Buffington et al., High Frequency Business Dynamics in the United States During the COVID-19 Pandemic (2021-03)
Summary
A Center for Economic Studies research paper, CES 21-06, dated March 2021, by Catherine Buffington, Daniel Chapman, Emin Dinlersoz, Lucia Foster and John Haltiwanger, on business dynamics during the COVID-19 pandemic. It draws on the Small Business Pulse Survey (SBPS) and Business Formation Statistics (BFS) of the U.S. "Census Bureau". The paper reports that existing small businesses showed sharp declines early in the pandemic and negative indicators through the first week of January 2021, with Accommodation and Food Services hit harder than Finance and Insurance. It states that about 74% of small businesses in the SBPS received PPP loans. It reports a surge in new business applications, especially in Retail Trade, and calls 2020 the highest year on record for new business formations.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“Buffington et al., High Frequency Business Dynamics in the United States During the COVID-19 Pandemic (2021-03).”
Original at www2.census.gov. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/buffington-et-al-high-frequency-business-dynamics-in-the-united-states-ca9d47565e/
, SHA-256 56de17433753f2a56e42b2016091b882eb6281d8bb95cfa70328439c81dee89c.Full text
High Frequency Business Dynamics in the United States
During the COVID-19 Pandemic
by
Catherine Buffington
U.S. Census Bureau
Daniel Chapman
U.S. Census Bureau and University of Maryland
Emin Dinlersoz
U.S. Census Bureau
Lucia Foster
U.S. Census Bureau
John Haltiwanger
University of Maryland
CES 21-06 March 2021
The research program of the Center for Economic Studies (CES) produces a wide range of
economic analyses to improve the statistical programs of the U.S. Census Bureau. Many of these
analyses take the form of CES research papers. The papers have not undergone the review
accorded Census Bureau publications and no endorsement should be inferred. Any opinions and
conclusions expressed herein are those of the author(s) and do not necessarily represent the
views of the U.S. Census Bureau. All results have been reviewed to ensure that no confidential
information is disclosed. Republication in whole or part must be cleared with the authors.
To obtain information about the series, see www.census.gov/ces or contact Christopher Goetz,
Editor, Discussion Papers, U.S. Census Bureau, Center for Economic Studies 5K038E, 4600 Silver
Hill Road, Washington, DC 20233, CES.Working.Papers@census.gov. To subscribe to the series,
please click here.
Abstract
Existing small businesses experienced very sharp declines in activity, business sentiment, and
expectations early in the pandemic. While there has been some recovery since the early days of
the pandemic, small businesses continued to exhibit indicators of negative growth, business
sentiment, and expectations through the first week of January 2021. These findings are from a
unique high frequency, real time survey of small employer businesses, the Census Bureau’s Small
Business Pulse Survey (SBPS). Findings from the SBPS show substantial variation across sectors
in the outcomes for small businesses. Small businesses in Accommodation and Food Services
have been hit especially hard relative to those Finance and Insurance. However, even in Finance
and Insurance small businesses exhibit indicators of negative growth, business sentiment, and
expectations for all weeks from late April 2020 through the first week of 2021. While existing
small businesses have fared poorly, after an initial decline, there has been a surge in new business
applications based on the high frequency, real time Business Formation Statistics (BFS). Most of
these applications are for likely nonemployers that are out of scope for the SBPS. However, there
has also been a surge in new applications for likely employers. The surge in applications has been
especially apparent in Retail Trade (and especially Non-store Retailers). We compare and
contrast the patterns from these two new high frequency data products that provide novel
insights into the distinct patterns of dynamics for existing small businesses relative to new
business formations.
*
*
Any opinions and conclusions expressed herein are those of the authors and do not represent the views of the
U.S. Census Bureau. All results in this paper use publicly available data from Census Bureau websites. The Census
Bureau has reviewed the public domain products for unauthorized disclosure of confidential information and has
approved the disclosure avoidance practices applied to this release (Approval ID: CBDRB-FY20-259, CBDRB-FY-20-
357, CBDRB-FY-20-214, CBDRB-FY21-094). We thank John Eltinge, Scott Ohlmacher, and Nick Orsini for helpful
comments.
1. Introduction
In the early stages of the COVID-19 pandemic in the United States, the Census Bureau
introduced new data products designed to capture its impacts on the American economy and
population (see Buffington et al. (2021)). These data products were designed to provide
geographically granular, timely, and high frequency information about the impact of the COVID-
19 pandemic on businesses and households. We describe the results from two new business data
products: the Small Business Pulse Survey (SBPS) and the weekly and monthly Business
Formation Statistics (BFS). 2
We report results from the first three phases of the SBPS (April-June, August-October,
and November-January) and there are distinct patterns in each of the three phases. The SBPS
shows that existing small businesses experienced very sharp declines in activity, business
sentiment, and expectations early in the pandemic. While the indicators of net growth, business
sentiment, and expectations have become less negative since the early weeks of the pandemic,
the indicators are still in a substantially negative range by the first week of January 2021. This
improvement has not been monotonic: conditions improve in the Summer of 2020 but worsen at
the end of 2020. 3 The patterns for new business formations from the BFS are in sharp contrast
to those for existing small businesses. After an initial decline, there has been a strong surge in
new business applications for both likely employers and nonemployers. For overall applications,
2020 is the highest year on record for new business formations with most of this surge occurring
in the second half of 2020.
Both the SBPS and BFS show substantial sectoral variation. Existing small businesses in
some service sectors have been especially hard hit. The SBPS provides insights into changes in
revenue, employment, and hours in these sectors. In order to illuminate these differences, we
focus on two extremes of the sectoral experience in the SBPS section of the paper: Finance and
Insurance and Accommodation and Food Services. We show differences in the changes in
revenue and employment as well as differences in their future needs and expectations of a return
to normal level of operations. Existing small businesses in Accommodation and Food Services
2
The SBPS public-use data are available at https://census.gov/businesspulsedata. The BFS public-use data are
available at https://www.census.gov/econ/bfs/data.html. Related recent research using the BFS can be found in
Dinlersoz et. al. (2021) and Haltiwanger (2021).
3
Section 2.1 discusses improvements in business sentiment and revenues; section 2.2 discusses improvements in
employment and hours.
2
h0ave been hit much harder than those in Finance and Insurance. This sectoral difference is
consistent with revenue published by the Census Bureau as part of the Monthly Retail Trade
Survey (MRTS) and Quarterly Survey of Services (QSS): by the end of 2020, Finance and
Insurance is close to pre-pandemic levels, but Accommodation and Food Services remain below
pre-pandemic levels. 4 This difference is also consistent with heterogeneity in key national
indicators published by Bureau of Labor Statistics: Finance and Insurance grew by about 50,000
jobs over 2020 while Accommodation and Food Services shrank by about 3 million jobs.5
However, existing small businesses in Finance and Insurance show no indicators of positive net
growth, business sentiment, and expectations. Apparently, even in the sectors that show
recovery for all types of businesses there has not yet been a recovery for existing small
businesses in those sectors.
The SBPS offers a rich range of indicators for the impact on small businesses. We find
that the adverse impact on revenue is greater than the impact on employment. Partly this is
accounted for by a decrease in hours rather than in employment, but the patterns suggest that
jobs have been less adversely impacted than revenue for existing small businesses. The
differential impact by sector and geography is also reflected in how existing small businesses
change their way of doing business. There has been a large increase in remote work activity, but
this varies by sectors (for example, businesses in Accommodation and Food Services exhibit the
smallest increase in remote work). Some small businesses pivoted to producing new goods and
services early in the pandemic (especially in Educational Services; Accommodation and Food
Services; and Information) and even more pivoted to new modes of delivery. Cash on hand is
limited, highlighting the precarious nature of small business operations. This is echoed in the
percentage of businesses noting the need to obtain financial assistance or additional capital in the
next six months. Federal assistance is an important source of support with about 74% of small
businesses in the SBPS receiving PPP loans. Reflecting these concerns and challenges, business
expectations for their return to their normal level of operations increasingly focus on the longer
horizon.
4
Finance and Insurance revenues in 2020:Q4 are 1,226,517 almost at the same levels as in 2019:Q4 at 1,227,080.
Food Services and Drinking Places (NAICS 722) revenues in December 2020 are 51,346 as compared to 65,085 in
December 2019. Revenue in Accommodation (NAICS 721) is 38,681 in 2020:Q4 as compared to 68,574 in
2019:Q4. (All in millions of dollars and not seasonally adjusted.)
5
BLS January 2021 Employment Situation.
3
Turning to business applications, there has been an enormous increase in applications in
the Retail Trade sector, particularly in Non-store Retail Trade, which includes e-commerce
business activity. 6 The patterns of sectoral changes differ for business applications that are
likely to lead to a business with employees (“employers”) and business applications that are
likely to lead to a business without employees (“nonemployers”). New applications in the Retail
Trade and Professional, Scientific, and Technical sectors have surged for both types of
applications. In the SBPS, existing small businesses in Retail Trade have been hit about as hard
as those in Accommodation and Food Services. The dispersion in response at the sectoral level,
as well as evidence that sectors hit hard for existing small businesses are seeing a surge in
applications for new businesses, suggest that the pandemic may lead to lasting structural changes
in the economy.
The pandemic has differential impacts on existing small businesses and new business
formations in terms of geography. For existing small businesses, the differential impact across
states diminishes over time. Moreover, there is less persistence in state rankings of business
sentiment than we see in sectoral rankings. The initial decline in business application activity is
concentrated in the Northeast and the West, but business application recovery early is
particularly emphasized in the South and the Midwest. The recovery was very uneven across
states, with the disparity in year-over-year business application growth rates becoming more
dispersed during the recovery.
The paper proceeds as follows. We provide results on small business using the SBPS in
Section 2. We provide results on business applications and potential new employer business
startups in 2020 using the BFS in Section 3. We conclude with some discussion of future
research areas.
2. Small Business Pulse Survey
The SBPS complements existing Census Bureau surveys conducted at higher frequencies
that are also intended to measure the impact of the pandemic on businesses. As its name
suggests, the SBPS focuses on small businesses. As context, there are about 5.3 million employer
firms in the U.S. in 2018 consisting of 7.1 million establishments and 130 million workers.
6
See North American Industry Classification System (NAICS) U.S. Census Bureau.
4
While the number of “small” businesses (defined here as less than 500 employees) dominates the
total number of establishments and firms (81% and more than 99%, respectively), the share of
workers at “small” firms so defined is about 47%. 7 Thus, the target pool of businesses is an
important part of the economy. Moreover, many COVID-19 relief policies targeted businesses of
this size.
The SBPS sample is drawn from the 2017 Economic Census for all single-location
businesses with 1-499 employees and $1000+ revenue that report an email address. 8 The survey
divides the sample equally across the nine weeks of each phase of the survey. Thus, each week
of the survey, about 100,000 emails are sent out to businesses asking them to participate in the
survey. Each of the nine subsamples is used only once each phase but are reused across phases
(thus some businesses may have responded three times, once in each phase). With a response
rate of about 25%, there are approximately 25,000 businesses responding each week. The survey
is designed to be representative and the results are re-weighted to be nationally representative. 9
The focus on single-location businesses allows the Census Bureau to publish detailed
geographic (at state and selected MSA levels) information. The focus on small single-locations
businesses makes it more likely that the business is operating in one sector which facilitates
publishing detailed sectoral (at 2-digit NAICS and selected 3-digit NAICS) information. It is
important to provide granular information to accurately capture the heterogenous impact of the
pandemic on businesses.
As noted above, the sample of the SBPS complements other existing Census Bureau
surveys conducted at higher frequencies. This is also true for its content. The trade-off is that the
SBPS is higher frequency and more timely, but less comprehensive in terms of coverage than the
Monthly Retail Trade Survey (which is not as frequent or timely). Many of the questions in the
SBPS have checkbox responses that capture only whether a business is impacted and not the size
of the impact. The survey has an internal narrative structure that leads the respondent through the
questions; the substantive section of the survey starts by asking about the overall effect of the
7
From 2018 Business Dynamics Statistics. There are also about 26 million non-employer businesses in 2018
(Nonemployer Statistics). County Business Patterns reports the following for 2018: 7.9 million establishments and
131 million employees. The BDS counts of establishments are lower due to the BDS’ use of longitudinal
information to handle potential double-counting.
8
As noted in Buffington et al. (2020) the reliance on email was due to the operational constraints due to the
pandemic.
9
See Buffington et al. (2020) and the SBPS website Small Business Pulse Survey Data (census.gov) for a more
complete discussion of the methodology.
5
pandemic on the business, then asks more detailed questions about operations, challenges,
finances (including requests for and receipts of federal assistance), and closes with a question
about the future analogous to the opening question.
The survey content was developed in partnership with eight other agencies and interested
stakeholders. 10 Their subject matter expertise was especially helpful in developing content useful
for policymakers as they develop programs intended to assist small businesses. The content of
the survey varies over Phase 1 and Phases 2-3, but there are about 20 questions on the survey (as
shown in Table 1). A guiding principle is to maintain consistency across phases where possible
to allow for longitudinal analysis. However, questions are edited, added, or dropped as analysis
of results and changing conditions warrants (these are discussed more below).
One critical feature of the SBPS is that it requests confirmation of the respondent’s
Employer Identification Number (EIN,) the last four digits of which are pre-printed in the
invitation to respond. If the provided EIN does not match that for the company, the respondent is
asked to provide the correct EIN. The EIN enables the use of existing Census Bureau
information on sector and geography, allowing survey information to be released by sector and
geography while minimizing respondent burden. This also means that the SBPS survey results
can be linked to other Census Bureau collections at the micro-level (which may allow us, for
example, to provide information concerning owner demographics collected on other surveys).
Another important feature of the SBPS is the time frame of the questions. Many of the
questions are designed to capture changes in behavior over the weeks of the pandemic and thus
ask about changes “in the last week.” For example, questions concerning changes in hours,
employment, and revenue use this time frame. This time frame allows us to capture flows of
changes, but it means that we will miss changes that precede the week in question. The survey
started in late April 2020, so some of the businesses’ changes in response to the pandemic may
have already taken place. 11 Other questions are designed to capture changes in behavior since the
start of the pandemic and thus ask about concepts “Since March 13, 2020.” Questions related to
10
The Small Business Administration, the Federal Reserve Board of Governors, International Trade Administration,
Minority Business Development Agency, Bureau of Transportation Statistics, National Telecommunications and
Information Administration, Office of Tax Analysis of the Treasury Department, and the Bureau of Labor Statistics.
See Buffington et al. (2020) for more information.
11
Bartik et al. (2020a) conducted a survey of about 5,800 small businesses March 28-April 4 and report 43% of
businesses had temporarily closed (and most were due to COVID-19) and on average businesses reported having
reduced their employment by 40% since January.
6
federal programs (requests for and receipt of federal assistance) use this time frame. Finally, the
two more holistic questions that bookend the survey have more open-ended time frames.
A number of other surveys have been conducted during the COVID-19 pandemic to
capture some of these same concepts. We highlight just a few here and cite to relevant results
throughout the empirical sections below. First, BLS conducted one phase of the Business
Response Survey (BRS) from July-September 2020 asking businesses about the impact of the
pandemic and their responses through seven questions over three areas: “1) business experiences
and payroll decisions, 2) worker benefits and their ability to telework, and 3) whether a business
received a loan or a grant from the government tied to the payroll.” 12 Second, Bartik et al. (2020
conducted a survey early in the pandemic of about 5,800 small businesses through Alignable
asking about impact, actions, and expectations. Third, the National Federation of Independent
Businesses (NFIB) has conducted fifteen waves of COVID-19 related surveys from a random
sample of their 300,000 small business members (usually with less than 1000 responses). In
addition to conducting surveys, other research relies upon novel uses of administrative,
commercial, and sensor data to track the impact of COVID-19 as we note in some examples.
As noted above, many of the empirical exercises focus on two sectors, Finance and
Insurance and Accommodation and Food Services, as exemplifying the differential impact of the
pandemic. As background, the latest available figures from County Business Patterns (for 2019)
show that there are about 476,000 establishments in Finance and Insurance employing 6.6
million people (see Table 2). There are about 741,000 establishments in Accommodation and
Food Services employing 14.5 million people. Since our sample is for small single-location
businesses, the table also provides some summary statistics for businesses with fewer than 500
employees (for 2018). More than 99% of establishments in either sector are small businesses;
69% of employees in Finance and Insurance work in small businesses and 95% in
Accommodation and Food Services do as well. More recent data from BLS shows that pre-
pandemic (January 2020), there were 6.5 million people working in Finance and Insurance and
13.9 million working in Accommodation and Food Services. As noted in the Introduction, the
number of workers in January 2021 increased by about 50,000 in Finance and Insurance and
decreased by about 3 million in Accommodation and Food Services.
12
See BLS’ website on the BRS: https://www.bls.gov/brs/methods/technical-notes.htm.
7
Before turning to the results, we first discuss caveats, some of which are general issues
with survey data and some of which are specific to the way the SBPS is conducted. Buffington et
al. (2020) discuss seven important considerations when analyzing the results of the SBPS.
Briefly these are: (1) non-response bias for business closures; (2) reliance on businesses that
have (and are willing to share) email addresses; (3) willingness of businesses to participate
through email invitation; (4) sample reliance on single-location businesses that might respond
differently than multi-locations businesses; (5) willingness of businesses that sought federal
assistance to respond to a survey from a federal source; (6) greater likelihood of responses from
businesses adversely impacted by the pandemic; and (7) representation of young businesses in
the sample. In the sectoral comparisons that follow we drop the Agriculture; Mining, Quarrying,
and Oil and Gas Extraction; Utilities; and Management of Companies and Enterprises sectors
due to the prevalence of suppressed cells.
We now turn to highlighting some results from the first three phases of the SBPS. For
ease of exposition, the discussion roughly follows the order in which the questions appear on the
survey. We focus on temporal and sectoral variation for a detailed analysis of the results in
subsections 2.1-2.6 before providing an analysis of results by geographic variation in subsection
2.7. 13
2.1 Business Sentiment and Revenues
The opening question asks the respondent, “Overall, how has this business been affected
by the Coronavirus pandemic?” The five checkboxes for responses range from large negative,
negative, no effect, positive, to large positive. The phrasing of the question is intended to capture
a holistic assessment of the impact and the time frame is open-ended. These responses are
summarized in the Overall Sentiment Index (OSI) with values ranging from -1 (most negative
impact) to +1 (most positive impact). 14 Figure 1 shows the OSI over all three phases of the SBPS
for the national average (panel A) and selected sectors (panel B). Starting with the national
13
A section in the appendix discusses the results on business closures in the SBPS. The results exhibit interesting
patterns over time and across sectors. However, analysis of non-response bias being especially important for
businesses that cease operations either temporarily or permanently is a concern that requires further research. It is
difficult to re-weight the data to correct for this type of non-response bias.
14
The five questions are assigned the following numerical values: large negative effect (-1), moderate negative
effect (-0.5), little or no effect (0), moderate positive effect (0.5), and large positive effect (1). See Index-
construction-for-the-Small-Business-Pulse-Survey.pdf (census.gov).
8
results, the OSI is everywhere negative reflecting the fact that while some businesses have seen a
positive impact (about 6.4% by the last week of Phase 3), these are in the minority. This
negativity decreases over time: the OSI index rises steadily over Phase 1, rises slightly less over
Phase 2, and flattens in Phase 3. Even by the first week of January 2021, the OSI remains
strongly in the negative range.
Turning to sectoral differences, panel B plots the OSI by sector. As with the national
numbers, these are everywhere negative but decreasing in negativity over time (especially in
Phases 1 and 2). Sectoral differences grow over time with the lines getting more spread out: the
average max-min difference rises from 0.35 in Phase 1 to 0.42 in Phase 3. The ranking of the
sectors is also relatively stable over time. 15 Finance and Insurance, Construction, and Retail
Trade are consistently at the top of the distribution over all three phases; Arts, Entertainment,
and Recreation; Educational Services; and Accommodation and Food Services are consistently at
the bottom. A common characteristic for the sectors at the bottom of the distribution is the
importance of in-person contact for their operations. We use Finance and Insurance (heavy black
line) and Accommodation and Food Services (heavy gray line) as examples of two sectoral
extremes in exercises that follow. For a sense of how these two sectors differ at the end of Phase
3, 64.9% of businesses in Accommodation and Food Services have a large negative effect as
compared to 13.7% of business in Finance and Insurance. 16
Another indicator of the status of businesses is changes in revenue. The SBPS asks “In
the last week, did this business have a change in operating revenues/sales/receipts, not including
any financial assistance or loans?” and has three checkboxes responses “Yes, increased,” “Yes,
decreased,” and “No.” 17 Figure 2 shows national and sectoral results over time. Starting with
national in Panel A, the figure shows the percent of businesses with increases in revenue (“Up” )
and decreases in revenue (“Down” ) in the previous week (“no change” is the residual of the sum
of the two lines). In the first week of Phase 1, 74.0% of businesses saw a decrease in their
revenue as compared to 5.9% who saw an increase. Over the rest of Phase 1, the share of
businesses with decreases falls and the share of increases rises, leading to the pincer-shape in
15
A notable exception is Real Estate and Rental and Leasing which moves up over time relative to other sectors.
16
These differences are significant at the 90%.
17
Another question asks, “In the last month, what were the total operating revenue/sales/receipts for this business,
not including any financial assistance or loans?” and provides ten checkbox responses. In Phase 1 weeks 1-3, the
question was phrased differently and referred to the prior week.
9
Phase 1. In the last week of Phase 1, 42.6% of businesses saw a decrease in their revenue and
19.7% saw an increase. In Phase 2, the shares level off with the share of businesses with
decreases exceeding those with increases. Finally, in Phase 3 there is a general worsening mostly
driven by increasing shares of businesses with declines. 18
As shown in Panel B, this pattern of a pincer movement, followed by flattening, and then
worsening over the three phases is repeated in Finance and Insurance and Accommodation and
Food Services sectors. 19 In the first week of Phase 1, 74.6% of businesses in Accommodation
and Food Services have decreases in revenue and 8.4% have increases. By contrast 67.8% of
businesses in Finance and Insurance have decreases in revenue and 3.2% have increases. In the
last week of Phase 3, Accommodation and Food Services is the sector with the highest
percentage of businesses with a decrease in revenues (56.1%) and 7.7% of businesses have an
increase in revenue. Finance and Insurance has one of the lowest percentages of businesses with
decreases (22.1%) and a smaller percentage of increases than does Accommodation and Food
Services (4.1%). 20 That is, more businesses in Finance and Insurance have no change in revenues
(73.8%) than is the case in Accommodation and Food Services (36.3%).
Finally, to put these two sectors in greater context of all sectors in our analysis, we plot
the percentage of businesses with declining revenue in Panel C. Again, the sectors follow the
same general pattern of a decline in Phase 1, flattening in Phase 2, and worsening in Phase 3.
Comparing the results from the Overall Sentiment Index for the two sectors, Finance and
Insurance (heavy black line) is consistently near the bottom of the distribution for revenue
declines (analogous to it being near the top for OSI). However, Accommodation and Food
Services (heavy gray line) starts off in the middle of the distribution in Phase 1 before moving to
the top in Phases 2 and 3. Notice also that the distribution becomes more disperse over Phase 2
and then Phase 3 (the average difference between the maximum and the minimum rises from 20
18
The BRS finds that 56% of establishments had decreases in demand for their goods or services (and 19%
experienced a government-mandated shutdown) but that 13% of establishments had increases in demand for their
goods or services.
19
The differences between Finance and Insurance and Accommodation and Food Services are significant over all
weeks over all questions except for the following: weeks 5 and 8 for decreasing revenue and weeks 20-23 for
increasing revenue.
20
Sectors with the least amount of businesses with decrease in revenues in the last week for the last week of Phase 3
are Management of Companies and Enterprises (16.6%) and Utilities (20.1%). As noted earlier, these sectors are not
included in the analyses in this paper.
10
percentage points in Phase 1 to 38 percentage points in Phase 3). It is worth noting that other
sectors with above average declining revenue include Retail Trade, Arts and Entertainment and
Other Services.
2.2 Changes in Employment and Hours
Now we turn to the margins of adjustment of hours and employment. In all three phases,
the survey collects the percentage of businesses that change employment and hours (and captures
whether the changes were increases or decreases). In Phases 2-3, questions were added about
rehiring furloughed workers and changes in hours of telework of paid employees.
Figure 3 shows temporal and sectoral variation in employment changes. To facilitate the
comparison with revenue changes, the figures repeat those from Figure 2 and add employment
changes (as dotted lines). Starting with the national results in panel A, in the first week of Phase
1, most businesses did not change employment (68.2%), 27.5% of businesses saw a decrease in
employment, and 4.2% saw an increase in employment. The percent of businesses with decreases
in employment falls over Phase 1 while the percent of businesses with increases rises slightly.
The percentages of businesses increasing and decreasing employment become more similar in
Phase 2 before dispersing in Phase 3. By the last week of Phase 3, 84.1% of businesses did not
change employment, 12.4% decreased and 3.4% increased employment. 21
Comparing revenue changes to employment changes (the solid lines to the dotted lines),
reveals that the average gap between the percentage of businesses with revenue increases and
employment increases falls over time from 6.4 percentage points in Phase 1 to 2.4 in Phase 2 to
1.9 in Phase 3. By contrast the average gap for businesses with revenue decreases and
employment decreases starts at 39.1 percentage points in Phase 1, falls to 20.7 in Phase 2, before
rising to 26.7 in Phase 3. These patterns are suggestive of labor hoarding at least in terms of jobs
but recall that we do not know the relative sizes of these decreases in revenue and employment.
Panel B of Figure 3 repeats the exercise for our two focus sectors: Finance and Insurance
and Accommodation and Food Services. While 89.6% of Finance and Insurance businesses have
no change in employment in the first week of Phase 1, only 45.4% of Accommodation and Food
21
The BRS found that 52% of establishments told employees not to work at some point during the reference period
but that slightly more than half of these continued to pay some or all of their employees while they were not
working.
11
Services business report no change. 22 The larger percentage of businesses reporting changes in
Accommodation and Food Services are both on the negative side (47.2% have decreasing as
compared to 8.7% in Finance and Insurance) and the positive side (7.4% versus 1.7%). 23 By the
last week of Phase 3, 26.7% of businesses in Accommodation and Food Services have declining
employment as compared to 5.5% of businesses in Finance and Insurance. Recall,
Accommodation and Food Services lost about 3 million employees while Finance and Insurance
gained about 50,000 employees over the last year (January 2020 to January 2021). 24
The overall picture is one in which businesses in Accommodation and Food Services
show more employment adjustments (and mostly on the decreasing margin) than do businesses
in Finance and Insurance. 25 We analyze these relationships further in Figure 4 which shows
scatterplots of revenue and employment declines by sector for Phase 1 (panel A) and Phases 2-3
(panel B). We have highlighted the two sectors with larger dots (Finance and Insurance in black
and Accommodation and Food Services in gray). In all three phases, the points are always above
a 45-degree line. The points are more clustered in Phases 2-3 than in Phase 1. In all three phases,
Finance and Insurance points are left-most (lower percentage of businesses reporting
employment changes) and Accommodation and Food Services are right-most. Table 3 shows the
pairwise correlations from these and confirms the differences between the phases. The
correlations are higher in Phase 1 than in Phases 2-3 for all sectors. Focusing on the two sectors,
the correlation for Finance and Insurance drops from 0.92 in Phase 1 to 0.23 in Phases 2-3. By
contrast, the correlation for Accommodation and Food Services hardly changes at all: dropping
from 0.98 in Phase 1 to 0.96 in Phases 2-3. A striking feature of Figure 4 is that all sectors in all
phases are above the 45-degree line implying that employment declines are smaller than revenue
declines.
22
Differences between the two sectors are significant except for increasing employment for most weeks in Phase 3.
23
There are some weeks in Phase 1 where Accommodation and Food Services has a larger percentage of businesses
with employment increases than employment decreases.
24
BLS reports that the number of employees in Accommodation and Food Services dropped from 13.9 million in
January 2020 to 11.0 million in January 2021. In contrast the number of employees in Finance and Insurance rose
slightly from 6.48 million in January 2020 to 6.53 million in January 2021 (not seasonally adjusted, from January
Employment Situation).
25
Using data from late March-early April, Bartik et al. (2020a) find a similar sectoral pattern. They note as
especially impacted: “retail, arts and entertainment, personal services, food services, and hospitality businesses all
reporting employment declines exceeding 50 percent. Finance, professional services, and real estate related
businesses have seen less disruption (p.4).”
12
Turning now to hours adjustments, Figure 5 shows the analogous plots for national (panel
A) and the two sectors (panel B). For ease of comparison, we repeat Figure 3 and now add in
hours changes (dotted lines). We see the same general patterns for hours that we saw for revenue
and employment (a pincer shape in Phase 1, flattening in Phase 2, and worsening in Phase 3)
with the percentage of businesses with changes in hours lying between the percentage of
businesses with changes in revenue and employment. 26 The sectoral results shown in Panel B
also follow this same general pattern. At the start of Phase 1, 31.2% of businesses in Finance and
Insurance have decreasing hours as compared to 57.6% in Accommodation and Food Services.
By the last week of Phase 3, 7.3% of business in Finance and Insurance have a decline in hours
as compared to 39.4% of businesses in Accommodation and Food Services (all differences
across sectors are significant except for increases in hours in weeks 21 and 23).
Adding up the indicators of employment adjustments and hours adjustment, the share of
small businesses that indicated they reduced employment or hours is 79 percent in the last week
of April which is slightly more than those that indicated a decline in revenue (74 percent).
Appropriate caution is needed in interpreting adding up these two statistics since many small
businesses likely reduced both employment and hours. Still this helps make the case that the
overall patterns of labor adjustments (hours or employment adjustments) echo those for revenue
(all three are copied in together in figure 1A in Appendix). For all three concepts and all three
samples, Phase 1 shows a rising share of businesses with increases in revenue, hours,
employment and falling share of businesses with decreases in these concepts in Phase 1 (yielding
the pincer shape). In Phase 2, the shares level off with the share of businesses with decreases
exceeding those with increases (the exception is employment in Finance). Finally, in Phase 3
there is a general worsening mostly driven by increasing shares of businesses with declines.
While the patterns are similar, the magnitudes are larger and Phase 3 deterioration is more
pronounced for businesses in Accommodation and Food Services. Importantly while there are
clear improvements on all of these margins relative to late April 2020, for all weeks through the
first week of 2021 there are more small businesses indicating declines than increases on each of
these margins.
26
The BRS finds that 30% of establishments reduced employee hours but that 5% of establishments increased
employee hours.
13
Furloughs and Rehires
Two questions added in Phase 2 focus on other ways in which businesses could adjust
their labor. The first question regards rehires, asking “In this last week, did this business re-hire
any paid employees who had been furloughed or laid off after March 13, 2020?” In addition to
yes and no responses, there is an option for businesses that did not lay off or furlough employees.
Consistently over Phases 2-3, businesses that did not furlough employees dominates the other
two responses (with about 55% for the national average). However, of those businesses that did
furlough employees, most did not rehire furloughed employees and the share that did rehire fell
over time (from 12.8% to 5.8%). 27 Figure 6 panel A shows all response categories for national
over time.
As shown in panel B of Figure 6, there is sectoral variation in the share of businesses who
furloughed workers (shown for the last week of Phase 3). However, these shares did not change
much over the time of the questions from early August to early January. No sector exceeds 20%
of businesses who have rehired furloughed workers in either time period. The two sectors of our
focus provide the largest percentage of businesses who furloughed employees (Accommodation
and Food Services with 72.3%) and the smallest percentages of businesses who furloughed
employees (Finance and Insurance with 21.6%).
Remote Work
The second question added in Phase 2 regards working from home, asking “In this last
week, did this business have a change in the total number of hours paid employees worked from
home?” Concerning working from home, the percentage of businesses with remote workers is
consistently below 50% for Phases 2-3 (starts at 46.8% and ends at 47.6%). Of those businesses
with remote work, most did not see a change in hours of remote work (in week 1 of Phase 2
86.3% did not have a change and 85.7% did not have a change in the last week of Phase 3). 28 For
those businesses that did change telework hours, there are roughly equal percentages of
businesses that increase and decrease telework hours in most weeks (for example, these are both
27
The raw numbers are: 5.7% and 2.6% reported rehiring and these are normalized to the population that was
furloughed (which is =100-population not furloughed).
28
The raw numbers are: 40.4% and 40.8% reported no changes in telework hours and these are normalized to the
population that have telework (which is =100-population no telework).
14
3.4% in the last week of Phase 3). 29 Figure 7 panel A shows all response categories for national
over time.
In terms of sectoral differences, panel B of Figure 7 plots the percentage of businesses
with remote work by sector for the last week of Phase 3. The sector with the largest percentage
of remote work in the last period of the survey is Information (79.6%) and Accommodation and
Food Services is the sector with the smallest percentage (21.7%). 30 The shares are relatively
stable over Phases 2-3 the exceptions are an increase in the percentage of businesses decreasing
telework in sectors Accommodation and Food Services and Other Services (from 10.3% and
8.8% to 17.5% and 12.4% respectively). The sector with the largest percentage of remote work
increasing in the final week of Phase 3 is Finance and Insurance (8.8%). The sector with the
largest percentage of remote work decreasing in the final week of Phase 3 is Accommodation
and Food Services with 17.5%. 31 Perhaps this latter result reflects the overall decrease in
employment in this sector at this time; recall that 26.7% of businesses have decreases in their
employment. These sectoral patterns are consistent with those found in the BRS. 32
2.3 Pivoting Goods/Services or Processes
As mentioned above, one way in which businesses could adapt to the pandemic is to
pivot to producing new goods and services. A number of stories in the popular press provided
examples, such as True Value retooling factory production of paint to produce hand sanitizers
(Ip, Wall Street Journal, April 16, 2020). Phase 1 includes a question intended to capture this
phenomenon: “In the last week, did this business shift to the production of other goods or
services?” The national percentage of businesses with shifts in production of goods or services
falls steadily from 6.8% in the first week of Phase 1 to 4.8% in the last week. A related question
about changes in the mode of delivery also appears only in Phase 1. The national percentage of
29
The BRS reports that 52% of establishments did not offer telework but that during the pandemic 31% of
businesses increased telework.
30
Dingell and Neiman (2020) use the detailed occupation data and O-Net data and estimate 76% of jobs in Finance
and Insurance and 4% of jobs in Accommodation and Food Services can be done at home. They estimate 72% of
jobs in Information can be done at home.
31
These percentages are recalculated to be conditional on having remote work.
32
According to the BRS, Accommodation and Food Services is the sector with the least provision of telework (91%
of establishments). On the flip side, Finance and Insurance is one of the top sectors in terms of increasing telework
(58%).
15
businesses changing the mode of delivery is 15.2% in the first week of Phase 1 and this falls to
6.6% by the last week in Phase 1.
We combine responses from both questions in Figure 8 to show the results by sector for
the first week of Phase 1 (“Early”) and the last week of Phase 1 (‘Late”). Changes to mode of
delivery dominate changes to goods/services, except in the Educational Services sector. Sectors
with the largest percentage of businesses with goods/services shifts in the first week of Phase 1
(and their percentages) are: Educational Services (19.7%), Accommodation and Food Services
(10.8%), and Information (10.1%). In terms of delivery changes, sectors with the largest
percentages of businesses are Accommodation and Food Services (42.2%) and Retail Trade
(31.8%). Comparing the Early and Late graphs, there are a smaller percentage of these
businesses making these changes late in Phase 1.
A related question concerns the changing use of online platforms. The question asks,
“Since March 13, 2020, has there been an increase in this business’s use of online platforms to
offer goods or services?” Similar to the remote work question, responses include “yes,” “no,”
and “this business does not use online platforms to offer goods or services.” Slightly more than
half of the businesses use online platforms to offer goods and services, and this share was
relatively flat over the survey (54.3% of businesses use online platforms to offer goods or
services in the first week of Phase 2 and 54.5% do in the last week of the Phase 3). However, in
both weeks, a large percentage of businesses who do use online platforms increased the intensity
of their use. In both the first week of Phase 2 and the last week of Phase 3, almost half of all
businesses that used online platforms increased their use (45.5% in both weeks). As we discuss
in the section on business applications, many new business applications are for online activity.
These patterns are consistent with both existing businesses pivoting and new businesses entering.
Turning to sectors, panel A of Figure 9 shows the percentage of businesses that use
online platforms for the first week of Phase 2 and the last week of Phase 3. Not surprisingly, this
varies by sectors from a low of 36.9% in Construction (23) to a high of 80.7% in Educational
Services (61) in the last week of Phase 3. Similar to the national results, the percentage of
businesses using online platforms is relatively flat over time. However, businesses intensity of
use of online platforms did increase in many sectors as shown in panel B of Figure 10. The
sector with the largest percentage of businesses increase their intensity of use of online platforms
16
is Educational Services (61) with more than three-quarters of businesses reporting an increase in
both time periods (76.4% and 81.2%).
The last question in this section concerns supply chain, production, delivery issues and
asks, “In the last week, did this business have any of following?” There are seven possible
responses: domestic supplier delays, foreign supplies delays, difficulty locating alternate
domestic suppliers, difficulty locating alternate foreign suppliers, production delays at this
business, delays in delivery/shipping to customers, and none of the above (and businesses can
select all that apply). About two-thirds of businesses have none of the above over both Phases 2
and 3 (an alternative version of this question was asked in Phase 1). Of those businesses with a
problem, domestic suppliers delays are the dominant response with 26.7% of businesses
identifying this as a concern in the last week of Phase 3.
Figure 10 shows the sectoral differences for the last week of Phase 3. Here we focus on
any problems (derived from 100-none of the above) and domestic supplier issues. In terms of
sectoral differences, Retail Trade is the sector with the largest percentage (48.0%) of businesses
with domestic supplier issues, while Finance and Insurance is the sector with the smallest
percentage (5.0%). While a comparable question is not available for Phase 1, in that phase Retail
Trade was also the sector with the largest percentage responding to the more general question
about disruptions in the supply chain (65.8%).
2.4 Operating Capacity and Challenges
We now focus on questions that measure the business’s current operating capacity
relative to one year ago (these questions appear in Phases 2-3). In the last week of Phase 3, about
half of all businesses note a decrease in capacity relative to one year ago: 33.4% report capacity
decreased by less than 50% and 17.1% report capacity decreased by more than 50%. 42.9% of
businesses note that there had been no change in operating capacity relative to a year earlier. The
remaining 6.6% of businesses report an increase in capacity (5.5% have a capacity increase less
than 50% and 1.1% have a capacity increase more than 50%).
The survey attempts to capture the reasons for changes in operating capacity with eight
possible responses: ability to hire (or re-hire) employees, availability of employees, ability of
employees to remote work, physical distancing of employees, physical distancing (or restrictions
on the number) of customers, availability of Personal Protective Equipment (PPE), availability of
17
other inputs used to provide goods or services, and none of the above. The category none of the
above dominates over Phases 2-3 at about 60% ending with 64% of businesses in the last week
of Phase 3. The response with the next largest percent of businesses is physical distancing of
customers or clients and/or limits on the number of concurrent customers or clients with 17.6%.
This was also the biggest capacity constraint at the start of Phase 2 (21%). While they shift their
rankings, the next most important challenges over Phases 2 and 3 are the availability of supplies
and availability of workers. 33
The changes in operating capacity for selected sectors for the last week of Phase 3 are
shown in Figure 11 panel A. In addition to our two example sectors, we include Retail Trade,
Educational Services, and Health Services. In this group, businesses in Retail Trade and Finance
and Insurance had the largest percent of businesses whose operating capacity has not changed
(45.4% and 66.2% respectively). Recall from Figure 10, however, that Retail Trade had
relatively large percentages of businesses reporting supply chain and delivery challenges
(relative to Finance and Insurance and Accommodation and Food Services). All three of the
sectors who perform services have the responses dominated by decreases (Educational Services
with 36.8% and Health Services with 54.2%) or large decreases (Food and Accommodation
Services with 45.0%). Overall, more than three-quarters of Food and Accommodation Services
business report a decline in operating capacity.
Factors impacting operating capacity for these five sectors for the last week of Phase 3
are shown in Figure 11 panel B. All five sectors also have the largest percentage of businesses
with “none of the above” (ranging from 47.4% for Health Services to 79.9% for Finance and
Insurance). Of the responses for specific issues, physical distancing of customers is particularly
important for all three services (Health, Educational, and Food and Accommodation) with about
one-third of businesses experiencing these issues in each of these sectors. Not surprisingly,
availability of PPE is especially important for Health Services (16.8%) while availability of
supplies is especially important for Retail Trade (17.7%). In contrast, Finance and Insurance has
33
Concerning the availability of workers, the Household Pulse Survey collects the “Main Reason for Not Working”
for adults not working when the survey is collected. Reasons include some directly related to COVID-19 (such as “I
was caring for someone with coronavirus symptoms” and “My employer closed temporarily due to the coronavirus
pandemic”) and some more general reasons (for example, “I am retired”.) In the week January 6-18, 2021, the top
three reasons are: retired (36%), some “other reason” (16%), and “I was laid off due to coronavirus pandemic” (8%).
Top reasons more closely related to availability during the pandemic are “I was caring for children not in school or
daycare” (6%), “I was sick with coronavirus symptoms” (5%), and “I was concerned about getting or spreading the
coronavirus” (5%). See Employment Table 3.
18
the smallest percentage of businesses for all categories except availability of telework (and that is
3.6%).
2.6 Finance
The finance questions on the SBPS include questions about cash on hand, missed loan
payments, missed other payments, requested assistance, and received assistance. The question
concerning cash on hand is, “How would you describe the current availability of cash on hand
for this business, including any financial assistance of loans?” There are seven checkbox
responses starting with no cash on hand, one week, 1-2 weeks, and so on to 3 or more months of
business operations (and don’t know). Figure 12 panel A returns to our two example sectors to
show the percentage of businesses cash on hand that would cover three or more months for
Finance and Insurance and Accommodation and Food Services. Both sectors show a rise in
Phase 1, flatten in Phase 2, and then a decline in Phase 3. Over all three phases, there is about a
20-percentage point difference between Finance and Insurance (around 40%) and
Accommodation and Food Services (around 20%).
A number of federal programs were developed to help small businesses during the
pandemic. Two related questions on the SBPS ask about requests for and receipts of financial
assistance from a variety of sources including some of these federal programs. Of the seven
federal programs listed, three have uptake rates greater than 10% at the national level by the last
week of Phase 3. These are the Paycheck Protection Program (PPP), Economic Injury Disaster
Loans (EIDL), and SBA Loan Forgiveness. Of these, the PPP has a much larger percentage of
businesses (73.6%) as compared to EIDL (23.4%) and SBA Loan Forgiveness (11.8%). All three
show a similar pattern that the receipt of financial assistance grew over Phase 1 and then held
mostly steady thereafter. 34
Figure 12 panel B shows the receipt of federal assistance for these three federal programs
by sector for the last week of Phase 3. The percentage of businesses receiving federal assistance
through the PPP program is larger than for the other two programs over all sectors. The top three
sectors in terms of percentages of businesses receiving PPP assistance are: Health Care and
Social Assistance (85.5%), Accommodation and Food Services (84.5%), and Manufacturing
34
Bartik et al. (2020b) find that receipt from the first tranche of PPP funding increases survival of small businesses
and has a positive but imprecise effect on employment based on analyses of surveys conducted in late April 2020.
Chetty et al. (2020) find PPP increased employment by 2% using private sector data.
19
(81.2%). Two of these sectors, Health Care and Social Assistance and Accommodation and Food
Services, are also on the upper end of percentages of businesses receiving EIDL and SBA Loan
Forgiveness assistance. Manufacturing is also on the upper end of percentages of business
receiving SBA Loan Forgiveness but is around the middle for EIDL assistance. By contrast,
Finance and Insurance is at the lowest end for all three types of assistance.
2.7 Forward-Looking Needs and Expectations
In this section we highlight results from questions that concern forward-looking
behaviors or expectations. The first question concerns future needs of the business, the second
question concerns business expectations. These provide an interesting complement to the
business formation series where future needs and /or expectations are inferred from
administrative data that may capture the intent to start a business.
Future Needs
The previous section highlighted federal financial assistance programs. Interestingly, the
need for financial assistance increased over Phases 2-3 of the survey. Phases 2-3 of the survey
include a question on future needs over the next 6 months. Respondents can select all that apply
from eight responses: obtain financial assistance or additional capital, identify new supply chain
options, develop online sales or websites, increase marketing or sales, learn how to better
provide for the safety of customers and employees, identify and hire new employees,
permanently close this business, and none of the above. Panel A of Figure 13 shows the
percentage of businesses for each response for the first and last weeks of both Phase 2 and Phase
3. While some categories are relatively constant over time, financial assistance gained in
importance (from 26.8% in first week of Phase 2 to 33.5% in the last week of Phase 3) while
providing for safety became relatively less important (from 19.3% in the first week of Phase 2 to
13.3% in the last week of Phase 3).
Figure 13 panel B shows the future needs for Finance and Insurance and Accommodation
and Food Services for the last week of Phase 3. Given the challenges and outcomes highlighted
above, it is not surprising that the percentage of businesses citing one of the seven future needs is
dominated by Accommodation and Food Services while none of the above is dominated by
20
Finance and Insurance. Interestingly, the future needs mostly closely aligned across the two
sectors are: “marketing or sales” and “online sales” where the differences between the two
sectors is less than 5 percentage points. In contrast, there is an almost 40 percentage point
difference between the two sectors for financial assistance (55.7% of businesses in
Accommodation and Food Services cite this as a future need).
Expectations
The final question on SBPS concerns expectations, asking: “In your opinion, how much
time do you think will pass before this business returns to its normal level of operations relative
to one year ago?” The responses include little/no effect, already back to normal, to varying
durations (1 month or less up to more than 6 months), to never and closed. Care must be taken
when comparing over all three phases because of changes introduced into the question and
responses over time. In the fourth week of Phase 1, the question is amended to add the phrase
“relative to one year ago” and the word “usual” was replaced with “normal.” At the start of
Phase 2, the categories “This business has returned to its normal level of operations” and “This
business has permanently closed” are added as responses.
Figure 14 panel A plots business expectations at the national level over all the weeks of
collection. To help distinguish the responses we have grouped them by short term (little or no
effect, already back to normal), medium term (2-3 months and 4-6 months), and long term (more
than 6 months, never, closed). The percent of businesses who see medium-term time periods
until a return to normal has fallen (from 51.8% in the first week of Phase 1 to 18.6% in the last
week of Phase 3). This reflects greater percentage of businesses with expectations in the short-
term and the long-term. The increases in percentage of businesses with short-term expectations
rises in Phases 1-2 before falling slightly in Phase 3; these changes are driven mostly by no
impact and already back to normal (solid lines). The percentage with long-term expectations
rises in Phases 1 and 3 (and holds flat in Phase 2); these changes are driven mostly by the rise in
6 months or more (dashed line). 35
35
In terms of expectations prior to the start of the SBPS, Bartik et al. (2020a) find that early in the pandemic (late
March-early April), 50% of their sample of small businesses expected that the crises would last until at least the
middle of June. Their survey asks respondents for “the most likely date” when the crisis would be over and also
captures of measure of their uncertainty. Another interesting dimension is differences over size and employer status.
Using their Stanford_Stripe sample, Bloom et al. (2021) find that in the fall of 2020, businesses with 20 or more
21
We show a similar grouping for sectoral differences in expectations in panel B. Here we
stack the responses from more positive (little/no effect, already back to normal), to varying
durations (1 month or less up to 6+ months), to never and closed. Finance and Insurance has
relatively more light part of its bar; they were either not impacted, little impact, or already back
to normal. Compare this to Educational Services, Arts and Entertainment, and Accommodation
and Food Services which have relatively lower percentages of businesses who had little/no effect
or are already back to normal. Health Services also has low little/no effect but has more robust
already back to normal.
In Panel C we narrow our focus to our two example sectors and examine changes over
time for first week of Phase 1 and the last week of Phase 3. The lighter bars are for the first week
of the survey and the darker bars are for the last week of the survey. For both sectors, the darker
bars are shorter for medium term. In a K-shaped recovery some groups are already moving on an
upward path in recovery; but others are on a downward or stalled path. Focusing on the last
week, 40.4% of businesses in Finance and Insurance expect a return to normal level of
operations over the shorter horizon, as compared to 7.3% of businesses in Accommodation and
Food Services. At the other extreme, 43.8% of businesses in Finance and Insurance expect a
return to normal to take place over the longer horizon, as compared to 76.8% of businesses in
Accommodation and Food Services. This suggests an expectation of a K-shaped recovery across
sectors. There is also likely a K-shaped recovery within sectors implied by the finding that for all
sectors in the SBPS that even through January 2021 there are indicators of negative growth,
business sentiment and expectations for existing small businesses. This persistent negativity
holds even in sectors where indicators of all business activity including large businesses have
shown a recovery. 36
We connect the responses for business sentiment and expectations in Figure 15 where we
plot Overall Sentiment Index against the Expected Recovery Index by sector for each of the three
employees were expecting a return to normal level of sales by 2021:Q2; smaller businesses expected a longer time
period until a return to normal and non-employer businesses had the most pessimistic expectations (see their Figure
A7). Another look at the long tail of expectations is provided by NFIB. The NFIB survey conducted in December
2020 finds that 4% of businesses are back to normal, 47% anticipate a return to normal in 2021 and 36% expect a
return to normal in 2022.
36
A Wall Street Journal article concerning restaurants notes “The coronavirus is splitting the restaurant industry in
two. Big, well capitalized chains like Chipotle Mexican Grill Inc. and Domino’s Pizza Inc. are gaining customers
and adding stores while tens of thousands of local eateries go bust.” October 12, 2020.
22
phases. The Expected Recovery Index captures the checkbox responses and converts them into
an index with values from -1 to +1. Business sentiment and expectations are positively related as
shown by the positive sloping lines. Over time, there are some shifts in the lines, but the lines
shift almost entirely in a parallel fashion. The change in the general level of business sentiment
(OSI) drives the shift down from the first week to the last week of Phase 1. The jump in the
fourth week of Phase 1 is likely due to the change in the expectations question; however, the
positive relationship between OSI and ERI is the same across weeks, that is, the slope does not
change.
2.8 Results by Geography
In examining results by geography, it is important to keep in mind that the pandemic
impacted local areas in differing intensities and at different times and that different state and
local policies also may impact business activity. It is beyond the scope of this paper to examine
the causes for these patterns. Goolsbee and Syverson (2020) find that fear of infection rather than
shutdown policies drives much of the economic slowdown for customer visits during the
pandemic. They find that these policies do lead to reallocation across businesses away from
nonessential businesses towards essential businesses and specifically cite as an example from
restaurants and bars towards grocery stores. 37 This suggests that we may expect to see changes
across states that evolve in a cadence more closely related to pandemic outbreaks than to state
policies. 38 On a more operational note, we also need to keep in mind that nonresponse
adjustments may not capture state differences in response. 39
As we did with the sectoral analysis, we start by examining the patterns in overall
sentiment and changes in revenue. We examine these in comparison with sectoral results. Two
37
Goolsbee and Syverson (2020) use data from SafeGraph which concerns customer visits only and hence does not
include activities, such as online sales, that do not require a visit to a physical location. Also, they only examine
industries for which consumer visits are a meaningful measure of economic activity. They examine policies at sub-
state levels.
38
Future research could look within states to see evidence of the type of reallocation that Goolsbee and Syverson
(2020) describe.
39
The SBPS initially weights businesses based on the number of eligible businesses within each state-subsector (by
state by 3-digit NAICS). Businesses are assigned into one of nine week groups by first sorting businesses using
MSA and 2018 annual payroll and then assigning them systematically across the nine week groups. The initial
weights are adjusted weekly for survey nonresponse to generate representative estimates at the national, sectoral,
and employment size class levels. Weights are not adjusted to state totals and therefore care should be taken in
interpreting state results.
23
things immediately stand out: differences in dispersion and persistence. We plot the Overall
Sentiment Index for the ten largest states (by population) in Figure 16 panel A. For these ten
states, New York (heavy black line) remains at the bottom over all three phases of the survey
while North Carolina (heavy dotted black line) remains relatively close to the top. Panel B of
Figure 16 shows the percentage of businesses with declining revenues by state over all three
phases of the pulse. We highlight two states of interest based on their relative position at the start
of the survey: New York (heavy black line) and South Dakota (heavy black dotted line). While
New York remains in the upper part of the distribution and South Dakota in the lower part, there
are many states that surpass them.
This evidence suggests there are differences in both dispersion and persistence between
sectors and states (compare to Figures 1B and 2C). We directly compare dispersion in states and
sectors in Figure 16 (OSI) and Table 4 (revenues). Figure 17 panel A presents box plots for OSI
over all three phases of the SBPS for sectors and states (excluding sectors 11, 21, 22, and 55 and
Puerto Rico). Comparing the two figures, the interquartile range for OSI is larger for sectors than
it is for states. Panel B shows the coefficient of variation for sectors and states and shows the
larger coefficient of variation for sectors (that increases over time).
Table 4 measures dispersion in the percentage of businesses with declining revenues
through the difference between the maximum and minimum sectoral values and coefficient of
variation (each of these are averaged for each phase). The max-min gap for sectors and states are
relatively close to each other in Phase 1 (about 20 percentage points for sectors and about 25
percentage points for states), but this gap widens for sectors (to 38 percentage points) and
narrows for states (to 23 percentage points) in Phase 3. Measuring dispersion using the
coefficient of variation tells basically the same story of dispersion being similar for states and
sectors in Phase 1 (0.10) but different in Phase 3 (0.27 for sectors and 0.13 for states).
We are also interested in whether there is persistence in the position of a sector or state in
the overall distribution of business sentiment. That is, we can see from the earlier plots that
Accommodation and Food Services and New York tend to be at the bottom of the OSI
distribution and Finance and Insurance and South Dakota tend to be at the top of the OSI
distribution. We look at persistence in a number of ways. First, we plot the first week of Phase 1
OSI against the last week of Phase 3 OSI for both sectors (panel A) and states (panel B) in
24
Figure 18. The sectors show a clear linear trend and the correlation between the two weeks is
0.90. There is less of a linear trend for states and the correlation between the two weeks is 0.54.
Another way to look at this is to group each of the weeks into quintiles and consider the
transition paths of sectors and states. We find that 60% of the sectors remain in the same quintile
as their starting place (for example, Finance and Insurance remains in the highest quintile in the
first and last weeks). By contrast, 27% of the states remain in the same quintile as their starting
place (for example, New York remains in the lowest quintile in the first and last weeks). Even so,
there is some stickiness for states, no state transitions from one extreme to the other extreme.
3. Business Formation Statistics
While existing small businesses have experienced persistent negative outcomes over the
course of pandemic through January 2021, the COVID-19 recession has also altered the
landscape for potential entrepreneurs in 2020, and into the foreseeable future as the economy
continues to adjust to potentially a new normal. While the amount of economic uncertainty and
restrictions introduced by the pandemic were apparently detrimental for new business
applications during March and April 2020, there has been an unprecedented surge in business
applications starting in May 2020. The time-series patterns for business applications have also
exhibited strong geographic and sectoral disparities.
What has made it possible to make timely statements about business formation activity
during 2020? Business formations can now be tracked on a high frequency, timely basis using
the Census Bureau’s Business Formation Statistics (BFS) program. The BFS offers accurate and
timely information on new business applications as well as actual and projected employer
business startups, The BFS was originally released as an experimental data product in 2018 as a
quarterly dataset for public use (see Bayard et. al (2018) for details about the development of the
BFS). The BFS uses applications for Employer Identification Numbers (EINs) obtained from the
Internal Revenue Service (IRS) to construct nationwide, industry-level and state-level time series
for applications made for new businesses, as well as actual and projected employer business
formations originating from these applications. Importantly the application process for new
EINs is largely online and automated so that even when IRS offices are not open new
applications can be processed almost immediately.
25
The BFS was released continuously at a quarterly frequency between 2018 and 2020,
with the earliest data pertaining to 2004q3. Up until 2020, the BFS time series included data on
only one recession: the Great Recession that started in 2007q4. As economic conditions rapidly
deteriorated with the onset of the pandemic, the demand by policymakers for near real-time data
sources for measuring economic activity once again surged. This demand presented a unique
opportunity for BFS to monitor business formation activity at a higher frequency. Recognizing
its potential as an early indicator of fast-evolving economic conditions during the COVID-19
recession, the Census Bureau started releasing the BFS at a weekly frequency on April 8, 2020.
Weekly frequency is the highest frequency with which the underlying administrative data
for the BFS is received by the Census Bureau. In many ways, the BFS was already prepared to
handle such high frequency, as the basic building blocks of the quarterly BFS data were weekly
batches of applications received on a timely, flow basis. Therefore, transition to a weekly release
did not require a lengthy setup phase, although it demanded substantial resources to carry out a
tight weekly production cycle that started with the ingest of new application data every Monday
and a public release of processed data every Thursday. The BFS has also initiated a monthly
release starting in January 2021. The monthly BFS builds on the same infrastructure and is
released within ten days of the end of the reference month. Taken together, the weekly and
monthly BFS offer unique high frequency, timely information on new business formations.
Analysis of the weekly and monthly data reveals several striking patterns, some of which
are summarized in this section. First, as emphasized by Dinlersoz et. al. (2021), the COVID-19
recession has been very different from the Great Recession in terms of the short-run dynamics of
business applications and projected formations. Second, there has been large variation across
states and sectors in the patterns of business applications. Third, the surge in new business
applications has been especially large for likely new nonemployer businesses but is also present
for likely new employer businesses.
3.1. Business Applications and Formations during the COVID-19 Recession
Leveraging the weekly BFS, Dinlersoz et al. (2021) investigate the weekly progression of
business applications and projected business formations from these applications during the
COVID-19 recession and compare it with that during the Great Recession. They show that both
the business applications and formations exhibit a slow and protracted decline during the Great
26
Recession (see their Figure 1). In contrast, they find that the drop in applications and projected
formations is very sharp but short-lived during the COVID-19 Recession. Applications and
projected formations both rebound very quickly, with the overall trajectory for the two series
exhibiting a V-shaped recovery.
While the transition rate from applications to employer businesses relative to the
reference period also fell during the Great Recession initially, it recovered rapidly. As shown in
the left panel of Figure 2 of Dinlersoz et al. (2021), after about 30 weeks of the onset of the
recession, the cumulative transition rate went back up to its level in the reference period and
exceeded that level thereafter. The projected transition rates for the COVID-19 Recession, on the
other hand, had not recovered around the same time mark from the recession. As discussed in
Dinlersov et. al. (2021), the decline in the transition rate in the COVID-19 recession largely
reflects a shift in the composition of applications towards those that are more likely to be new
nonemployer businesses than new employer businesses. By construction, applications for likely
nonemployers will have a low transition rate to becoming employers.
These patterns are illustrated in Figure 19 which uses the monthly BFS. Overall
applications are decomposed into likely employers (High-Propensity Business Applications) and
likely nonemployers (the difference between overall applications and High-Propensity Business
Applications or NHBA=BA-HBA). High- Propensity Business Applications (HBA) are defined
using information provided on the EIN application including indicating hiring employees;
providing a date for first-wages paid or planned; providing a NAICS code in the Manufacturing,
a portion of Retail Trade, Health Care, or Accommodation and Food Services sectors. 40 Bayard
et. al. (2018) show that more than one-third of HBA transition to a new employer business but
only 3 percent of the nonemployer applications transit to a new employer business. Haltiwanger
(2021) provides evidence that variation in HBA closely tracks employer startups (which is
evident in Figure 19) and variation in NHBA closely tracks variation in nonemployers.
Overall applications and likely nonemployer applications are at the highest for the entire
series in 2020 with the surge occurring in the second half of 2020. Interestingly, the surge
continues into January 2021. For likely employers, there has also been a surge in 2020 but it is
more muted. Still applications for likely employers have had their highest year since 2006. This
40
See https://www.census.gov/econ/bfs/definitions.html. Also, applications for corporations are classified as HBA.
The component of Retail Trade classified in HBA is that dominated by bricks and mortar stores.
27
surge in applications for likely employers is important given the disproportionate role that new
employer businesses play in job creation, innovation and productivity growth (see, Decker et. al.
(2014)). The surge in applications for likely nonemployers is potentially an acceleration of the
shift towards the gig economy in the last decade or so as emphasized in Abraham et. al (2021).
3.2. Sectoral Differences in Applications during the COVID-19 Recession
The COVID-19 Recession has also been accompanied by large shifts in sector-level
application activity. Initially, nearly all sectors experienced a sharp decline in business
applications in the early phases of the recession (see Figure 20). However, many sectors
rebounded relatively quickly in application volumes, with a particularly large rise in retail
(NAICS 44-45). A major component of that rebound in retail was accounted by non-store
retailers (NAICS 454), which includes mainly internet-based retail (part of e-commerce) activity,
as shown in Figure 21. By mid-2020, the number of applications in this sub-sector exhibited an
increase of more than 200% from its low in early 2020, before subsiding in the later part of the
year. The surge of applications in this sub-sector far exceeded the other sub-sectors in retail
(Figure 21).
The surge in non-store retailer applications is consistent with the push towards remote
interactions between businesses and consumers in the pandemic. It is also broadly consistent
with the pre-pandemic trend of increasing e-commerce. In this respect, this shift towards new
businesses in the non-store retailer sector in the pandemic may reflect an acceleration of pre-
pandemic trends. An open question is how many of these new non-store retailer businesses will
be employers or nonemployers. Most of the businesses in non-store retail sub-sector have
historically been non-employers - in 2018 over 90% of non-store retailers were non-employer
businesses, based on combined Nonemployer Statistics (NES) and County Business Patterns
(CBP) data. The public domain BFS includes the decomposition of BA into HBA and NHBA at
the 2-digit NAICS level. 41 Further research (and time) is needed to understand the outcomes for
the surge in applications for new non-store retailers.
Another open question is the extent to which some of the surge in EINs are existing businesses adding a new line
41
of business. It might be for example that existing businesses add a new online line of business and apply for a new
EIN for this activity. However, there is no requirement that adding a new line of businesses requires a new EIN for
an existing business. Moreover, Decker et. al. (2016) find that multiple EIN firms are the exception and not the rule
28
The monthly data provide insights into the sectoral patterns decomposed for applications
for likely nonemployers and likely employers. Haltiwanger (2021) shows there has been a surge
in Retail Trade and Professional, Scientific and Technical Services in both types of applications.
Existing small businesses in both sectors have had substantial declining revenue during the
pandemic (see Figure 2). The differing patterns for existing small versus new businesses suggest
that the pandemic has induced considerable restructuring of business activity. An open question
is how persistent these changes will turn out to be and whether any resulting businesses will have
differing characteristics, such as scale, from those started during non-pandemic times.
3.3. Geographic Disparity in Application Activity during the COVID-19 Recession
The weekly BFS has been released for each state and Census Region, in addition to
nationwide. The week-to-week evolution of business applications exhibits substantial
heterogeneity across the geography. Figure 22 looks at the year-over-year percentage growth in
weekly application series by region for 2020. All regions experience a sharp decline in the earlier
part of 2020 as the COVID-19 crisis hits. The decline is most pronounced in the Northeast,
where, as of week 13, applications fell by as much as 50% compared to the same week in 2019.
Starting in week 14, however, there is a steady increase in the growth rates for all regions, with
the South leading the way, followed by the Midwest. A peak in growth rates occurs in week 27
for all regions. This peak is notably pronounced for the Midwest and the South. Observe also that
the Northeast gradually catches up with the Midwest between week 13 and 27, even though it
experienced the sharpest decline in growth rates initially. All regions experience a decline in
application growth rate following week 27. The South and the Midwest maintain high growth
rates for the rest of the year. The Northeast also does relatively well in this period, exhibiting
higher weekly year-over-year growth rates than the Midwest.
and are concentrated among large, mature firms that operate in multiple states. Large, mature firms that operate in
many locations account for a substantial share of employment and revenue but a very small share of firms. Viewed
from that perspective, it is very unlikely that the share of multiple EIN firms has surged to the degree needed to
account for surge in new EIN applications. It is also possible that the surge in new EINS reflects in part existing
businesses undergoing a change in ownership. An interesting area for future research is to explore the impact of the
pandemic on ownership change of existing businesses.
29
The heterogeneity across states is even more pronounced. Figure 23 shows the path of
weekly year-over-year growth rates for business applications for selected states in 2020. New
York, which experienced an early outbreak during the pandemic, exhibits a large decline in
growth rate of applications heading into week 13. California, Texas, Illinois and Florida also
went through declines initially, though less severe than in the case of New York. Georgia
experienced much milder decline, a pattern that generally holds for other southern states. The
recovery in applications was quite uneven. Illinois and Georgia had skyrocketing application
growth, whereas the other states had more muted rebounds. Even New York, which experienced
the largest decline initially, went from a 50% deficit in applications in week 13 to a nearly 75%
surplus in week 27, relative to the same weeks in 2019. As in the case of the regions, application
growth generally recedes for all states featured in Figure 23 after week 27. Nevertheless, Georgia
and Illinois maintained high rates of growth compared to the other states in Figure 23.
Figure 24 offers more insight to the variation across states in weekly year-over-year
growth rates. It plots the weekly dispersion in growth rates across all 50 states and District of
Columbia included in BFS. The dispersion across states is quantified using three different
measures: standard deviation (SD), interquartile range (IQR), and the difference between 90th
and 10th percentiles (90-10). All three measures exhibit similar trends over the weeks in 2020.
Before week 13, the dispersion measures exhibit little trend, except for some decline heading
into week 13 as all states experience a sharp drop in growth rates nearly simultaneously.
However, starting with week 13, the dispersion measures start to increase, culminating in a peak
in dispersion in week 27 that coincides with the peak in application volumes for states in general.
All three measures indicate that there was an increasingly large divergence in growth rates
between week 13 and 27. While all dispersion measures subside after week 27, the dispersion
remains high compared to the initial part of 2020, hinting at potentially long-lasting differences
in business application growth brought about by the individual experiences of states during the
pandemic. It is perhaps no surprise that states followed diverging paths, given the heterogeneous
nature and timing of state-level measures taken to abate the impact of the COVID-19 crisis.
3.4. Did the PPP contribute to the surge in new business applications?
Given the importance of the PPP for existing small businesses, a natural question to ask is
whether the PPP contributed to the surge in new business applications in the BFS starting in May
30
2020 and continuing through January 2021. This is a complex question that requires more
research in part because the PPP and other fiscal stimulative programs may have impacted new
business formation indirectly via the impact on the overall economic outlook. Haltiwanger
(2021) has an extensive discussion of this issue presenting evidence that the direct impact of the
PPP program on the surge in new business applications was likely modest.
The core reason for the likely modest direct impact of the PPP on new business
applications is that the PPP program is designed to provide financial support to existing
businesses. Specifically, rules for obtaining a PPP loan require the business be in operation in
February 2020. For existing employer businesses that received a PPP loan such businesses
already had an EIN in February 2020 well before the surge in new business applications. Even
for existing nonemployers that received a PPP loan, many also likely had an EIN. The primary
group that may have been impacted are sole proprietor, nonemployers that did not have an EIN
prior to applying for a PPP loan. Using SBA data on all loans, about 600K of the 5.2 million
PPP loans by August 2020 went to sole proprietor, nonemployers. 42 This can be interpreted
(see Haltiwanger (2021) for more discussion) as an upper bound on PPP loan recipients that
applied for an EIN directly as part of participation in the PPP program. 43 This upper bound is
relatively small compared to almost 4 million new business applications from May 2020 to
January 2021. It is also worth emphasizing that the surge in new business applications is not just
in overall applications but also in applications for likely new employers (HBA and WBA). New
employer businesses that began operations after February 2020 are not eligible for PPP loans.
4. Conclusion and Future Work
The COVID-19 pandemic highlights the need for timely, high frequency, and
geographically and sectorally granular socio-economic data. We focused on two economic data
products introduced by the Census Bureau during the pandemic. The Small Business Pulse
42
For SBA data on all loans see https://www.sba.gov/funding-programs/loans/coronavirus-relief-options/paycheck-
protection-program/ppp-data#section-header-2.
43
It would also be interesting to obtain data on loan applications by employer and nonemployer status. As discussed
in Haltiwanger (2021) there was no incentive for existing nonemployer, sole proprietors as of February 2020 to
apply for an EIN in order to apply for a PPP loan (especially in the first round in 2020). It is more likely that some
of the nonemployer, sole proprietors that obtained loans applied for an EIN after receiving the loan. A business
bank account facilitates the reporting requirements of a PPP loan and an EIN is needed for a business bank account.
31
Survey provides information about an important part of the economy which was especially
impacted by COVID-19: small businesses. The weekly and monthly Business Formation
Statistics provide forward-looking information about an important driver of economic growth:
employer business startups, which contribute to jobs and productivity. These two products
capture temporal, sectoral, and spatial variation in the impact of the pandemic on the U.S.
economy.
One important area missing from these new data products that still needs better
measurement is business deaths. The SBPS is not well designed to capture busines deaths since
survey nonresponse is likely greater for closed businesses. Existing research suggests that,
similar to business births, business death may be best captured as a process resulting eventually
in the event of death. There may be extended periods in which a business is dormant but then
becomes active again. Crane et al. (2020) highlight many different approaches to measuring
business deaths including the use of credit card data, clock-in/out tracking data, mobile device
location data, online platform rating data, energy use data, vacancy rates, survey data (such as
the SBPS), and business administrative data. To this, we would also add administrative data on
workers (such as LEHD data).
The empirical exercises in this paper could be extended to consider more relationships
between concepts captured in the SBPS. For example, exercises could focus on relationships
between revenue changes and the ability to pivot to new goods and services or new ways of
delivering these goods and services. It may also be possible to disentangle some of the sectoral
and geographic impacts. For example, to look at state differences in customer shifts away from
businesses in Accommodation and Food Services towards businesses in Retail Trade (that is,
from restaurants to grocery stores). The SBPS website provides data for state by sector which
could be used for this type of analysis. Another area of exploration would be to use the size class
information provided by the SBPS to examine any differential patterns based on size. The size
class information is based on the number of employees and there are three classes: 1-4, 5-19, and
20 or more.
Finally, as of the writing of this paper, the SBPS is now in Phase 4 of collection. Phase 4
retains the core concepts of overall sentiment; changes in revenue, employment, and hours; and
expectations. Content was reviewed for continued relevance and new concepts were added to
capture the changing nature of the pandemic. New content added includes business requirements
32
for COVID-19 testing or vaccinations for onsite employees, changes in the physical footprint,
capital expenditures, business travel, and updates to the federal assistance programs. With new
content added, less relevant concepts were removed to maintain the overall response burden.
Moving beyond the publicly available data to the underlying micro level data, future
research could more deeply explore the patterns that we illuminate. It is likely that there are rich
and complex dynamics at play over time, geography, and sector that can only be captured using
the micro data. The underlying micro data for these two products will be available to qualified
researchers on approved projects in the Federal Statistical Research Data Center network. We
hope this paper inspires others to use these rich sources of information.
33
References
Abraham, Katharine G., John C. Haltiwanger, Kristin Sandusky, and James R. Spletzer (2021)
“Measuring the Gig Economy Current Knowledge and Open Issues” in NBER/CRIW
Conference Volume on Measuring and Accounting for Innovation in the 21st Century,
(forthcoming).
Bartik, Alexander W. Marianne Bertrand, Zoë B. Cullen, Edward L. Glaeser, Michael Luca,
Christopher T. Stanton. 2020a. “How Are Small Businesses Adjusting to COVID-19?
Early Evidence from a Survey,” NBER Working Paper No. 26989, April.
Bartik, Alexander W. Marianne Bertrand, Zoë B. Cullen, Edward L. Glaeser, Michael Luca,
Christopher T. Stanton, Adi Sunderam. 2020b. “The Targeting and Impact of Paycheck
Protection Program Loans to Small Businesses,” NBER Working Paper No. 27623, July.
Bayard, Kimberly, Emin Dinlersoz, Timothy Dunne, John Haltiwanger, Javier Miranda, and
John Stevens. 2018. “Early Stage Business Formation: An Analysis of Applications for
Employer Identification Numbers,” NBER Working Paper, 24364.
Bloom, Nicholas, Robert S. Fletcher, and Ethan Yeh. 2021 “The Impact of COVID-19 on US
Firms” NBER Working Paper, 28314, January.
Buffington, Catherine, Carrie Dennis, Emin Dinlersoz, Lucia Foster, and Shawn Klimek. 2020.
“Measuring the Effect of COVID-19 on U.S. Small Businesses: The Small Business
Pulse Survey” Center for Economic Studies Working Paper, 20-16.
Buffington, Catherine, Jason Fields, and Lucia Foster. Forthcoming. “Measuring the Impact of
COVID-19 on Businesses and People: Lessons from the Census Bureau’s Experience.”
AEA Papers and Proceedings.
Chetty, Raj, John N. Friedman, Nathaniel Hendren, and Michael Stepner. 2020 “How did
COVID-19 and Stabilization Policies Affect Spending and Employment? A New Real-
Time Economic Tracker Based on Private Sector Data,” NBER Working Paper No.
27431, November.
Crane, Leland D., Ryan A. Decker, Aaron Flaaen, Adrian Hamins-Puertolas, and Christopher
Kurz. 2020. “Business Exit During the COVID-19 Pandemic: Non-Traditional Measures
in Historical Context,” Finance and Economics Discussion Series 2020-089. Washington:
Board of Governors of the Federal Reserve System,
https://doi.org/10.17016/FEDS.2020.089.
Decker, Ryan, John Haltiwanger, Ron Jarmin and Javier Miranda. 2016. “The Role of
Entrepreneurship in US Job Creation and Economic Dynamism” Journal of Economic
Perspectives, 28(3), 3-24.
Decker, Ryan, John Haltiwanger, Ron Jarmin and Javier Miranda. 2016. “Where has all the
Skewness Gone? The Decline in High Growth (Young) Firms in the U.S.?” European
Economic Review, 86, 4-23.
34
Dingel, Jonathan I. and Brent Neiman. 2020. “How Many Jobs Can be Done at Home?” Becker
Friedman Institute White Paper, June. https://bfi.uchicago.edu/working-paper/how-
many-jobs-can-be-done-at-home/
Dinlersoz, Emin, Timothy Dunne, John Haltiwanger, and Veronika Penciakova, 2021. “Business
Formation: A Tale of Two Recessions,” AEA Papers and Proceedings (forthcoming).
Goolsbee, Austan and Chad Syverson. 2020. “Fear, Lockdown, and Diversion: Comparing
Drivers of Pandemic Economic Decline 2020,” Journal of Public Economics, 193.
https://doi.org/10.1016/j.jpubeco.2020.104311.
Haddon, Heather. 2020. “McDonald’s, Chipotle and Domino’s are Booming During Coronavirus
While Your Neighborhood Restaurant Struggles” The Wall Street Journal, October 12.
Haltiwanger, John. 2021. “Entrepreneurship During the COVID-19 Pandemic: Evidence from
the Business Formation Statistics,” Paper for presentation at the NBER Entrepreneurship,
Innovation and Economic Policy Conference, April.
Ip, Greg. 2020.“Shoes to Masks: Corporate Innovation Flourishes in Coronavirus Fight” The
Wall Street Journal, April 16.
National Federation of Independent Businesses, 2020. “Covid-19 Small Business Survey (14):
PPP, EIDL, the Economy and the Vaccine. December 6-11.
https://assets.nfib.com/nfibcom/Covid-19-14-Questionnaire-Write-up-FINAL-Web.pdf
35
Table 1: Overview of SBPS Questions
Concept Question
Phase 1 Phases 2-3
Subjective well Overall, how has this business been Unchanged
being affected by the COVID-19 pandemic? [5-
bins]
Operating In the last week, did this business Added “revenue/sales/receipts, not including any
revenue experience a change in operating financial assistance or loans?”
revenues? [3 checkboxes]
Operating In the last week, what was the total Added “revenue/sales/receipts, not including any
revenue operating revenue for this business? [10 financial assistance or loans?”
bins]
Operating In the last week, did this business In the last week, did this business do any of the
status temporarily close any of its locations for following? [4 checkboxes]
at least one day? [2 checkboxes]
Number of In the last week, did this business have a Unchanged
employees change in the number of paid employees?
[3 checkboxes]
Re-hire In the last week, did this business re-hire any
employees paid employees who had been furloughed or laid
off after March 13, 2020? [3 checkboxes]
Number of In the last week, did this business have a Unchanged
hours worked change in the total number of hours
worked by paid employees? [3
checkboxes]
Work from In the last week, did this business have a change
home in the total number of hours paid
employees worked from home? [4checkboxes]
Supply chain In the last week, did this business have In the last week, did this business have any of the
disruptions in its supply chain? [2 following? [7 checkboxes]
checkboxes]
Product/service In the last week, did this business shift to
offerings the production of other goods or services?
[2 checkboxes]
Mode of In the last week, did any of this business’s
product/service locations adopt pickup/carry-out/delivery
delivery as their only means of providing goods
and services to their customers? [2
checkboxes]
Factors that In the last week, was this business's operating
affect capacity capacity affected by any of the following? [8
checkboxes]
Capacity How would you describe this business's current
change operating capacity relative to one year ago? [5
bins]
36
Cash flow How would you describe the current Added “including any financial assistance or
availability of cash on hand for this loans?”
business? [7 checkboxes]
Loan defaults The White House declared a national Unchanged
emergency because of the COVID-19
pandemic on March 13, 2020. Since then,
has this business missed any loan
payments? [2 checkboxes]
Other payment Since March 13, 2020, has this business Unchanged
defaults missed any other scheduled payments,
not including loans? [2 checkboxes]
Application for Since March 13, 2020, has this business Unchanged but now with [14 checkboxes]
financial requested financial assistance from any of
assistance the following sources? [10 checkboxes]
Receipt of Since March 13, 2020, has this business Unchanged but now with [9 checkboxes]
financial received financial assistance from any of
assistance these programs from the federal
government? [5 checkboxes]
Online Since March 13, 2020, has there been an increase
platforms in this business's use of online
platforms to offer goods or services? [3
checkboxes]
Future plans In the next 6 months, do you think this business
will need to do any of the following? [8
checkboxes]
Subjective In your opinion, how much time do you Unchanged but now with [8 checkboxes]
return to think will pass before this business
normal returns to its usual level of operations? [6
operations checkboxes].
In week 4 of Phase 1, this was changed
to: In your opinion, how much time do
you think will pass before this business
returns to its normal level of operations
relative to a year ago?
37
Table 2: Finance and Insurance and Accommodation and Food Services
Panel A: Establishments and Employment
Sector Number of Establishments Number of Employees
2019 total
Finance and Insurance 476,369 6,554,437
Accommodation and Food Services 741,351 14,502,449
2018 total
Finance and Insurance 477,562 6,499,208
Accommodation and Food Services 733,134 14,345,140
2018 small (less than 500)
Finance and Insurance 475,861 4,467,617
Accommodation and Food Services 732,415 13,571,550
Panel B: Employment
January 2020 January 2021
Finance and Insurance 6,476.3 6,525.0
Accommodation and Food Services 13,871.5 10,967.8
Sources:
Panel A: County Business Patterns
Panel B: Table B1 of Employment Situation, February 5, 2021 (NSA)
Table 3: Correlation between Declines in Revenue and Employment by Sector
Sector Phase 1 Phases 2-3
Construction 0.91 0.82
Manufacturing 0.87 0.10
Wholesale Trade 0.94 0.11
Retail Trade 0.98 0.21
Transportation and Warehousing 0.81 0.59
Information 0.92 0.34
Finance and Insurance 0.92 0.23
Real Estate 0.79 0.13
Professional, Scientific, Technical 0.90 0.35
Administrative 0.89 0.78
Educational Services 0.94 0.25
Health Care and Social Assistance 0.98 -0.03
Arts, Entertainment, and Recreation 0.96 0.72
Accommodation and Food Services 0.98 0.96
Other Services 0.98 0.62
Source: Authors’ calculations.
38
Table 4: Sectoral and Geographic Dispersion for Percentage of Businesses with Declining Revenues
Phase1 Phase2 Phase3
Max-Min
Sectors 19.9 20.5 38.0
States 25.1 25.6 22.5
Coefficient of Variation
Sectors 0.10 0.20 0.27
States 0.10 0.16 0.13
Notes: Sectors excludes sectors 11, 21, 22, and 55. States excludes Puerto Rico. Average of the Max-
Min and Coefficient of Variation for each phase.
Source: Authors’ calculations from Small Business Pulse Survey.
39
Figure 1: Business Overall Sentiment
Panel A: National
Panel B: Sectoral
Source: Authors’ calculations from Small Business Pulse Survey.
40
Figure 2: Revenue Changes
Panel A: National
Panel B: Two Sectors
Panel C: All Sectors (Declining Only)
Source: Authors’ calculations from Small Business Pulse Survey.
41
Figure 3: Employment (and Revenue) Changes
Panel A: National
Panel B: Two Sectors
Source: Authors’ calculations from Small Business Pulse Survey.
42
Figure 4: Declines in Revenue and Employment
Panel A: Phase 1
Panel B: Phases 2 and 3
Source: Authors’ calculations from Small Business Pulse Survey.
43
Figure 5: Hours Changes
Panel A: National
Panel B: Two Sectors
Source: Authors’ calculations from Small Business Pulse Survey.
44
Figure 6: Furloughs and Rehires
Panel A: National
Panel B: Sectors
Source: Authors’ calculations from Small Business Pulse Survey.
45
Figure 7: Remote Work
Panel A: National
Panel B: Sectors
Source: Authors’ calculations from Small Business Pulse Survey.
46
Figure 8: Pivoting to New Goods/Services and Delivery Mechanisms
Source: Authors’ calculations from Small Business Pulse Survey.
47
Figure 9: Online Platforms and Changes in Online Platforms
Panel A: Online Platforms by Sector
Panel B: Changes in Online Platforms by Sector
Notes: Panel B is percentage of businesses conditional on being in Panel A.
Source: Authors’ calculations from Small Business Pulse Survey.
48
Figure 10: Production, Supply Chain, and Delivery Problems
Source: Authors’ calculations from Small Business Pulse Survey.
49
Figure 11: Operating Capacity and Factors Impacting Capacity
Panel A: Change by Sector
Panel B: Factors by Sector
Source: Authors’ calculations from Small Business Pulse Survey.
50
Figure 12: Financial
Panel A: Cash on Hand: Two Sectors
Panel B: Received Assistance from Three Federal Programs by Sector
Source: Authors’ calculations from Small Business Pulse Survey.
51
Figure 13: Future Needs
Panel A. National
Panel B. Two Sectors
Source: Authors’ calculations from Small Business Pulse Survey.
52
Figure 14: Expectations
Panel A: National
Panel B: Sectoral
Panel C: Two Sectors
Source: Authors’ calculations from Small Business Pulse Survey.
53
Figure 15: Business Overall Sentiment and Expectations
Source: Authors’ calculations from Small Business Pulse Survey.
54
Figure 16: Business Sentiment and Revenue by State
Panel A: Overall Sentiment Index (Ten Largest States by Population)
Panel B: Revenue Changes (Declining Only)
Source: Authors’ calculations from Small Business Pulse Survey.
55
Figure 17: Dispersion by Sector and State in Overall Business Sentiment
Panel A: Box Plots
Panel B: Coefficient of Variation
Source: Authors’ calculations from Small Business Pulse Survey.
56
Figure 18: Persistence by Sector and State in Overall Business Sentiment
Panel A: Sectors
Panel B: States
Source: Authors’ calculations from Small Business Pulse Survey.
57
Figure 19: Monthly New Business Applications and (employer) Business Formations (indices=1
in 2006)
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
7/1/2004 3/1/2005 11/1/2005 7/1/2006 3/1/2007 11/1/2007 7/1/2008 3/1/2009 11/1/2009 7/1/2010 3/1/2011 11/1/2011 7/1/2012 3/1/2013 11/1/2013 7/1/2014 3/1/2015 11/1/2015 7/1/2016 3/1/2017 11/1/2017 7/1/2018 3/1/2019 11/1/2019 7/1/2020
Likely Employers Likely Nonemployers Business Formations
Source: Monthly BFS. Notes: Indices are relative to average in 2006 of each series. Likely
employers = HBA, Likely nonemployers=BA-HBA, Business Formations is the spliced actual
and projected series over the next eight quarters.
58
Figure 20. Business Applications by Sector, Weekly, 2019:w1 to 2020:w40
Source: Special Projects, Business Formation Statistics (https://www.census.gov/econ/bfs/projects.html)
59
Figure 21. Business Applications by Retail Sub-sector, 2019:w1-2020:w30
Source: Special Projects, Business Formation Statistics (https://www.census.gov/econ/bfs/projects.html)
60
Figure 22. Weekly Business Applications by Region, Year-over-Year Growth Rates (%), 2020
Source: Business Formation Statistics Public Data
61
Figure 23. Weekly Business Applications by State, Year-over-Year Growth Rates (%), 2020
Source: Business Formation Statistics Public Data
62
Figure 24. Weekly Dispersion in Year-over-Year Growth Rates across States, 2020
Source: Business Formation Statistics Public Data
Notes: SD, 90-10, IQR refer to standard deviation, difference between 90th and 10th percentiles, and interquartile
range, respectively.
63
Appendix A: Additional Figure
Figure A1: Revenue, Hours, and Employment Adjustments over Time
64
Appendix B: Location and Business Closures
We see evidence of adjustments at the extensive margin especially in the early weeks of
the survey collection period. Recall that the survey targets single location businesses so that
questions concerning closing a location may be closely related to closing the business. 44 The
survey actually collects both pieces of information. The survey asks about locations closures in
all three phases but with some important differences. In Phase 1, the question is “In the last
week, did this business temporarily close any of its locations for at least one day?” with yes or no
as possible responses. The question is expanded for Phases 2-3 to allow for reopening and
permanent closures of a location, but the response for temporarily closing a location is retained.
The survey collects information on permanent business closures in Phases 2-3 as part of the
response to the question about expectations. In contrast to the location closure question, the time
frame here is open-ended. Finally, the survey also collects information on expected permanent
business closures in Phases 2-3 as part of the response to the question about future needs over the
next six months.
Figure B.1 plots the time series patterns of location and business closures. Panel A shows
national results for location closures (left) and business closures (right). Starting with the
location closures, 41.4% of businesses temporarily closed a location in the first week of the
survey (late April). The percentage of businesses drops steadily over Phase 1 (ending at 17.9%)
and Phase 2 (ending at 1.8%) before rising slightly Phase 3 (ending at 3.8%). A response
category for permanent location closures was added in Phase 2; the national percentage of
permanent location closures hovers between 1.1-1.5% over both phases. To provide some
context for these figures, the establishment exit rates for firms smaller than 500 employees from
the Business Dynamics Statistics (BDS) are 9% in 2018 (the latest year available) and 12% in
2008 during the Great Recession.
The percentage of businesses responding to question about permanent business closure
are slightly higher than those responding to permanent location closure; nevertheless, the
difference between the two never exceeds 1 percentage point. (The figure with these is scaled 0-
44
The question asks about “locations” rather than “location” to allow for the possibility that even though our target
sample is small single-location businesses, a respondent may be answering for a business with multiple locations.
For example, the business may have added a location in the intervening time between our administrative data and
the survey outreach.
65
10 to accommodate the small magnitudes.) In the final week of Phase 3, 1.8% percent of
businesses have permanently closed. The annual firm closure rate for firms less than 500
employees is 8.2% in 2018 according to the BDS. Thus, the closure rate from SBPS is low
relative to BDS. The BDS relies on administrative records while the SBPS depends on closed
businesses responding to the survey. The expected permanent business closure measure does not
depend upon closed business responding to the survey. The expected permanent business
closures move between 4.5-5.5% over both Phases of the survey starting at 5.5% in the first
week of Phase 2 and ending at 4.5% in the last week of Phase 3. Crane et al. (2020) discuss these
expectations results and note “In aggregate, this pace of business closure [5% for 9/20-9/26
week] may not be markedly elevated relative to historical patterns, the annual exit rate for firms
with fewer than 500 employees averaged about 7.8 percent during 2015-2018 (p.21).”
Panel B repeats these exercises for the two sectors Finance and Insurance (blue) and
Accommodation and Food Services (green). Starting again with location closures, temporary
location closures (solid lines) start off at 54.6% for businesses in Accommodation and Food
Services and 31.1% for business in Finance and Insurance. These fall over Phase 2 before rising
very slightly in Phase 3, but the percentage for Accommodation and Food Services is always
higher than that for Finance and Insurance (and always significantly). By the end of Phase 3 in
early January, the percentage of businesses with temporary location closures in the last week are
9.4% for Accommodation and Food Services and 1.7% in Finance and Insurance.
The percentage of businesses with permanent location closures (dotted lines) are much
smaller and the differences between the two sectors are rarely significant. Again, the figure for
business closures is scaled 0-10 to accommodate small magnitudes. The percentage of businesses
that closed is about 1% for Finance and Insurance over Phases 2-3. This compares to about 3-4%
for businesses in Accommodation and Food Services (and these differences are significant in
every week except 9/13-9/19). In the final week of Phase 3, 4.0% of businesses Accommodation
and Food Services are permanently closed and 1.3% in Finance and Insurance. In terms of the
expected permanent business closures, over Phases 2-3, about 2-3% of businesses in Finance and
Insurance expect to permanently close in the next six months as compared to 10-11% in
Accommodation and Food Services. To put these numbers in context in 2018, the annual
establishment exit rates for firms smaller than 500 were: Finance and Insurance (8.9%) and
66
Accommodation and Food Services (9.5%). Crane et al. (2020) provide a sectoral comparison of
historical exit rates and expected exit rates. Their Figure 11 shows that Finance and Insurance is
below historical average, while Accommodation and Food Services is above.
Figure B.1: Location and Business Closures
A. National
B. Two Sectors
67
File and source
- File
- buffington-et-al-2021-ces-wp-21-06-high-frequency-business-dynamics-covid.pdf
- Size
- 2,636,474 bytes
- SHA-256
- 56de17433753f2a56e42b2016091b882eb6281d8bb95cfa70328439c81dee89c
- Original
- www2.census.gov