In these difficult times, we’ve made a number of our coronavirus articles free for all readers. To get all of HBR’s content delivered to your inbox, sign up for the Daily Alert newsletter.

The greatest challenge we face in the current crisis is striking the right balance between preventing not only the damage inflicted by Covid-19 but also that which comes from a severe contraction of economic activity.

This is a matter of both life and death, and prosperity or poverty, as a deep recession is likely to cause widespread hardship, including increased mortality rates for reasons other than the virus. We must, therefore, find smart strategies for reopening businesses while minimizing health risks.

Further Reading

As economists, we have explored a variety of options: a long lockdown, relaxation with rules, stop-and-go lockdowns, and phased-in lockdowns by age, geographic region within countries, and by country in sectoral waves. Our analyses suggest that, assuming that minimum epidemiological standards are met (declining new cases for one week or two weeks, adequate intensive care unit capacity and testing and tracing protocols), the latter approach — sequential opening of sectors starting with those least likely to generate a substantial resurgence of the virus — is the most promising.

Here’s why:

The Long Lockdown

While it is difficult to be precise about the economic cost of continued lockdowns, we know the effects are already massive. In Switzerland, for example, the loss of value added is between CHF 0.7 and 1.4 billion per day. In Spain, the cost is slightly above €3 billion per day. In the United States, it is a staggering $14 billion per day. There may well be other less visible but potentially significant non-monetary costs, including depression and other mental health problems, domestic violence, the saturation of hospitals impacting the treatment of other diseases, etc.

Ideally, societies would find an epidemiologically risk-free approach for reopening, for example by waiting for a vaccine to be discovered. However, this would take as much as 12 to 24 months. We cannot remain in full lockdown that long.

Another risk-free option in medical terms would be to make the release of the lockdown conditional on a serological immunity test. But the WHO recently warned that the presence of antibodies (IgG) does not necessarily offer comprehensive and lasting protection. And, again, we have no visibility into when such testing will be widely available. Moreover, it is likely that immunity in the population is still too low: blood tests conducted in Santa Clara County, California, estimate that only between 2.5% and 4.2% are infected. In Geneva, one of the hardest hit parts of Switzerland, the estimated prevalence is 5.5%. Herd immunity is very far from being achieved.

Relaxation with Rules

Could we lift the lockdown on a large scale by imposing only strict rules on the wearing of masks, social distancing, and other sanitary rules applied in every business? This is what many corporate leaders clamor for, but it faces limitations: Will there be enough masks? Will people follow the guidelines, for example, by installing privacy preserving contact tracing applications on their smartphones?

More importantly, relying on individual responsibility here creates obvious conflicts between personal and collective interests. Will a retailer who doesn’t have the space for of social distancing be prepared keep her shop closed when her direct competitor with a larger floor is allowed to open? And, importantly, it will be difficult to control the human density outside their stores, offices and production sites, particularly in city centers and on mass transit systems.

Stop and Go

A fast and large-scale lifting of the lockdown would likely lead to other lockdowns in the near future. There are several important drawbacks to this stop-and-go approach.

First, governments and health organizations are still not able to closely monitor the spread of the virus or lock down at great speed. Data lags and political decisions take time. This creates the risk of generating new Covid-19 explosions, which eventually will make a stop-and-go option even more costly — medically and economically.

Second, a stop-and-go period would greatly increase uncertainty, curbing business investment and economic growth beyond the effects of the pandemic. This will be not only about the dates when the economy starts or locks down again, but also about how workers, suppliers, consumers, and the markets will react.

Last, there are fixed costs involved in restarting businesses, including payroll adjustments and commercial effort reactivation, which will have to be incurred irrespective of whether the “go” period is short or long.

Phased-in Re-opening by Age Group

A tempting option for a selective lifting of the lockdown would be to proceed by age group; for example, by first releasing those under 45. Our computations, based on Swiss data, suggest that, compared with a release of the entire population, the risk of overcrowding intensive care facilities with this approach would be reduced by 80% for the obvious reason that younger people are less susceptible to complications from infection. However, not all are; many have already fallen victim to Covid-19.

And such an approach would raise other issues. Besides the fact that it is inherently discriminatory, it ignores the ubiquitous complementarity between workers of different generations. Indeed, with a sequential release by age group, firms with younger workforces would find themselves at an advantage over their competitors, but they would also face major management and monitoring difficulties.

Phased-in Re-opening by Region Within Countries

Another option would be to design the lockdown release by geographical area. Countries such as France and Spain seem to envisage taking this route, with the degree of openness and freedom to commerce depending on the severity of the pandemic in a specific region. The United States also seems to be adopting such a strategy, letting states decide when they reopen.

The efficacy of such an approach will depend on the coincidence between the epidemiological and the economic realities. First, regional differentiation makes sense only to the extent that there is a clear difference in the prevalence of infections. Second, unless in the case of clearly distinct market units — an island or distant state like Alaska, for example — an approach by region ignores the difficulty to restructure intra-state/region supply chains and connected product markets. It distorts competition. It is not compatible with labor mobility. And it relies to a great extent on personal responsibility, since enforcement would require a control of regional borders.

Phased-in Re-opening by Country in Sectoral Waves

We believe the best option is a gradual release by sector — in several waves, within countries — with the objective of avoiding a congestion of hospitals and especially intensive care units. Our analysis suggests that this is actually possible. The sectors to be released from lockdown as a matter of priority could be chosen based on the following analysis:

  • Inability to conduct core business from home. Industries such as accommodation and food services, construction, healthcare, transportation, and warehousing could be the targets of a first wave of lockdown release because they are less suitable for virtual work. Some organizations in these sectors have already been deemed essential and continued operations through the pandemic.
  • Importance to the national economy. In some countries, manufacturing accounts for a larger percentage of GDP than accommodation and food services (e.g., China or Ireland). In others it is the opposite (e.g., Australia or Norway).
  • Value added per worker. All else equal, it makes sense to release fewer (less risk of contagion) but more productive (smaller economic loss) people.
  • Business viability. One must also take into account the viability of certain sectors. For example, retailers are facing severe hardships and the business model makes it harder for state support to compensate.

A caveat: Within each sector, certain functions are either dispensable in the short term or easier to perform remotely. Companies reopening would therefore determine which jobs could remain virtual or begin later. Think of support staff, coordination, and reporting jobs, certain maintenance and service tasks, and management and non-operational departments within large firms.

As an example, we apply our model to the case of Switzerland. We start by constructing an indicator of “value loss in the lockdown,” which combines the ease of switching to remote work with the relative weight of a sector in the economy. For some sectors (construction, retail) we acknowledge that, irrespective of their contribution, they are mostly represented by small businesses whose return to economic activity is vital.

Confronting this measure with an evaluation of the risk of contamination — that is, the density of the workplace and the resulting ability of workers to operate while observing requirements of social distancing and hygiene — we conclude that, for Switzerland, employees in the health (those not already directly involved with treating Covid-19 patients), construction, and manufacturing sectors could be released in a first wave (1.8 million people, out of an active population of 5.5 million). The second wave would concern workers in retail and wholesale trade (0.6 million). Third, and in the absence of a surge of the epidemic (a process that will have to be followed with all the available means), the 1.2 million in finance, administration, hotels, restaurants, and transportation could be released.

Some logistical complementarities would still have to be considered during this transitional period; for example, the reopening of day-care centers and primary schools, and a cautious return to work of certain catering and transport functions. Our analysis suggests that this proposed plan would progressively put businesses back into full action at a pace that would have to be adapted to national circumstances and with reduced risks of overwhelming hospitals.

We have done a similar analysis for Spain, which would suggest that, in the first wave, manufacturing, construction, and health services are released. In a second stage, retail and real estate would be allowed to operate. In a final wave, the rest of sectors (mainly tourism, financial services, and administrative and professional services) could reopen. The comprehensive analysis can be found here.

This strategy has the virtue of being adaptive — as data is gathered following each sector-wide reopening, adjustments can be made concerning the timing of subsequent phases, and protective measures adopted in previously released sectors can be copied and improved as more is learned about the epidemic. If adopted by national governments, plans for phased-in re-openings by sector would offer companies some much needed forward guidance and allow them to plan accordingly.

There is no perfect strategy, at least not with the information we have today. But we think this approach will allow us to get through this once-in-a-century pandemic with as little loss of life and life opportunities as possible.

If our content helps you to contend with coronavirus and other challenges, please consider subscribing to HBR. A subscription purchase is the best way to support the creation of these resources.