Keeping Business Alive: The Government Will Pay (Emmanuel Saez and Gabriel Zucman)

Summary

A policy paper by UC Berkeley economists Emmanuel Saez and Gabriel Zucman, Revised March 16, 2020, arguing that government should act as payer of last resort during the coronavirus shutdown, paying idle workers and the necessary maintenance costs of hibernating businesses to prevent layoffs and liquidations. They estimate that a nation-wide lockdown would cut direct output in the United States by around 30%, a 7.5% loss of annual GDP if it lasts a quarter, and that special unemployment insurance for idle workers plus payment of rent, utilities, interest and health insurance would cost around 3.75% points of GDP, financed by public debt. Loans, they argue, only smooth losses, and $1,000 checks to each household are poorly targeted; the program should be limited in time, e.g., 3 months. They cite Denmark and President Macron’s address of March 16, 2020 as moves in this direction.

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

Keeping Business Alive: The Government Will Pay
Emmanuel Saez and Gabriel Zucman, UC Berkeley
Revised March 16, 2020

The coronavirus threatens the world’s economic life. The most important message that
needs to come from heads of state immediately, even before any new law or complete
implementation details are provided, is: “Do not lay off your workers or liquidate your business.
Government will pay your idle workers and your necessary maintenance costs while you are
shutdown. Government money is coming soon.”1 This is crucial to stanch the flow of mass
layoffs and business destruction that is already starting. Why is such a measure necessary, and
how to implement it in practice?

Social distancing measures, essential to fight the epidemic, are sharply reducing demand
in many sectors such as transportation, restaurants, hotels, and entertainment. This direct
output loss is expected to be short, probably a few months. It is possible to roughly estimate
this loss by summing up output in sectors that are going to shut down (as of writing, nation-
wide lockdowns have already been decided in Italy, Spain, and France, among other countries).
In the United States, we estimate that the direct output loss will be around 30%. If this direct
loss lasts a quarter, the annual GDP loss will be 7.5%—comparable to a very severe recession.

The government cannot undo this direct output loss, but it can alleviate economic
hardship during the epidemic and prevent the direct output loss from causing lasting damage to
the economy. In other words, the government can prevent a very sharp but short recession
from becoming a long-lasting depression. Absent government actions, the direct output loss
will create large losses for businesses and will lead to mass layoffs. Many businesses and
workers do not have enough liquidity to weather dramatic shortfalls in demand. The risk is to
see many businesses liquidate, severely affecting workers’ families. The death of a business has
long-term costs: the links between entrepreneurs, workers, and customers are destroyed and
often need to be rebuilt from scratch; laid off workers need to find new jobs.2 Keeping
businesses alive through this crisis and making sure workers continue to receive their paychecks
is essential—even for businesses and workers that have to remain idle due to social distancing. 3

In the context of this pandemic, we need a new form of social insurance, one that
directly helps both workers and businesses. The most direct way to provide this insurance is to
have the government act as a payer of last resort so that hibernating businesses can keep
paying their workers (instead of laying them off) and can keep paying their necessary bills such
as rent, utilities, interest, etc. (instead of going bankrupt).

1 Indeed, as we were writing this article, several European countries have moved in this direction, most notably

Denmark with its temporary wage compensation scheme and France with President Macron’s address on March
16, 2020.
2 This point on direct losses and indirect losses through feedback effects is well explained in Gourinchas, Pierre-

Olivier “Flattening the Pandemic and Recession Curves”, March 13, 2020.
3 Standard economic models assume zero transaction costs for hiring workers, finding customers, deploying

capital, etc. and hence cannot capture the issue at hand well.
In practice, the key step is to make sure that cash flows to idle workers and businesses
immediately. Payments should be set in the simplest form. Idle workers should immediately
start receiving special unemployment insurance benefits so that they are no longer a cost to
their employers—even though they stay formally employed—and no re-hiring process is
needed once they can come back to work. The unemployment insurance system is already up
and running. This makes it possible to compute and deliver compensation to idle workers. Self-
employed individuals (such as gig workers) could report themselves as idle and be eligible for
this special unemployment insurance. In case of partial idling, unemployment insurance
benefits would be prorated. Unemployment insurance benefits are progressive, since they
replace a higher fraction of earnings for low-paid workers. This is a desirable feature, as low-
paid workers are more likely to be affected by the lockdown (i.e., less likely to be able to work
from home) and less likely to have savings to replace a temporary loss in earnings.

In the payer-of-last-resort program we envision, businesses on lockdown would report
their monthly necessary costs of maintenance and receive payment from the government.
Necessary costs are rent, utility payments, interest on debt, health insurance of idle workers,
and other costs that are vital for the maintenance of the business even if the business is no
longer operating. For partially shutdown sectors, the government would pay a fraction of the
maintenance costs. The amounts don’t need to be exact; verification and correction can take
place once the lockdown is over. Any excess government payment could be transformed into
an interest-free interest loan that the government could recoup over several years.

The key advantage of this policy is that businesses can hibernate without bleeding cash
and hence without risking bankruptcy. The reason why such a policy would work in the case of
the coronavirus pandemic is twofold. First, it is clear what is driving the shock: a health crisis
that has nothing to do with any business’s decision and will be temporary. Second, different
industries are affected differently. That’s in contrast to normal recessions, where the drop in
demand is widely spread and has no clear timeline.

Providing liquidity—in the form of interest-free loans, for example—can help businesses
and laid off workers weather the storm, but this policy is insufficient. Loans do not compensate
businesses and workers for their losses; loans just allow them to smooth costs over a longer
time horizon. In the case of the coronavirus crisis, however, it makes sense for the government
to compensate businesses and workers for their losses through social insurance so that each
business can re-emerge almost intact after the hibernation due to social distancing ends, rather
than loaded with a heavy coronavirus debt burden.

How much would such a payer-of-last-resort program cost? Based on national account
statistics by industry, we estimate that with a nation-wide lockdown, up to 30% of aggregate
demand could evaporate in the US over the next 3 months, leading to a 7.5% drop in annual
GDP. Compensating idle workers and necessary business maintenance costs would involve
government payments of around half of this total. Unemployment insurance replaces about 50-
60% of wages, and essential maintenance costs of businesses are probably less than half of
their normal operating costs (for example, non-flying planes do not burn fuel). The total cost for
the government would be around 3.75% points of GDP, financed via an increase in public debt.
The direct output loss from social distancing measures would in effect be put on the
government’s tab, i.e., socialized.

Current proposals to deal with the economic consequences of the pandemic do not go
far enough or are not well targeted to the ailing sectors. Business loans help businesses but do
not compensate them for their losses. Postponing tax payments helps with liquidity but is not
well targeted, since it also benefits individuals and businesses not directly affected by the
pandemic. Direct payments to individuals (such as $1,000 checks to each household) help
alleviate temporary economic hardship but this policy is poorly targeted as well: it’s too little
for those who lose their jobs, and it is not needed for those who don’t. During social distancing,
the goal should not be to increase aggregate demand, since people can no longer spend on
many goods and services. Unemployment insurance and paid sick leave policies come closest to
helping laid off workers and those unable to work, but they do not prevent layoffs and do not
help businesses.

A payer-of-last-resort program will work if it is limited in time (e.g., 3 months), so that
the cost remains manageable and business decisions are not affected. It would not fully offset
the economic cost of the coronavirus. No matter what governments do, there will be real
output losses. Even if airlines workers are paid, the plane rides won’t happen. For other sectors,
supply-chains distortions will happen no matter what, due, e.g., to quarantine measures. But a
payer-of-last-resort program would alleviate the hardship on workers and businesses. It would
maintain the cash flow for families and businesses, so that the coronavirus shock has no
secondary impacts on demand—such as laid-off workers cutting down on consumption—and a
quick rebound can take place once demand comes back. Business activity is on hold today, but
with an intravenous cash flow, it can be kept alive until the health crisis is over.

File and source

File: raw.pdf
Size: 118,330 bytes
SHA-256: b51abef38850c6a5ce4326334e05049e52b77d91e5978b52385aaac55c744507
Our copy: raw.pdf
Original: gabriel-zucman.eu