What is the economic logic of using test and trace to fight Covid-19?

Targeted policies such as test and trace can help to control the spread of Covid-19 while limiting the economic and mental health costs associated with lockdowns. But the incentives need to be right to ensure fairness and compliance.

The combination of widespread testing and contact tracing – often shortened to ‘test and trace’ – can be a cost-effective way to fight the pandemic. It allows for restrictions that are targeted at those who are most likely to be infectious and as a result causes less economic damage than general restrictions (such as nation-wide lockdowns).

Further, by keeping infection numbers down, test and trace makes it possible not only to replace more expensive measures now, but also to prevent them from becoming necessary in the future.

But test and trace is not cheap. The UK government spent more than £13.5 billion on its much-maligned system. Citizens also face the costs, as they may repeatedly be told to stay at home by contact tracers or apps, on the chance that they might be infected.

So is test and trace worth it? And how can governments get it right?

What do anti-epidemic strategies try to do?

The most basic goal of any anti-epidemic strategy is to limit the number of people that each infected person infects – what’s referred to as the R number. This is true whether the aim is to make the virus more or less disappear from a population (as has been pursued in Australia, New Zealand and Taiwan for much of the pandemic) or to keep the number of infections at a level where the health system can cope (as has been the case in much of the rest of the developed world, insofar as there was a strategy).

The main difference between these approaches is the reduction in the R number that governments are seeking to achieve. The former approach requires keeping R below 1 constantly, while the latter is more dependent on the current status of the epidemic. For example, many countries have been following a strategy that implements measures in response to local infection numbers or hospitalisation numbers.

In either case, governments can rely on two broad strategies to control the spread of the virus.

First there are general (or ‘dumb’) policies, which treat everyone as if they might be infected and impose general restrictions. Mask mandates, venue capacity restrictions and lockdowns fall into this category.

Since these policies apply to whole populations, they can be hugely costly, especially where freedom of movement is concerned, and schools and businesses are closed. A study from the Czech Republic estimates that each week of school closures wipes out between £1 billion and £2 billion in future income, even if schools switch to online learning (Jann et al, 2021).

This number is likely to be higher by a factor of ten or more in the UK, where average wages are higher (and the population is larger). Other studies have shown that even in countries with very good online learning infrastructure, the learning losses from school closures are immense (Engzell et al, 2021).

Targeted (or ‘smart’) practices, on the other hand, adjust the severity of measures according to the probability that a person is infectious. People with a positive test are told to isolate, as are those who have been in contact with them , although not as strictly given that they have a lower probability of being infectious.

Why are targeted measures cheap (even if they cost a lot)?

In general, these more targeted measures have a much lower economic cost than the general ones. While contact tracing might put many thousands in isolation, even at the height of a pandemic wave, it is unlikely to affect more than a few percent of the population.

A simple back-of-the-envelope calculation explains how this is efficient. Imagine that a test and trace programme that puts 1% of citizens in quarantine lowers R to the same level as a set of society-wide restrictions. Then the programme can cost 100 times as much per affected person as the per person cost of the general measures, and still achieve the same effect at the same cost. Since in reality, the direct and indirect costs of test and trace are likely to be much lower, an effective programme can therefore be very cheap.

But unlike the general measures – the effectiveness of which is mostly independent of the number of infections – the targeted measures work best at relatively low infection levels. At higher numbers, health authorities cannot keep up, contact tracers are overwhelmed and many people never get tested or are only told to quarantine by the time they have already infected others.

This is why all counter-epidemic measures must solve an ‘intertemporal’ problem, in which costs and benefits today are traded off against the same tomorrow, next week or next month. If infection numbers are allowed to grow for a while, any health system will reach its limit.

This will be the case whether almost nobody is immune (as was the case for the whole world for most of 2020), or whether a large part of the population is recovered or vaccinated. In this way, targeted measures can help to prevent the future costs of general measures.

How do we get the incentives right in contact tracing?

These cost-benefit calculations only apply if test and trace actually works. This requires test results to arrive quickly and tracers to work fast.

But there is also an internal logic of incentives that any programme must follow. Those who test positive or are quarantined need to follow the orders and guidance of the health authorities. But in the UK, studies suggest that only around 50% of contacted people comply with quarantine orders and around 80% share data of close contacts (Smith et al, 2021).

Often, the economic incentives are quite simple. If those who are quarantined only get statutory sick pay (as they do in England), compliance is a question of economic means – some will be more able to afford the time off than others.

Other countries have provided more generous rules for those who stay at home. In Germany, for example, employees continue to receive their full salary if they are put in mandatory quarantine by contact tracers.

Since compliance with isolation or quarantine orders is often seen as a moral issue – it was called a ‘civic duty‘ by Matt Hancock, the UK’s then secretary of state for health – many are quick to call for police controls and sanctions. This may range from monetary fines all the way up to the five years in prison for breaking quarantine orders that a bill before the Hungarian parliament introduced.

But such draconian enforcement is likely to be counterproductive. If people know that quarantine orders are harshly enforced, they will not get tested if they experience symptoms, and provide less information about whom they have been in contact with if they test positive.

Cooperation with contact tracers is almost impossible to compel by force. In the UK, a third of young people have already deleted their contact tracing apps and even if one had to prove that the app is installed, it is easy to render ineffective.

Fines, on the other hand, also have the effect of giving some people the impression that breaking Covid-19 rules is a commodity that one can pay for, rather than a transgression against the rules of society.

In a famous experiment, Israeli day-care institutions introduced fines for parents who arrived late to pick up their children (Gneezy and Rustichini, 2000). The study showed an increase in the number of late pick-ups. The awkwardness of arriving late had been replaced with a simple monetary transaction for those who can afford it.

Similar effects can be expected if wealthy people decide it is worthwhile simply to risk a modest fine on the low chance that their quarantine adherence is subject to supervision.

The incentives, as well as the messaging, around contact tracing are therefore hard to get right. It probably helps to rely on simple messages such as: if someone is told to stay at home, they are endangering others if they go out, even if they feel fine. Such simple messages have, unfortunately, become mangled in many governments’ communications around the world.

But just as importantly, a person needs to be able to stay at home and cooperate with the health authorities without suffering great economic or social disadvantages.

How do things change when many people are vaccinated?

Not all that much. Societies will still need to limit transmissions for a while, since otherwise even an epidemic in a mostly vaccinated population can overwhelm the health services. Indeed, this is what several Western countries are now experiencing.

Targeted measures, including contact tracing, remain one of the cheapest and most sustainable ways of controlling the spread of the virus.

Where can I find out more?

Covid-19: Better contact tracing could help save lives and the economy: Report on how contact tracing can be improved based on a community pilot project.
Fostering participation in digital contact tracing: Dominik Rehse and Felix Trehmöhlen discuss the future of digital contact tracing, reporting that the technology needs to be widely adopted to be most effective.
Optimizing and implementing contact tracing through behavioral economics: Report on how to enhance contact tracing with control measures such as isolation and quarantine.
Measuring the scientific effectiveness of contact tracing: Evidence from a natural experiment: Thiemo Fetzer and Thomas Graeber explore the effectiveness of contact tracing on infections and deaths from Covid-19.