Axios-Ipsos Coronavirus Index: Protesters fear the spread

Eight in 10 Americans worry that mass demonstrations around George Floyd's killing, police brutality and structural racism could trigger new coronavirus infections, in Week 12 of the Axios-Ipsos Coronavirus Index.

Why it matters: More than one in 10 people surveyed has an immediate family member or close friend who's participated — and 2% say they've taken part themselves. That puts tens of millions of people in close contact with protesters.

• It may be weeks before we fully understand the impacts of the protests on infections. But they're not the only example of Americans easing up on social distancing: 45% of respondents say they've visited in person with friends or relatives in the last week.
• Even as protesters decide the stakes are worth the risks, they're taking steps to avoid spread: 87% say they wore masks, 35% wore gloves and one third say they maintained a six-foot distance.

By the numbers: Most Americans — 86% — say protesting poses a large or moderate risk to their health.

• That's huge compared with how Americans assess the risks of activities such as returning to work as normal, retail shopping, getting a haircut or even going out to dinner — all of which have declined significantly over the past week.
• While 79% of Americans worry about a spike in cases, it's a matter of degree — an even 50% said they were either extremely or very concerned (29% were only somewhat concerned).
• Democrats (60%) were more likely than independents (51%) or Republicans (37%) to be extremely or very worried about an increase in cases. Women and older people also were more likely to worry.

Between the lines: Age, education, party ID and proximity to a city — not race — appear to correlate most to participation in protests. Overall, 11% of respondents say someone in their immediate circle of family and friends has protested. But here's how that breaks down:

• For people under 30, 20% are close with someone who has protested.
• For people over 65, it's just 4%.
• 16% of Democrats are close with someone who has protested, compared to 12% of independents and just 3% of Republicans.
• 16% of people with a bachelor's degree are close with someone who has protested, but that's only true for 7% with a high school education or less.
• People who live close to a city were more likely to be close with someone who protested: 14% of people in urban areas, 10% in the suburbs and just 6% in rural areas.
• The 2% of respondents who said they had personally protested is too small a sample from which to draw many conclusions.

The big picture: The findings in the latest installment of our weekly national survey come as states move forward with phased reopening of their economies and as more Americans are getting braver about testing how they rejoin society.

• Setting the protests aside, just 21% of respondents say they're very concerned about a coronavirus outbreak, the lowest share since mid-March. And only 13% say they're very concerned about the possibility of getting sick.
• Nearly one in four went out to eat last week, the highest percentage since mid-March.
• And 43% see a large or moderate risk in going to grocery store, down from a high of 70% in mid-April.

What they're saying: The protests may actually be having the effect of driving down people's perceptions of other risks, said pollster Chris Jackson, senior vice president for Ipsos Public Affairs: "If something else comes along that's so much more exceptional, people will gravitate to that and that will reset how they evaluate everything else."

• Some perceptions of risk around returning to more mundane activities such as shopping and dining may be easing as people test the waters and don't immediately get sick.
• But the searing images of thousands in the streets with masks on also may be making other activities seem less risky by comparison — and that could give people a false sense of security about a return to normal.

Methodology: This Axios/Ipsos Poll was conducted June 5-8 by Ipsos' KnowledgePanel®. This poll is based on a nationally representative probability sample of 1,006 general population adults age 18 or older.

• The margin of sampling error is ±3.3 percentage points at the 95% confidence level, for results based on the entire sample of adults.