A Timely Agent Based Model

I’m going to play economic imperialist and declare that epidemiology is partly within the domain of economics. I’m not sure how big the overlap is, but certainly there are economic questions that bear epidemiological type modeling. The movement of ideas is, in the long run, as important as the movement of a virus is in the short run. In any event, economists should be using this sort of methodology more often.

Never reason from a fatality rate


Richard Epstein has produced several posts and a video interview arguing that the mainstream media is overreacting to the Coronavirus pandemic. Richard understands the potential seriousness of this situation and the proper role of government. He recognises the value of the Roman maxim Salus populi suprema lex esto – let the health of the people be the highest law. In public health emergencies, many moral and legal claims resulting from individual rights and contracts are vitiated, and some civil liberties suspended.

Nevertheless, along with Cass Sunstein, Richard claims that this particular emergency is likely to be overblown. His justification for this is based on data for infection and fatality rates emerging from South Korea and Singapore that appear (currently) under control with only a relatively small proportion of their population infected. This was achieved without the country-wide lockdowns now being rolled out across Europe. Extrapolating from this experience, Richard suggests that the Coronavirus is not too contagious outside particular clusters of vulnerable individuals in situations like cruise ships and nursing homes.

The line of argument is vulnerable to the same criticism that one should never reason from a price change. The classic case of reasoning from a price change is reading oil prices as a measure of economic health. When oil prices drop, it could herald an economic boom or, paradoxically, a recession. If the price dropped because supply increased, when OPEC fails to enforce a price floor, then that lower price should stimulate the rest of the economy as transport and travel become cheaper. But if the price drops because economic activity is already dropping, and oil suppliers are struggling to sell at high prices, then the economy is heading towards a recession. The same measure can mean the opposite depending on the underlying mechanism.

The same logic applies to epidemics. The transmission rate is a combination of the (potentially changing) qualities of the virus and the social environment in which it spreads. The social environment is determined, among other things, by social distancing and tracking. Substantial changes in lifestyle can have initially marginal, but day after day very large, impacts on the infection rate. When combined with the medium-term fixed capacity of existing health systems, those rates translate into the difference between 50,000 and 500,000 deaths. You can’t look at relatively low fatality rates in some specific cases to project rates elsewhere without understanding what caused them to be the rate they are.

Right now, we don’t know for sure if the infection is controllable in the long run. However, we now know that South Korea and Singapore controlled the spread so far and also had systems in place to test, track and quarantine carriers of the virus. We also now know that Italy, without such a system, has been overrun with serious cases and a tragic increase in deaths. We know that China, having suppressed knowledge and interventions to contain the virus for several months, got the virus under control only through aggressive lockdowns.

So the case studies, for the moment, suggest social distancing and contact tracing can reduce cases if applied very early on. But more draconian measures are the only response if testing isn’t immediately available and contact tracing fails. Now is sadly not the time for half-measures or complacency.

I believe that Richard’s estimated fatality rates (less than 50,000 fatalities in the US) are ultimately plausible, but optimistic at this stage. Perversely, they are only plausible at all insofar as people project a much higher future fatality rate now. People must act with counter-intuitively strong measures before there is clear and obvious evidence it is needed. Like steering a large ship, temporally distant sources of danger must prompt radical action now. We will be lucky if we feel like we did too much in a few months’ time. Richard believes people are more worried than warranted right now. I think that’s exactly how worried people need to be to adopt the kind of adaptive behaviors that Richard relies on to explain how the spread of infection will stabilise.

Coronavirus and takings

City governments are flirting with a ban on evictions during the coronavirus pandemic. I doubt, however, that doing so comports with the Constitution’s takings clause or, perhaps, the contracts clause.

San Jose has introduced legislation that will ban evictions due to un/underemployment resulting from coronavirus. Seattle’s socialist firebrand, Kshama Sawant, calls for similar action. Her letter, though, betrays the truth behind many proposed emergency measures–she’s leveraging the crisis to further her political agenda, particularly her hatred of capitalism. In the letter, she froths: “The status quo under capitalism is deeply hostile to the majority of working people, and it would be unconscionable to place the further burden of the Coronavirus crisis on those who are already the most economically stressed.” Never mind that the status quo in the absence of capitalism would be grinding poverty.

But, in any case, the proposal to ban evictions and force landlords to renew leases as the pandemic sweeps across the states raises serious constitutional concerns. Even in times of crisis, observance of constitutional norms remains essential. In part, this is because laws passed as emergency measures tend to hang about long after the emergency subsides. New York rent control began as a wartime measure, for instance, and that curse still plagues the New York rental market. The other reason, of course, is that the Constitution is built for just these moments. The pressure to invade rights, after all, comes when things are not going well. As Justice Sutherland once put it, “If the provisions of the Constitution be not upheld when they pinch, as well as when they comfort, they may as well be abandoned.”

Forcing landlords to either renew leases or forego eviction for lease violations likely raises at least two constitutional problems: takings and impairment of contractual obligations. While such laws don’t literally seize property, they effectively impose a servitude on landlords’ property, stripping them of control over the disposition and occupation of their land. When an essential attribute of property ownership is destroyed by regulation in this manner, the government must offer compensation. We already know this compensation requirement applies during national emergencies. During World War II, for instance, the Supreme Court held that the United States had to compensate property owners and leaseholders when it temporarily seized factories for wartime production.

The contract clause problem is also straightforward: barring landlords from enforcing lease terms impairs obligations under pre-existing contracts. The contracts clause, though, has been severely undermined in recent decades, such that a showing of a compelling interest like mitigating the impact of the pandemic may well satisfy the flaccid demands of the modern contracts clause.

It may seem profoundly harsh to impose constitutional constraints on governments trying to resolve a crisis. But three things ought to be kept in mind.

First, an emergency certainly means that some will face a heavy burden, but that fact tells us nothing about how that burden should be allocated. Why should landlords bear the costs? Indeed, As the Supreme Court said in Armstrong v. United States, the takings clause exists to avoid imposing societal burdens on specific individuals: “The Fifth Amendment’s guarantee that private property shall not be taken for a public use without just compensation was designed to bar Government from forcing some people alone to bear public burdens which, in all fairness and justice, should be borne by the public as a whole.”

Second, we should keep in mind that lease agreements already account for risk. That’s baked into the price and terms that give rise to a mutually agreeable arrangement between parties. To simply allow one party to slip out of the terms of the lease distorts that arrangement.

Third, the takings clause does not bar emergency measures, including the seizure of property, but only upon just compensation. No exigency should excuse cities like San Jose or Seattle from compensating for the costs they’re hoisting upon landlords. And in the case of the contracts clause, the government could still honor existing leases by acting as a guarantor for tenants who can’t pay the rent.

All of these points apply to a world in which landlords do not voluntarily exercise leniency. But I think we’ll find that most landlords are forgiving during a temporary crisis. Most landlords have an extreme aversion to evicting tenants–it’s the nightmare, last-ditch option that they try hard to avoid. That, plus the simple dose of compassion that many landlords will feel inspired to offer, may do more toward helping see us through than any emergency measures.