COVID-19 is an upper respiratory tract infection similar to other coronaviruses. Its main symptoms are fever, cough, and shortness of breath. The fact that the symptoms are non-specific is what makes it very difficult (if not impossible) to identify without having a test. The other thing that makes it difficult to track or contain is that most people who contract COVID-19 either have mild symptoms or none at all, and even if they exhibit symptoms, it could take several days before doing so. On top of that, the rate of transmission (R0) is higher than the flu, making it about twice as likely to catch COVID-19 than the flu.
When I discussed how much we should worry about coronavirus about three weeks ago, the percent of mild cases was estimated at 80 percent. A recent CDC report put the hospitalization rate at 12 percent, which is higher than the initial estimated 20 percent. The number of asymptomatic or mild cases makes it difficult to determine what the fatality rate is, especially with a lack of testing in the United States. The crude death rate, which is calculated by number of deaths divided by number of confirmed cases (key word is "confirmed), is 1.5 percent. However, given that so many mild and asymptomatic cases are going uncounted, epidemiologists generally believe the fatality rate is lower. How much lower? That one is tricky to answer, although some are putting it at 1.0 percent, which is ten times higher than the flu. Conversely, a study from the New England Journal of Medicine (Fauci et al., 2020) projected that the fatality rate could be as low as 0.1 percent, which would be more on par with the flu.
A March 16 report from the Imperial College in London doesn't help with calming fears (Ferguson et al., 2020). That report predicted a worst-case scenario of 2.2 million deaths in the United States unless extreme disease-suppression measures (e.g., self-imposed quarantines, school closures, major social distancing) are taken for months. This worst-case scenario assumes that nothing is done, which is clearly not in line with the currently reality. Another Imperial College study suggests that getting COVID-19 is the equivalent of packing in a year's worth of risk into a couple of weeks (Walker et al., 2020). That sounds pretty dire, doesn't it?
Thankfully, worst-case scenarios are uncommon. When NIAID Director Dr. Anthony Fauci when asked by CNN reporter Jake Tapper about how many COVID-19 cases there would be, Fauci said the following:
"To be honest with you, we don't really have any firm idea. There are things called models. And when someone creates a model they put in various assumptions. And the model is only as good and as accurate as your assumptions. And whenever the modelers come in, they give a worst-case scenario and a best-case scenario. Generally, the reality is somewhere in the middle. I have never seen a model of the diseases that I have dealt where the worst-case scenario actually came out. They always overshoot. So when you use numbers like a million, a million-and-a-half, two million [deaths], that almost certainly is off the chart. Now, it's not impossible, but very, very unlikely."
Fauci proceeded to estimate that with our current trajectory, there would be 100,000 to 200,000 deaths, which is much lower than the CDC's worst-case scenario of 1.7 million U.S. deaths. A study from University of Washington School of Medicine published last week models a significantly less dire scenario (Murray, 2020; see March 30 estimate updates here). What is scariest about this report is that we are already approaching the peak in which hospitals become overwhelmed and there is a lack of beds and ventilators (see below). The good news is this peak going to be short-lived. More than that, the University of Washington has a baseline estimate of about 83,967 deaths (range estimate of 36,614 to 152,852 deaths).
This model is well within Fauci's estimate, as well as the estimate of 100,000-240,000 deaths, which was released just yesterday from the White House Coronavirus Task Force. Aside from University of Washington being a Top 20 medical school, the University of Washington model is more believable than the Imperial College one because it is a trajectory based on current mitigation and suppression policies, not on an unrealistic scenario of "doing nothing." I admit that it would be nice to have no deaths whatsoever. However, we are in the midst of a pandemic, so evading an even scarier number of 2.2 million like the Imperial College's worst-case scenario provides a silver lining.
But again, we don't know for certain how this situation will play out because the situation is evolving. That uncertainty certainly contributes to the fears and anxieties people have on COVID-19. The virus' properties are not fully understood, and projections could change as we get more information. Are we looking at a single peak and we're done with COVID-19? Will it subside because of warmer weather and come back with a vengeance in the fall? Also, the role of those who are asymptomatic is not clear. Oxford University released a model earlier this month (Lourenco et al., 2020) suggesting that half of the population of United Kingdom has already been infected. Is this a rosy projection or is the percentage of those who are asymptomatic higher than anticipated? We won't know the answer until there is more serology testing for coronavirus antibodies, but if true, it would imply that COVID-19 is much less deadly than initially anticipated and that we could remove the lockdowns. Is there immunity once exposed to COVID-19 or is it possible to catch it again? What is the rate of infection? There are enough unanswered, vital questions that it causing stress on a global level.
Looking at Economic Impact
Keep in mind that these previously mentioned statistics have a U.S. focus since that is where I live. As of 10:12 EST today, the global death toll is at 44,156 (Johns Hopkins). However, what is scary on the public health front is the worst-case scenario that millions could die globally while obliterating healthcare systems throughout the world. I think it is easier for people to think of the health-related effects of stopping a pandemic because a pathogen is most commonly associated with health. Conversely, COVID-19, or rather the response to COVID-19, is wreaking havoc elsewhere: the economy.
Most countries in the Western world, and indeed a number of countries in the developing world, have chosen drastic measures to slow the spread of COVID-19, whether that is closing down multiple industries or enacting either shelter-in-place, a nationwide lockdown, or a quarantine. When large swathes of the economy are being shut down and millions are not working, economic downturn is all too predictable.
When people hear about the economy, there are those who think it comes down to a callous reduction of dollars and cents. It's much more than that. Macroeconomic and microeconomic fundamentals have direct impact on one's livelihood. A livelihood is what is required to put food on the table, a roof over one's head, buy clothes, pay for education and healthcare...you get the idea.
What makes this situation particularly nerve-racking is that coronavirus is not just a health crisis. It is also an economic crisis. At this point, it seems that we are already heading for. Whether we look at economic forecasts from Goldman Sachs (also see here), J.P. Morgan, Morgan Stanley, Morningstar, Deutsche Bank, Harvard Business Review, Standard and Poor's, Fitch, Pacific Investment Management (also see here), Oxford Economics, or the OECD, it's not looking good. The International Monetary Fund (IMF) Director is already saying that 2020 is going to be a global recession with a comparable impact to the Great Recession, but is expected to recover in 2021. Former Federal Reserve Chair Ben Bernanke says that the shutdowns are more analogous to a natural disaster than they are the Great Depression.
The unknown nature of coronavirus is tied to the unknown nature of how the economic downturn is going to play out in the upcoming months. Much like there are unknowns with COVID-19, there are unknowns with the economy. Most of the current projections I have seen are showing a sharp economic contraction in Q2 (April-June), but are to start recovering either in the latter half of 2020 or early 2021. This is what is referred to as a V-shaped recovery.
A V-shaped recovery is looking to be the baseline, but there remain unknowns. How will households react? What impacts will monetary and fiscal policy have? How big of a hit will capital formation, labor participation, supply chains, and productivity growth take? How long will lockdowns, travel bans, and cancellations of large events last? While a V-shaped recovery is still the baseline scenario, we could be looking at a more prolonged U-shaped recovery, or even worse, an L-shaped recovery.
We already have started to see certain indicators get worse than initially projected. To cite a couple of examples: Last month, economic forecasting predicted that the GDP would fall, but still not so badly that it would automatically result in a recession. Now we find ourselves on a downward trajectory that will most likely result in a recession. Consensus estimates for jobless claims in the U.S. were supposed to be at 1.6 million claims last week, but ended up being 3.3 million. That number is scary because with all the people staying at home, the unemployment number has only begun to rise. Pew Research survey results already show that a third of U.S. citizens have either had been laid off or had their pay cut.
While it is possible that the situation gets worse and worse, the truth is that we do not know what the depth or the duration of the economic downturn will be here. The magnitude of quarantines and lockdowns taking place worldwide are historically unprecedented. But if we're looking at worst-case scenarios for public health, we should look at the worst-case economic scenario for fairness' sake. What are some of the more salient points to be made about how bad the economic situation can be?
- High unemployment rates. As already stated, livelihood is a vital means to not just survive, but to live life to the fullest. The more people that are deprived of a livelihood, the more people feel the hurt. The St. Louis Federal Reserve President James Bullard said that the unemployment rate could hit as high as 30 percent. A research paper released from the St. Louis Federal Reserve last week included a back-of-the-envelope estimate, and said that unemployment for the second quarter (April-June) could be 32.1 percent. To contextualize, the Great Depression reached 24.7 percent at its peak in 1933, which means we could be looking at an unprecedented amount of unemployment.
- Much like with the public health aspect, it is likely that we would reach 32.1 percent unemployment. Plus, it would be likely that as lockdowns are lifted, a good number of people would resume working, and much of the remainder would eventually be working as we recover from the recession. At the same time, it does underscore what happens when you shut down "non-essential" businesses.
- Disproportionate economic pain. Given the nature of the economic downturn, our response to fighting coronavirus will have uneven economic harm. As the Brookings Institution brings up, this will disproportionately affect low-wage workers. Small- and medium-sized businesses will feel more pain than larger enterprises. Those who have salaried pay, have paid sick leave, and can work remotely will fare better. While such industries as healthcare, grocery, and streaming services are doing better than ever, other such industries as restaurants, travel, hospitality, and entertainment are taking a huge hit.
- Economic loss. The Government Accountability Office estimated that U.S. GDP output losses associated with the Great Recession range from several trillion to over $10 trillion. Given that the low-end estimation on aforementioned Federal Reserve back-of-the-envelope calculations was the approximately the same unemployment rate as the Great Recession, it would not be at all surprising that if the economic loss was significantly worse.
- Increased social unrest. A 2013 report from the International Labor Organization found that the unemployment rate is the single largest determinants in social unrest: "A one standard deviation increase in unemployment raises social unrest by 0.39 standard deviations, while a one standard deviation increase in GDP growth reduces social unrest by 0.19 standard deviations (p. 14)." Jobs are how societies thrive, so it is no surprise that a significant increase in unemployment leads to more social unrest.
- Health impacts of economic downturns. The National Institutes of Health (NIH) released a literature review entitled Health Impacts of the Great Recession: A Critical Review (Margerison-Zilko et al., 2016). Traffic fatalities and alcohol consumption decreased because of the Great Recession. However, the majority of indicators took a downward turn, including fertility, self-related health, morbidity, psychological distress, and suicidal behavior.
- Instability in the developing world. If COVID-19 is having this much impact on developed nations, it would be a reasonable inference that it will be even more jarring for developing nations, especially given the correlation between GDP per capita and longevity (e.g., Swift, 2010). Emerging markets take on further debt, thereby increasing the likelihood of default and high levels of inflation. Argentina was already dealing with these issues before COVID-19. I can only imagine how their economy fares after this. India, Argentina, and Turkey are already experiencing capital flight (New York Times). Furthermore, individuals within these developing nations have minimal financial resilience, which is going to exacerbate both poverty and capability to fight COVID-19. Fragile states, such as Algeria and Nigeria, could easily destabilize, thereby creating real potential for instability in the MENA region. It's not just the economies would feel the strain. By extension, health care systems in developing nations would also feel the strain, which could also kills thousands of people, depending on the severity of the economic damage.
This list is by no means complete, and it does not account for the human costs of the social isolation incurred, but it should help to give you an idea of what sort of havoc is unleashed on the world if we completely ignore the economic piece of the puzzle.
How Do We Strike a Balance?
What we face is scary. On the public health side, we face a virus that could kill thousands, if not millions of people globally. On the economic side, we face the potential for Great Depression 2.0. We have to get the public health aspect of this right, no doubt about that. After all, the economic outcome is so interconnected with how quickly we mitigate the public health issues behind coronavirus.
The majority of expert economists at the University of Chicago's Booth School in a recent poll show that abandoning lockdowns prematurely when there is a high likelihood of resurgence would cause greater economic damage. If we take the University of Washington model at face value, now would be a bad time to stop because we are making our way up to the peak. It varies from state to state, but things are supposed to start dying down around late April-early May, according to the model.
Additionally, researchers from the Federal Reserve released a paper using the 1918 Spanish flu as a case study to show that short-term non-pharmaceutical interventions can be so effective that they could potentially result in longer-term economic benefit (Correia et al., 2020). There seems to be a case for certain interventions, at least in the short-run. At the same time, putting the economy on hold for months on end until scientists come up with a vaccine is not feasible. The Left-leaning Vox acknowledges we need an endgame. Last week, New York Governor Andrew Cuomo mentioned finding a solution that balances both health and economic concerns. However, timing on lifting the lockdowns can be tricky, as Hoover Institute research fellow Paul Gregory points out:
"If the release turns out to be premature, accusations of sacrificing precious lives for economic gains will abound. If the release is judged as too delayed, the charge will be the destruction of the economy. Our political leaders will probably never face such a momentous decision in their political careers. They will have none if they come down on the wrong side."
Perhaps the lockdowns are the right short-term decision to quell the transmission of COVID-19, and is thus the "least worst option." Even if that were the case, we should not have these measures in place a second longer than necessary. I can imagine decision-makers justifying lockdowns for much longer than necessary, either by saying "look how well this is working, let's keep doing it" or "imagine how much worse it would be without these lockdowns." Not even an economy as developed as that of the United States has unlimited resources. As individuals lose jobs and businesses close, we will realize just how limited those resources really are. Logically speaking, there is a juncture during which the cost rises too high and we need a new gameplay because costs exceed benefit. That is the whole point of a cost-benefit analysis: to estimate the strengths and weaknesses of current policy, as well as comparing the strengths and weaknesses of alternative policies. We need to compare current policy to realistic alternative scenarios, not an unrealistic "incomplete action" worst-case scenario like the Imperial College study does.
When governors are implementing lockdowns, "stay at home" orders, or "shelter-in-place," some of the most questions being answered. What is the net impact of closing "non-essential business?" What metrics are we looking at to determine when it's "safe to come out?" At what point will social distancing become ineffective? What sort of risk assessment was done to determine which jobs are "essential?" At what point does the economic harm exceed the virus? How much do we value a human life in the cost-benefit analysis? The average life insurance policy is $168,000. The median compensation for 9-11 survivors was $1.7M [in 2005 dollars]. The Environmental Protection Agency (EPA) puts the value at $8.4M in 2014 dollars (or $9.6M in current dollars). Such a wide range on the value of a human life has major policy implications in either direction. What other costs and benefits are to be valued in the cost-benefit analysis?
That is why I am glad that the American Enterprise Institute made an attempt to conduct a cost-benefit analysis (Scherbina, 2020). Her best-case scenario was a suppression period of seven to eight weeks, although her calculation was heavily caveated in Section IV of her analysis because of all the unknowns surrounding COVID-19. Although this cost-benefit analysis was released last week, it scares me that decision-makers are not taking these considerations into account when making their decisions. We're giving into panic and succumbing to the precautionary principle on steroids without asking ourselves whether this is ultimately the right decision or how long we should use such suppression measures.
On the one hand, I applaud AEI making an effort to do so because a cost-benefit analysis is the sort of thing we need to have to make an informed decision. On the other hand, I would still consider it preliminary. The reason for that is that there is a too broad of a range of estimated fatalities. As Dr. Jay Bhattacharya, who is a Professor of Medicine at Stanford University and a research associate at the National Bureau of Economic Research, points out (please listen to interview below), the policy measures for preventing two-million-plus deaths in the U.S. is different from preventing 50,000 or 100,000 deaths. Until we conduct a study on a broad population have an estimate of how many people have developed antibodies for COVID-19, we really don't know the lethality of COVID-19. Assistant Secretary of HHS Brett Giorir calls this a "denominator problem." Without accurate information on the prevalence and lethality of COVID-19 vis-à-vis the case fatality rate (CFR), we are flying blind.
What Does an Endgame Look Like?
In order to have a proper cost-benefit analysis, we need proper data. In order to have proper data, we have to have more testing. Testing is vital. Without testing, we cannot identify hotspots or identify key demographic trends, which means we cannot come up with a viable strategy. Thankfully, the FDA approved a molecular test from Abbot this weekend that can give test results in about five minutes and a 15-minute serological test from BioMedomics, so that should help things along. What else can be part of an endgame? While compliance seems to be high in the U.S. (Gallup), it will eventually wane with the growing financial and mental stress that come with the lockdowns. We need a more targeted approach than the blunt instrument of shutting down large swathes of the economy, but here are a few ideas:
- It is not only molecular tests that we need to figure out who currently has COVID-19. We also need serological tests to see who has developed antibodies to COVID-19. Those who have developed antibodies could go back to work since Dr. Fauci thinks that reinfection seems unlikely. Not only that, we have a better sense about such things as rate of transmission, what percent have severe symptoms (versus being mild or asymptomatic), what the fatality rate is, and whether we have reached herd immunity or not.
- We need to expand treatment capacity, including temporary hospitals, making more masks and ventilators, and doing what we can (within reason) to speed up the production of a vaccine.
- We know that COVID-19 disproportionately affects the elderly and the immunocompromised. As such, there should be stricter social distancing measures for those demographics.
- We can establish "safe zones" and gradually reopen establishments in the reverse order in which we closed them. Depending on what a cost-benefit analysis would have to say, I would also say that we could gradually allow for larger crowds as the transmission of COVID-19 decreases. This, of course, depends on whether we are passed the peak and infections are going down.
- I covered monetary policy last week, and anticipate covering the Coronavirus Aid, Relief, and Economic Security Act (CARES Act) in future posts.
These are some preliminary ideas, but I will end with this. Asking how long people put their livelihoods on hold, when we all can return to our lives, and what an endgame looks like is far from crazy. It's natural. The longer it takes, the more pressing the urge to answer these questions becomes. Ultimately, I hope we can come up with a plan to minimize coronavirus-related deaths and make sure as few people as possible lose their livelihood.
No comments:
Post a Comment