Why We Need to Start Lifting Coronavirus Lockdowns Sooner Rather than Later

You know that we continue to live in an age of political polarization when even a pandemic is politicized. During this pandemic, there are two main sides on the lockdown debate: those who want to keep the lockdowns going ("Closers") and those who want to lift the lockdowns ("Openers"). Openers are concerned about the economic damage being done to the millions unemployed and worry that the benefits don't justify the costs of the lockdown. Closers are concerned that opening the economy would result in a major uptick in COVID-19 cases, overburden the system, and make a bad situation worse.

For those who hold a polarizing view, one side has the tendency to see the other side not only as wrong, but as evil and morally deficient. There are Openers who think the Closers are either too oblivious to notice the economic effects or are too privileged to care about the newly unemployed because they have the convenience of being able to stay at home without their jobs or livelihoods being at risk. It can come off as callous and supporting the suffering of millions. There are Closers who think that Openers are too selfish, with little to no disregard for others (especially the elderly), and only care about money. The more belligerent Closers view Openers as the moral equivalent of murderers.

In addition to vilification of an opposing side, fear has also become even more widespread during the pandemic, which is saying something. The fear of Openers is that the economy will be so damaged that it will take years to recover and that millions will suffer in poverty and struggle simply to survive. This phenomenon will end up being more pronounced in the developing world. The fear of Closers is that they could catch COVID-19 and die. Alternatively, Closers fear that because many infected with COVID-19 are asymptomatic, they can spread the virus to others without knowing it, which could unknowingly and unintentionally harm another. Closers also have the fear that opening the economy at this moment would create an even worse problem than what we are going through right now.

During my time on the high school debate team, one has to be able to argue the other side in addition to one's own. If you cannot do so, you truly do not understand the topic at hand. I have been tracking the severity of the virus to make sure that my reaction to COVID-19 was neither an overreaction nor an under-reaction (see my March and April synopses). What I know is that COVID-19 is an upper respiratory infection with non-specific symptoms. It is not particularly fatal relative to past pandemics (see below), but it started off with an adequately high rate of transmission. To make it more challenging to fight, it can take two weeks to show symptoms (assuming one is not asymptomatic). Unlike SARS, the world was unable to contain COVID-19, which has in part led to the mess we are in.

While I understand the concerns of Closers and while I understand that each jurisdiction should have a different policy prescription based on their local reality regarding COVID-19, I consider myself more of an Opener who thinks that gradually opening the economy for most jurisdictions is something we need. You might be wondering how I came to that conclusion. Well, here is the list of reasons and considerations I came up with for gradually reopening the economy:

1. The worst-case scenario death toll has not come to pass. As of date, there have been about 70,000 recorded COVID-related deaths in the U.S. People dying in a pandemic is tragic, and I'm not here to minimize those who lost loved ones. At the same time, let's remember what predictions we were looking at that scared us to go into lockdown mode in the first place. The CDC's worst-case scenario was a death toll of 1.7 million. The Imperial College model was even scarier by estimating 2.2 million deaths in the U.S. By the way, this is the same model that scared the United Kingdom so badly that it went from a herd immunity approach to lockdown. The Imperial College's inflexible model has not been updated, made a hair-raising prediction with the unrealistic assumption that we would sit on our hands and do nothing, and has since been debunked. 
• Contrast that to the University of Washington model, which is the model that the White House uses, is updated regularly, and much more accurately predicted the amount of deaths than Imperial College. This model shows that 45 states are past the peak of this COVID-19 wave (i.e., all but Arkansas, Utah, Kansas, and the Dakotas). Up until yesterday, the University of Washington was projecting 72,433 deaths. They have since increased their estimate to 134,475 deaths. In spite of the upward revision, 134,475 deaths would be 92 percent lower than the CDC's worst-case scenario. If a second wave of the same magnitude were to hit, that would still be 84.2 percent lower than the CDC's worst-case scenario. It might come off as callous to view the death toll in such terms, but the reality is that we're in the middle of a pandemic. The goal is not zero deaths because we are well past that point. The goal is "the fewer deaths, the better." 
2. Early estimates likely overestimated COVID-19's fatality. Early estimates put the mortality rate of COVID-19 somewhere between 2 percent to 3.4 percent (see Rajgor et al., 2020). Using the case fatality rate (CFR), which is the number of confirmed cases divided by the number of deaths, in this instance skews the fatality rate upward because it only counts those who have been tested. The current testing comes with a selection bias by omitting asymptomatic or mild cases of COVID-19, which makes the fatality rate seem significantly higher than it actually is. Thankfully, expansion of molecular testing (seeing who currently has COVID-19) and use of serological testing (seeing who had COVID-19) vis-à-vis randomized testing has provided a more accurate picture with an infection fatality rate (IFR). With the preliminary IFR data available, the IFR ranges from 0.1 percent to 0.6 percent. While COVID-19 is likely to be more deadly than influenza, which has a fatality rate of 0.1 percent, it is nowhere near as deadly as initially predicted. COVID-19 is also not as deadly as other coronaviruses. SARS has a fatality rate of 8.6 percent, whereas MERS has a fatality rate of 30 to 40 percent. The lockdowns were made with an assumption about a fatality rate being anywhere between three and thirty-four times higher than what current data are showing. Knowing that COVID-19 is not as fatal as initially perceived weakens the argument for continued lockdowns. 
3. Flattening the curve has succeeded. The idea of "flattening the curve" is not to eliminate the disease. Flattening the curve is supposed to slow the rate of infection and spread it out over time in order to not overwhelm our healthcare systems, as well as prevent deaths specific to there not being enough available medical care to fight COVID-19. As such, the main metric in making sure that the curve is flattened is hospital capacity (e.g., having adequate number of staff, hospital beds, ventilators). As long as hospitals have the capacity to deal with the influx of patients, the argument for the lockdowns becomes diminished. If one moves the goalpost from "flatten the curve" to "we can't lift the lockdowns until there are very few or no deaths" or "we need a vaccine before doing so," all that has been proven is that the lockdowns were enacted under false pretenses. 
• You can see data from health analytics firm Definitive Healthcare or from the University of Washington for those metrics.  
• Side note: If we look at the state of New York, the U.S. state that was most heavily hit by COVID-19, Governor Andrew Cuomo was asking Trump for 40,000 ventilators. The University of Washington estimated that only 6,000 ventilators would be needed in New York. More interestingly, Cuomo actually sent ventilators to Maryland and Michigan because there were so many. 
4. The lockdowns are killing the economy. While the number of deaths has remained well below what doomsayers and their epidemiological models were predicting, the economic damage that I saw coming from a mile away has become abundantly clear in the past few weeks. The extent of the economic damage depends on the length and severity of the lockdowns. Each day that we spend in lockdown causes billions of dollars and increases the likelihood that millions of people around the world will permanently lose their jobs, thereby subjugating them to poverty-stricken conditions. 
• There have been over 30 million who filed unemployment claims in the past six weeks. While some will be able to go back to work once the lockdowns are over, the Congressional Budget Office's (CBO) latest projection shows that unemployment will be at an annual average of 10.1 percent in 2021.
• The real GDP for Q1 2020 decreased at an annualized rate of 4.8 percent. Since the lockdowns were not taking place until the end of Q1, we are really not going to see the full economic damage until Q2 figures are released. Although I don't have Q2 figures (obviously since we are in Q2), it is a safe bet that we are in a recession. How long and how deep of a recession will depend on how long these lockdowns last.  
• Federal Reserve President James Bullard estimates that the U.S. economy is hemorrhaging $25 billion a day, which means a 40-day quarantine would cost the U.S. $1 trillion. 
• Manufacturing output is at its lowest 5.4 percent in March, which is the biggest plunge in over 70 years.  This will hit small businesses in particular. A 2016 study from the J.P. Morgan Chase Foundation found that the median amount of cash reserves that a small business has would only last 27 days.
• A Rand Corporation study calculates that 73 percent of workers are going to experience an income decline of some sort (Strong and Welburn, 2020).
• If people cannot find jobs in the regular economy, they will go to the black market to make sure they can feed their families. During the Great Recession, the underground economy accounted for about 20 percent of the Spanish and Italian GDP. The increase of crime, violence, social unrest, and human rights abuses that take place in the black market should be enough to warrant avoiding this outcome.
5. Higher unemployment among health care workers. This seems counterintuitive since we are in a pandemic. You would think there would be a high enough demand for health services to keep health care workers employed. At the same time, the ironic truth is that healthcare workers are being furloughed or laid off. As of April 7, 230 hospitals already started furloughing. I can only imagine there being more layoffs and furloughs since then. But how can this be? We did not end up overwhelming the healthcare system in most places in the U.S. There was actually plenty of hospital capacity. What ended up happening is that non-essential medical procedures were postponed. Because there is such an excess capacity and hospitals are financially hurting because they cannot bring in additional revenue with the "non-essential" medical procedures, hospitals have needed to cut staff. If we keep going, not only do we undermine health care capability to fight off other diseases, but we make it more difficult to fight COVID-19 both in this wave and in the possibility of a second wave. 
6. We are neglecting non-COVID healthcare at our own risk. Other health care problems don't take a vacation simply because COVID-19 exists. A few things to point out regarding the healthcare-related effects of a lockdown and the resulting recession:
• As a meta-study from the National Institutes of Health (NIH) found, health indicators that took a downward turn during the Great Recession include morbidity, suicidal behavior, psychological distress, fertility, and self-related health (Margerison-Zilko et al., 2016). The reason I bring this up is that because we are in a lockdown-induced economic downturn that is most probably worse than the Great Recession. 
• Another study from the Lancet found that higher unemployment due to the 2007-2009 financial crisis caused 260,000 excess cancer deaths worldwide (Maruthappu et al., 2016).
• Calls to mental health hotlines have increased almost 900 percent since the shutdowns. Given the extended isolation measures caused by the lockdowns, this is sad but unsurprising.  
• Hospitals need revenue to keep operating. As an example, Ohio hospitals are expected to lose $1.2B a month. Hospitals shutting down would only put further strain on the health care system, which would be especially bad in the event of a second wave. 
• In addition to the millions struggling financially, an estimated 12.7 million have lost their employer-sponsored health insurance.  
• The tradeoff is not "lives versus the economy," but rather lives lost to COVID-19 versus lives lost trying to reduce COVID-19 deaths. Making decisions about lives saved versus lives lost is arduous, but it is not unprecedented. The question is whether our decision leads to a net-positive outcome. 
• May 8, 2020 Addendum: A recently released study from mental health foundation Well Being Trust touches upon many of my previously made points regarding non-COVID health. The alarming finding this study is that 75,000 additional people could end up dying due to the suicides, alcohol abuse, and drug abuse caused by the unemployment, social isolation, and hopelessness caused by the lockdowns. If this study ends up being remotely accurate, it severely undermines the argument that the lockdown saves lives. Again, it is not about "lives versus economy." It is about the lives saved by the lockdown versus lives taken as a result of the lockdown.
• May 14, 2020 Addendum: The Lancet released a paper a couple of days ago estimating the effects of the lockdowns' disruption of food chains and health services in the next 6 months. The least severe scenario is 253,500 additional child deaths and 12,200 additional maternal deaths (Roberton et al., 2020). 
7. How much longer can lockdowns last, and can we voluntarily do social distancing? As human beings, we generally have a need for social interaction. As concerned as we are about the pandemic, quarantine fatigue is rearing its ugly head. Especially in a democratic society, you can keep people under lockdowns and stay-at-home orders for so long. Gallup polling from April 30 is showing that U.S. citizens are becoming less vigilant about social distancing. Location data from social networking service firm Foursquare shows that fast food and gas station visits are at pre-COVID levels in certain regions. Auto shops, big box stores, and convenience stores are seeing similar trends. Trails, hardware stores, and liquor stores are more frequently visited. Apple's mobility trend reporting shows that people are moving around more. 
• Last month, I wrote on whether we can voluntarily social distance. Between the data I found then and the data from Foursquare and Apple, there is a clear trend that considerable social distancing was taking place prior to the lockdowns. This voluntary behavior that predated the lockdowns would limit the effects of the lockdown, and any subsequent study trying to measure the effects of the lockdowns would need to account for this voluntary behavior. 
• The data are showing quarantine fatigue and less vigilant social distancing. As I will elaborate upon in the end, perhaps we need a different approach to deal with the inevitability that social distancing is going to become more lax, regardless of whether there is a government mandate. The question is whether we can minimize the severity of the social distancing while still addressing the public health concerns. 
8. COVID-19 doesn't affect everyone equally. According to a study at the Lancet, about 1 percent of those aged 30 or under will have severe symptoms (Verity at al, 2020). COVID-19 has a disproportionate effect on the elderly, as well as those with a pre-existing condition. While this does not provide comfort for those in either of those categories, this does have public policy implications because there is a clearly defined population at risk. Instead of requiring everyone to stay at home, state governments could implement an approach more targeted towards those vulnerable groups. 
• Another point to be made is that not all states and jurisdictions are equally affected. New York state accounts for 33.1 percent of COVID-19 deaths, although its 19.5 million citizens account for 5.9 percent of the U.S. population. What is taking place in New York is not reflective of the rest of the country. This is why each governor should craft a recovery plan that takes into account the reality of COVID-19 within their state. 
9. There is not a guarantee of a vaccine. While a vaccine is certainly possible, anyone banking on the hope of one is optimistic, dare I say bordering on pollyannish. It is unrealistic to bank on a vaccine, least of all because having a vaccine in 12 to 18 months would be an optimistic timeframe. This assumes that we end up finding a vaccine. We have not found vaccines for other such coronaviruses as SARS and MERS. We have been looking for an HIV vaccine for 40 years, and guess what? Still nothing. It took a decade to find a vaccine for the deadly avian flu H5N1. We still haven't even found a cure for the common cold! While vaccines can provide considerable protection, they do not offer complete protection. The CDC recently found that the current flu vaccine is 45 percent effective. We have learned how to deal with other deadly diseases without a vaccine. While a vaccine would certainly help, we should also prepare ourselves for living our lives without the existence of a vaccine. As such, a vaccine should not be a precondition for lifting the lockdowns. 
10. Large-scale contact tracing might prove infeasible. There are Fourth Amendment considerations with potential violations to privacy for contract tracing, particularly if it is not voluntary. If it passes the constitutional hurdle, the public may or may not support it. Kaiser Family Foundation survey data suggest that support for contact tracing depends on implementation and addressing privacy concerns. As the Left-leaning Center for American Progress points out, there are the logistic challenges of implementing contact tracing in such a large country: adequate contract tracing personnel, only traces person-to-person contact (as opposed to contact with surfaces), a decentralized health system, a minimum of 60 percent cooperation rate, and other technical and procurement issues. 
11. Simply because a lockdown could be rendered safe does not make it wise policy. If we were only concerned with safety, we would have some weird outcomes, both in terms of public policy and our individual behaviors. Automobiles would need to be banned because over 30,000 people in the United States die from motor vehicle crashes (see NHSTA data here). 650,000 people die from heart disease. Exercise reduces the likelihood of heart disease, yet we don't mandate a daily exercise regimen for every U.S. citizen. About 5,000 people in the U.S. die from choking on food, but we are not going to mandate puréed food. By choosing to eat fatty and/or salty foods, smoke, or not exercise, people trade years of their life with less-than-healthy lifestyle choices because of their own perception of risk. This is not a call to throw all precaution in the wind or to say that we should not take any safety precautions (more on that below). 
• What I am getting at is the following. First, death is an inevitable part of life. There have been over 200,000 COVID-19 deaths globally [as of 7:57am on 5/5], but there have also been about 20.19 million deaths total this year. COVID-19 deaths have accounted for about 1.2 percent of all deaths this year so far. Second, if we did not tolerate at least some risk, we would not live our lives. 
• Epidemiological expertise in a pandemic is significant, but it is not to be the ultimate arbitrator in how we live. This expertise has to compete with other considerations and trade-offs, and it has to convince us of the severity of COVID-19, which brings me to my next point...
12. Burden of proof is on governments to justify lockdowns, especially now. In the initial onset, a strong case could be made for a near-term lockdown as a safety precaution. We could not have known in February or even March if COVID-19 would end up as deadly as SARS, MERS, or the many other lethal diseases that have existed. If it were that deadly and we did not exercise caution, the death toll could have been staggering. The thing is that we're past the point of "better safe than sorry." The outbreak was not nearly as grim or dire as many early models suggested. In the meantime, the economy and the livelihoods of millions are taking a big hit. More COVID-19 data is being collected and disseminated, which means we can better project what is going to happen. Especially in a relatively free society that values "presumption of innocence," the burden of proof lies upon the governments to provide such measures are necessary or that there are not more moderate measures that could help. Until governors can use data and accurate models to back up their positions, their lockdowns are unjustifiable and should be gradually lifted. 
13. Are lockdowns an evidence-based practice? If you are going to deprive people of civil liberties, shut down large swathes of the economy, and try to delay the spread of COVID-19, we should ask whether there is at least some basis showing that it is a winning strategy. Lockdowns are a novel policy response. St. Louis had the most severe non-pharmaceutical interventions (NPI) during the Spanish flu pandemic, but none of those NPIs were lockdowns (Hatchett et al., 2007). During the Spanish Flu, St. Louis did not have a stay-at-home order, nor did it have a business closure that exceeded 48 hours. There was a very limited lockdown during the Ebola crisis, but nothing empirically definitive from that experience. Aside from that, we would need to use data from the current pandemic to prove or disprove its efficacy. There are two sub-questions that could be asked here.
1. Do lockdowns slow the rate of transmission? A study on COVID-19 in Italy estimated that the lockdowns reduced transmission by about 45 percent, thereby averting 200,000 hospitalizations (Gatto et al., 2020). Going back to Point 3, the main goal of flattening the curve is to slow the rate of transmission. In the Gatto et al. study, the rate of transmission was still above 1.0, which means that the disease would not die on its own. At the same time, Italy's health care system was overwhelmed. If accurate, this study could be enough to justify a lockdown in an Italian context. However, since most places in the U.S. never exceeded capacity, I could easily argue that the lockdowns were or continue to be unjustifiable in the U.S., especially since University of Washington modeling shows that most states have passed their peak. 
2. Do lockdowns reduce the rate of deaths? I am more skeptical about lockdowns preventing deaths than I am in lockdowns reducing the transmission rate. Flattening the curve would, at best, have a minimal effect and would be contingent upon exceeding hospital capacity (See Point 3). In its analysis on lockdowns, the American Enterprise Institute (AEI) did not find correlation between lockdowns and change in daily deaths from all causes on a given date in 2020 versus a year earlier. AEI thus concludes that there is a lack of evidence for lockdowns for death prevention, and that other public health measures are more effective.
• May 7, 2020 Addendum: Economists from the University of California-Davis found that "non-pharmaceutical interventions [i.e., stay-at-home and social distancing orders] may have been effective in slowing the growth rate of confirmed cases of COVID-19, but not in decreasing the growth rate of cumulative mortality"(Lin and Meissner, 2020, p. 2).
• Side observation: I found a meta-study on the NIH website covering physical interventions for respiratory diseases. Across the 67 studies, it was found that a) hand-washing and wearing surgical or N95 masks were the most effective physical intervention reducing the spread of a respiratory infection, and b) there is limited evidence for social distancing (Jefferson et al., 2011).

The video above is an interview with Johan Giesecke, who is a senior epidemiologist who is a former advisor to the Director-General of the WHO. 

Postscript
As we receive more data, we see that COVID-19 is nowhere near as lethal as initially projected. All the while, economic damage has predictably been gargantuan. What makes that scarier is that we have not begin to see the damage and unintended consequences that the lockdowns have caused, whether to the economy, individual health, civil liberties, or the social unrest it will end up causing. Panic needs to replaced with data-informed decisions. While COVID-19 is serious, it is also true that fear has been misleading the public in terms of creating a proportionate response.

What does that response entail? I brought up a few ideas last month, but increasing testing capacity to isolate hotspots is vital. Expanding medical capacity would also help. In terms of gradually opening the economy, we should go beyond thinking in terms of "essential" and "non-essential" jobs. We should also think of it in terms of level of risk and use standard risk management to think of it in terms of "low, medium, and high risk." Laws requiring masks in public indoor venues, capacity limits in restaurants and bars, and a recommendation for keeping a distance of six feet also seem like good ideas. There are policy options that exist between "lock everything down" and "do nothing." States are starting to realize that. Such states as Texas, Florida, and Georgia have started with a partial reopening. Even the Governor of New York, which was the most heavily hit state, realizes that lockdowns are not an indefinite solution, which is why he released a recovery plan earlier this week. Now is the time to start implementing those policy alternatives so we can minimize both the public health and economic effects.

For more insight from health experts on the ineffectiveness of lockdowns, you can look at the analyses of Johan Giekse (former Chief Scientist at the European Center for Disease Control and advisor to the Director-General at the WHO), John Ioannidis (Professor of Medicine at Stanford University), John Lee (former Professor of Pathology at Hull York Medical School), Martin Kulldorf (Professor at Harvard Medical School), Michael Levitt (Nobel Prize-winning Professor of Structural Biology at Stanford) and Knut Wittkowski (Former Biostatistician at Rockefeller University).