Monitoring Population Movement as States Reopen

The population map is a result of a collaboration between Facebook Data for Good and COVID-19 Mobility Data Network, of which Direct Relief is a founding member.

With about half of states lifting their shelter-in-place orders and reopening to varying degrees, many Americans are stepping out of quarantine for the first time in weeks. Restaurants are serving in-house diners, storefronts are opening their doors to shoppers, and people are going back to their workplaces, creating scenes of pre-pandemic life and stirring hopes for a return to ‘normal.’ But without a vaccine, the threat of transmitting Covid-19 still remains, despite policy cues otherwise.

“This is a very, very dangerous time to be changing these policies” says Direct Relief’s Andrew Schroeder, who has been using anonymized data to track how people are moving during the pandemic. While the curve has flattened nationally, it has “plateaued at a very high level,” and rates of infection continue to climb in previously low-risk areas.

Meanwhile, people have started moving about at significantly higher rates, even in states where restrictions have yet to be lifted, like California. According to Schroeder, this uptick in mobility could have dire consequences.” We’re seeing this high rate of correlation rate between the mobility rate and the death rate,” he says.

In this episode of the podcast, we speak with Schroeder about how people are moving during this new phase of reopening and what it means for the projected course of the Covid-19 pandemic.

Click here to explore the Covid-19 Mobility Data Network dashboard.


Transcript:

ANDREW SCHROEDER: Americans are for sure moving around a lot more and that’s true in states that have sort of officially reopened, where there’s a directive from the governor or something that has said that they’re reopened.

It’s also true in places that have not officially reopened, like California, which is signaling that they’re going to be relaxing some of the restrictions, but officially speaking, the orders remain on the books. Even there, you’ve seen actually pretty significant increases in the degree to which people are moving around and especially at certain kinds of sites.

Parks and recreation areas in particular, in a place like California.

AMARICA RAFANELLI: So in those states that have officially lifted their shelter-in-place orders, what kinds of trends have you observed?

SCHROEDER: Well there are the trends that you can see right now, in terms of the last week, that areas that are more rural, so meaning they are low-density population, areas which are hard to work from home, and did not actually see nearly the same degree of decline in mobility before restrictions were being lifted — those places have in a number of areas already returned to baseline. And baseline for us is considered to be February. So that’s at the level of the degree to which people are moving around in public on any given day.

Even if you’ve seen movement restrictions lifted, it doesn’t necessarily mean that places have come back to baseline. Georgia’s a good example. Atlanta, Georgia –even though Georgia itself has changed the policies around what you’re allowed to do people have not necessarily come back to suddenly pack into restaurants or movie theaters or even walk around on the streets at a level that was comparable to say, February. So that would mean that it’s uneven. People are using their judgment and just the issuing of the order doesn’t completely transform the situation.

RAFANELLI: Are there any similarities between the states that have lifted their shelter-in-place orders? What kinds of factors might influence a state to ease Covid-related restrictions sooner than others?

SCHROEDER: Well, to be quite honest, I just think that lifting of the restrictions is a bad idea, at this point, based on public health. So I would say that most of the places that are moving towards rapid repeal of these policies are making bad decisions, just to be quite blunt.

I would say that there are a number of factors that are leading people to make bad decisions in these cases.

Economics appears to be the leading, explicit cause. So, there’s no question that, shelter-in-place orders of all kinds have had an enormous impact on the economics of people throughout the country. Even though the stock market’s pretty high still at 24,000, the unemployment rate is through the roof and continuing to rise. And there’s a lot of pressure on states to allow for a business to resume so that you can relieve the economic pressure.

I would say there’s also a lot of political pressure that has been mounting opposition to what is seen as a fairly restrictive set of government focused policies.

And that has added up towards a rapid move towards repeal of these movement restrictions. The fact of the matter is that the United States isn’t close to having Covid-19 under control — so the curve is flattened nationally in terms of cases, but it has plateaued at a very high level. And it looks as though there is no point on the horizon at which you can foresee under any of the epidemiological models, a significant reduction in the death rate. In fact, quite the opposite, if you take New York City out of the data. New York City has turned a corner in a number of respects. It was one of the earliest and clearly the largest epicenter and so it has an outsized impact on the overall data. New York has been on a relatively reasonable decline. Most other places though are on a pretty steep ascent and the geography of the virus is expanding to include a large number of places that were not really experiencing significant outbreaks, just say a few weeks ago. So, that means that this is a very, very dangerous time to be changing these policies.

We’re still pretty far out from a vaccine. We’re still pretty far out from anything like a reasonable treatment. And those were the things that flattening the curve was intended to buy time in order to have available as tools, given that we don’t have them.

It’s still the case that the only real tool we have is social distancing. That’s what makes the current moment so dangerous.

RAFANELLI: Are there any States or localities that have experienced a relative surge in cases and started lifting restrictions?

SCHROEDER: Yeah, for sure. Georgia is a really good example.

Actually, Georgia has seen a significant increase in cases over the last week. They’re averaging over 900 cases a day. That seems to be gradually going up as the mobility rate has increased in Georgia, and they were one of the first to sort of roll back government level policies.

So yeah, there’s a number of factors that have played out, you know, that are kind of similar to that around the country. We’ve heard a lot about the rise of the caseload in the plain States. South Dakota is probably the most famous one of these, uh, where the governor made a pretty big point of saying that the distancing restrictions were largely a matter of personal choice.

And South Dakota, relative to population density, has had one of the most severe outbreaks in the country, outside of Sioux falls, largely associated with large meat processing plants, and then the communities of people that are the families and coworkers of those that work in these large meat processing plants.

That has been a trend that’s kind of echoed across many other parts of the country.

RAFANELLI: So, with that said, is mobility data being used by policymakers to inform their decisions about when and if to lift social distancing restrictions?

SCHROEDER: I think the answer is yes, it is by some, but not necessarily by all.

I mean I don’t have visibility to everyone, so I can only comment on what I can comment on. But I would say that through the mobility data network that we’ve been working on, which has been focused in a few states, like in California, New York, Massachusetts, Illinois, and Michigan for instance, there has been a lot of interest in using day-to-day mobility data.

So, looking at intersections between mobility, data and race and ethnicity has been one of the big questions that I think has been raised by a lot of this information lately, where we found that in areas where you have demographics that tilt towards large numbers of African Americans and Latinos you have had significant challenges with reducing rates of mobility, not just now as a result of lifting restrictions, but kind of all along. And that aligns with job structure. African Americans and Latinos by and large tend to work in what are now being described in the press as quote, unquote, essential occupations, which are people that work in services, food, they work in meatpacking, in kind of middle and lower tiers of healthcare, in a lot of occupations, like warehousing and logistics, that you just have to show up for work. You can’t do them remotely.

And so you’ve had much higher rates of mobility for weeks now and that has really caught up in terms of rates of infection and then rates of hospitalization, acute complications and deaths. And that’s, I think, one of the trends that’s really not going away.

RAFANELLI: What can, you know, how people move today tell us about the projected course of this pandemic?

SCHROEDER: You know, I think the main thing to bear in mind is that the rate of mobility in, at least in the focus areas around cities so far, where some of the early epidemiological studies have been done, appears to align very well with the death rate. Meaning that, as you change the mobility rate up and down in different areas over time, this factor aligns better than any other factor with the rate at which people are dying from Covid-19.

The reason that it’s important to focus on the death rate is that the death rate doesn’t have the same kind of problem with the denominator or the total number of people that are infected as the case rate does. So the case rate tells us how many people have expressed and been diagnosed. It doesn’t tell us how many people have the disease or how many people have actually been exposed to the disease. But the death rate is much more of a solid, stable number. And in that case, we can tell that the most likely predictor of where the death rate is going is likely to be the rate of mobility.

As you increase rates of mobility, particularly in areas that have large amounts of indoor space — there’s something really significant about, at least there appears to be so far, not just concentrations of people and not even just movement — although movement is very significant, but movement in areas with large amounts of indoor space tend to be much more at risk than others. Leading towards things like high rates of death.

I think that’s something that the public health community has been saying for some time now, actually, in different forms. I don’t actually think that’s a new message.

I think it’s something that there’s just a lot more evidence for right now because we’ve seen of different changes in waves of the virus and there has been a lot of data collected — a lot of data sources have come online — so we have a lot of evidence that we can point to, but the message is not fundamentally different. So, we’re seeing this high rate of correlation between the death rate and the mobility rate.

Just as a side note to that, it does not mean that if you leave your house, you’re going to die of Covid-19. That’s just not the case. But it does mean that, on average, the best predictor of the social death rate is whether or not your area is seeing high rates of mobility or not.

So we really need to get people to stick with a lot of these policies, even as we may relax some specific pieces in order to test out how to bring society back online, not in a blanket way, but in a very targeted way.

Transcript has been edited for clarity.

Exit mobile version