How Did U.S. Consumers Use Their Stimulus Payments? — NBER Working Paper 27693 (August 2020)
Summary
NBER Working Paper No. 27693, dated August 2020, by Olivier Coibion, Yuriy Gorodnichenko and Michael Weber, using a large-scale survey of U.S. consumers to study how CARES Act transfers affected consumption, saving and labor-supply decisions. The abstract says most respondents primarily saved or paid down debts, with only about 15 percent reporting that they mostly spent the transfer, and that individuals report spending or planning to spend only around 40 percent of the total on average. It reports no meaningful effect on labor-supply decisions, except for twenty percent of the unemployed who say the payment made them search harder. The paper states that NBER working papers have not been peer-reviewed and that the views are the authors' own.
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
- How Did U S Consumers Use Their Stimulus Payments
- Case
- How Did U S Consumers Use Their Stimulus Payments
Cite this document
“How Did U.S. Consumers Use Their Stimulus Payments? — NBER Working Paper 27693 (August 2020).”
Original at www.nber.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/how-did-u-s-consumers-use-their-stimulus-payments/
, SHA-256 2daac389f165409d1bbe1edeea4d95d46e13d2554bc76b95d91a7d3d59e1a7d7.Full text
NBER WORKING PAPER SERIES
HOW DID U.S. CONSUMERS USE THEIR STIMULUS PAYMENTS?
Olivier Coibion
Yuriy Gorodnichenko
Michael Weber
Working Paper 27693
http://www.nber.org/papers/w27693
NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
August 2020, Revised August 2020
This research was funded in part by National Science Foundation grant 1919307. 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 They have not been
peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies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.
© 2020 by Olivier Coibion, Yuriy Gorodnichenko, and Michael Weber. 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.
How Did U.S. Consumers Use Their Stimulus Payments?
Olivier Coibion, Yuriy Gorodnichenko, and Michael Weber
NBER Working Paper No. 27693
August 2020, Revised August 2020
JEL No. E2,E3,E6
ABSTRACT
Using a large-scale survey of U.S. consumers, we study how the large one-time transfers to
individuals from the CARES Act affected their consumption, saving and labor-supply decisions.
Most respondents report that they primarily saved or paid down debts with their transfers, with
only about 15 percent reporting that they mostly spent it. When providing a detailed breakdown
of how they used their checks, individuals report having spent or planning to spend only around
40 percent of the total transfer on average. This relatively low rate of spending out of a one-time
transfer is higher for those facing liquidity constraints, who are out of the labor force, who live in
larger households, who are less educated and those who received smaller amounts. We find no
meaningful effect on labor-supply decisions from these transfer payments, except for twenty
percent of the unemployed who report that the stimulus payment made them search harder for a
job.
Olivier Coibion Michael Weber
Department of Economics Booth School of Business
University of Texas at Austin University of Chicago
2225 Speedway 5807 South Woodlawn Avenue
Austin, TX 78712 Chicago, IL 60637
and NBER and NBER
ocoibion@gmail.com michael.weber@chicagobooth.edu
Yuriy Gorodnichenko
Department of Economics
530 Evans Hall #3880
University of California, Berkeley
Berkeley, CA 94720-3880
and IZA
and also NBER
ygorodni@econ.berkeley.edu
I Introduction
Amidst the rising spread of COVID-19 and the pervasive imposition of lockdowns in March 2020,
the U.S. Federal government passed the CARES Act on March 27th, 2020. This stimulus package
was exceptional both in size (over $2 trillion in allocated funds) and in the speed at which it was
legislated and implemented. A major component was a one-time transfer to all qualifying adults
of up to $1200, with $500 per additional child. How effective were these transfers in stimulating
the consumption of recipients? While the 2001 and 2008 fiscal stimulus payments provide some
guidance, the unprecedented nature of the COVID-19 shock, the associated uncertainty about the
length and severity of the pandemic, and the widespread prevalence of lockdowns which restrict
in-person shopping make it ex-ante unclear how individuals used their payments.
Using a large-scale survey of U.S. households, we document that only 15 percent of
recipients of this transfer say that they spent (or planned to spend) most of their transfer payment,
with the large majority of respondents saying instead that they either mostly saved it (33 percent)
or used it to pay down debt (52 percent). When asked to provide a quantitative breakdown of how
they used their checks, U.S. households report having spent approximately 40 percent of their
checks on average, with about 30 percent of the average check being saved and the remaining 30
percent being used to pay down debt. Little of the spending went to hard-hit industries selling large
durable goods (cars, appliances, etc.). Instead, most of the spending went to food, beauty, and other
non-durable consumer products that had already seen large spikes in spending even before the
stimulus package was passed because of hoarding.
These average responses mask significant differences across households. For example,
lower-income households were significantly more likely to spend their stimulus checks, as were
households facing liquidity constraints. Individuals out of the labor force were also more likely to
spend their checks than either employed or unemployed individuals, consistent with motives of
consumption smoothing and hand-to-mouth behavior. Other groups that were more likely to report
spending most of their checks were those living in larger households, men, Hispanics and those
with lower education. In contrast, African-Americans were much more likely to report using their
checks primarily to pay off debt, as were older individuals, those with mortgages, unemployed
workers and those reporting to have lost earnings due to COVID. For those who did not wish to
spend their stimulus payment and had to decide whether to pay off debt or save their checks,
higher-income individuals were more likely to save than pay off debts, those with mortgages or
2
renters were much more likely to pay off debts instead as were financially constrained individuals.
These heterogeneous responses are broadly in line with recent models with assets of different
levels of liquidity in which wealthy-hand-to-mouth behavior arises (see, e.g., Kaplan and Violante,
2014). To the contrary, those that lost financial wealth show no differential willingness to spend
compared to other households, consistent with a low marginal propensity to consume out of stock
market wealth (Lettau and Ludvigson, 2004; Di Maggio, Kermani, and Majlesi (forthcoming)).
Our survey also asked respondents whether transfer payments were likely to affect their
labor supply decisions. Ninety percent of employed workers who had received a stimulus check
reported that the transfer had no effect on their work effort (as opposed to e.g. searching harder for
new work) while eighty percent of those employed workers who did not qualify for a check
reported that receiving such a check would not affect their work effort. People out of the labor
force similarly report that the extra income had not affected their search effort either qualitatively
(e.g. by starting or stopping job search) or quantitatively (in terms of their search intensity). Only
for unemployed workers do we find some evidence that one-time fiscal transfers affect labor
supply: approximately 20 percent of those who received a payment claim that this made them
search harder for a job, while two-thirds report that it had no effect. To the best of our knowledge,
we are the first to report how one-time stimulus payments affect labor supply decisions, an
important margin in many macroeconomic models used to characterize the effects of fiscal policy.
These results possibly also inform the debate on how additional unemployment benefits affect
labor supply decisions during recessions.
We add to the literature studying how one-time fiscal stimulus payments affect saving,
consumption, and debt payments of individuals and hence estimate the partial equilibrium effect
on these outcome variables from the anticipated, one-time provision of liquidity (Parker, 2011).
Consistent with an earlier literature for the 2001 tax rebates such as Shapiro and Slemrod (2003),
we find that little more than 20% of survey respondents would use the rebates mainly to increase
spending. Johnson, Parker, and Souleles (2006) documents that in the first quarter post-payment,
households spent between 20%-40% of the rebate which increased to about two thirds over the
course of the six months. They also find that low-income households and individuals with low
levels of liquid wealth showed a larger consumption response. Agarwal et al. (2003) use credit
card data and find many households initially used the rebates to pay down debt but spent about
40% of the average rebate over the following nine months.
3
Studying the 2008 fiscal stimulus payments, Parker et al. (2013) find that the average
household spent between 12% and 30% in the quarter after receiving payments on non-durables.
Durable purchases and especially car purchases also increased substantially in the quarter after
receiving the payments increasing the total consumption response to 50% to 90% of the stimulus
payment which are substantially larger than the estimates for the 2001 stimulus payments and our
own estimates for 2020. Broda and Parker (2014) use the Nielsen Homescan Panel data and find
that individuals increase their grocery spending by about 10% in the week after receiving payments
with larger effects for low income and wealth households but do not change their spending upon
announcement of the program. Parker (2017) also uses Nielsen data and documents that persistent
household characteristics such as impatience and lack of financial planning determine the spending
response to the payment of the rebates. Parker and Souleles (2019) compare the estimates of the
willingness to spend due to the arrival of liquidity elicited from surveys (reported effects) and from
spending data (revealed preference) and find that both methods result in similar average
propensities. Sahm, Shapiro, and Slemrod (2010) use data from Michigan Survey of Consumers
and also find that only 20% of respondents used the stimulus checks to mainly increase spending,
while more than half of the respondents mainly wanted to pay off debts. They find an average
MPC of about one third over the following twelve months. Sahm, Shapiro, and Slemrod (2012)
compare the marginal propensities to consume of one-time payments to a flow of payments from
reduced tax withholdings and find that the one-off stimulus check results in a marginal propensity
to consume of twice the size compared to the reduced withholdings. Baker et al. (2020) study the
response of the 2020 stimulus checks using data from a FinTech app. They find spending increased
by 29% of the stimulus amount in the 10 days after receiving payment, mainly driven by food and
non-durables. Their sample is tilted towards low-income and liquidity-constrained households
with the average income in the sample being $25,000 and the average account balance being $141,
possibly partially explaining the large immediate spending response.
II The Survey
Our survey was run on individuals participating in the Nielsen Homescan panel. This is a
representative panel of 80,000-90,000 individuals who track their purchases on a daily basis. We
have been running repeated quarterly surveys of these households, as described for example in
Coibion, Gorodnichenko and Weber (2018) and Coibion, Georgarakos, Gorodnichenko and Weber
4
(2019). The results presented in this paper focus on the July 2020 wave of the survey, in which
approximately 46,000 individuals were invited and 12,000 responded to the survey. This
corresponds to a response rate of 26%, given that many individuals participating in the Nielsen
Homescan opt out of surveys over time, reducing the population that can potentially be reached.
The response rate compares favorably to other private surveys. For example, Qualtrics estimates
that their average response rate is between 5% to 10%. The number of respondents is significantly
larger than the Survey of Consumer Expectations (~1,500 respondents per wave) or the Michigan
Survey of Consumers (~500 respondents per wave). Sampling weights are provided to correct for
possible imbalances in the panel so that our estimates are representative for the U.S. population.
The survey includes a wide range of questions covering the macroeconomic expectations
of respondents, questions on their spending and investment patterns, and questions on their labor
market status (see Coibion, Gorodnichenko and Weber, 2020 a,b,c for details). In this specific
wave, we introduced additional questions designed to assess how the stimulus payments in the
CARES Act affected consumer behavior.1 For example, survey participants are asked to report
income from their latest tax return to determine the eligibility and timing of the payments. The
survey also includes a very brief description of this portion of the CARES Act:
“In response to the COVID-19 crisis, the Congress approved a $2 trillion package
to stimulate the economy, including one-time $1,200 check per person (plus another
$500 per child) to persons with annual income less than $75,000. Couples who filed
jointly and made less than $150,000 will get a one-time $2,400 check (plus another
$500 per child). The Federal government calls it an Economic Impact Payment.”
They were then asked a number of follow-up questions. First, we asked whether they had received
any such payment or expected to do so. If they had received it, they were asked about whether it
was in the form of a direct deposit or check. They were then asked how much they had received
and when. Next was a qualitative question about how they had used the income (or planned to use
the income for those who expected to receive but had not yet received their check). Specifically,
they were asked: “How did you use the payment from the Federal Government?” and had three
options to select between: “Mostly to increase spending,” “Mostly to increase saving,” and
“Mostly to pay off debt.” Those who said they had not and would not be receiving a check were
asked the same question as a hypothetical, assuming they were to receive a $1,000 payment. This
1
This block of questions is reported in Appendix A.
5
qualitative question closely follows that used by Shapiro and Slemrod (2003) and Sahm, Shapiro
and Slemrod (2010).
Following this qualitative description, respondents were asked to assign dollar amounts to
each of the following categories for how they used the transfer:
[] Saving,
[] Debt payments (mortgages, auto loans, student loans, etc.),
[] Durable goods such as houses/apartments, cars, large appliances and electronics,
[] Food, health/beauty aids, and household products,
[] Medical care (including health insurance, out-of-pocket medical bills and prescription
drugs),
[] Other consumer spending.
For those individuals who expected to receive a check but had not yet received it, they were asked
equivalent questions about how they expected to spend their payments once they arrived, while
those who would not receive a payment were asking the equivalent question in hypothetical form.
After these questions on spending and saving decisions, respondents were asked several
questions about how the stimulus payments would affect their labor supply. For those who initially
reported being employed, they were asked “How did the payment from the Federal Government
influence your work effort? Because of the payment…” after which they could select among the
following options:
[] I started to work fewer hours,
[] I started to work more hours,
[] I started to work an additional job,
[] I quit some of my jobs,
[] I quit working,
[] The payment did not influence my work effort.
Those respondents who initially reported not working but looking for work (the unemployed) were
instead asked the following: “How did the payment from the Federal government influence your
decision to look for a job or not? Please choose options that apply to you. Because of the payment,
[please choose all relevant options]” The options for them to choose amongst were:
[] I stopped looking for a job,
[] I started looking for a job,
[] I delayed looking for a job,
[] I am looking for a job but put less effort into it,
[] I am looking for a job and put more effort into it,
[] The payment did not influence my decision.”
6
Again, those individuals who reported that they had not yet received a payment from the Treasury
but expected to do so were asked equivalent questions in expectation. Those who did not expect
to receive any payment were also asked an equivalent question as a hypothetical. To the best of
our knowledge, these are the first survey questions that attempt to measure the labor supply
response of households to stimulus payments.
This structure of the survey and the broad coverage of respondents improves upon existing
studies in several dimensions. First, we collect not only the reported qualitative use of stimulus
payments (i.e., most spend, mostly save, etc.) but also quantitative measures of actual spending.
Second, while credit card data provides incredible details for consumer spending, our survey
provides much richer demographic information, collects macroeconomic and microeconomic
expectations, covers parts of the population who are less financially savvy and may have no bank
account or who use other means of payments (cash, check, gift cards, etc.) and includes individuals
who receive their stimulus payments via checks rather than direct deposits only. Third, when
scanner data for consumer spending is released by Nielsen, we can combine our survey with
higher-frequency measures of consumer spending, which is hard to do using typical surveys.
III Reported Treasury Payments in the Survey
We first consider household responses about whether and when they received payments. Eligibility
for receiving stimulus payments was based in part of Adjustable Gross Income (AGI) from 2019
(or 2018 if a tax return for 2019 was not filed yet; the deadline for tax filing was extended to July
15, 2020). For example, single filers earning more than $99,000 or joint filers with no children
earning more than $198,000 were ineligible for any check. Single filers earning less than $75,000,
heads of households earning less than $112,500, and joint filers earning less than $150,000 all
qualified for payments of $1,200 per adult and $500 per dependent minor. For those earning
between these thresholds, stimulus payments decline with income by $5 for every $100 in additional
AGI. As a result, we begin by describing the reported taxable income of households in our survey.
The distribution of respondents’ income is plotted in Figure 1. Less than 15 percent of
respondents say that they do not remember their taxable income from 2018 or 2019 or prefer not
to answer the question. The vast majority (71 percent) of those who did answer the question report
an income of less than $80,000. For comparison, in 2017, the most recent year for which
comparable information is available, the Internal Revenue Service (IRS) reports that 73 percent of
7
tax returns had an AGI of $75,000 or less. The IRS also reports that approximately 5 percent of all
tax returns had an AGI of $200,000 or more. In our survey, approximately 2.5 percent of
respondents report such high-income levels, a slightly lower number which could reflect that
higher-income respondents are more likely to refuse to answer this question. But since these high-
income respondents are ineligible for the stimulus payment, an under-representation of these
individuals in the survey is not an issue. In short, we interpret respondents’ answers about their
incomes as broadly in line with what the IRS reports.
Respondents’ answers about how much they received (if anything) in stimulus payments
are also consistent with the guidelines governing the amount of payment to be received. For
example, most individuals below the thresholds described above should receive either $1,200 (for
single tax filers) or $2,400 (for joint filers) augmented with $500 for each minor dependent. As
documented in Figure 2, respondents to our survey report answers that are largely consistent with
these amounts. More than fifty percent report having received $1,200, and ≈15 percent report
having received $2,400. There are also clear spikes at increments of $500, such as $1,700, $2,200,
etc… For respondents living in one-member households and earning less than $75,000, the median
payment received was $1,200 (the average was $1,150), very close to the $1,200 they all should
have received. For one-member households earning more than $99,000 (the upper threshold for
receiving payments), the median payment is zero (the average is $300) and less than a quarter
report having received a stimulus payment. These facts suggest that respondents are providing
answers regarding their stimulus payments that are consistent with the guidelines governing who
should receive how much.
The survey also asked about when participants received their transfers. About half report
having received their checks by mid-April, and most had received them by mid-May, as shown in
Figure 3. This timing is consistent with reports from the IRS. For example, the IRS claimed to
have distributed nearly 90 million payments as of April 17th and nearly 130 million payments as
of May 11th, out of over 150 million to be paid out.2 There are only a small number of respondents
(6 percent) who claim to have received payments prior to April, which is impossible given that the
2
See https://www.irs.gov/newsroom/why-the-economic-impact-payment-amount-could-be-different-than-
anticipated and https://www.irs.gov/newsroom/treasury-irs-deliver-89-point-5-million-economic-impact-payments-
in-first-three-weeks-release-state-by-state-economic-impact-payment-figures
8
CARES Act did not pass until March 27th. But overall, the reported distribution of timing of
payments is consistent with what the IRS reports about when payments were sent.
In addition, the survey asked respondents to report whether they had received their payment
in a direct deposit (bank transfer) or as a check. As reported in Table 1, the majority of respondents
who received a payment did so via direct deposit (80 percent). While stimulus payments for these
respondents can be easily observed in credit/debit card data and similar data from financial
aggregators, check payments (20 percent) pose a potential problem for these data. Fortunately, we
observe check payments and thus can provide a more complete picture of how stimulus payments
affected households’ consumption and saving. The direct-deposit vs. check distinction is also
potentially important for saliency of stimulus payments (Sahm, Shapiro, and Slemrod, 2012).
The IRS processed payments initially via direct deposit to all those who were eligible and
had filed income taxes and received direct deposits for their refunds in 2018 or 2019, whereas checks
were mailed out only after this initial process had taken place. Consistent with this, we observe in
Figure 3 that most recipients of payments via direct deposit received them in early to mid-April,
whereas recipients of checks were staggered over a later period. We interpret this as more validation
of the quality of the data on stimulus payments received by households in our survey.
IV Effects of Stimulus Payments on Behavior
We now turn to how stimulus payments actually affected the behavior of U.S. households. We
first focus on qualitative effects on spending, saving and labor supply. We then turn to quantitative
magnitudes. Finally, we consider heterogeneity across households.
4.1 Qualitative Overview of Individual Responses to Stimulus Payments
We report in Table 2 the response of households to the qualitative question of how they used their
stimulus payment: mostly spent, mostly saved, or mostly to pay off debt. Column 1 focuses on the
90 percent of households in the survey who have already received their payment. Only 15 percent
report that they mostly spent their transfer, even lower than the corresponding values found by
Shapiro and Slemrod (2003a, 2003b) and Sahm, Shapiro and Slemrod (2010) after the 2001 and
2008 transfers respectively. One third report that they primarily saved the stimulus money, leaving
over half of respondents answering that they primarily used the Treasury transfers to pay off debt.
Column 2 presents equivalent shares for survey participants who anticipate receiving a check but
9
have not received one yet. They report very similar plans as those who have already received their
payments: only twelve percent plan to mostly spend it while more than half plan to use it primarily
to pay off debt. Even those who do not qualify for a stimulus payment report similar answers when
asked hypothetically what they would do if they received $1,000 from the Federal government: 14
percent report that they would primarily spend the money. However, a larger share of these
individuals report that they would save the money rather than pay off debts, likely reflecting the
fact that most of those who are ineligible for stimulus payments are higher-income and are less
likely to be constrained by debt levels. Jointly, these qualitative responses suggest that the stimulus
payments had only limited effects on spending.
4.2 The Marginal Propensity to Consume out of Stimulus Payments
In addition to qualitative questions, the survey asked participants to assign specific dollar values
to different ways they used (or would use) their stimulus payments, including saving, paying off
debt and different categories of spending. We report the resulting allocation of stimulus funds (as
a fraction of total checks received) by individuals in columns (1) and (2) of Table 3. On average,
households report having spent approximately 40 percent of their stimulus checks, with the
remaining sixty percent split almost evenly between saving (27 percent of stimulus) and paying
off debts (31 percent of stimulus). Relatively little of the spending went to large durable goods or
medical care (7 and 6 percent respectively). Instead, most of the spending was on food and personal
care products (16 percent) or other consumer products (13 percent).3
The large cross-sectional standard deviations in Table 3 illustrate that there was significant
variation across individuals in terms of how they allocated their stimulus funds. One way to see
this is in Figure 4, which plots the distribution of marginal propensities to consume, i.e. the fraction
of stimulus payments that was spent. Approximately 30 percent of respondents had an MPC of
one: they spent the entire stimulus check. Almost 40 percent had an MPC of zero: they did not
spend any of their stimulus payment. The remaining 30 percent of respondents had a mix of
spending and saving.
Figure 5 shows that this heterogeneity in marginal propensities extends to most categories
of using stimulus funds. For example, a little over twenty percent of respondents saved all of their
stimulus checks while over sixty percent saved none. Similarly, ≈15 percent of individuals reported
3
We report the distribution of use shares by state in Appendix Table 2.
10
that they had used all of their payment on paying down debt, while nearly sixty percent used none
at all for this purpose. Spending categories are similarly dispersed: almost 90 percent report that
they spent none of their stimulus payment on large durable goods, and twenty percent claim that
they did not spend any extra money on medical care or other consumer goods.
Another way to see this heterogeneity is to examine the ways in which stimulus payments
were allocated by households, conditional on their qualitative description of what they did (i.e.
mostly spent, mostly saved, or mostly paid off debt). We report this in columns (3)-(8) in Table 3.
The results vindicate the consistency between respondents’ qualitative and quantitative answers
while also providing some nuance. For example, individuals who claimed to mostly spend their
stimulus payment subsequently report having spent 84 percent of their transfer on average, with
most of that spending going to food/beauty/personal products and other consumer goods.
Similarly, consumers who claimed to have mostly saved their Treasury payment subsequently
reported that they used 73 percent of that payment on savings, with almost all of the remainder
going to spending on food/beauty/personal and other consumer products. Of those who claimed to
mostly pay off debt, we find that they subsequently report having used 53 percent of their transfers
to pay off debt, with most of the remaining spending going to food/beauty/personal and other
consumer products.
In short, we find that the marginal propensity to consume out of stimulus payments was
relatively low on average, a little over 40 percent. But this relatively low number masks large
variation across households: about twenty percent saved essentially all of their check, about twenty
percent used essentially all of their check to pay down debt, and around forty percent spent all of
their payment, leaving only about twenty percent who did a mix. Understanding the underlying
heterogeneity in these patterns (i.e., who spent their payment?) is therefore important for the design
of future stimulus measures.
4.3 Heterogeneity in MPCs
Which individuals tend to spend their stimulus payment? The question of identifying consumers
whose MPCs are higher matters not just for the effectiveness of stimulus packages but more
generally for understanding economic dynamics, as emphasized in Kaplan and Violante (2014)
and Kaplan, Moll and Violante (2018). To characterize the heterogeneity in propensities to spend,
11
save or pay back debt, we begin by regressing indicator variables for each on a wide range of
observable characteristics of respondents, i.e.
𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞
𝐷𝐷𝑖𝑖 = 𝛼𝛼 + 𝜷𝜷𝑿𝑿𝒊𝒊 + 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 (1)
𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞
where 𝐷𝐷𝑖𝑖 is an indicator variable equal to one when an individual responds in the qualitative
question that they mostly spent their stimulus payment and zero otherwise. We report results in
column (1) of Table 4, with columns (2) and (3) showing results where indicator variables are set
equal to one when individuals respond that they mostly saved their payment and mostly used them
to pay off debt respectively. Coefficients in the table are multiplied by a hundred and can be
interpreted as the percentage increase in the probability of mostly spending the payment (in column
1) associated with that characteristic.4
There are a number of observable factors that make it more likely that a respondent mostly
spent their stimulus payment. For example, men are a little more likely to spend their checks than
women, although the difference is small in economic terms. African Americans are much more
likely than whites to primarily pay off their debts, whereas Hispanics are less likely to pay off
debts and more likely to spend the payment. Larger households are also more likely to spend their
checks and less likely to save them, while older respondents are a little more likely to pay off debts
and a little less likely to save their payment than younger households. We also find that more
educated individuals are more likely to save their payment. Additional factors that make
individuals more likely to spend their checks include being out of the labor force and neither
owning a home nor renting (e.g. living with parents). Finally, those who received larger payments
(even after conditioning on household size) are less likely to spend their payments and more likely
to save them. Given how large the COVID-related stimulus payments were relative to those in
2001 or 2008, it is consistent with consumer theory that a larger fraction would be saved compared
to previous experiences.
Some characteristics have little effect on the likelihood that an individual says they will
mostly spend their payment but instead matters more for whether they will save their stimulus
check or use it to pay down debt. Most strikingly, liquidity constrained individuals5 are no more
likely to say they will mostly spend their transfer than non-liquidity constrained individuals.
4
Descriptive statistics for regressors are reported in Appendix Table 3.
5
Respondents are identified as liquidity constrained if they answer “no” to the following question: “Suppose that you
had to make an unexpected payment equal to one month of your after-tax income, would you have sufficient financial
resources (access to credit, savings, loans from relatives or friends, etc.) to pay for the entire amount?”
12
Instead, being liquidity constrained makes respondents much more likely to choose to pay off debts
than save the check. This result is particularly striking because liquidity constraints are typically
viewed as one of the major determinants to generating high MPCs in macroeconomic models.
Contrary to this, we find that being financially constrained primarily determines whether the extra
income is saved or used to pay off debt.
We can also perform a similar analysis using the quantitative answers about how the
payments were spent. In this case, we run:
𝑠𝑠𝑖𝑖𝑐𝑐𝑐𝑐𝑐𝑐 = 𝛼𝛼 + 𝜷𝜷𝑿𝑿𝒊𝒊 + 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 (2)
where 𝑠𝑠𝑖𝑖𝑐𝑐𝑐𝑐𝑐𝑐 is the share (scaled to be between 0 and 100) of individual i’s stimulus payment that
went to a specific category, where the possible categories include saving, paying off debt, spending
on food/beauty/personal products, spending on durables, spending on medical services, or
spending on other consumer products and services. We also report results for aggregate spending
share (the sum of the four consumer spending shares). We report the results of these regressions
in columns (1)-(7) in Table 5.6
Because these results break spending into different categories and have quantitative
measures of amounts actually saved and used to pay off debt, they provide a more nuanced view
of how different characteristics of individuals related to their use of the stimulus payment. For
example, they confirm that black households save less of their transfer on average (5% less) and
pay off more debt (3% more) than white households. They also reallocate some spending toward
food/beauty/personal products and away from other consumer products. Similarly, we see that the
increased saving of richer households relative to lower-income households reflects the fact that
they spend relatively less on food and durables.
Some of the largest differences are associated with housing, i.e. whether individuals own
a house with or without a mortgage or are renters. For example, owners with a mortgage spend
15% more of their stimulus on paying off debt than owners without a mortgage. The latter instead
save more and purchase more food/personal products as well as other consumer products.
Another source of economically large differences is liquidity constrains. The results in
columns (1) and (2) confirm that liquidity constrained individuals use less of their stimulus checks
for saving and more toward paying off debt relative to unconstrained individuals and the differences
6
We estimate specification (2) using OLS. The results are similar when we use Tobit, see Appendix Table 1.
13
are large: 22% less for saving and 12% more for paying off debt. The net difference implies 10%
more spending for liquidity constrained individuals, which is allocated mostly to additional spending
on food/beauty/personal products and medical care. These results therefore indicate that liquidity
constrained households do ultimately have higher MPCs than non-liquidity constrained households.
We also see that those who lost earnings due to the COVID pandemic use about 5% more of the
stimulus to pay off debt and 4% less to increase savings, and also spend relatively more on medical
care but less on other consumer products, suggesting that their lost earnings were preventing them
from getting all the medical services that they needed. In contrast, there is no discernible effect for
those reporting lost financial wealth due to the COVID epidemic, which is consistent with small
wealth effects on consumption. Being in lockdown in April is associated with lower spending out of
stimulus payments. Receiving stimulus payments in the form of a check rather than a bank transfer
is associated with a somewhat higher propensity to consume (consistent with a greater saliency of a
payment) but the economic magnitude is relatively small. Finally, we see that those who received
larger payments ultimately spent a somewhat smaller share of that income (on food/personal and
other consumer products) and relatively more on buying back debt and saving. This suggests that
stimulus payments to households may run into diminishing returns: households save increasingly
larger fractions of additional dollars provided by the government.
We find little role for individuals’ macroeconomic expectations (which are highly
correlated with perceptions of current macroeconomic conditions) in determining how they use
their stimulus payments. Coefficients on their perception of the current unemployment rate are
small and generally not significantly different from zero. For inflation expectations, we find no
effect of higher inflation expectations on the marginal propensity to consume, although individuals
who anticipate higher inflation are more likely to choose to pay off debts than save their extra
income. But these effects are relatively small in economic terms. In a similar spirit, expected
mortgage rate and time before conditions return to normal have relatively small economic effects
on consumer decisions. For example, a one-standard deviation increase in expected mortgage rate
is associated with a two-percentage-point higher share of stimulus payment spent consumer goods.
4.4 Work and job-search effort
Respondents were also asked about how stimulus affected (or would affect) their labor supply
decisions. To the employed, we asked whether having received the payment made them work more
14
or fewer hours (intensive margin) or led them to start or stop actively looking for new work
(extensive margin). As shown in Panel A of Table 6, the vast majority (90 percent) of those who
have already received a stimulus payment report that it had no effect on their labor supply decision,
either along the intensive or extensive margin. The same qualitative outcome obtains for those who
were still waiting for their checks or those who did not anticipate getting a check. Panel C of Table
6 also documents a similar finding for those who are out of the labor force: over 90 percent report
that stimulus payments did not induce them to start or stop looking for work nor did it change the
intensity of any job search they might have done.
We asked a similar question to the unemployed to assess whether stimulus payments affected
their job search behavior, again looking at both the intensive margin of job search (e.g. I would look
harder for a job) or the extensive margin of job search (e.g. I would stop looking for a job). Of those
who had already received their stimulus payment, about two-thirds reported that it had no effect on
their job search decision, while over 20 percent report that the stimulus check led them to search
harder for a job. Similar qualitative patterns hold for those still waiting for their stimulus payments
or for those who did not expect to receive any stimulus. In short, we find evidence that some
unemployed workers increase their job search effort in response to stimulus payments but otherwise
find no meaningful evidence of labor supply effects from the CARES Act one-time transfers.
These results have important implications for policy. A vast literature in labor economics
shows that extensions in unemployment benefits and increases in payments reduce the search
efforts of individuals (Hagedorn, Manovskii and Mitman, 2019) during normal times. Yet, little is
known about how these payments affect labor supply decisions and search efforts during a deep
recession. To the best of our knowledge, no research exits studying the effects of stimulus
payments on labor supply. Even though the replacement rates of unemployment benefits have been
well above 100% for the median household (Ganong, Noel, and Vavra 2020), our findings suggest
that additional one-time payments to households mainly stabilize aggregate demand by spurring
consumption and reducing household debt. The latter should increase the resilience of the financial
system and help reduce the risk of future financial crises. Consistent with our results, US consumer
credit card debt plunged in recent months.7
7
https://www.nytimes.com/2020/08/06/business/credit-card-debt-plunges-driving-a-decline-in-overall-household-
debt.html https://www.newyorkfed.org/microeconomics/hhdc.html
15
V Conclusion
Sending money directly to households has been one of the main components of U.S. stimulus
packages in the last three recessions. How effective is the most recent program in terms of
stimulating consumption? We provide some of the first estimates for transfers to U.S. households
put in place under the CARES Act. We find evidence that is broadly consistent with the 2001 and
2008 experiences, albeit with somewhat lower marginal propensities to consume: U.S. households
only spent around 40 percent of their stimulus payments but there is significant heterogeneity in
terms of how different individuals respond. Many spent their entire stimulus payment, and just as
many saved their entire check or used it to pay off debt.
Why were the stimulus payments not more successful in spurring consumer spending? One
possibility may be that it reflects the presence of the pandemic that caused the need for spending
in the first place. Few restaurants are operating at full capacity, many bars and shopping outlets
are closed, recreational activities are curtailed, and travel options are limited, so there is less scope
for spending on the part of consumers. Furthermore, the closing of offices and widespread
lockdowns reduce the need for transportation, which may help explain why so little spending went
to larger durable goods like cars. To the extent that the pandemic will ultimately end, it suggests
that future stimulus payments to households may be more effective in future crises.
Another, less optimistic, interpretation is that the stimulus payments were less effective
because they were larger than previous ones. As the size of one-time transfers to households rises,
diminishing returns induces individuals to consume smaller fractions of their temporarily higher
income. This suggests that there is a bound on how much stimulus can be generated through direct
transfers to households. In the face of large crises, government may want to consider a broad range
of policies targeting aggregate demand, with direct transfers being only a part of the fiscal policy
response. For example, direct government purchases of goods and services can provide an effective
stimulus, since they translate into purchases in a one-for-one manner. Another strategy can be to
increase transfers to cash-strapped local and state governments to help prevent them from cutting
the services that they provide in the midst of crisis, much as was done in the American Recovery
and Reinvestment Act of 2009.
16
References
Agarwal, Sumit, Chunlin Liu, and Nicholas S. Souleles, 2007. “The Reaction of Consumer
Spending and Debt to Tax Rebates-Evidence from Consumer Credit Data,” Journal of
Political Economy 115(6): 986-1019.
Baker, Scott R., R.A. Farrokhnia, Steffen Meyer, Michaela Pagel, and Constantine Yannelis, 2020.
“Income, Liquidity, and the Consumption Response to the 2020 Economic Stimulus
Payments,” manuscript.
Broda, Christian and Jonathan A. Parker, 2014. “The Economic Stimulus Payments of 2008 and
the aggregate demand for consumption,” Journal of Monetary Economics 68(Supplement):
S20-36.
Coibion, Olivier, Yuriy Gorodnichenko and Michael Weber, 2018. “Monetary Policy
Communications and their Effects on Household Expectations,” manuscript.
Coibion, Olivier, Yuriy Gorodnichenko and Michael Weber, 2020. “The Cost of the COVID-19
Crisis: Lockdowns, Macroeconomic Expectations, and Consumer Spending,” manuscript.
Coibion, Olivier, Yuriy Gorodnichenko and Michael Weber, 2020. “Labor Markets During the
COVID-19 Crisis: A Preliminary View,” manuscript.
Coibion, Olivier, Yuriy Gorodnichenko and Michael Weber, 2020. “Does Policy Communication
During COVID work?” manuscript.
Coibion, Olivier, Dimitris Geogarakos, Yuriy Gorodnichenko and Michael Weber, 2019.
“Forward Guidance and Household Expectations,” manuscript.
Di Maggio, Marco, Amir Kermani, and Kaveh Majlesi (forthcoming). “Stock Market Returns
and Consumption,” Journal of Finance.
Ganong, Peter, Pascal Noel, and Joe Vavra, 2020. “US Unemployment Insurance Replacement
Rates During the Pandemic,” University of Chicago, Becker Friedman Institute for
Economics Working Paper No. 2020-62.
Hagedorn, Marcus, Iourii Manovskii and Kurt Mitman, 2019. “The Impact of Unemployment
Benefit Extensions on Employment: The 2014 Employment Miracle?” manuscript.
Johnson, David S., Jonathan A. Parker and Nicholas S. Souleles, 2006. “Household Expenditure
and the Income Tax Rebates of 2001,” American Economic Review 96(5): 1589-1610.
Kaplan, Greg and Giovanni L. Violante, 2014. “A Model of the Consumption Response to Fiscal
Stimulus Payments,” Econometrica 82(4): 1199-1239.
17
Kaplan, Greg, Benjamin Moll and Giovanni L. Violante, 2018. “Monetary Policy According to
HANK,” American Economic Review 108(3): 697-743.
Lettau, Martin and Sydney C. Ludvigson, 2004. “Understanding Trend and Cycle in Asset Values:
Reevaluating the Wealth Effect on Consumption,” American Economic Review 94(1): 276-
299.
Parker, Jonathan A., Nicholas S. Souleles, David S. Johnson and Robert McClelland, 2013.
“Consumer Spending and the Economic Stimulus Payments of 2008,” American Economic
Review 103(6): 2530-2553.
Parker, Jonathan A. and Nicholas S. Souleles, 2019. “Reported Effects vs. Revealed-Preference
Estimates: Evidence from the Propensity to Spend Tax Rebates,” American Economic
Review: Insights 1(3): 273-290.
Parker, Jonathan A., 2011. “On Measuring the Effects of Fiscal Policy in Recessions,” Journal of
Economic Literature 49(3): 703-18.
Parker, Jonathan A., 2017. “Why Don’t Households Smooth Consumption? Evidence from a $25
million experiment,” American Economic Journal: Macroeconomics 9(4): 153-183.
Sahm, Claudia R., Matthew D. Shapiro, and Joel Slemrod, 2010. “Household Response to the 2008
Tax Rebates: Survey Evidence and Aggregate Implications,” Tax Policy and the
Economy 24 (2010): 69-110.
Sahm, Claudia R., Matthew D. Shapiro, and Joel Slemrod, 2012. “Check in the Mail or More in
the Paycheck: Does the Effectiveness of Fiscal Stimulus Depend on How It Is Delivered?”
American Economic Journal: Economic Policy 4: 216–250.
Shapiro, Matthew D., and Joel Slemrod, 2009. “Did the 2008 Tax Rebates Stimulate Spending?”
American Economic Review Papers and Proceedings 99 (May 2009): 374-379.
Shapiro, Matthew D., and Joel Slemrod, 2003a. “Did The 2001 Tax Rebate Stimulate Spending?
Evidence From Taxpayer Surveys,” in Tax Policy and the Economy, ed James
Poterba. Cambridge: MIT Press, 2003.
Shapiro, Matthew D., and Joel Slemrod, 2003b. “Consumer Response to Tax Rebates,” American
Economic Review 93: 381-396.
18
fraction
Le 0 .02 .04 .06 .08 .1
ss
th
$1 an
0, $1
00 0,
1 00
$2 -$ 0
0, 20
00 ,0
$3 1 00
-$
0, 30
00 ,0
$4 1 00
-$
0, 40
00 ,0
$5 1 00
-$
0, 50
00 ,0
$6 1 00
-$
0, 60
00 ,0
$7 1 00
-$
0, 70
00 ,0
$8 1 00
-$
0, 80
00 ,0
$9 1 00
0, -$
00 90
$1 1 ,0
00 -$ 00
,0 10
$1 01 0,
00
10 -$
11 0
,0
$1 01 0,
00
20 -$ 0
,0 12
01 0,
$1 00
30 -$
13 0
,0 0,
$1 01 00
40 -$
14 0
,0
$1 01 0,
00
50 -$
15 0
,0
$1 01 0,
00
60 -$
16 0
,0
$1 01 0,
00
70 -$
17 0
,0
$1 01 0,
00
80 -$
18 0
,0
01 0,
Figure 1. Distribution of income reported on the most recent tax return.
$1 00
90 -$
19 0
,0
01 0,
$1 00
98 -$
19 0
,0 8,
pr 01 00
ef an
d 0
er
Notes: the figure plots the distribution of income reported in the most recent tax return (2019 or 2018).
no ab
tt ov
do o e
no an
tr sw
em er
em
be
r
19
Figure 2. Distribution of stimulus payments from the U.S. Treasury.
.6
.4
fraction
.2
0
0
60 00 00 00 00 00 00 00
12 17 22 24 29 34 39
Notes: The figure plots the distribution of stimulus payments received by survey respondents. The maximum amount is censored at $4,200.
20
Figure 3. Distribution of the date when U.S. Treasury stimulus payments arrive.
.15 by mail
by bank transfer
.1
fraction
.05
0
1 1 1 1 ne 1
Feb
ar pr ay 1 July
M A M Ju
Notes: The figure plots the distribution of dates when a payment is received by type of payment (check or direct deposit). Stimulus payments by check account for
approximately 20 percent of all payments.
21
Figure 4. Distribution for propensity to spend (aggregate across spending categories).
.4
.3
fraction
.2
.1
0
0 20 40 60 80 100
Notes: the figure reports the distribution for the share of stimulus payment spent on durables, food, medical supplies, and other consumer products.
22
Figure 5. Expenditure share for stimulus payments (effectively marginal propensities).
Save Pay down debts Spend: durable goods
.6
.6 .8
fraction
.4
fraction
.4 fraction .4 .6
.2 .2
.2
0 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Spend: food, health/beauty aids, and HH products Spend: medical care Spend: other
.6 .8 .8
.6 .6
.4
fraction
fraction
fraction
.4 .4
.2
.2 .2
0 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Notes: each panel in the figure reports the distribution for the share of stimulus payment used for saving, paying off debts, and consumer spending.
23
Table 1. The distribution for how/if a respondent received a stimulus payment.
How/if a check was received Percent
Yes, I received it as a check by mail 16.52
Yes, I received it as a direct deposit into my bank account 64.68
No, but I expect to receive it (my payment is scheduled, my payment is delayed, etc.) 3.14
No (my household is not qualified, I don’t have the right paperwork, etc.) 10.55
Not sure, don’t know 5.11
N obs 11,793
Notes: the table reports the distribution for how/if a respondent received a stimulus payment.
Table 2. Distribution for the use of stimulus payment, qualitative.
Those who Those who expect to Those who did not receive
received a check receive a check and do not expect to
(actual spending) (planned spending) receive a check
(hypothetical spending)
(1) (2) (3)
Mostly to increase spending 15.04 12.97 14.17
Mostly to increase savings 32.80 38.38 45.76
Mostly to pay off debt 52.17 48.65 40.07
N obs 9,966 336 1,491
Notes: the table reports the distribution of qualitative responses for the use of stimulus payment.
24
Table 3. Reported shares for the uses of received stimulus payments, qualitative response.
Qualitative response
Any Mostly spend Mostly save Mostly pay debts
Spending item mean st.dev. mean st.dev. mean st.dev. mean st.dev.
(1) (2) (3) (4) (5) (6) (7) (8)
Saving 0.27 0.42 0.07 0.20 0.73 0.39 0.04 0.16
Debt payments (mortgages, auto loans, student loans, etc.) 0.31 0.40 0.10 0.23 0.06 0.19 0.53 0.41
Durable goods such as houses/apartments, cars, large 0.07 0.22 0.17 0.33 0.03 0.15 0.07 0.21
appliances and electronics
Food, health/beauty aids, and household products 0.16 0.28 0.31 0.37 0.08 0.21 0.17 0.26
Medical care (including health insurance, out-of-pocket 0.06 0.18 0.05 0.16 0.03 0.14 0.08 0.20
medical bills and prescription drugs)
Other consumer spending 0.13 0.30 0.31 0.41 0.07 0.24 0.11 0.27
Memorandum:
Total spending 0.42 0.43 0.84 0.29 0.21 0.36 0.42 0.41
Notes: the table reports the distribution for uses of stimulus check. Columns (1) and (2) report statistics for the full sample. Columns (3)-(8) report statistics for subsamples of
respondents reporting that a particular use was dominant.
25
Table 4. Propensity to spend (save) the stimulus payment, qualitative responses.
Mostly spent Mostly saved Mostly paid down debt
(1) (2) (3)
Male (gender of the respondent) 2.348** -0.190 -2.158
(1.096) (1.271) (1.352)
Hispanic -4.398*** -2.266 6.664***
(1.524) (1.905) (2.153)
Race: Black -3.668** -4.285** 7.953***
(1.424) (1.801) (2.029)
Race: Asian -0.984 5.224* -4.239
(2.409) (3.149) (3.206)
Race: Other non-white -1.535 -2.365 3.901
(1.929) (2.503) (2.765)
A child under 6 years old -2.973* 0.377 2.596
(1.763) (2.108) (2.382)
A child between 7 and 12 years old -1.402 2.021 -0.619
(1.643) (1.908) (2.159)
A child between 13 and 17 years old -1.084 -3.483* 4.567**
(1.606) (1.893) (2.183)
Household size 1.700*** -2.562*** 0.862
(0.581) (0.675) (0.747)
Age (of the respondent) -0.055 -0.045 0.100*
(0.043) (0.052) (0.056)
Household head’s education: Some college -3.718*** 1.370 2.349
(1.254) (1.469) (1.613)
Household head’s education: College+ -1.619 2.941* -1.322
(1.344) (1.596) (1.709)
Log(Household income) -0.161 4.212*** -4.051***
(0.770) (0.888) (0.977)
Housing: own, have mortgage -4.101*** -8.346*** 12.446***
(1.187) (1.517) (1.539)
Housing: rent -1.102 -6.751*** 7.852***
(1.346) (1.613) (1.707)
Housing: other arrangement 6.046** -3.551 -2.495
(2.934) (3.046) (3.418)
Employed (of the respondent) -4.108*** -0.685 4.794***
(1.099) (1.391) (1.460)
Unemployed (of the respondent) -5.076*** -4.638** 9.714***
(1.826) (2.036) (2.442)
Liquidity constrained 0.734 -26.895*** 26.162***
(1.099) (1.290) (1.469)
Liquidity constrained (prefer not to answer) -2.637* -18.457*** 21.094***
(1.590) (2.092) (2.336)
Financial wealth lost due to COVID, flag 1.760 0.220 -1.980
(1.522) (1.858) (1.902)
Earnings lost due to COVID, flag -3.089** -5.875*** 8.964***
(1.223) (1.495) (1.669)
Current unemployment rate -0.035 -0.057* 0.092***
(0.025) (0.031) (0.034)
Expected inflation over the next 12 months -0.047 -0.206** 0.253**
(0.085) (0.103) (0.115)
No stay-at-home order as of April 7, 2020 2.191 -1.140 -1.051
(2.436) (2.852) (3.004)
Stimulus payment received as a check 1.137 0.145 -1.281
(1.146) (1.398) (1.456)
Log(stimulus payment) -2.217** 3.551*** -1.334
(1.022) (1.216) (1.296)
Time to normal conditions (months) 0.007 0.022 -0.028
(0.048) (0.059) (0.064)
Expected mortgage rate -0.031 -0.006 0.037
(0.047) (0.058) (0.065)
R-squared 0.018 0.138 0.140
Notes: The table reports OLS estimates of specification (1) where the dependent variable is the part of stimulus payment used for the purposes indicated
in the column title. All coefficients are multiplied by 100. The sample includes those who received a check. Robust standard errors are in parentheses.
***, **, * denote statistical significance at 1, 5 and 10 percent. The number of observations is 9,868.
26
Table 5. Expenditure shares for the stimulus payment by spending/saving use, quantitative responses.
Food, Other
Debt Durable health/beauty Medical Total
Saving aids, and HH consumer
payments goods care spending
products spending
(1) (2) (3) (4) (5) (6) (7)
Male (gender of the respondent) 0.903 0.421 1.256* -1.837** -1.226** 0.483 -1.324
(1.167) (1.144) (0.686) (0.773) (0.495) (0.896) (1.226)
Hispanic -3.466** 4.070** -1.475 -1.890 1.046 1.716 -0.604
(1.632) (1.899) (1.003) (1.274) (0.797) (1.404) (1.894)
Race: Black -4.470*** 3.036* -0.340 4.372*** -0.040 -2.559** 1.433
(1.523) (1.669) (0.915) (1.322) (0.705) (1.082) (1.698)
Race: Asian -2.141 -3.550 0.531 6.274*** 1.484 -2.597 5.691**
(2.719) (2.600) (1.582) (2.037) (1.091) (1.836) (2.808)
Race: Other non-white -0.819 3.471 -0.846 -1.123 -0.480 -0.202 -2.652
(2.170) (2.369) (1.234) (1.535) (0.985) (1.830) (2.436)
A child under 6 years old 2.077 -0.534 -0.148 -1.529 -0.291 0.425 -1.542
(1.809) (1.960) (1.099) (1.433) (0.722) (1.317) (1.981)
A child between 7 and 12 years old 1.305 -0.288 -0.463 0.230 0.395 -1.179 -1.017
(1.608) (1.808) (0.968) (1.299) (0.717) (1.200) (1.825)
A child between 13 and 17 years old -3.668** 0.884 1.171 0.287 -0.231 1.557 2.784
(1.582) (1.817) (0.940) (1.258) (0.732) (1.304) (1.826)
Household size -1.679*** -0.358 0.582* 1.752*** -0.289 -0.008 2.037***
(0.590) (0.613) (0.340) (0.463) (0.251) (0.461) (0.641)
Age (of the respondent) 0.024 -0.112** -0.050* -0.017 0.089*** 0.065* 0.087*
(0.045) (0.046) (0.026) (0.032) (0.020) (0.035) (0.049)
Household head’s education: Some college 1.510 2.500* 0.082 -1.832* -0.070 -2.190** -4.010***
(1.287) (1.330) (0.774) (0.973) (0.598) (1.065) (1.451)
Household head’s education: College+ 3.127** 1.617 -0.515 -2.436** 0.339 -2.132* -4.744***
(1.408) (1.408) (0.800) (1.011) (0.619) (1.137) (1.536)
Log(Household income) 4.125*** 1.180 -1.233** -2.513*** 0.012 -1.571** -5.305***
(0.769) (0.795) (0.526) (0.619) (0.360) (0.617) (0.874)
Housing: own, have mortgage -5.803*** 15.130*** -0.375 -4.236*** -0.653 -4.064*** -9.327***
(1.373) (1.291) (0.711) (0.885) (0.601) (1.036) (1.408)
Housing: rent -3.582** 5.206*** 1.813** 1.286 -1.563** -3.160*** -1.624
(1.404) (1.378) (0.837) (1.034) (0.679) (1.129) (1.538)
Housing: other arrangement -3.875 -0.515 0.044 4.775** -2.067** 1.639 4.390
(2.631) (2.612) (1.520) (2.392) (1.040) (2.288) (3.106)
Employed (of the respondent) -1.769 5.513*** -0.036 -1.561* -0.531 -1.616 -3.744***
(1.243) (1.222) (0.680) (0.815) (0.547) (0.984) (1.308)
Unemployed (of the respondent) -7.376*** 8.565*** 0.300 3.360** -0.705 -4.144*** -1.189
(1.670) (2.131) (1.151) (1.581) (0.852) (1.435) (2.151)
Liquidity constrained -21.770*** 11.948*** 0.689 6.517*** 2.850*** -0.234 9.822***
(1.095) (1.242) (0.696) (0.885) (0.563) (0.871) (1.277)
Liquidity constrained (prefer not to answer) -16.657*** 9.716*** 1.178 2.680* 2.309** 0.774 6.941***
(1.722) (1.986) (1.179) (1.403) (0.971) (1.452) (2.054)
Financial wealth lost due to COVID, flag -1.712 -1.030 0.807 2.264** -0.332 0.003 2.742
(1.656) (1.600) (0.861) (1.122) (0.632) (1.170) (1.705)
Earnings lost due to COVID, flag -4.423*** 4.952*** 0.020 0.132 1.847*** -2.528*** -0.530
(1.322) (1.405) (0.767) (0.920) (0.598) (0.939) (1.411)
Current unemployment rate -0.058** 0.037 -0.018 0.033 0.013 -0.007 0.021
(0.025) (0.029) (0.016) (0.021) (0.014) (0.021) (0.030)
Expected inflation over the next 12 months -0.270*** 0.275*** 0.022 -0.092 0.023 0.043 -0.005
(0.091) (0.097) (0.057) (0.066) (0.043) (0.069) (0.099)
No stay-at-home order as of April 7, 2020 -2.555 -0.955 1.281 -2.518* -0.048 4.796** 3.511
(2.411) (2.366) (1.511) (1.517) (0.934) (2.323) (2.772)
Stimulus payment received as a check 0.922 -2.520** -0.338 0.350 -0.221 1.807* 1.598
(1.280) (1.198) (0.663) (0.874) (0.528) (1.012) (1.347)
Log(stimulus payment) 1.692*** 1.538*** -0.649 -0.741 -0.260 -1.581** -3.230***
(0.591) (0.556) (0.506) (0.482) (0.282) (0.649) (0.694)
Time to normal conditions (months) 0.007 -0.013 -0.007 -0.010 0.023 -0.001 0.005
(0.052) (0.055) (0.033) (0.038) (0.025) (0.043) (0.058)
Expected mortgage rate -0.047 -0.139*** 0.103*** 0.116*** 0.035 -0.068* 0.186***
(0.043) (0.049) (0.035) (0.042) (0.023) (0.039) (0.054)
R-squared 0.126 0.082 0.017 0.078 0.016 0.025 0.079
Notes: The table reports OLS estimates of specification (2) where the dependent variable is the share of stimulus payment used for the purposes
indicated in the column title. The sample includes those who received a check. Robust standard errors are in parentheses. ***, **, * denote statistical
significance at 1, 5 and 10 percent. The number of observations is 9,868 in each column.
27
Table 6. Work/job-search effort as a function of stimulus payment.
Expect to Did not
Received a
receive a receive a
check
check check
(1) (2) (3)
Panel A. Employed
I would start to work fewer hours 0.05 0.03 0.01
I would start to work more hours 0.04 0.11 0.04
I would start to an additional job 0.02 0.07 0.05
I would quit some of my jobs 0.00 0.05 0.01
I would quit working 0.00 0.00 0.00
The payment would not influence my work effort 0.89 0.77 0.90
N obs 5,724 179 827
Panel B. Unemployed
I would stop looking for a job 0.01 0.00 0.01
I would start looking for a job 0.07 0.08 0.09
I would delay looking for a job 0.03 0.13 0.04
I would be looking for a job but put less effort into it 0.04 0.07 0.05
I would be looking for a job and put more effort into it 0.21 0.23 0.24
The payment would not influence my decision 0.65 0.58 0.61
N obs 773 58 196
Panel C. Out of labor force
I would stop looking for a job 0.01 0.01 0.01
I would start looking for a job 0.01 0.04 0.03
I would delay looking for a job 0.01 0.00 0.01
I would be looking for a job but put less effort into it 0.01 0.02 0.02
I would be looking for a job and put more effort into it 0.01 0.08 0.06
The payment would not influence my decision 0.97 0.85 0.88
N obs 4,440 129 615
Notes: the table reports the distribution of responses for how stimulus payment affected work effort and intensity of job search. Panel
A included the employed. Panel B includes the those who do not have a paid job but are looking for a job. Panel C includes those who
do not have a paid job and do not look for a job.
28
Appendix A. Survey questions about stimulus payments
29
TS0. Please choose an amount that includes income that you reported on your latest tax return (2018 or 2019). If you do
not remember, please estimate and choose an appropriate amount. (Select one)
() Less than $10,000
() $10,001 - $20,000
() $20,001 - $30,000
() $30,001 - $40,000
() $40,001 - $50,000
() $50,001 - $60,000
() $60,001 - $70,000
() $70,001 - $80,000
() $80,001 - $90,000
() $90,001 - $100,000
() $100,001 - $110,000
() $110,001 - $120,000
() $120,001 - $130,000
() $130,001- $140,000
() $140,001 - $150,000
() $150,001 - $160,000
() $160,001 - $170,000
() $170,001 - $180,000
() $180,001 - $190,000
() $190,001 - $198,000
() $198,001 and above
() prefer not to answer
() do not remember
TS1. In Response to the COVID-19 crisis, the Congress approved a $2 trillion package to stimulate the economy,
including one-time $1,200 check per person (plus another $500 per child) to persons with annual income less than
$75,000. Couples who filed jointly and made less than $150,000 will get a one-time $2,400 check (plus another $500 per
child). The Federal government calls it an Economic Impact Payment.
Did you receive a payment from the Federal government?
() Yes, I received it as a check by mail
() Yes, I received it as a direct deposit into my bank account
() No, but I expect to receive it (my payment is scheduled, my payment is delayed, went to a wrong account/address, etc.)
() No (my household is not qualified, I don’t have the right paperwork, etc.)
() Not sure, don’t know
;ASK IF TS1= “YES”; SHOW TS2A AND TS2B ON THE SAME PAGE
TS2A. How much did you receive? Please give an approximate amount if you do not remember the exact amount.
____ dollars
TS2B. When did you receive it? Please give an approximate date if you do not remember the exact date.
____ day
____ month
;ASK If: TS1= “Yes”
;RANDOMIZE
30
TS4. How did you use the payment from the Federal government? (Select one)
() Mostly to increase spending
() Mostly to increase savings
() Mostly to pay off debt
;ASK If: TS1= “Yes”
;RANDOMIZE
TS5A. Did you spend your payment from the Federal government on any of following?
[] Saving
[] Debt payments (mortgages, auto loans, student loans, etc.)
[] Durable goods such as houses/apartments, cars, large appliances and electronics
[] Food, health/beauty aids, and household products
[] Medical care (including health insurance, out-of-pocket medical bills and prescription drugs)
[] Other consumer spending >ANCHOR
;PIPE RESPONSES SELECTED AT TS5A INTO TS5 IN SAME RANDOM ORDER
TS5. Out of XXX dollars (REPORTED IN TS3A), how much did you spend on the items you just chose? Please report
dollar amounts for each of the individual components listed below. The total should be approximately equal to XXX
dollars.
[] Saving
[] Debt payments (mortgages, auto loans, student loans, etc.)
[] Durable goods such as houses/apartments, cars, large appliances and electronics
[] Food, health/beauty aids, and household products
[] Medical care (including health insurance, out-of-pocket medical bills and prescription drugs)
[] Other consumer spending
Total ____ [automatically calculated]
;ASK If: TS1= “yes” and “has paid job”= “No”
TS10. How did the payment from the Federal government influence your decision to look for a job or not? Please choose
options that apply to you. Because of the payment, [please choose all relevant options]
[] I stopped looking for a job
[] I started looking for a job
[] I delayed looking for a job
[] I am looking for a job but put less effort into it
[] I am looking for a job and put more effort into it
[] The payment did not influence my decision >EXCLUSIVE
;ASK If: TS1= “yes” and “has paid job”= “Yes”
TS11. How did the payment from the Federal government influence your work effort? Please choose options that apply to
you. Because of the payment,
[] I started to work fewer hours
[] I started to work more hours
[] I started to work an additional job
[] I quit some of my jobs
[] I quit working
[] The payment did not influence my work effort >EXCLUSIVE
31
;ASK If: TS1= “No, but I expect to receive it”. Show TS3A and TS3B on the same page
TS3A. How much do you expect to receive?
____ dollars
TS3B. When do you expect to receive it? Please give an approximate date if you do not remember the exact date.
____ day
____ month
;ASK If: TS1= “No, but I expect to receive it”
;RANDOMIZE
TS7. How do you plan to use the payment from the Federal government? (Select one)
() Mostly to increase spending
() Mostly to increase savings
() Mostly to pay off debt
;ASK If: TS1= “No, but I expect to receive it” and “has paid job”= “No”
TS12. How would the payment from the Federal government influence your decision to look for a job or not? Please
choose options that apply to you. Because of the payment,
[] I would stop looking for a job
[] I would start looking for a job
[] I would be looking for a job
[] I would be looking for a job but put less effort into it
[] I would be looking for a job and put more effort into it
[] The payment would not influence my decision >EXCLUSIVE
;ASK If: TS1= “No, but I expect to receive it” and “has paid job”= “Yes”
TS13. How would the payment from the Federal government influence your work effort? Please choose options that apply
to you. Because of the payment,
[] I would start to work few hours
[] I would start to work more hours
[] I would start to an additional job
[] I would quit some of my jobs
[] I would quit working
[] The payment would not influence my work effort >EXCLUSIVE
;ASK If: TS1= “No” OR TS1= “Do not know”
;RANDOMIZE
TS8. If you were to receive a one-time $1,000 payment from the government, how would you use it? (Select one)
() Mostly to increase spending
() Mostly to increase savings
() Mostly to pay off debt
32
;ASK If: (TS1= “No” OR TS1= “Do not know” ) and “has paid job”= “No”
TS14. If you were to receive a payment from the Federal government, how would it influence your decision to look for a
job or not? Please choose options that apply to you. Because of the payment,
[] I would stop looking for a job
[] I would start looking for a job
[] I would delay looking for a job
[] I would be looking for a job but put less effort into it
[] I would be looking for a job and put more effort into it
[] The payment would not influence my decision >EXCLUSIVE
;ASK If: (TS1= “No” OR TS1= “Do not know” ) and “has paid job”= “Yes”
TS15. If you were to receive a payment from the Federal government influence, how would it influence your work effort?
Please choose options that apply to you. Because of the payment,
[] I would start to work few hours
[] I would start to work more hours
[] I would start to an additional job
[] I would quit some of my jobs
[] I would quit working
[] The payment would not influence my work effort >EXCLUSIVE
33
Appendix B. Additional Figures and Tables
34
Appendix Table 1. Expenditure shares for the stimulus payment by spending/saving use, quantitative responses, Tobit, marg. effects.
Food, Other Total
Debt Durable health/beaut Medical spending
Saving y aids, and consumer
payments goods care
HH products spending
(1) (2) (3) (4) (5) (6) (7)
Male (gender of the respondent) 3.757*** 3.392*** 5.174*** 1.345* 1.167** 4.192*** -0.482
(1.002) (0.998) (0.731) (0.738) (0.519) (0.859) (1.073)
Hispanic -0.323 4.047*** 0.706 0.194 2.625*** 4.724*** 0.121
(1.520) (1.453) (0.945) (1.046) (0.821) (1.334) (1.520)
Race: Black 0.170 5.429*** 3.698*** 7.200*** 3.181*** 1.956* 2.512*
(1.410) (1.336) (0.935) (1.140) (0.761) (1.122) (1.370)
Race: Asian 0.760 -2.122 2.553 7.871*** 3.041** -0.788 5.530**
(2.193) (2.125) (1.631) (1.993) (1.318) (2.036) (2.407)
Race: Other non-white 0.482 3.881** 0.978 1.566 0.470 1.800 -0.972
(2.040) (1.815) (1.212) (1.271) (0.917) (1.593) (1.962)
A child under 6 years old 3.489** 0.396 0.596 -0.182 1.042 2.376* -1.009
(1.750) (1.432) (0.907) (1.095) (0.821) (1.401) (1.589)
A child between 7 and 12 years old 1.867 0.545 0.772 0.688 1.417* 0.747 -0.563
(1.584) (1.344) (0.874) (1.016) (0.761) (1.237) (1.466)
A child between 13 and 17 years old -2.334 1.622 2.100** 2.116** 1.221 3.464*** 3.027**
(1.488) (1.336) (0.915) (1.030) (0.767) (1.326) (1.462)
Household size -1.718*** -0.161 1.006*** 2.084*** 0.068 0.084 2.184***
(0.562) (0.493) (0.302) (0.375) (0.238) (0.434) (0.547)
Age (of the respondent) -0.046 -0.173*** -0.126*** -0.045 0.077*** 0.052 0.068
(0.040) (0.039) (0.024) (0.028) (0.019) (0.032) (0.042)
Household head’s education: Some college -2.900** -0.161 -2.663*** -2.541*** -1.791*** -4.812*** -3.006**
(1.185) (1.110) (0.606) (0.745) (0.486) (0.799) (1.192)
Household head’s education: College+ -4.617*** -4.013*** -7.686*** -7.685*** -4.910*** -10.052*** -5.974***
(1.231) (1.185) (0.654) (0.825) (0.537) (0.884) (1.306)
Log(Household income) 7.010*** 3.212*** 1.188*** -0.661 1.614*** 1.146** -4.451***
(0.792) (0.683) (0.448) (0.488) (0.368) (0.558) (0.737)
Housing: own, have mortgage -5.182*** 16.386*** -0.558 -3.147*** -0.380 -3.633*** -8.449***
(1.096) (1.219) (0.784) (0.847) (0.554) (0.910) (1.255)
Housing: rent -2.327* 7.277*** 1.936** 2.333** -0.687 -2.048** -1.546
(1.227) (1.311) (0.825) (0.909) (0.591) (0.949) (1.324)
Housing: other arrangement -2.817 3.318 0.809 4.610** -1.192 3.764* 2.534
(2.508) (2.415) (1.461) (1.919) (0.966) (1.922) (2.597)
Employed (of the respondent) -1.549 5.716*** -0.692 -2.044*** -1.117** -2.743*** -4.655***
(1.085) (1.039) (0.666) (0.732) (0.500) (0.880) (1.138)
Unemployed (of the respondent) -8.338*** 7.399*** 0.502 2.687** -0.832 -3.535*** -1.864
(1.807) (1.754) (0.995) (1.233) (0.744) (1.258) (1.758)
Liquidity constrained -21.666*** 12.135*** 0.930 7.960*** 4.335*** 0.068 9.958***
(0.935) (1.039) (0.641) (0.778) (0.556) (0.833) (1.079)
Liquidity constrained (prefer not to answer) -13.113*** 9.711*** 0.766 3.610*** 2.196** 0.195 6.848***
(1.288) (1.757) (1.041) (1.282) (0.990) (1.327) (1.773)
Financial wealth lost due to COVID, flag -0.174 -0.090 1.734* 1.881* 0.737 0.394 1.989
(1.366) (1.341) (1.001) (1.110) (0.665) (1.246) (1.505)
Earnings lost due to COVID, flag -4.939*** 3.982*** -0.571 1.730** 2.788*** -2.196** 0.857
(1.203) (1.114) (0.704) (0.820) (0.625) (0.960) (1.198)
Current unemployment rate -0.069** 0.028 -0.006 0.037** 0.017 -0.014 0.032
(0.027) (0.023) (0.015) (0.016) (0.011) (0.019) (0.026)
Expected inflation over the next 12 months -0.250*** 0.246*** 0.006 -0.035 0.034 0.084 0.002
(0.085) (0.079) (0.051) (0.056) (0.039) (0.066) (0.085)
No stay-at-home order as of April 7, 2020 -2.715 -2.253 1.824 -1.885 -0.109 3.898* 4.203*
(2.149) (2.038) (1.548) (1.470) (0.981) (2.110) (2.506)
Stimulus payment received as a check 0.225 -3.157*** -0.445 -0.235 -0.445 1.399 1.207
(1.125) (1.084) (0.675) (0.804) (0.504) (0.924) (1.209)
Log(stimulus payment) 2.788*** 0.727 -1.415** -2.240*** -0.325 -0.895 -3.989***
(0.985) (0.896) (0.566) (0.771) (0.470) (0.820) (1.042)
Time to normal conditions (months) 0.024 -0.004 0.008 -0.004 0.030 0.008 -0.014
(0.049) (0.046) (0.031) (0.033) (0.022) (0.038) (0.050)
Expected mortgage rate -0.060 -0.076* 0.089*** 0.082*** 0.041** -0.043 0.136***
(0.053) (0.040) (0.022) (0.028) (0.019) (0.036) (0.044)
Observations 9,868 9,868 9,868 9,868 9,868 9,868 9,868
Notes: The table reports marginal effects for Tobit estimates of specification (2) where the dependent variable is the share of stimulus payment used
for the purposes indicated in the column title. The sample includes those who received a check. Robust standard errors are in parentheses. ***, **, *
denote statistical significance at 1, 5 and 10 percent.
35
Appendix Table 2. Reported shares for uses of stimulus payments, qualitative response.
Food,
Other
Debt Durable health/beauty Medical Total
State Saving consumer
payments goods aids, and HH care spending
spending
products
(1) (2) (3) (4) (5) (6) (7)
Alabama 0.26 0.24 0.07 0.24 0.06 0.13 0.50
Arizona 0.23 0.32 0.10 0.16 0.06 0.13 0.45
Arkansas 0.26 0.32 0.10 0.14 0.04 0.15 0.43
California 0.27 0.31 0.06 0.17 0.05 0.14 0.42
Colorado 0.27 0.29 0.07 0.12 0.06 0.19 0.44
Connecticut 0.31 0.37 0.05 0.12 0.05 0.09 0.32
Delaware 0.30 0.31 0.06 0.20 0.03 0.11 0.39
District of Columbia 0.40 0.41 0.05 0.04 0.00 0.08 0.18
Florida 0.28 0.33 0.06 0.16 0.06 0.10 0.39
Georgia 0.29 0.32 0.06 0.16 0.04 0.13 0.39
Idaho 0.29 0.20 0.08 0.14 0.11 0.19 0.51
Illinois 0.29 0.33 0.07 0.15 0.06 0.09 0.38
Indiana 0.24 0.32 0.08 0.16 0.06 0.14 0.44
Iowa 0.24 0.33 0.09 0.09 0.04 0.21 0.43
Kansas 0.33 0.25 0.11 0.13 0.08 0.10 0.42
Kentucky 0.24 0.24 0.10 0.22 0.04 0.15 0.52
Louisiana 0.17 0.33 0.09 0.26 0.03 0.12 0.50
Maine 0.30 0.34 0.02 0.15 0.04 0.15 0.36
Maryland 0.36 0.23 0.08 0.16 0.05 0.12 0.41
Massachusetts 0.31 0.28 0.06 0.20 0.04 0.10 0.41
Michigan 0.32 0.27 0.08 0.14 0.07 0.13 0.41
Minnesota 0.27 0.27 0.07 0.13 0.07 0.20 0.46
Mississippi 0.13 0.35 0.07 0.23 0.07 0.15 0.52
Missouri 0.28 0.29 0.07 0.18 0.06 0.12 0.42
Montana 0.33 0.21 0.07 0.18 0.11 0.10 0.46
Nebraska 0.24 0.33 0.07 0.19 0.07 0.10 0.43
Nevada 0.24 0.33 0.12 0.17 0.08 0.07 0.43
New Hampshire 0.20 0.38 0.07 0.18 0.04 0.13 0.42
New Jersey 0.29 0.36 0.05 0.15 0.04 0.10 0.34
New Mexico 0.30 0.27 0.07 0.15 0.06 0.15 0.44
New York 0.29 0.31 0.07 0.19 0.04 0.10 0.40
North Carolina 0.24 0.30 0.07 0.20 0.07 0.12 0.46
North Dakota 0.43 0.08 0.08 0.14 0.03 0.25 0.49
Ohio 0.27 0.28 0.10 0.14 0.05 0.16 0.45
Oklahoma 0.23 0.30 0.08 0.12 0.10 0.17 0.47
Oregon 0.30 0.32 0.06 0.11 0.08 0.13 0.38
Pennsylvania 0.28 0.30 0.07 0.15 0.05 0.16 0.42
Rhode Island 0.15 0.42 0.07 0.16 0.06 0.13 0.43
South Carolina 0.30 0.30 0.04 0.15 0.10 0.11 0.40
South Dakota 0.20 0.34 0.04 0.19 0.11 0.11 0.45
Tennessee 0.28 0.34 0.07 0.15 0.06 0.10 0.38
Texas 0.27 0.34 0.06 0.16 0.06 0.10 0.39
Utah 0.35 0.32 0.13 0.11 0.03 0.06 0.34
Vermont 0.23 0.32 0.05 0.24 0.07 0.09 0.45
Virginia 0.22 0.37 0.07 0.15 0.06 0.13 0.41
Washington 0.23 0.30 0.09 0.18 0.05 0.14 0.47
West Virginia 0.21 0.33 0.07 0.21 0.08 0.09 0.46
Wisconsin 0.33 0.33 0.05 0.13 0.05 0.11 0.34
Wyoming 0.30 0.26 0.25 0.13 0.06 0.00 0.44
Notes: the table reports the distribution for uses of stimulus check by state. Each column shows the share spent on the use indicated in
the column title.
36
Appendix Table 3. Descriptive statistics.
Mean St.Dev.
Male (gender of the respondent) 0.33 0.47
Hispanic 0.14 0.35
Race: Black 0.13 0.33
Race: Asian 0.05 0.22
Race: Other non-white 0.08 0.27
A child under 6 years old 0.11 0.31
A child between 7 and 12 years old 0.14 0.35
A child between 13 and 17 years old 0.14 0.35
Household size 2.49 1.31
Age (of the respondent) 50.79 15.15
Household head’s education: Some college 0.31 0.46
Household head’s education: College+ 0.42 0.49
Log(Household income) 10.86 0.82
Housing: own, have mortgage 0.37 0.48
Housing: rent 0.31 0.46
Housing: other arrangement 0.05 0.21
Employed (of the respondent) 0.54 0.50
Unemployed (of the respondent) 0.08 0.27
Liquidity constrained 0.32 0.47
Liquidity constrained (prefer not to answer) 0.09 0.29
Financial wealth lost due to COVID, flag 0.13 0.34
Earnings lost due to COVID, flag 0.18 0.38
Current unemployment rate 22.45 19.46
Expected inflation over the next 12 months 2.79 5.32
No stay-at-home order as of April 7, 2020 0.04 0.18
Log(stimulus payment) 7.30 0.53
Time to normal conditions (months) 11.08 9.14
Expected mortgage rate 8.27 11.44
37
File and source
- File
- how-did-u-s-consumers-use-their-stimulus-payments.pdf
- Size
- 345,454 bytes
- SHA-256
- 2daac389f165409d1bbe1edeea4d95d46e13d2554bc76b95d91a7d3d59e1a7d7
- Original
- www.nber.org