Since the advent of Covid-19, there has emerged a critique of urban density as deadly in fighting the current and future pandemics, and that the answer lies in dispersed development, otherwise known as sprawl. But a new study on density related to coronavirus infection and mortality rates released by the Johns Hopkins University School of Public Health challenges much of this “conventional wisdom.” The study was led by Shima Hamidi, Bloomberg Assistant Professor of American Health in Environmental Challenges and a planner, urban designer, and Smart Growth advocate. Her research examined data from more than 900 metropolitan counties across the U.S. and concluded that there is no correlation between density and infection rates, but higher death rates in places described as “sprawl.” I sat down with Hamidi to discuss the findings of her research and the implications for community and urban design.
MJC: Michael J. Crosbie
SH: Shima Hamidi
What prompted you to take on the question of density and Covid infection and death rates?
As an urban planner and designer, I know the profession has advocated for the development of dense, mixed-use, walkable, transit-oriented community design—the opposite of sprawl. There are many studies over the past decades on how compact development can improve the quality of life: from economic aspects, to environmental sustainability, to public health. The professional consensus in favor of compact development has been challenged by the emergence of Covid-19, reflected in discussion among scholars and practitioners, and in media outlets, about a fear of density. We noticed that these media reports and professional discussions all have one thing in common: they aren’t based on hard evidence, but instead on generalized observations about the number of infections and deaths primarily in one city—New York—and then extrapolated to the entire U.S. The harm is that such conclusions could change current community design and practice. We wanted to take a step in contributing to evidence-based conversations. So, instead of just looking at New York City, we studied data from 913 counties located in metropolitan areas across the country to come to more generalized conclusions.
After an analysis of the data, you found that there is no significant relationship between density and coronavirus infection rates. Did this surprise you?
Density or sprawl is not a risk factor in terms of the infection rate. But we did find that sprawling areas have a significantly higher death rate. Two concepts that are confounded make this finding unexpected. Often in discussions about the pandemic, when people talk about density they’re actually talking about crowding or overcrowding. They’re not the same. Crowding is defined as a large number of people gathered closely together, such as in bars, restaurants, sports events, concerts, on beaches. Crowding takes place both in sprawling and dense areas. Crowded places are the source of transmission of the disease, not the density of a city or county. This fact is not articulated in discussions about density. For example, in his April 13 coronavirus press conference, New York Governor Andrew Cuomo stated: “It’s very simple, it’s about density. It’s about the number of people in a small geographic location, allowing that virus to spread. Dense environments are its feeding grounds.”
But he’s actually talking about crowding, not density. Our research wanted to distinguish between the two and control for them. More confusion arises from a focus on the number of people infected, not the per-capita infection rate—a much more informative measure. Our study controls for crowding versus density, per-capita infection rates, as well as socio/demographic/economic profile, racial disparities, healthcare infrastructure, and other factors. And we found that density is unrelated to the Covid-19 infection rate. This was unexpected. Our initial hypothesis was that we would find a relationship between infection rate and density, but we didn’t.
Why do you think this is the case?
One possibility is that density plays two conflicting roles, which could cancel each other out. On the one hand, density increases the number of contacts between people and the likelihood of transmission of the disease in crowded places. However, we also know that people in compact areas are more equipped to follow social-distancing advisories due to better access to online grocery shopping and home delivery services, for example, than those who live in sprawling areas. People in dense areas are also more likely to voluntarily follow social-distancing practices and are more cautious of crowded places than their counterparts in suburban and exurban areas. So we think that these factors might cancel each other out.
You also found that metropolitan areas with large populations and a high number of “tightly connected” counties had higher infection rates. Describe what this “internal connectivity” is between counties, and how it might factor into infection rates.
Two other concepts often confounded are density and size. We define the size of a metropolitan area as population. The U.S. Census Bureau defines a metro area as a collection of counties that have social, economic, and commuting relationships with each other. They are clustered together in the metro area, sharing significant transportation, movement, and commuting relationships—this is what we describe as “internal connectivity.” Pandemics spread with the movement and interaction of people, which are more frequent in large populations. This is not related to density. The disease might start at the urban core, then spread to suburban, exurban, and rural areas, places that are not as well equipped with healthcare infrastructure as the core. This results in higher death rates in these sprawling areas.
Your study considered a lot of factors: population size, education levels, age, race, and healthcare infrastructure, among others. Were there factors that seemed to have a higher impact on infection rates and death rates?
Three were most significant: race and age disparity, education, healthcare infrastructure. We found a higher correlation between percent of Black population and infection and mortality rates, likely because of underlying health conditions, less access to workplace policies that allow work at home and social distancing. The proportion of senior population was very significant. In more than 23 states, more than half the deaths are in nursing homes. We found that it was one of the most significant predictors of the mortality rate. In terms of education, workers with college degrees rely less on public transportation, are more able to work remotely, and have a better understanding of the virus and social-distancing policies. For healthcare infrastructure, in terms of ICU bed rates, we found that death rates decline with a higher number of ICU beds so that healthcare facilities are not overwhelmed.
Your study considered the “activity density” of a place. What is “activity density,” and what impact does it have?
The level of activity in an area is driven by population and level of employment. We wanted to control for both of them in the model, so we used a formula: population plus employment divided by land area. If you have higher population or employment opportunity, or both, these can generate activity. We found no relationship between activity density to infection rates, and in fact saw lower death rates in higher activity density, basically because healthcare infrastructure is better. This was probably the most surprising finding of the study.
The study’s conclusions appear counterintuitive, especially to many city dwellers concerned about remaining in metropolitan areas. How would you advise a city dweller about whether to move to a less-dense place or not?
A national survey in May that found that 27 percent of adults, and 43 percent of Millennials, are thinking of moving to the suburbs. Businesses are already moving to suburb offices. A report from Commercial Real Estate Services found in the first quarter of 2020, nine out of ten of the largest office markets in the U.S. reported increased vacancy rates in downtowns. Our study suggests that dynamic relationships and density are very complex—there are so many players and factors that change results and expectations. We hope the study’s findings offer a new perspective and pause trends we are observing of movements to suburban and exurban areas. Such trends are often propelled by media outlets. I’ve seen a study that refers to density as a “risk factor” or “enemy” of people. Another described sprawl as a “secret weapon” against Covid-19. None of these studies are based on any real evidence.
What does your study mean for planners and architects trying to incorporate what we are learning about the pandemic into concrete planning and design decisions?
Our findings suggest that architects’, planners’, and urban designers’ roles in addressing the pandemic crisis are not in changing the paradigm of compact community design, since we found no evidence that sprawling counties are more immune to the disease than compact ones. In fact, we found the opposite: pandemics are deadlier in low-density and isolated areas, which have less access to quality healthcare infrastructure. Design professionals should continue to practice and advocate for compact and dense development rather than sprawl. But also be cautious about larger metropolitan areas (in terms of population) that could be more vulnerable to pandemic outbreaks than smaller metro areas. In terms of community design, density is unrelated to infection rates, and inversely related to death rates, which is an important and profound finding for community design, transportation expenditure, tax policy, congestion pricing, affordable housing, and nearly every other key issue that is important to planners and architects.