It took awhile for Americans to be worried about the coronavirus. In February, the United States had very few recorded cases. Over the course of March, the number began to increase — slowly, and then quickly. By April 1, the nation had about 200,000 known cases, a seemingly unfathomable number.

But it soon got worse. Cases surged in April — as did public concern about contracting the virus. Polling from Quinnipiac University in March showed about half the country was very or somewhat concerned about either contracting the virus or knowing someone who might. By April, that figure jumped to 85 percent, with more people then saying they were very concerned than indicated any concern at all the prior month.

As the rate of new cases declined, though, so did concern. Particularly among Republicans, worry about contracting the virus or knowing someone who might slid in May and June. Then, in July, it shot right back up. Quinnipiac’s most recent poll shows about half the country is again very concerned about contracting the virus or knowing someone who might.

Why? Because of the recent surge in new cases. In July, the average number of new cases each day has neared 60,000, more than twice the average for the prior two months.

Of course, that also means Americans are more likely to know someone who actually has contracted the virus. In the most recent Quinnipiac poll, more than half of Americans said they know someone who has contracted it, up from about 45 percent the prior two months.

That adds an interesting layer of complexity to the question about concern. A person who knows someone who contracted the virus ought to express more concern about someone they know contracting the virus; after all, it has already happened. Sure enough, those who know someone who contracted the virus are about 15 points more likely to say they are at least somewhat concerned about the possibility.

Something you may have noticed in the graphs above showing views by party is the views among Republicans continue to lag well behind the overall population, as Democrats are more likely to express both concern and a relationship with an infected person.

What’s particularly interesting about Quinnipiac’s findings is the level of concern and the likelihood of knowing someone who contracted the virus has been mostly flat among Republicans for the past two months. Even though new cases are far more likely to be in more-Republican areas than they were at the beginning of the pandemic, Republicans still show about as much concern as they did in May.

This leads to an interesting finding. Concern about contracting the virus is more closely related to political party than to actually knowing someone who contracted it.

Although Americans overall are 15 points more likely to say they are worried about personally contracting the virus or it infecting someone they know if, in fact, they know someone who had contracted it, there’s not a significant difference among Democrats. Democrats who say they don’t know anyone infected by the virus are just as concerned about it as those who say they do. Among independents and Republicans, the difference between those groups is about the same as the overall difference.

Overall, though, Republicans are far less concerned than Americans overall about being infected or knowing someone who might be infected. They are 20 points less concerned, while Democrats are 16 points more concerned than Americans overall — a 36-point spread.

That partisan gap holds even when considering whether respondents know someone who contracted the virus. Democrats who know an infected person are 28 points more likely to express concern than Republicans who do. Democrats who don’t know an infected person are 40 points more likely to express concern than Republicans who don’t.

In other words, if someone tells you they know someone who has contracted the virus, that gives you less of a sense of how concerned they are about the virus than if they tell you what political party they belong to.

There are a lot of complicating factors here, of course, that polling can’t necessarily capture. The data do nonetheless serve to bolster the idea that the country’s view of the pandemic is shaped by politics more than experience.