Survey Evidence and Aggregate Implications,” in “Tax Policy and the Economy, Volume
24” NBER Chapters, National Bureau of Economic Research, Inc, November 2010, pp. 69–
110.
, , and , “Consumer Response to the Coronavirus Stimulus Programs,” Presenta-
tion November 11, 2020. http://macromomblog.com/wp-content/uploads/2021/02/Rebates-2020-
111020.pdf, 2020.
43
Schild, Jake and Thesia Garner, “Consumer Response to Economic Impact Payments
during the COVID-19 Pandemic and the Role of Subjective Assessments of Well-Being:
A View from the U.S. Using a Rapid Response Survey,” Bureau of Labor Statistics Working
Paper. https://www.bls.gov/osmr/research-papers/2021/pdf/ec210060.pdf, 2020.
Shapiro, Matthew D. and Joel Slemrod, “Consumer Response to the Timing of Income:
Evidence from a Change in Tax Withholding,” American Economic Review, 1995, 85,
274–83.
and , “Did the 2008 Tax Rebates Stimulate Spending?,” American Economic Review,
May 2009, 99 (2), 374–79.
Souleles, Nicholas S, “The Response of Household Consumption to Income Tax Refunds,”
American Economic Review, 1999, 89 (4), 947–958.
Splinter, David, “Stimulus Checks: True-Up and Safe Harbor Costs,” Joint Committee on
Taxation Working Paper., June 2022.
Stephens, Melvin, ““3rd of the Month”: Do Social Security Recipients Smooth Consump-
tion between Checks?,” American Economic Review, 2003, 93 (1), 406–422.
Sun, Liyang and Sarah Abraham, “Estimating dynamic treatment effects in event studies
with heterogeneous treatment effects,” Journal of Econometrics, December, forthcoming
2021, 225 (2), 175–199.
US Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Sur-
vey: Interview Survey [computer file], Washington, DC: US Department of Labor, Bureau of
Labor Statistics, 2020.
Wooldridge, Jeffrey, “Two-Way Fixed Effects, the Two-Way Mundlak Regression, and
Difference-in-Differences Estimators,” SSRN, 2021.
44
File and source
- File
- economic-impact-payments-and-household-spending-during-the-pandemic.pdf
- Size
- 497,262 bytes
- SHA-256
- 5e1b395c7c654aa9cebbd7d245ca5aacc8c343ba703d73dddd7fcd9b5b69217c
- Original
- www.nber.org