Oi Tracker Summary 2020 06
Summary
A five-page non-technical research summary, The Economic Impacts of COVID-19: Evidence from a New Public Database Built Using Private Sector Data, by Raj Chetty, John N. Friedman, Michael Stepner and the Opportunity Insights team. It describes the Opportunity Insights Economic Tracker, built from anonymized private-company data such as credit card processors and payroll firms. The summary reports six findings on spending by high-income households, small business revenue and job losses by ZIP code, state re-openings, stimulus payments and Paycheck Protection Program loans. It states that PPP had little effect on small business employment and estimates a cost per job saved of more than $370,000. It closes with policy implications, including targeted support for low-income workers.
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
Full text
N O N-TECH NIC AL RESE ARCH SU M M ARY N OVEM BER 2020
The Economic Impacts of COVID-19:
Evidence from a New Public Database Built Using Private Sector Data
RAJ CHETTY, JOHN N. FRIEDMAN, NATHAN IEL HENDREN, MICHAEL STEPNER,
AND THE OPPORTUNITY INSIGHTS TEAM
How has COVID-19 affected our economy KEY FINDINGS
and what policies will foster a recovery for
• As COVID-19 infections increased in March, high-income
all Americans? households sharply reduced their spending, primarily
Government surveys of households and businesses show that on services that require in-person interactions.
COVID-19 reduced GDP and increased unemployment sharply. • Because of this reduction in spending by high-income
These sources, while critical for measuring the scope of the crisis, consumers, businesses in the most affluent neighborhoods
are more limited in their capacity to inform policy decisions. In in America lost more than half of their revenue.
particular, national surveys are neither frequent nor large enough to
reveal how the crisis has affected specific areas or subgroups. • As these businesses lost revenue, they laid off their
employees, particularly low-income workers. Nearly
In response to this challenge, we created the Opportunity 50% of low-wage workers working in the highest-rent
Insights Economic Tracker, a freely available interactive website ZIP codes lost their jobs, compared with 30% in the
that measures economic activity at a granular level in real time. lowest-rent ZIP codes.
The tracker is built using anonymized data from several private
companies, such as credit card processors and payroll firms. • Policy efforts to date — stimulus payments to
From this data, we construct statistics on consumer spending, households and Paycheck Protection Program loans
employment rates, and other indicators by county, industry, and to small businesses — have not led to a rebound in
(pre-crisis) income level. These new statistics allow us to study how spending at the businesses that lost the most revenue.
COVID-19 has affected the economy with unprecedented precision. As a result, they have had a limited impact on the
employment rates of low-income workers.
Because government statistics reveal that almost all of the reduction
• In the long-term, the only way to drive economic
in GDP came from a reduction in consumer spending, we begin by
recovery is to invest in public health efforts that will
studying the drivers of this sharp drop in spending. We then examine
restore consumer confidence and spending.
the impacts of spending reductions on businesses and workers.
Finally, we analyze the effects of policies enacted to mitigate these • In the meantime, providing and extending targeted
economic impacts and discuss what our findings imply for policy assistance to low-income workers impacted by the
going forward. economic downturn (such as through unemployment
benefits) is critical for reducing hardship and addressing
disparities in COVID’s impacts.
The Economic Impacts of COVID-19: Evidence from a New Public Database Built Using Private Sector Data | PAGE 1
FIGURE 1: Consumer Spending Changes During COVID-19 Crisis, by Income Group
Consumer Spending Per Day ($ Billions)
10 This graph plots spending
for households in the top
vs. bottom 25 percent of the
8
income distribution in 2019
$-1.0 Billion and 2020. Income is imputed
(54% of Agg.
Spending Decline) based on the ZIP code where
6
households live.
Data Source: Affinity Solutions. Click here for
4 up-to-date data.
$+0.1 Billion
2
Feb 1 Mar 1 Apr 1 May 1 Jun 1 Jul 1 Aug 1 Sep 1 Oct 1
2019 Top Income Quartile 2020 Top Income Quartile
2019 Bottom Income Quartile 2020 Bottom Income Quartile
FINDING 1 FIGURE 2: Small Business Revenue Losses from
Jan to Apr 2020 by Zip Code in NYC
High-income households accounted for most of the
reduction in spending.
Most of the reduction in consumer spending resulted from reductions
by high-income households. As of May 31, more than two-thirds of
the total reduction in credit card spending since January had come
from households in the top 25 percent of the income distribution.
Meanwhile, households in the bottom 25 percent continued to
spend at the same levels they had before the crisis, as illustrated in
Figure 1.
High-income households cut spending primarily because of health
concerns rather than a loss of income or purchasing power.
Spending fell most on services that require in-person interaction and
thereby carry a risk of COVID-19 infection, such as transportation
and food services.
The pattern of spending reductions during this recession differs
sharply from that of prior recessions, during which spending on
services remained essentially unchanged while spending on durable
goods (e.g., new appliances or cars) fell sharply.
FINDING 2
Small business revenues declined most in affluent areas.
Next, we examine the impacts of the consumer spending shock
on businesses, recognizing that the sectors in which spending fell 73% 54% 13%
most consist of goods produced by small, local businesses (e.g.,
restaurants).
Small businesses in the most affluent ZIP codes — which tend to cater
to high-income customers — lost more than 50% of their revenue This map shows changes in small business revenues by
when COVID-19 hit, as compared with 30% in the least affluent (low
ZIP code. Red areas show places where businesses lost
rent) ZIP codes. This pattern is illustrated in the map of New York City
in Figure 2 below; businesses on the Upper East Side of Manhattan more revenue.
lost far more revenue than those in Harlem or the Bronx. Data Source: Womply
The Economic Impacts of COVID-19: Evidence from a New Public Database Built Using Private Sector Data | PAGE 2
FINDING 3 FIGURE 3: Reductions in Employment Rates of Low-
Income Workers by ZIP Code in NYC
Job losses at small businesses have been largest in
affluent areas.
As businesses lost revenue, they laid off their employees. In the highest-
rent ZIP codes, more than 50% of low-wage workers at small businesses
were laid off within two weeks after the COVID-19 crisis began; by
contrast, in the lowest-rent ZIP codes, fewer than 30% lost their jobs.
The map in Figure 3 illustrates this result by showing changes in
employment rates of low-income workers by ZIP code in New York. Low-
income people working in rich areas of the city were most likely to have
lost their jobs, mirroring the pattern of small business revenue losses.
Businesses in more affluent areas not only laid off more low-
wage workers but are also posting fewer jobs to hire new workers,
suggesting that the recovery may take longer in such areas.
EVALUATING POLICY RESPONSES
The government has implemented a number of policies in an effort
to mitigate the economic effects of the pandemic. How successful
were these efforts, especially in raising the employment levels of the
low-income workers who have experienced the largest job losses?
FINDING 4
State-ordered re-openings of economies had small
effects on economic activity. 75% 52% 13%
Some states began to re-open non-essential businesses as early as April
20, while others waited until the end of May. By comparing the trajectory
of early-opening states to similar states that remained closed, we find
that re-openings increased spending and revenues only modestly. This map shows changes in employment rates for low-
For example, Figure 4 shows that consumer spending patterns in wage workers by the ZIP code of their employer. Red
Colorado and New Mexico were nearly identical from February areas show places where workers were more likely to
through May despite the fact that Colorado began re-opening select lose their jobs.
businesses on May 1 and New Mexico on May 16. We also find little or
Data Source: Earnin
no impact of earlier re-openings on employment rates.
These findings show that it is the fear of COVID-19 itself, not
executive orders restricting business activity, that are the primary
cause of reduced economic activity and job loss.
FIGURE 4: Effects of Re-Opening on Consumer Spending: Colorado vs. New Mexico
New Mexico Colorado New Mexico Begins
20% Closing Closing Re-Opening
Colorado Begins
Re-Opening
Change from Pre-Crisis Level
0% This graph shows trends
in consumer spending
in Colorado and New
Mexico around the times
-20%
of the stay-at-home and
re-opening orders.
-40% Data Source: Affinity Solutions
New Mexico
Colorado
-60%
Feb 1 Feb 15 Feb 29 Mar 14 Mar 28 Apr 11 Apr 25 May 9 May 23 Jun 6
The Economic Impacts of COVID-19: Evidence from a New Public Database Built Using Private Sector Data | PAGE 3
These findings show that it is the fear of COVID-19 itself, not executive orders restricting
business activity, that are the primary cause of reduced economic activity and job loss.
FINDING 5
Stimulus payments increased spending substantially, particular, small businesses in affluent areas — received relatively
especially among low-income households. But they did little of the revenue from this surge in consumer spending. Perhaps
not lead to large gains for the businesses most affected as a result, employment growth has significantly lagged spending
by the crisis or to increases in employment. growth, leaving employment rates recovering at slow rates,
especially in affluent areas (Figure 5).
The CARES Act allocated nearly $300 billion in direct payments to
households, the majority of which arrived on April 15. We find that The national employment rate may rise as firms begin to rehire
spending increased sharply immediately following these deposits, workers; however, employment is likely to remain depressed in
especially among low-income households. more affluent counties, where local business revenues have not
recovered significantly despite the stimulus.
However, most of the additional spending induced by the stimulus
went to goods that require no in-person contact (e.g., orders of
durable goods). The businesses most affected by the crisis — in
FIGURE 5: Impact of Stimulus Payments on Business Revenue and Employment
20%
Change from Pre-Crisis Level
0%
1.0% This graph plots total
spending for high– and
-14.2% low-income consumers as
-20% well as employment levels
-21.5%
in high– and low– rent
-30.2%
areas around the time
-40% of the stimulus payment
provided through the
CARES Act.
Jan 15 Feb 1 Feb 15 Mar 1 Mar 15 Apr 1 Apr 15 May 1 May 15 Jun 1 Jun 15
Data Source: Earnin and Womply
Small Bus. Employment – Rent Q1 Small Bus. Revenue – Rent Q1
Small Bus. Employment – Rent Q4 Small Bus. Revenue – Rent Q4
FINDING 6
Loans to small businesses have had little impact on in Figure 6) and changes in payroll are very similar for smaller and
employment rates. larger firms, implying that the PPP had little effect on small business
employment to date. This may be because the businesses that took
Congress also devoted more than $500 billion to small business
up PPP assistance were in sectors that were less affected the crisis
loans as part of the Paycheck Protection Program (PPP), so named
and already intended to keep most of their workers on payroll.
because the loans do not need to be repaid if businesses maintain
Because of its limited effectiveness, we find that each job saved
employment at pre-crisis levels.
through the PPP program cost taxpayers more than $370,000.
Firms with fewer than 500 employees were eligible for PPP loans. In
The small increases in employment that have occurred at businesses
Figure 6, we assess the program’s effect on employment of low-wage
regardless of PPP eligibility appear to be attributable to the increase
workers by comparing employment patterns at firms above and
in consumer spending that resulted partly from the stimulus and
below the 500-worker eligibility cutoff. Both hours worked (shown
perhaps more broadly from receding health concerns.
The Economic Impacts of COVID-19: Evidence from a New Public Database Built Using Private Sector Data | PAGE 4
FIGURE 6: Impact of Paycheck Protection Program Loans on Employment
PPP Program Begins PPP Employment Impact: +1.78 p.p.
0% April 3
Cost Per Job Saved = $377K
This graph shows
Change from Pre-Crisis Level
changes in hours
-5%
worked for businesses
of various sizes around
-10% the time of the PPP
loan disbursements.
-15% Firms with less than
501-800 Employees 500 employees were
eligible for PPP.
-20%
100-500 Employees Data Source: Paychex and Earnin
-25%
Feb 15 Mar 15 Apr 15 May 15 Jun 15 Jul 15 Aug 15 Sep 15
WHAT THIS ANALYSIS MEANS FOR POLICY GOING FORWARD
A large initial reduction in spending by high-income households driven by health concerns about The only path to
COVID-19 has cascaded through to a loss of business for firms that cater to high-income customers,
leading to layoffs of low-wage workers at those businesses. Given this sequence of events, the only full economic
path to full economic recovery in the long run is to restore consumer confidence by focusing on health
policies that will address the virus itself. Traditional economic tools — loans to firms or blanket stimulus recovery in the
payments to households — may have weaker effects on restoring employment in the sectors and areas long run is to
where it fell most when the fundamental constraint on spending is health concerns. Relatedly, payroll
tax reductions also may not increase revenues among businesses that were hit hardest since they do not restore consumer
target relief to households that have lost the most income.
confidence by
In the meantime, it is critical to focus on supporting the many low-income individuals who have lost their focusing on health
jobs to limit hardship and further economic losses. For instance, extending unemployment benefits or
other programs that provide support specifically to those who have lost income may be more valuable policies that will
than making further stimulus payments to all households or loans to small businesses.
address the
Our findings also suggest that it may be useful to consider assistance targeted specifically to low-income
virus itself.
people who are employed (or were previously employed) in areas that have suffered the largest losses –
such as affluent, urban areas – since prior experience suggests that relatively few people move to other
labor markets to find new jobs after recessions, leading to long-term income losses in hard-hit areas.
Of course, these results could change over time: the recession may turn into a more traditional economic
shock as time passes, in which case tools such as stimulus and liquidity could become much more
impactful in areas with depressed spending. The tracker constructed here can be used to monitor
economic activity and evaluate policy impacts on an ongoing basis in this crisis and beyond, providing a
new tool to support economic policy in the age of big data.
Want to learn more?
Read the Paper
Explore the Data
All materials are freely available for use with citation
Based at Harvard University, Opportunity Insights identifies barriers to economic opportunity and develops scalable solutions that will empower
families throughout the United States to rise out of poverty. opportunityinsights.org
The Economic Impacts of COVID-19: Evidence from a New Public Database Built Using Private Sector Data | PAGE 5
File and source
- File
- oi-tracker-summary-2020-06.pdf
- Size
- 1,505,160 bytes
- SHA-256
- 7c67ce09aefae866c6b0c341366831ce792259c5701c28ab7d8398921b951fe8
- Our copy
- oi-tracker-summary-2020-06.pdf
- Original
- opportunityinsights.org