In the pandemic today, the United States and many European countries are experiencing the worst of both worlds. Businesses are only partly open, causing widespread unemployment, many schools remain closed, and our personal lives are greatly curtailed — yet the spread of the novel coronavirus is accelerating rather than receding. On Sunday, the seven-day average for daily infections in the United States reached 69,000, the highest figure yet.

Nobody wants to return to the full-scale open-ended shutdowns of the early days of the pandemic. And yet muddling forward is not working. As a compromise, some European leaders have embraced the idea of “circuit breakers” — fairly intense closures lasting a limited time — that would interrupt viral spread and bring case counts down without the long-lasting social and economic pain of extended lockdowns. Britain’s Labour Party leader, Keir Starmer, has called for a circuit-breaker of two to three weeks, for example, to prevent a “sleepwalk into … a bleak winter.” Circuit breakers can be one-off interventions or regular occurrences (say, a three-week “reset” every two months).

The case for circuit breakers rests in part on the failures of our current approach. Phased reopenings aim to allow as much activity as possible, consistent with keeping cases at controlled levels through social distancing, masking and other restrictions. But they have a number of disadvantages in practice. First, they are hard to maintain over extended periods as compliance fatigue sets in, standards relax — and cases tick up. Furthermore, disparities in infection rates, both between and within states, make sacrifice inefficient. Many towns and subpopulations have already sacrificed more than enough to eliminate internal spread, but a constant threat of outside reinfection prevents them from relaxing and enjoying their triumph.

The underlying idea of the circuit breaker approach makes intuitive sense: After cutting oxygen off from a fire to reduce it to embers, for instance, restoring airflow intermittently doesn’t produce large flames. But add a steady continuous oxygen supply, even at a low level, and the blaze quickly rages out of control.

One study making use of epidemiological models — as yet, not peer-reviewed — found that a two-week circuit breaker would halve the number of deaths in the United Kingdom between now and the end of the year. And our own preliminary research suggests that if society wants to allow a certain amount of social and economic activity, doing so within select windows, followed by short-term shutdowns, will lead to fewer cases than allowing the same amount of activity to occur across unbroken stretches of time.

Using mathematical techniques, our work explored questions like this one: Suppose a town wishes to allow 1,000 small but risky activities like haircuts, ballgames, academic classes and small social gatherings over the course of a month. Is it better to put all the activities in one half of the month or space them out evenly? The mathematical answer is that concentrated sacrifice, followed by a period of relative openness, beats sacrifice that is spread out consistently over time. The size of the impact depends on many factors, but a typical analysis finds that 10 to 20 percent more activity might be possible (for the same amount of disease spread) when the activity is more concentrated.

Of course, not all activities can be shifted in time, so necessary events like urgent medical procedures would continue during a circuit-breaker shutdown. Additionally, when a month of activity is compressed into two weeks, businesses may have to extend opening hours to maintain appropriate physical distancing while keeping up with customer demand. Avoiding crowding is paramount. But even if you account for such issues, the benefits of intermittence remain.

There are other points to consider in designing an effective circuit breaker. A raging outbreak in one subpopulation (college students, for instance, or migrant workers) can prevent a whole community from stamping out the disease. Public health officials should therefore devote extra resources — such as testing, tracing and information campaigns — to the most infected populations. It may be necessary to prolong the circuit breaker (perhaps three weeks instead of two) in hot zones or among highly infected groups.

According to our mathematical models, one of the least efficient ways to fight the disease is to reduce activity solely among subpopulations that already have low prevalence. If a circuit-breaker policy exempts too many of the people at highest risk (such as service workers whose jobs put them in close contact with others) it could inadvertently end up doing exactly that.

Last spring, nearly half of workers were defined as essential; such exemptions made the shutdowns somewhat more palatable but reduced their impact. Because circuit breakers are short, they require fewer exemptions. This time around, many more factories and retail establishments should close, making up the time by opening for extra hours during the subsequent weeks. On the other hand, businesses that have essentially eliminated their transmission risk (through contactless supply chains, personal protective equipment and other innovations) might continue. In general, the open periods following the shutdown would resemble current rules, under which masks and sensible caution combine with relative social freedom.

If successful, a circuit breaker could reduce infections to levels that give the United States a second chance to implement an effective testing and tracing regimen. Public and private entities across the country are testing at far higher levels than earlier in the pandemic, but they can’t keep up with the rising cases; a circuit breaker would allow a reset. The same is true for tracing, which can range from mild interventions (phone calls to high-risk contacts) to robust approaches (detective work on possible sources, rapid testing and subsidized quarantines). The robust responses are more effective, but only feasible when case numbers are already low — a situation a circuit breaker might bring about.

Critics rightly point out that even limited-duration shutdowns are difficult both for individuals and for businesses: New federal subsidies, now stalled in Congress, would therefore be essential.

It would be wonderful if giving up public activities was unnecessary — if masks, contact tracing and common-sense distancing were enough. But they’re not. Curtailing social and business activity is still essential: The point is to make the sacrifices as efficient as possible. The amount of sacrifice required to keep the disease under control is greatly reduced when it is coordinated and concentrated. The current approach — uneven, indefinite sacrifice that wanes over time — is clearly not working.