The COVID States Project has been probing American behavior (and attitudes) during the pandemic since March 2020. In 12 survey waves, an array of researchers — from Northeastern, Northwestern, Harvard and Rutgers — have polled some 185,200 Americans about subjects ranging from social distancing practices to attitudes toward vaccines and judgments about state politicians’ leadership. The consortium’s 55th report, issued this month, looked at social isolation — something many of us have experienced over the past year and a half. The focus was on the most isolated people: those who said they had one or zero people to turn to for help in various kinds of crises. One of the more striking findings, as David Lazer, a professor of political science and computer science at Northeastern, explained in a recent interview, is that even as vaccines have arrived and society has opened up, some measures of loneliness have not abated. The interview has been edited for length and clarity.

Q: You have four measures of social isolation. People are asked to say how many people they could turn to for help if they fell ill; how many people might lend them money; how many people they could talk to if they had a personal problem or felt sad; and how many people could help them if they needed a job. If they answer “one” or “none” to those questions, they are considered socially isolated. When you average those four measures, roughly 37 percent of respondents were socially isolated in late April 2020. That’s a strikingly high percentage. Do you have a figure from before the pandemic, to put that number into context?

A: If I were to give my biggest regret, it’s that I wish we had done a survey in January — but what can you do? We weren’t planning on there being a pandemic in March. We can compare dynamics during the pandemic, but we don’t have a great baseline from before it.

Q: Let’s focus on that composite figure for a minute: Roughly 37 percent of people report that they have zero or one person to turn to — on average, across the four measures — in late April 2020. But then something counterintuitive happens. The figure falls to about 33 percent by late May. That’s at the height of the social shutdowns, yet the proportion of socially isolated people was dropping.

A: I need to be properly modest here about some of the patterns we see and can only offer speculation in some cases. There may have been a certain amount of solidarity in that particular, extreme moment, early in the pandemic. Then we sort of entered the long slog of social isolation, which had a cumulative effect on our relationships with one another.

Q: In general, the data shows isolation rising after that first burst of solidarity and peaking in September 2020. Then things generally get better after that. But the measure of one specific kind of isolation — emotional isolation, the question of whether you have someone to talk to about personal problems — just gets worse and worse.

A. Yes, the high point of social isolation for the emotional measure is in the survey from June that we just completed.

Q: What could possibly explain why people would feel lonelier over the past couple of months than they did a year ago?

A: Relationships are something that you have to cultivate. And so maybe there was someone that you would have felt comfortable calling up or reaching out to at the beginning of the pandemic, but then you haven’t spoken to them in six months — and do you feel as comfortable reaching out? The negative spin I can give these numbers is that those relationships may have decayed somewhat, because you need that experience of casually getting together. You don’t get together on Zoom just to chat and gossip for 20 minutes. But of course, in real life, we do that all the time. You bump into someone at work, you bump into someone in the street, and you get together with friends for dinner. Those serendipitous collisions sustain our connections, and without them relationships may have decayed, and the “reopening” we’ve experienced the last few months doesn’t immediately bring our relationships back to where they were. Indeed, the reopening may make more apparent what we’ve lost.  

 But I’ll also put a more positive spin on it. It could be that people are revisiting some relationships. We’ve recently seen increases in people quitting their jobs. You don’t want to quit a job at the height of a pandemic, but now maybe you think there are opportunities out there. The trends we’re seeing could involve people purposefully rewiring their relationships — which has a short-term cost. Maybe some of your old relationships have decayed, but you’re building new ones. And yet those new friends aren’t “old friends” you can confide in yet. Maybe, on the positive side, that’s the pivot that a lot of people are making.

Q: There has been a lot of discussion of the racial dimensions of the pandemic. Were Black Americans more likely to be isolated than White Americans?

A: It depends a little bit on the measure we’re looking at. It does look like on the emotional side, White Americans were less likely to be socially isolated than other racial groups. That comes out pretty clearly in the data. They were also less likely to be isolated in terms of having someone to care for them. But then, notably, they were more likely to be isolated in terms of the economic side. So White respondents were more likely generally to be socially isolated with respect to someone who would lend them money or might help find them a job.

 This could be because there are stronger norms of informal networks of social insurance among Hispanics and African Americans that involve, let’s say, lending and borrowing money or helping people find jobs. It could be that different groups may also have different norms about what one can ask for or request help for. We don’t really have a firm answer on that from just these data.

Q: What were some of the characteristics that protected people from social isolation?

A: Gender stands out. Men were more likely to be emotionally isolated, which fits with both stereotypes and literature, I’m afraid. And we found that religiosity really popped out as something that was protective: Our measure was whether people attended services “once a month or more” or “a few times a year or less.” Attending services is something that you could continue to do during the pandemic, even if it was on Zoom. It’s an interesting question whether we would have found the same result if we’d asked people if they were in a bowling league. We did not ask about other group activities. But it’s clear that religiosity, as measured by attendance at religious services, was strongly, strongly protective.

Q: Loneliness can be dangerous, this report suggests. This population you have homed in on is at significant risk of depression — especially people who are emotionally isolated. The data show that a third or more of the group that has zero or one friend to talk to about problems reported signs of moderate depression.

A. Yes, emotional isolation has by far the strongest connection to depression. We have to be careful there: We can’t say whether that’s causal or not because, you know, all these things come together as a package. Depressed people might be less likely to have friends, for example. But there is a very strong correlation. The rate of depression in this group fluctuates around 36 percent or more — about 50 percent higher than the rate for people who are not emotionally isolated.

Q: And for context, about a third of respondents reported being emotionally isolated, as of June.

A: That’s right. And, in addition, there are much higher levels of depressive symptoms relative to levels before covid. That combination of high depressive symptoms and high levels of emotional isolation is deeply worrying. The question is: How long do our social networks take to bounce back? There really aren’t good analogues in the research on social networks that provide guidance as to what is next.

Twitter: @cshea4