The Pandemic or the Recession: Which is More Costly?

Published April 13, 2020

The United States has taken unprecedented actions to halt the spread of the novel coronavirus and the illness it causes, COVID-19.  Nonessential businesses have been closed, travel severely restricted, and sports and recreation halted.  Most Americans cannot leave their homes except for government approved purposes.  These actions have been spurred by a virus that has sickened over 1.2 million worldwide and caused at least 68,000 deaths.

Yet the economic costs of the responses to the pandemic are enormous; in the past two weeks, almost 10 million new unemployment claims have been filed.  The Federal Reserve Bank of St. Louis projects that continuation of the current shutdown could result in 30% unemployment and a 50% decline in GDP by June, depths which exceed the Great Depression. The restrictions on business and life may need to be kept in place for years, not weeks, to prevent a full out pandemic.

Which would be worse, an unmitigated pandemic or the economic depression required to contain the pandemic?  Economics offers a way to compare these costs using dollars as the unit of measurement.  I will offer some hopefully easy to follow calculations, using the worst worst-case projections I have seen regarding the pandemic and the economic shutdown.  To preview, the costs of either are staggering.

The primary impact of the pandemic will be premature or avoidable deaths.  We must put the tragedy of a life lost into dollars to compare this with lost GDP.  Economists do this using the value of a statistical life (VSL).  The concept uses the many different choices people make in their lives between a risk of death and money, or risk and time.  People might say that human life is priceless, and yet our actions reveal a different attitude.  Almost everybody will sometimes take a turn at faster than the recommended speed on the road sign, accepting a slightly higher risk of death to arrive at our destination sooner (which we probably already did by choosing to drive the first place).

Each one of these tradeoffs provides insight on an implicit value people place on their lives.  A little math helps us clarify the tradeoff.  If spending $50 can avoid a 1 in 100,000 risk of death, the person who avoids the risk implicitly values their life at $5 million or more (which is 50/.000001), while the person taking the risk implicitly values their life at $5 million or less.  There is no “correct” value of life and different people will value life at different levels.  For government actions where lives are in the balance but we do not know the identity of the specific individuals at risk, the VSL provides probably the best way to value the lives.  Safety policies saving lives at a cost of less than the “average” VSL people use in their personal decisions are good investments; policies with a cost per life saved exceeding this value are poor investments.  The Federal government has adopted the VSL as a way to value lives saved in benefit-cost analysis, using a value selected from reviewing dozens of studies.  I will apply a VSL of $10 million, which rounds up the value used by the Federal government.

The worst of the worst-case models of the pandemic is a study from Imperial College in London.  They project 2.2 million deaths in the U.S. in a five month pandemic with no social distancing in place.  Applying the $10 million VSL gives a cost of the pandemic of $22 trillion.

The mortality cost of COVID-19 should be adjusted for the age of the victims.  The values of life people implicitly reveal in marketplace decisions reflect values for adults throughout the age distribution.  The $10 million VSL probably reflects a value applied by a person at the median age in the U.S., 38 years.  The young and old have different numbers of expected years of life at stake in safety decisions.  Alternatively, a public policy preventing the deaths of children will save more years of life than an action that prevents the deaths of retirees.  This is an unavoidable reality.  The elderly face a much greater risk of death from COVID-19;  the fatality rate used in the Imperial College study for persons 80 and older is over 4,500 times higher than that for children under age 10.  A large portion of deaths will be among persons 70 and older.  One approach would apply a value of a statistical year of life saved and use remaining years of life expectation for victims of different ages.  More simply we may choose to apply a somewhat lower VSL to reflect the age vulnerability.  A reduction of the VSL applied lowers the mortality cost of the pandemic proportionally.

Please note: Adjusting for expected years of life lost does not mean that every death from COVID-19 is not tragic.  Nor would this imply that the elderly are “expendable” in any way.

On the cost side we’ll consider the reduction in GDP due to reduced economic activity

plus the lost consumer surplus on the consumption and entertainment that we do not enjoy.  Consumer surplus is the difference between a person’s value for a good and what they pay for it.

Let’s take the St. Louis Federal Reserve Bank’s projection of a 50% decline ind GDP.  GDP in 2019 was $21.4 trillion, so the decline would be $10.7 trillion over a full year.  The labor force is 165 million so a 30% unemployment rate implies about 50 million persons out of work, up from 6 million prior to the pandemic.  A rough estimate is that consumer surplus on the first units of a good is probably equal to the price; if the price of shoes is $100, the consumer surplus on the first pairs sold would be about $100.  If we assume linear demand, the consumer surplus is a triangle and its area is (since surplus equals price on the first units) half of GDP.  Consumer surplus might be $10.7 trillion for a year.  With a triangle, we might infer that the surplus lost by losing half of the economic activity (the 50% reduction in GDP) would be the thin end of the triangle, or perhaps a quarter of the total.  But the quantity of many activities, like concerts, sports, zoos, museums, and dining at restaurants, has been reduced to zero.  We are losing first unit consumer surplus on many goods and particularly forms of entertainment.  So let’s say we would lose half of the consumer surplus, or about $5.3 trillion annually.  We have an economic cost per year of $16 trillion.

The Imperial College analysis shows that a shutdown must remain in place until a vaccine is developed to avoid the worst case pandemic, which would be at least 18 months.  The study does not claim that an economic shutdown of this length will occur, but flattening the curve retains nearly all of the pandemic potential even after 6 or 12 months.  The only way to prevent the unmitigated outbreak is continued social distancing.  The economic shutdown must be continued to prevent as opposed to delay the 2.2 million deaths.  The economic cost in lost GDP and consumer surplus of an l8 moth shutdown would be $24 trillion.

These are just the headliner costs on each side.  There are, unfortunately, many others.  I will just list them here.  The costs of the pandemic would include:

  • 260 million total cases of COVID-19 in the U.S.;
  • over 10 million hospitalizations with an average stay of 10 days;
  • over 3 million cases requiring intensive care;
  • medical costs of all the treatment; hospitalization costs $10,000 or more per night;
  • fear and emotional pain of patients and loved ones suffering from COVID-19, waiting to see if their case will worsen into life threatening pneumonia;
  • lost productivity from persons unable to work while ill, which could amount to 20% of the workforce sick at one time.

Significant other costs of shutting down the economy and society include:

  • the social and psychological costs of unemployment, which include stress, depression, substance abuse, domestic abuse, marital dissolution, and even suicide;
  • thousands of businesses, representing the hopes, dreams, and hard work of the founders, closed by government order, many to never reopen after an 18 month hiatus;
  • Deaths from poverty.  Poor people are less able to afford safety- and health-related purchases like car repairs, fruits and vegetables, and health insurance.  Research shows that $50 million in government imposed costs might produce one extra death; lost GDP during the shutdown would yield 300,000 extra deaths;
  • Americans ordered not to leave their homes have lost a significant element of their personal autonomy.  Each coercively imposed decision generates a cost apart from the consequences of the action.


What are the takeaways here?  The quantified costs on each side of the ledger could be $20 trillion.  Add in the other costs and the total on either side might reach $30 trillion.  This is a really, really horrible tradeoff.  The governors considering stay at home, isolation and quarantines orders face a horrific decision; for the largest states, the costs could easily be $1 trillion either way.  I have compared worst cases, so one consolation may be that the option we choose will not turn out to be this costly.

The case for saving lives over the economy is not obvious.  The costs are enormous from either the unmitigated pandemic or year and a half long shut down of the economy.  I would not favor strict cost-benefit analysis on important questions; I recommend using the numbers as a starting point.  The headliner costs are pretty similar.  Reasonable people will easily disagree on which prospect seems less worse.

I have considered the two extreme cases of an unmitigated pandemic and the shutdown currently in place being continued for a full 18 months until a vaccine might be available.  These extremes are corner solutions.  There are a list of intermediate policies that could produce an outcome between the two extremes; there are probably at least 20 and realistically 100 dimensions of action on which we could choose.  Crafting an intermediate policy would hopefully allow us to avoid most of the potential deaths and the bulk of the economic costs.  For example, after a couple of additional months of extreme social distancing, we may have the capacity to test more widely, have more masks available, have expanded critical care capacity, and knowledge about which drugs most effectively treat COVID-19.  We might also selectively protect the vulnerable population while reopening the economy.  Our goal should be to minimize the costs from COVID-19, which are the mortality and health costs plus the economic and societal costs.  There’s a saying in economics which has been attributed to Frank Knight that to call a situation hopeless is to call it ideal.  Our terrible tradeoff may appear hopeless, but we need to focus on policies which will hopefully allow us to avoid the bulk of these health and economic costs.