Skip to main content
SearchLoginLogin or Signup


Published onApr 09, 2020

“It is not easy for a free community to organize for war,” wrote John Maynard Keynes in 1940.1 He was commenting on something very obvious: people do not like to be told what to do. Keynes was frustrated by the inability of political leaders to lucidly explain to the public what needed to be done. Resources had to be allocated to the war effort and, after that, a clear statement of how the remainder would be shared amongst the public had to be made. Instead, politicians were glossing over both issues with superlatives and no clear plan. Writing, as I am in March 2020, as politicians announce today what they claimed was unthinkable just yesterday (and I mean that literally), I get where Keynes is coming from even if the magnitude of the problem seems comfortably lower.

Keynes was particularly concerned that the decisions that needed to be made were numerous and inter-related with one another.

Is it better that the War Office should have a large reserve of uniforms in stock or that the cloth should be exported to increase the Treasury’s reserve of foreign currency? Is it better to employ our shipyards to build warships or merchant-men? Is it better that a 20-year-old agricultural worker should be left on the farm or taken into the army?

He pointed out the obvious. A start was to think about which margin to fix — the standard of civilian life or the war effort — leaving the other as the residual. This had to be decided one way or the other. In our present conundrum, when asked, people would surely say that we should fight the pandemic first and adjust the rest. The fact that, in 1940, Keynes was pointing out that it was not obvious what Britain had decided should give us pause.

It is for this reason that having a clear and resolved approach to holding the line on health is warranted. In its absence we may have a situation where we end up at a point where we have the opportunity to improve the economy and public health. At that point, political leaders have tried to limit the damage on the economy but, in the process, have failed to contain the virus and, thus, have chosen a lower point with regard to public health. That situation calls for strong measures to move back to what might be possible.

It is a striking fact that even the most market-loving, capitalist nations quickly abandoned the decentralized process of allocating resources in the face of World War II. To be sure, no one expected the military to use markets to decide where to deploy troops and equipment but the fact that the rest of the economy moved to a war footing in this way is useful to reflect upon. In particular, for the most part, even though they may have flirted for a day with relying on strong advice to citizens, governments in the COVID-19 pandemic realized that was insufficient to their ends and ended up with strict and, in some cases, very strictly enforced policies. More authoritarian regimes were a little quicker to act initially, but the lag between could be measured in days for most countries.

Why central planning?

Economists claim that markets are the most efficient way of allocating resources and solving the age-old problem of who gets what if there isn’t enough to go around. To be sure, markets are quite amazing in this regard and every economist has their moment of wonder that it works. Here is Thomas Schelling:

Most people, whether they drive their own taxis or manage continent-wide airlines, are expected to know very little about the whole economy and the way it works. They know the prices of the things they buy and sell, the interest rates at which they lend and borrow, and something about the pertinent alternatives to the ways they are currently earning their living or running their business or spending their money. The dairy farmer doesn’t need to know how many people eat butter and how far away they are, how many other people raise cows, how many babies drink milk, or whether more money is spent on beer than on milk. What he needs to know is the prices of different feeds, the characteristics of different cows, the different prices farmers are getting for milk according to its butter fat content, the relative costs of hired labor and electrical machinery, and what his net earnings might be if he sold his cows and raised pigs instead or sold his farm and took the best job for which he’s qualified in some city he is willing to live in.

Somehow all of the activities seem to get coordinated. There’s a taxi to get you to the airport. There’s butter and cheese for lunch on the airplane. There are refineries to make the airplane fuel and trucks to transport it, cement for the runways, electricity for the escalators, and, most important of all, passengers who want to fly where the airplanes are going.2

It is a miracle and we should appreciate it as such. The problem is that it doesn’t always get the job done.

When the job to be done is urgent and resources need to be reallocated quickly, the system can gum up. The issue is not markets per se but the problems of relying on a decentralized process whereby everyone allocates the resources they control on the basis of their own information and preferences. Instead, the problem we face in a time of war (or pandemic) is that resources, currently controlled by individuals, need to all be applied to a new end and the task of convincing everyone to choose to do so is unlikely to work out well.

This notion was captured in a 1990 paper by economists Patrick Bolton and Joe Farrell.3 Imagine a situation where we need one factory to produce face masks and another to produce ventilators but we don’t know which will be able to do each task at the lowest cost. In a market economy, each factory owner might look at the situation and try to work out what to do. One option is that they both jump in and start producing the product they think they will provide most efficiently. They retool for that purpose but there is a chance that they will end up both choosing the same thing and we will end up with too many face masks and too few ventilators or vice versa. Another option is to wait and see what the other factory chooses to do and then do the opposite. But in this world, we have both factories waiting to see what happens and there is a consequent delay. In other words, decentralization either will not get the job done or it will cause it to be delayed.

The alternative is for someone to choose who does what. This is the role of central planning. This prevents both duplication and delay but opens up another problem: the government may make the wrong choice. They may end up producing both goods at a higher cost than otherwise. Suffice it to say, at times of crisis we do not let the perfect be the enemy of the good and so comfortably resort to centralized resource allocation and wear the potential productive inefficiency.4

There are three areas where in the COVID-19 pandemic, market processes were abandoned in favor of planning. These included the mobilization of resources to dramatically expand health care system capacity, the institution of price controls for certain important goods and services and the use of blanket restrictions on movement of people. Each of these will be discussed in turn.

Surfing the Curve

The initial responses from governments to the pandemic were to institute progressively strong forms of social distancing in the hope of reducing R0 (the number of people infected people themselves infect). Those responses had the goal of what came to be known as ‘flattening the curve.’ In a scenario where this needs to be done once, this involved a scenario such as depicted in Figure 3-1. The task was not so much as to reduce the total number who became infected but to spread them out over time to economize on health care system resources.

Figure 3-1: Flattening the Curve

The problem was determining how flat we needed to go. The flatter the goal, the harder it is to achieve and, moreover, the greater are the consequent costs of prolonged economic harm, social isolation, and the possibilities that there could be a subsequent re-emergence of the pandemic causing us to do it over again.

The health care system capacity is likely way lower than the diagram is showing beyond what flattening the curve can actually achieve. This differs by country. Japan has 13 beds per capita while the United States has under 3 beds per capita. And this is just a guide. There were large national differences in key inputs such as ICU beds, ventilators, hospital protective equipment and health care workers. Nonetheless, in most cases, it was clear that policies aimed at reducing R0 had happened too late to prevent health care system capacity from being reached. In Italy, doctors were having to perform heart-breaking triage operations determining which patients would get scarce resources. This was beyond their usual option of providing those resources even when they had a low probability of being required. From an economics perspective, health care resources were going to have demand that far outstripped supply. What is more, there was no prospect or desire to use higher prices to deal with the shortage. As Keynes noted for World War II, a plan for rationing was required but no plan was being formulated.

Given this, it is somewhat surprising that more was not being done to dramatically increase the capacity of the health care system. It is a policy option that both reduces the cost of overwhelmed capacity and reduces the amount of flatness of the curve and its associated costs. To be sure, in March 2020, calls were being made for more ventilators and other equipment.5 But countries had not done what China had done earlier in Wuhan by building entirely new hospitals in just over a week. Everyone marvelled at this. I heard: “wow, we can’t do that.” And this was mostly from the health care industry whose basic message for years is how hard it is to provide more. They have had expansion beaten out of them by years of a scarcity mindset.

While flattening the curve could take place and reduce the required capacity expansion, what was required was to surf the curve (see Figure 3-2). In this situation, health care system capacity would be temporarily expanded so as to cover the unflattened portion of the curve.

Figure 3-2: Surfing the Curve

Building out that capacity requires a new mindset and it requires it quickly. The nature of the problem was obvious on the ground. This is Dr Daniel Horn (a physician at Massachusetts General Hospital in Boston):

In the face of a global shortage, American industries can step up and quickly produce ventilators. All week, I have been receiving text messages and emails that say things like “By the way, my company makes parts for G.E. ventilators. We just got a big order that we are pushing through as fast as we can.” The General Motors chief executive, Mary T. Barra, announced that G.M. was working closely with Ventec Life Systems, one of a few ventilator companies based in the U.S., to rapidly scale up production of their critically important respiratory products. My colleagues at the nation’s top hospitals are getting phone calls from tech leaders asking for ventilator specs.

Such stories give me hope. But we need the federal government, too. … We need a plan.6

Sound familiar? This is precisely the coordination problem as outlined by Bolton and Farrell. Hospitals alone cannot procure what they need. Some factories can make some parts better than others. And then there is the issue of which hospitals to send them too. There was no information present and, in the US, despite having the powers to do so, no central action was being taken.

This highlighted the need for a war-like resource allocation mindset. Someone needed to take control and, when it came to fast and rapid capacity of health care, most countries had an obvious candidate — the military. Mobile Army Surgical Hospital or M*A*S*H was not just a TV show but also a capability that armed forces needed. It just had to be moved to civilian ends. In some countries, this happened with the military preparing and/or building facilities in Switzerland, Colombia, the Netherlands, Italy and France. The US also redirected hospital ships to California and New York to handle patients with other conditions who might be pushed out of those systems.

The numbers involved, however, suggest that a more comprehensive and aggressive solution would likely be required. Not only military provision but also a means of diverting manufacturing effort to the cause. Much of this was lying idle due to social distancing. The need was for a centralized process to unlock that potential and ensure timely provision. In World War II, businesses quickly retooled for military production. The same would be required here. Moreover, there would likely be a need for additional health care workers. This too could be a mobilization effort (perhaps even supported by conscription). The good news is that those resources were idle. The better news was that, unlike in wartime, no one was asking those helping to kill others.

Price Controls

Hand sanitizer and toilet paper went first. Hand sanitizer made sense. It was a genuine surge in demand as people expected to use more, much more, of it and were advised to do so. Toilet paper came as a surprise. This was not just a surge in demand but hoarding. But why? As Justin Wolfers argued, showing that economists were unafraid to tell it like it is, “[o]nce they have more toilet paper, people aren’t going to poo more.”7 He saw it like a bank run. People saw that toilet paper supplies were dwindling and bought more because they were concerned about supplies down the track. This created a run on the product just like a bank run. Suffice it to say, as it turned out, toilet paper was back in the store after the initial rush before people found that they didn’t have a square to spare.8

Hand sanitizer and other products that might be subject to real shortages were another matter. One story involved a couple of entrepreneurs who bought up a huge supply of hand sanitizer right after the first US death on March 1.9 They had intended to sell their stock of 17,700 bottles at a large mark-up on Amazon. Before they could do so, Amazon cracked down preventing them and others from selling the items that were in high demand. Ebay followed suit. In the end, the bottles were donated to hospitals.

Price gouging is given an ugly name because, of course, it is associated with people taking advantage of shortages in times of crisis to make a profit. At normal times, we usually like high prices because they signal to others were demand is high and there are profitable opportunities to produce more. In other words, they are part of a market process for resource allocation. When we outlaw or otherwise try to provide a cap on prices, what we are doing is accepting a shortage.

As John Kenneth Galbraith, who headed up the US World War II office of price controls, noted, this is an acceptance of a ‘disequilibrium system’ where demand persistently outstrips supply. This meant that items subject to controls needed to be rationed. As Galbraith noted, the outrage at this process tended to involve surprising items (just as we saw with toilet paper):

[F]or some reason ceilings on fur coats inspired them to special anger. On several occasions I found myself contending with new colleagues (and once with a new administrator) who were enthusiastic about dropping price controls on fur coats. When they saw that this action would put a premium on high-priced coat manufacture, would draw materials (“trim”) away from cheaper lines, they soon reversed themselves. In doing so they adopted a position entirely consistent with a broad theoretical pattern for allocating resources and equalizing incentives. Of the existence of any such theoretical pattern they were totally unaware.10

Given this, it is instructive to consider why we happily resort to price controls in times of crisis.

The reason is that it may well do a better job at resource allocation. According to research by Piotr Dworczak, Scott Kominers and Mohammad Akbarpour,11 who we want to get hand sanitizer can be different from who the market will allocate it to. In particular, had the price gougers got their way, only the wealthy would have got their hands on it and, in the process, protected themselves and the people they interacted with from infection.12 But those are the very people who have the best access to health care, don’t live in more densely-populated neighborhoods, or who can easily work from home. What makes more social sense is for the poorer members of the community to be allocated the hand sanitizer. Price controls give them a fighting chance. What might even be better is directly allocating hand sanitizer supply according to where it can have the greatest impact.

Restrictions on Movement

Here is the calculation. If the R0 for COVID-19 is 2 then an infected person will cause 2 others to be infected. But it obscures the magnitude of the problem. Those 2 people will infect 2 more people and so on. After ten rounds of this that adds up to 1,024. (If R0 = 3, it is 59,049!) If one percent of those people die from the disease, one infected person has been responsible for 10 deaths. Suffice it to say, that puts mass shootings in perspective. It is no surprise we want to keep infected people isolated.

The problem is how to do that. For starters, you have to know you are infected and with COVID-19, as was already explained in Chapter 1, the majority of infected people don’t know it. Moreover, once you do know it, using claims such as I have just done that you might be responsible for between ten and six hundred deaths, means that being infected carries a social stigma. Laura Derksen and Joep van Oosterhout found this when examining the spread of AIDS in Africa.13 They found that when people were asked to opt for testing or AIDS-related care, there were few takers because people feared the stigma of being seen to be concerned. People have to be generally and publicly knowledgable about the social benefits of these actions otherwise they might choose to cover up symptoms fearing discrimination. If refusing to go out is seen as a cover, that may be a problem.

Ordinarily, if you were going to restrict movement of people, you would try and target the individuals both at risk and who are likely to be people with a higher individual-R0. For COVID-19, some physicians wondered if a more targeted approach could be achieved. For instance, as those over the age of 70 were more likely to require the higher end of health costs (including death) associated with the virus, would it be better to isolate them leaving the rest of the population to circulate?14 Doing this would greatly mitigate the economic costs from a broader policy.

The problem with targeting is that there are real doubts it would work. If a large proportion of those under 70 still become sick and need hospitalization, resources could still be overwhelmed. Moreover, with large numbers of infected younger people, we lack the people to support the elderly being isolated. As Alex Tabarrok agues, there are internal contradictions that may well render it impossible for a more ‘surgical’ approach to social distancing.15

Targeted policies are also hard to enforce. When there are restrictions on movement, it is very easy for the authorities to see whether people are moving or not. In their absence it is harder to tell and can require more resources.16 For these reasons, to deal with the costs associated with COVID-19 transmission, governments have opted for blanket policies akin to martial law in wartime or other times of emergency. These may be supported by penalties for violations but nothing like the type of taxes that economists would otherwise recommend so that exceptions can be made at the discretion of individual decision-makers willing to bear a taxation cost. Instead, a heavy-handed approach is used without much room for nuance.

At the time of writing, the length of time for social distancing has been of the order of a few weeks. What remains unknown is precisely how long it both needs to and can last. I leave that as an open question that may already be answered by the time you read this book.

Steve Petrov: Has anyone actually measured how difficult manufacturing facemasks and ventilators is? (setting aside the fact that it turns out ventilators can do more harm than good and the key resource is the service provided by skilled Drs in deciding when to and when NOT to use this last resort technology.)But in short, the facts suggest that know how is the key for producing complex medical equipment moreso than central planning and that it is mostly a matter of existing suppliers scaling up. See for example:
Steve Petrov: Some of us grew up thinking that this is EXACTLY where marketplaces come in handy. See also the efficient market theory re info flow. Markets are better at making really complex things like cell phone networks and computers and sub-contracting the bits out to the most efficient and highest quality producers. Ventilators are probably fairly simple compared to a lot of what private industry produces.
Steve Petrov: In the USA there’s no scarcity of privately employed bureaucrats, clerical workers, health care marketing executives etc. New insurance office buildings can be built quite quickly. I don’t know what the explanation for this gross inefficiency is but it’s not a lack of resources. It’s a gross lack of mis-allocated resources.A market purist would argue that it’s due to government subsidising the insurance industry (government covers old people and the disabled via Medicare and SSDI leaving industry to cherry pick from the young, healthy and employed resulting in a duopoly and inefficient market. When was the last time someone was able to bargain with a healthcare provider over the price of a service. But I’m going off track here… Suffice it to say that the US healthcare market is anything but efficient, private or rational.
Steve Petrov: Hey! How does one reconcile this with the idea that competition leads to greater efficiency?! If governments are bad at choosing winners when they have time to carefully consider options why can they be expected to magically become efficient and wise decision makers (relative to the marketplace) when decisions must be made quickly?Who’s more likely to be first to ID an opportunity or unmet need or want?
Steve Petrov: This is really interesting because the facts indicate that the highly consolidated nature of the food supply chain has resulted in great inefficiencies and lack of agility in pivoting from institutional bulk food supply to consumer food supply. See for example: related issue is that a highly linear supply chain is more vulnerable to break downs eg one giant meat plant suffering worker infections has a disproportionate affect. Kind of the antithesis to the military’s famed decentralised, independent packet design of the Internet and it’s resilience to attack. A small trade off in network inefficiency for a gain in robustness.
Steve Petrov: The history is interesting. FDR sought to award large government contracts to industry to provide supplies to the Soviets in their fight against the Nazis and met with great resistance from the captains of industry for a range of reasons (resistance to central resource allocation, disdain for the Nazis, admiration for the “Hitler Miracle” etc.
Steve Petrov: Going back to my earlier comments about national economies and their comparative closeness to their production possibilities curves Britain was likely in a much different situation to the US.
Steve Petrov: I’m not sure where this fits in best but.. the war metaphor is superficially appealing but also fraught with imprecision. For example, a striking difference between a war effort where government takes a command and control role and COVID 19 is the lack of government/centralised consumption. In wartime government consumes vasts amounts of GDP/production capacity. But with a pandemic government uses central authority to suppress economic activity rather than consume it. (I suspect that the government requisition of face masks and ventilators is a tiny % of GDP compared to a war of attrition whereby government requires battleships, tanks, aeroplanes and general supplies en mas.)Perhaps this means that an economy has a better chance of a V shaped recovery when emerging from a shut down than it does emerging from a war? Perhaps it means that government stimulus and inflationary risks are fundamentally different in a shut down versus a central consumption scenario? I’m not sure what the answer is but I reckon as an economist you might have some thoughts about this.
William Janeway: The excellent news is that, unlike general mobilization for world war, the number of discrete items is very limited. See
Mike Abernethy: argues
Gray Newman: Perhaps to make it clearer: “In its absence we may have a suboptimal situation….”
Michael Jones: Not with a sufficiently high fine for violating the policy.
Michael Jones: Pun intended? There are stories of literal fights over grocery products.
Frank Weiss: The contrast between central planning and the market is of course fine. However, the US has neither in the relevant health product markets! E.g. price control of masks and but central allocation. Other countries seem to be less concerned about price control, but I can’t be sure. In any case, US evidnece is not in favor of central allocation, for a race has not been run.
Frank Weiss: but NOT central allocation
Kingsley Lewis: The discussion of targeted policies and herd immunity is extremely weak, which is a weakness of the from the whole book. Estimates of death rates for less vulerable groups are coming down dramatically and do not suggest that health resources would be overwhelmed. Protection of the vulnerable could be voluntary (no enforcement problem). Protection of the vulnerable would require vigorous government persuasion and financial support but this would be peanuts compared with the costs of general lockdowns. Families and other people dealing with the vulnerable could be educated on how to minimise transmission risks. Sure there are difficulties but these are manageable and there are no “internal contradictions”. A targeted strategy results in herd immunity, a massive and permanent advantage compared with general lockdowns. (OK, permanent immunity is unproven but it is likely, as would also be the case for a vaccine). Sadly there would be deaths for a small fraction (<0.2%) of the less vulnerable. But this is far less than the deaths, mental illness and despair caused by prolonges lockdowns.
Eric Rasmusen: Unclear. With price controls, you get some people buying huge quantities because it is so cheap, for themselves or for illegal resale. The price control needs to be combined with a quantity limit per buyer.
Eric Rasmusen: Wouldn’t it be only the hospitals would get their hands on it? They surely are willing to pay more than almost all individuals, especially if we increase funding to hospitals during the crisis.
Eric Rasmusen: The toilet paper shortage is over in my town. An interesting sideline on this, though, is that the lockdown has changed the *type* of toilet paper from commercial to home, which we are told are quite different. I guess the market is responding well to the shift. See
Gray Newman: The shelves in my town are still largely empty of toilet paper for residential use. Worth considering the market segmentation argument made above.
Eric Rasmusen: Note that the first requirement is that someone say they’re willing to buy the ventilators, etc. The US government didn’t place any orders, apparently. To get more quantity supplied needs a demand shift. If the US government demanded more tanks in WW2, firms would supply them, but not if it doesn’t place the orders. Then, maybe government coordination is needed too, but the first step is to express willingness to pay for more ventilators. This is especially true if there is a risk that if a company increases supply, the government will simply commandeer it and the company will go bankrupt.
Eric Rasmusen: It prevents duplication, but not delay. We used central planning for covid-19 tests. The FDA and CDC decided there should be just one test, to avoid duplication. They banned other tests and delayed disastrously until political pressure was put on them.
William Janeway: US mobilization for WW II was both duplicated and delayed due to coordination failure in supposedly centralized decision-making process. Competition between claimants on resources (Army, Nacy) generated “priorities inflation” until finally massive increase in supply eliminated shartages. See Eliot Janeway, The Struggle for Survival (1951) and Mark Wilson, Destructive Creation (2016)
Eric Rasmusen: Wouldn’t the owners of the two factories just phone each other up and coordinate?
Frank Weiss: May be illegal.
+ 1 more...
Dermot Crean: UK Nightingale hospitals?
Dermot Crean: Per 1000 pop?
Samuel J. Klein:
Samuel J. Klein: