Covid-19: Measuring the Urgency to Act

The scale of the Covid-19 problem is beyond what any of us have had to deal with in our lifetimes. It is an unfamiliar situation, and will require unfamiliar actions to solve. This virus is a threat that moves faster than our human institutions are set up to respond to.

If we cannot come together decisively and intelligently, millions will die. I do not pull punches on these numbers because I want everyone to be aware of the unprecedented scale of the situation.

In a few weeks, Covid-19 has completely destroyed wide swathes of the nation’s economy. Restaurants, hospitality, transportation — representing 20% of American workers — gone.

The Imperial College of London’s report states that 2.2M Americans are projected to die if we do nothing. 1.2M will die if we “flatten the curve” to treat people without overwhelming the health care system. I think their report is flawed in that it doesn’t consider cheaper mechanisms of suppression once case numbers are low enough (see this response from the New England Complex Systems Institute). However, it’s the most authoritative report I’ve seen, and is being used to guide policy decisions.

If we do nothing, from the Imperial College report

Every country in the world is faced with Hobson’s moral choice of staggering proportion. Do we

A) Allow this virus to run through our population? An uncontrolled spread would overwhelm hospital capacity 30 times over and kill millions. Even a controlled spread (per the Imperial College study) would overwhelm hospitals by a factor of 8 and still result in over a million deaths.

B) Go into societal lockdown to hold off the epidemiological tsunami? Tens of millions lose their jobs, the economy melts down, we all suffer from social isolation. We can maybe relax the lockdown after a couple months and maintain that posture more cheaply through testing and contact tracing (see Coronavirus: The Hammer and the Dance). But some form of lockdown will be needed for around 18 months (time to a vaccine).

How does one begin to navigate this unfamiliar moral terrain? What do the choices look like if we put some numbers to them? Specifically, what is the financial cost of shutting down? And what is the cost of a death?

We Should Pay Almost Anything to Shorten the Shut Down

Goldman Sachs projected that the US economy will shrink 24% next quarter. Our GDP last year was $21.4 trillion. 24% of that is a little over $5 trillion. That’s what this lockdown is costing us. About $14 billion every day.

I hope that we can relax the lockdown after a month or two, and then carefully open up the economy. If the case numbers are low enough, we should be able to keep the numbers down. We will need to do widespread testing and contact tracing. But whether the current intense lockdown lasts for 1 month or 18, the daily cost is astronomical. We should spend whatever it takes to shorten it.

For example, IndieBio is a biotech accelerator. They are currently seeking 8 companies working on Covid-19. They have committed to giving each company $250K. Should the government multiply these funds by a factor of 10? That would cost $20M. Divide by $14B per day, and the answer is “yes” if we expect this investment to shorten the duration of the shut down by merely 2 minutes.

Economic Slowdown also Costs Human Lives

Most of us recoil at having to place a value in dollars to a human life. However, this question is at the heart of choosing the right path forward. The flu kills around 50,000 people per year, and we don’t halt the economy. But we do for 2 million people. There’s an implicit value attributed to life in these decisions. For clearer thinking, we should make it explicit.

How much is the value of a human life in dollars? This question gets into some existential quicksand, but it is an important question that arises in many policy questions. Should we reduce the speed limit from 65 to 55? That depends on the balance of economic productivity gain from less time driving with an increase in highway fatalities. The U.S. Office of Management and Budget puts the value of a human life in the range of $7 million to $9 million. Let’s use the midpoint — $8M — for the rest of this post.

Let’s look at some implications.

The amount we are spending every day to keep this virus from exponentially spreading ($14B) is equal to 1750 lives. Note: these are statistical lives. These are 1750 people who would not die if our economy were fully functional. They’re people who drown because there wasn’t a life guard, commit suicide because of depression from poverty, die from a heart attack because there was no ambulance to pick them up.

The $14B we spend every day holds off potentially millions of deaths. But it comes at a cost that can be denominated in dollars or lives. By using lives, we are at least comparing apples to apples in our choices. In terms of lives, we are currently sacrificing 1750 souls a day to buy us time to build masks and ventilators, and to research vaccines.

In contrast, around 220 Americans died every day during WW2. So this present emergency is about 8x more acute than the second world war. In this light, Trump’s refusal to use war powers to mobilize industry seems misguided.

This is what 1000 people look like, economically equivalent to 14 hours of shut down

Here’s another implication. 2.2M deaths at $8M value per life is $17.6 trillion. That means we are morally better off locking the country down — at the cost of 1750 lives per day — so long as a vaccine is deployed within 3.5 years. Ouch.

Edit Mar 28, 2020: if we use a framework like Disability Adjusted Life Years, the cost of a Covid-19 death is likely much lower. The average age of the deceased are in their late 70s, so a better estimate would be 10 years of life lost at $100K per year, equally $1M vs. $8M. However, we should account for the number of survivors with permanent lung damage, but we don’t have enough data yet to make a meaningful estimate.

By the way, this post talks only about American lives and American economic harm. America is 5% of the world’s population and 15% of the world’s GDP. This is a global problem, and the cost of a global shut down is around $100B per day. That’s our entire Defense spending in a week.

We Really Are All in this Together

There’s a level of interconnectedness in this pandemic that is new. If a neighbor can’t afford insulin and dies from diabetes, it does not directly affect others. But if this neighbor can’t get tested for Covid-19, or can’t afford to shelter at home, and spreads the disease in our community — we are all directly affected. In this increased vulnerability, it’s very much in our interests for our taxes to support UPS drivers with the coronavirus to stay home instead of them delivering packages out of economic necessity.

Some of the implications of these lessons are:

The Feds Should Pay

This $5T annual cost of shutting down does not equal a $5T market that private enterprise can tap. Rather, it reflects an opportunity for savings from reducing the duration of the lockdown that goes to all of society.

As such, private industry is not well positioned to bear this cost. The free rider problem removes the incentive for any particular company to spend money to shorten the shut down for everyone.

What’s needed is strong action from the Federal Government. The feds should:

• Spend the money to requisition billions of N95 masks and PPE, and a few million ventilators — and then the free market can make them.
• Offer multi-billion dollar prizes to whoever comes up with a revolutionary vaccination or therapeutic.
• Identify all the top biotech institutions (such as funds and research universities), and provide each one with a billion dollars and a phone number to call to get any regulatory obstacle removed.

A rational society (or a rational government as a proxy for society) should be throwing almost unlimited resources at any company, any technology, any process that has a credible chance of reducing the duration of our lockdown.

We Need Coordination from the Federal Government

America is filled with latent, decentralized brilliance. There is a vast desire to chip in and solve this problem.

But we need a coherent plan from the top. Do we close borders, both internationally as well as between states? Do we need to lockdown everywhere, or just around a few hot spots? And when will these lockdowns end? What has to be in place in terms of testing and tracing before we can relax the pressure? Is there a central location that collects data on infections and movements so that epidemiologists can mine the data and make intelligent recommendations? These decisions must be coordinated and enforced at the Federal level.

For better or worse, we need to fight this battle in America and with the Trump administration. They have squandered a lot of trust by downplaying the virus until recently. We can’t afford to waste any more time. We should demand a plan from them that will give citizens and private industry sufficient structure to figure out how they can best contribute to eliminating Covid-19.

Get Rid of Regulations that Slow Us Down

The scale of this problem — in dollars and lives — is so vast that nothing is more important than speed. The FDA needs to offer a blanket “right to try” for all Covid-19 therapeutics. We should relax HIPAA requirements to allow seamless data sharing among doctors on treatments.

Since a vaccine is the most likely way out of this situation that doesn’t involve over a million deaths, let’s look at how we can make one appear as soon as possible.

The FDA requires vaccinations to go through 3 phases of clinical trials.

Phase I administers the candidate vaccine to a small group (less than 100 people) with the goal of determining whether the candidate vaccine is safe and to learn more about the responses it provokes among test subjects.

Phase II, which includes hundreds of human test subjects, aims to deliver more information about safety, immunogenicity, immunization schedule and dose size.

Phase III, which can include thousands or tens of thousands of test subjects, continues to measure the safety (rare side effects sometimes don’t appear in smaller groups) and effectiveness of the candidate vaccine.

In the normal world, this makes sense. But we’re in a bizzaro Covid-19 world. How do these phases look like when each day of delay costs 1750 lives?

Let’s say we skip phases 1 and 2 and go straight to 3, and we use a large population of 100,000 subjects. And let’s say we really messed up and killed 10% of these subjects. In the normal world, that would be colossally, criminally negligent.

However, in the new reality of Covid-19 moral calculus, killing 10,000 people in search of a vaccine is a reasonable tradeoff if it brings this epidemic to an end 6 days earlier. A real nightmare.

This doesn’t mean we should rush any vaccine into widespread distribution. Before we give it to billions of people, we need to make sure it is effective and doesn’t cause harm worse than the disease. So there is some minimum time to do research. However, the potential harm to the Phase 3 candidates is insignificant compared to the benefit of releasing a working vaccine even days earlier.

And normally, industry wouldn’t begin to scale for production until after Phase 3. Let’s suppose it takes 3 months and costs $1B to scale for production. If we wait until after phase 3, those 90 days cost us $1.26 trillion. So if there’s even a 1 in 1000 chance that a certain vaccine could end this disease, we should spend the billion now to tool for production.

100 Manhattan Projects

The WW2 project to develop the atomic bomb cost $23B in today’s dollars. If a shut down costs us $5T a year, and let’s say we’re willing to spend half that to end it, that’s $2.5T. That’s over 100 Manhattan Projects.

We should demand from the government a coherent plan for getting out of this plague state. This plan should have multiple parallel tracks, in case of unforeseen failure. We can’t afford another misstep, like the CDC’s test kits not working and we had no backup. We should do critical path analysis of these plans, and throw a Manhattan Project at every bottleneck.

It is possible to have a functional city under Covid-19. China has brought their new infections to zero, and this video shows what life there is like.

We need Manhattan Projects to

• Make all the equipment hospitals need.
• Develop hundreds of millions of tests (both to see if you have it as well as serontological tests to see if you have recovered from it).
• Sends teams to China, South Korea, Taiwan, Singapore and Hong Kong to learn how to control the spread.
• Create infrastructure to allow us to open the economy as much as possible while keeping R0 under 1.
• Develop working therapeutics.
• Produce an effective vaccine.
• Set up systems to prevent the next Covid-19, without creating a dystopian surveillance state.

We can do this, but the scale of the problem is immense. We need federal resources to do it, and we have a moral imperative to act now.