In the first phase of the coronavirus pandemic, many countries were forced to embrace extreme guidelines for physical distancing to break the back of uncontrolled transmission. Locking down society proved effective at flattening the growth curve of new cases. But it came at the high cost of shutting down the economy.

Today, governments face a more complex—and, in some respects, even more daunting—challenge: to begin relaxing physical-distancing guidelines in order to restart the economy and do so in a way that doesn’t spark new outbreaks of infection. The stakes could not be higher. Making the right choices will allow economic and social activity to resume while keeping transmission at manageable levels. Making the wrong choices, however, could not only result in a renewed surge of cases and fatalities but also wreak further economic harm by dashing consumer and business confidence. Indeed, some state and regional governments around the world have already had to slow the reopening process or even institute new lockdowns in the face of spikes in the number of people infected.

To address the complex tradeoffs of restarting the economy while also managing virus transmission levels, governments should look to the process that they typically use to make decisions about public spending. Elected officials routinely make tough choices about how to use limited financial resources to maximize social and economic well-being. There are always more good ideas than available resources to fund them. The fiscal budgeting process is the mechanism by which public officials assess the benefit-to-cost ratio of various spending choices and make critical spending decisions and tradeoffs.

The same approach can be applied to reopening the economy. To make decisions about relaxing physical-distancing guidelines, public officials should assess the benefits and costs of various types of social interactions to make informed tradeoffs that will allow more economic and social activity while also minimizing the risk of uncontrolled transmission. In other words, governments should prepare a social interactions budget.

To be sure, setting a social interactions budget is, for the moment, far more difficult than preparing a fiscal budget. Few, if any, governments currently have an established and agreed-upon metric to measure interactions or a dedicated government department to track interactions accounts. Even more challenging, the dynamic relationships between government decisions on physical distancing and the behavioral responses of citizens, and their combined impact on transmission rates, are far less predictable—and less well understood—than the links between government programs and the expenditures. A social interactions budget will also need to be dynamic. The coronavirus is novel, it has a relatively high infection-fatality rate for vulnerable groups, and it has the potential for exponential growth. In other words, the consequences of overspending—of allowing more social interaction than is prudent—are far more severe than a fiscal deficit and harder to recover from.

Nevertheless, framing the challenge of reopening the economy in terms of a social interactions budget is a useful way for policymakers to surface the key tradeoffs they are facing and to begin to address them. There are four steps to get started.

Adopt a Framework for Comparing Activities

All things being equal, more social interactions will lead to more cases. But the relationship isn’t linear. Some types of interactions are safer than others.

Social and economic activities differ in terms of the number of unique interactions they involve (that is, the number of interactions with different people, as opposed to recurring interactions with the same people) and in terms of the relative risk of those interactions. The degree of risk is a function of the duration and proximity of contact among individuals and the location in which the contact occurs. For example, activities that involve sustained, close contact with large numbers of people indoors—in nightclubs, say, or at wedding receptions—are potential superspreader events. They are more likely to lead to higher rates of transmission than activities such as golfing or tennis that involve smaller numbers of people who are at more of a distance from one another and that take place outdoors. (See Exhibit 1.)

Therefore, the first step that governments should take is to evaluate the relative transmission risk posed by each activity against the economic and social benefits of allowing it to occur. What will be the economic impact of easing restrictions on a particular activity in terms of reduced unemployment or increased consumer spending? What will be its social impact in terms of, for instance, reduced social isolation and a resulting decrease in the use of mental health resources? For any given activity, what is the potential to use personal protective equipment, adopt good hygiene practices, and modify other behaviors to reduce the impact on transmission rates? And how might easing restrictions on some activities affect the likelihood of widespread citizen compliance with other, more important, restrictions?

This matrix of considerations allows policymakers to make structured choices about which restrictions to lift, which to retain, and which to modify. For example, an Australian state government is assessing each step in its reopening process against three dimensions: its potential economic benefits in terms of jobs and economic value creation, its potential social benefits in terms of improved mental health and social equity, and the degree of increased health risk from the kind of social interactions that are likely to occur.

It is worth noting that, in this context, transmission risk refers only to the risk of the infection being passed along within the population, not the impact of infection on any particular individual. The goal of the social interactions budget is to minimize the broader social risk that the transmission rate gets out of control. The health risk for specific individuals is a separate question, one that varies substantially across population segments. Individuals who have a higher risk of a poor outcome should be encouraged to restrict their interactions in order to minimize their personal risk, and governments and private-sector organizations should develop specific policies to protect these vulnerable groups.

Budget Exponentially, Not Arithmetically

In the early phase of the pandemic, many nations experienced exponential case growth that threatened to rapidly overwhelm the national health system. That’s an important lesson to keep in mind when planning the reopening phase as well. As restrictions on social interactions are unwound, the degree of interaction in a community also grows exponentially.

Consider this simple illustrative example of two communities: in one, people average two new unique contacts per day; in the other, people average four new unique contacts per day. At the end of single week, each member of the first community will be part of a cluster of 128 people in unique contact. But members of the second community will be able to trace a contact to 16,384 people—a 128-fold difference. (See Exhibit 2.) The point: each additional source of interaction adds dramatically to the degree of contact in a community and, hence, to the risk of uncontrolled transmission.

The implication for governments is that since each incremental easing of restrictions on social interaction adds exponentially to risk, the resulting economic and social benefits need to be correspondingly greater.

In general, citizens should be encouraged to stay within their local bubbles of home and work or school. Restarting education—especially for children in preschool or primary or middle school—is critical, both to ensure that an entire generation of students does not fall behind and to free up parents to return to work. And a great deal of local business activity should have a relatively modest cost in terms of social interaction. For example, given that most interaction in quick-service restaurants occurs between groups of people that either live or work together, the cost to the social interactions budget will be relatively low if physical-distancing requirements are observed.

But governments should be particularly cautious about easing restrictions on activities that encourage interactions beyond the local community. For instance, youth or school sports leagues can create significant new unique contacts (both among players on the court or field and among parents and supporters on the sidelines, many of whom would not otherwise be in contact with each other) and should probably be delayed or discouraged. Many activities connected to travel and tourism—flying, taking a cruise, and visiting an amusement park, for example—will have a much higher cost in terms of new unique contacts and should be carefully assessed against the potential economic and social benefits.

Such tradeoffs are, of course, extremely complex, involving tough decisions that must balance competing goals not only of economics and health but also of social equity and personal freedom. All the more reason for governments to develop explicit frameworks in a social interactions budget to address these tradeoffs in a systematic and accountable fashion.

Monitor Interaction Levels Dynamically

Interaction levels are rarely static; they are likely to fluctuate, as will their impact on economic growth and virus transmission. Therefore, it’s important to take a control tower approach that draws from multiple live (or nearly live) data sets to continuously monitor interaction levels. This will allow governments to develop a holistic view of the level of social interaction in the community at a given moment in time and to assess the impact of these levels on economic activity and on the rates of virus transmission. 

Relevant data sets that governments can use to develop this dynamic view include:

• Interactions data sets. These include mobility data from telecommunications companies, crowding data derived from CCTV feeds, transit statistics, traffic flow data, and store transaction data.
• Economic data sets. These include anonymized private-sector business data from credit card processors, payroll firms, and financial services institutions. The data can be analyzed to assess consumer spending patterns, employment rates, business revenues, and other economic indicators.
• Virus-transmission data sets. These include data from sentinel testing, sewage testing, hospital admissions, and excess-deaths analysis. Governments can use these data sets to track the rate of infection spread and the virus-related health outcomes.

Such data sources are becoming increasingly available to track the impact of the coronavirus pandemic. In the US, a group of economists, led by Raj Chetty at Harvard University and with the support of the Bill & Melinda Gates Foundation, has created the Opportunity Insights Economic Tracker to monitor the economic impact of the pandemic by county and income group in nearly real time. And in Australia, a state government is combining data feeds on public-transport payments, bank payments, smartphone mobility, school attendance, and road traffic with leading health indicators (such as the number of ambulance calls for patients reporting breathing problems) to develop a holistic picture of the consequences of reopening the economy at a statewide and regional level. 

When a government is actively monitoring such data, it is in a position to respond to behavioral changes and fine-tune restrictions on the basis of shifts in economic activity or virus transmission levels. Active monitoring also makes it possible for governments to use artificial intelligence techniques, such as machine learning, to identify patterns and meaningful relationships in the data.

Engage the Public to Manage Individual Interactions Budgets

When it comes to managing the coronavirus, what’s good for the individual is also good for the nation. Governments should engage with the public to help people understand that while restrictions are lifting, the virus will be best controlled if we all limit our interactions to those activities that will produce the greatest economic and social benefit. We should do the essential things: for example, commute to work (if we must be there in person), return to school, go to the store for food and other critical goods, and care for relatives and friends. But we should limit inessential interactions—particularly those that greatly expand the risk of virus transmission while producing minimal positive economic or social impact.

The public should be encouraged to keep a log of daily contacts (a widely followed practice in New Zealand during the early phase of the pandemic). A log makes people aware of their own level of interaction and helps contact tracers track down contacts, should that become necessary. Any mobile apps used in contact tracing could also be adapted to inform users of the number of contacts or new unique contacts that they have had each day, helping individuals track and manage their own behavior.  

Governments should also explain the risk frameworks that they are using (so people can make their own judgments about situations to avoid) and publish the data they are collecting about overall interaction levels, economic impacts, and virus transmission rates. Community and business leaders should be enlisted to amplify messages that encourage people to manage their own personal interaction budgets. And private-sector organizations should be encouraged to provide the flexibility that makes it possible for individuals to choose the level and type of interactions that are most appropriate for them.

If we want to reopen our economies safely, we must keep transmission levels under control. Although we cannot return to the levels of social interaction that caused coronavirus cases to grow exponentially, we can live within a meaningful social interactions budget that allows more economic and social activity while also helping to manage the virus.