Updated February 17, 2021

The full impact of COVID-19 on the New York-New Jersey-Connecticut metropolitan region will not be known for months, if not years, but the scope of the pandemic and who is most impacted is becoming clear. The crisis is not only stretching the response capabilities of our local, state, and federal governments, but is further exposing gaping inequities in our society. COVID-19 and threats from future pandemics also raise questions about the benefits associated with urban density, the underlying reason that New York and other large metropolitan areas have attracted people and businesses for centuries.

We should not give up on density as a means to sustainably expand economic opportunity and a high quality of life in a diverse society. Rather, we need to assess how the benefits from density can best be reconciled with its risks during a pandemic.

To better understand the implications of this crisis, Regional Plan Association has been collecting various data repositories that are keeping track of the pandemic. This includes mapping the daily number of COVID cases and deaths and overlaying these on maps of population and other characteristics. As more information becomes available, we will use this data to help identify correlations between COVID-19 morbidity and social, built environment, and policy response factors. Below, we explore how cases have grown in different parts of the region and where those who are most vulnerable to debilitating health or economic effects live in the region.

Preliminary Insights on Correlations with COVID-19

After integrating social variables with the reported number of COVID-19 deaths, some patterns are beginning to emerge within our region (see merged dataset and technical notes here).

Given that the curve is still evolving within the Tri-State area and throughout the country, it is difficult to make any hard conclusions or provide concrete insights. Also given the high numbers of deaths in Queens, Brooklyn, and the Bronx, the correlations are highly influenced by NYC factors. In spite of this, some patterns are starting to become clear. Some of the emerging correlations within our region are confirming anecdotal evidence and journalistic reporting. Up until now the strongest emerging correlations between the number of COVID-19 deaths with social and built environment factors are the following:

• Deaths by County and Percent of Persons of Color (all persons except white, non-Hispanic)
• Deaths by County and Percent of persons (age 5+) who speak English ​“less than well”
• Deaths by County and Percent of Persons Below Poverty Line (also inverse correlation: number of deaths decreases as per capita income increases)
• Deaths by County and Percent of Crowded Occupied Housing units (more people than rooms)

To date, however, an expected correlation between COVID-19 and age or group quarters is not shown as having a positive correlation:

• Deaths by County and Percent of Senior Population
• Deaths by County and Percent of Population living in Group Quarters

The lack of correlation is likely explained by the relative homogeneity across counties when measured by these two factors. This does not indicate that such populations are less vulnerable. As more detailed COVID-19 reporting is confirming, age in particular is correlated with deaths. Instead, this highlights the need for more site specific and detailed data.

Who is vulnerable in a health crisis?

Two types of vulnerability are important to consider. Some groups may be more vulnerable to becoming infected with the virus or suffer more serious health consequences. But social vulnerability also refers to those who are more likely to suffer from the secondary impacts of isolation, lost income, homelessness and other consequences even if they don’t become sick themselves. There is often an overlap between populations that experience these different forms of vulnerability.

The Centers for Disease Control and Prevention Social Vulnerability Index (CDC SVI) is the most comprehensive national dataset for identifying populations that are most vulnerable to the social and economic impacts of disease and other disruptions. The CDC SVI was developed prior to the COVID-19 outbreak, and while not intended to be used as a predictive model, there is growing evidence suggesting significant correlations between the social factors used by the SVI with confirmed numbers of COVID-19 cases and deaths. Here at RPA, we’re also integrating additional variables to the CDC SVI dataset.

The numbers from these datasets show that the tri-state region is not only the national epicenter of the pandemic, it also has millions of socially and economic vulnerable residents:

• Six million residents, mostly in poorer urban and rural communities, live in census tracts that rank high on the CDC SVI.
• Two million residents, the majority of whom are not poor, have no health insurance.
• About 30% of workers in the region have jobs that require a high level of physical proximity to others, including many essential health care, food, delivery and transportation workers who live paycheck to paycheck.

National Insights

Most counties at the national level are in earlier phases of ​“the curve”, so it’s difficult, if not impossible to find these types of associations for the country as a whole. But as COVID-19 inevitably continues to advance, more clear patterns will begin to emerge at the national level.

It’s important to emphasize that these insights are all preliminary. We will need to wait for the situation to evolve further, and obtain more complete and granular data. In the meantime we are planning to incorporate additional variables to the CDC SVI, tracking data repositories, and continue mapping relevant data.

RPA will continue to analyze this data to help answer questions presented by the current crisis. Advancing this type of research could help identify insights and valuable lessons for creating a more resilient region, one that is better equipped to confront future pandemics over the long term.

Tracking COVID-19

About half of the confirmed cases and deaths from the entire country are being reported within the Tri-State Region: almost 196,000 confirmed cases and 7,600 deaths as of April 8th.

For the region as a whole, there are tentative signs that the number of daily new confirmed cases is stabilizing. But more time needs to pass for this trend to become more evident. Within the subregions, New York City stands out with the highest share of accumulated cases and deaths. The curves for Long Island, Hudson Valley, and Northern New Jersey are somewhat comparable, but substantially smaller when compared to that of NYC. The numbers in Southwestern Connecticut have remained much lower when compared to other subregions within the Tri-State area.

It is important to mention that this data comes with a lot of caveats. Given that the testing ability has been limited, confirmed cases in some communities might be overrepresented, while in some others actual cases might be higher than what is being documented. This also extrapolates to deaths associated with COVID-19, with many official counts not including fatalities occurring in homes.

In order to have a more complete picture, we will need to wait for the COVID-19 curve to evolve further, both within our region and beyond. We will also need more complete and granular COVID-19 data. This type of data should also include demographic and socioeconomic variables.