By Derek Thompson

One month ago, the CDC published the results of more than 20 pandemic forecasting models. Most projected that COVID-19 cases would continue to grow through February, or at least plateau. Instead, COVID-19 is in retreat in America. New daily cases have plunged, and hospitalizations are down almost 50 percent in the past month. This is not an artifact of infrequent testing, since the share of regional daily tests that are coming back positive has declined even more than the number of cases. Some pandemic statistics are foggy, but the current decline of COVID-19 is crystal clear.

What’s behind the change? Americans’ good behavior in the past month has tag-teamed with (mostly) warming weather across the Northern Hemisphere to slow the pandemic’s growth; at the same time, partial immunity and vaccines have reduced the number of viable bodies that would allow the coronavirus to thrive. But the full story is a bit more complex.

1. Behavior: Maybe Americans finally got the hang of this mask and social-distancing thing.

“If I were ranking explanations for the decline in COVID-19, behavior would be No. 1,” says Ali Mokdad, a global-health professor at the University of Washington, in Seattle. “If you look at mobility data the week after Thanksgiving and Christmas, activity went down.”

Other officials have pointed to Google mobility data to argue that Americans withdrew into their homes after the winter holidays and hunkered down during the subsequent spike in cases that grew out of all that yuletide socializing. New hospital admissions for COVID-19 peaked in the second week of January—another sign that social distancing during the coldest month of the year bent the curve.

Our cautious behavior evidently requires the impetus of a terrifying surge. In the spring, southern and western states thought they had avoided the worst of the early wave, and governors refused to issue mask mandates. Then cases spiked in Texas, Florida, and Arizona, and mask-wearing behavior in the South increased. When cases came down again, people relaxed, cases went up again, and the awful do-si-do continued.

The lesson is not to let today’s good news become tomorrow’s bad news, again. Until much of the population is vaccinated, don’t interpret the decline in cases as a green light to resume your pre-pandemic behavior.

2. Seasonality: The coronavirus was perhaps destined to decline this time of year.

Behavior can’t explain everything. Mask wearing, social distancing, and other virus-mitigating habits vary among states and countries. But COVID-19 is in retreat across North America and Europe. Since January 1, daily cases are down 70 percent in the United Kingdom, 50 percent in Canada, and 30 percent in Portugal.

This raises the possibility that SARS-CoV-2, the virus that causes COVID-19, is seasonal. Last year, a meta-study of coronaviruses such as SARS-CoV-2 found that they typically peak in the Northern Hemisphere during the winter, with the most common peak months being January and February. “The apparent seasonality of human coronaviruses across the globe suggests that this phenomenon might be mined to produce improved understanding of transmission of COVID-19,” the authors concluded.

The notion of seasonality is both obvious and mysterious. We know that many respiratory viruses are less virulent in the summer, accelerate in the closing months of the calendar year, and then recede as the days grow longer after December. But as the Harvard epidemiologist Michael Mina told New York magazine, “We don’t fully appreciate or understand why seasonality works.”

What we call seasonality seems to be a combination of environmental factors and the things people do in response to them. Many viruses fare best in cold and dry conditions; they’re not well designed to thrive in warmer, sunnier, and more humid outdoor areas. Each virus is a bundle of genes and protein encased in a fatty lipid molecule. This fatty shell breaks down more easily in warmer environments. You can see this for yourself when you try washing a smear of butter off your hands with cold water versus warm water.

But seasonality isn’t just “the air” or “the weather,” because the coronavirus doesn’t thrive in the air nor does it live inside the weather. It thrives and multiplies in our bodies, and people do different things with their body when the temperature changes. When the temperature drops, we trade the outdoors, where the virus can struggle to multiply, for the indoors, where we clump together on chairs and couches, making it easier for the virus to pass from one host to another. We close our windows and constrict circulation and ventilation, which give airborne viruses another advantage.

Natural adaptation to cold weather could also make us vulnerable to respiratory viruses such as SARS-CoV-2. Our blood pressure rises during the winter, and our vitamin D levels dip as the days shorten and the sun hangs at a lower angle in the sky. While feeling cold can’t give you a cold, cold and dry air can indeed suppress the local immune response in our nasal passages, which makes us easier targets for, say, an airborne respiratory virus that binds with enzymes commonly found in our noses.

Still, February is pretty cold. In many parts of the country, it’s been colder and drier than it was in late December, and unexpected places are experiencing a cold snap. So trotting out a meta-study of seasonality and saying the decline in cases was simply inevitable won’t satisfy skeptics. Something else may be happening too.

3. Partial immunity: Is the virus running out of bodies?

The coronavirus needs bodies in order to survive and replicate, and it now has access to fewer welcome hosts. Fifteen to 30 percent of American adults have already been infected with COVID-19, according to CDC estimates. Since people recovering from COVID-19 typically develop lasting immunological protection for many months (at least), the number of antibodies swirling around the U.S. population may naturally constrict the original coronavirus’s path forward.

America’s seroprevalence—that is, the number of people with coronavirus antibodies from a previous infection—is not randomly distributed across the country. Instead, immunity is probably concentrated among people who had little opportunity to avoid the disease, such as homeless people, frontline and essential workers, and people living in crowded multigenerational homes. It might also include people who were more likely to encounter the virus because of their lifestyle and values, such as risk-tolerant Americans who have been going to eat at indoor restaurants.

What I’m describing here is not herd immunity. Nothing is herd immunity, really. But it is partial immunity among the very populations that have been most likely to contract the disease, perhaps narrowing the path forward for the original SARS-CoV-2.

The emphasis here is on the word original, because we cannot forget the variants. The virus mutations from South Africa and Brazil in particular may elude the immunological protection among COVID-19 survivors, according to Mokdad. “In studies of seropositive and seronegative people”—that is, people with and without antibodies—“it didn’t seem to make a difference for the South Africa variant; everybody got the new variant equally,” he told me. “So we have to be looking at variants that previous infections are not protecting.”

This fact heightens the importance of accelerating vaccinations before these variants take off in the United States. And, as it happens, vaccinations are the last piece of our explanatory portfolio.

4. Vaccines: The shots work.

COVID-19 cases started falling in January, when almost nobody outside of the health-care industry had been vaccinated. So vaccines probably don’t help us understand why the plunge started. But they can tell us a bit more about why the decline in hospitalizations has accelerated—and why it’s likely to continue.

The vaccines—especially the synthetic-mRNA vaccines from Pfizer-BioNTech and Moderna—are highly effective at preventing infection. But preventing infection is not all they do. Among those infected, they also reduce symptomatic illness. And among those with symptoms, they reduce long-term hospitalization and death to something like zero. A vaccine is not just one line of immunological defense, but several—a high wall protecting a castle and, to fight the few who bypass the wall, a group of castle defenders holding vats of searing-hot tar to pour all over the invaders. (Research indicates that some vaccines, such as AstraZeneca’s, lose their efficacy in the presence of coronavirus variants, but others, such as Pfizer’s, seem to provide potent protection. More research is necessary to say anything certain about how the vaccines protect against serious illness caused by the more contagious new strains.)

A bit of back-of-the-envelope math shows why this period of declining hospitalizations should keep going. Let’s assume the CDC is correct that about 25 percent of adults have COVID-19 antibodies from a previous infection. Let’s add to that number the 10 percent of adults who have received vaccine shots since December, assuming an overlap of 3 percent. That would mean one-third of adults currently have some sort of protection, either from a previous infection or from a vaccine. At our current vaccination pace, we’re adding about 10 million people to this “protected” population every week. We’re accelerating toward a moment, sometime this spring, when half of American adults should have some kind of coronavirus protection. And we should be particularly optimistic about severe illness among older Americans, since the vaccines are disproportionately going to people over 50, who have accounted for 70 percent of all hospitalizations.

That’s a lot of messy arithmetic. But the upshot is simple: Even if the rise of new variants slows the decline in cases, it is unlikely to lead to a sharp rise in mortality and hospitalizations. Although the pandemic isn’t over, we have perhaps reached the beginning of the end of COVID-19 as an exponential, existential, and mortal threat to our health-care system and our senior population.