Balloux suggested three possible (and possibly overlapping) answers. First, that some degree of “residual social distancing” is still preventing COVID-19’s spread. This resonates with the argument made by Lyman Stone, the health data analyst I interviewed in April about why lockdowns may not be the best option for a robust and effective public health response. It also fits with some recent U.S. poll data suggesting most Americans are (or were, pre-police protests) still being more cautious than usual even as lockdowns end and their movement increases. There’s reason to believe rising mask usage is making a difference, too: A study published in The Lancet on Monday used data from 16 nations to show a dramatic drop in COVID-19 transmissibility when masks are worn.

Second, Balloux continued, is the much-debated subject of seasonality: Other coronaviruses follow predictable seasonal patterns, with infections increasing when it’s cold and dry and decreasing when it’s warm and humid. Maybe summer is giving us a respite more than anything we’re doing or neglecting. The George Floyd protests will function as an enormous natural experiment here: If there is no protest-linked resurgence, that alone may reasonably establish seasonality (though the rising caseload in hot, humid India isn’t encouraging). The severe outbreak in Brazil — which now has the second-highest caseload worldwide as the Southern Hemisphere moves into winter — may also help to bring clarity.

Balloux’s third idea was cross-immunization: “A proportion of the population might have pre-existing immunity to #SARSCoV2, potentially due to prior exposure to ‘common cold’ coronaviruses.” If that were true, it could explain why there might be an appearance of herd immunity even as antibody testing for this specific virus returns fairly few positives.