Since the Commonwealth’s first reported cases in early March, COVID-19 has quickly and dramatically affected Massachusetts in unprecedented ways. Crises tend to shine a light on existing inequalities in society, and the COVID-19 outbreak is certainly no exception. Overall, COVID-19 has disproportionately affected the most vulnerable populations in the Commonwealth. While the general population is aware that COVID-19 poses particular risks for the elderly and people with pre-existing conditions, the virus has also revealed by its impacts several elements of racial and socioeconomic inequality in Massachusetts. The entire economy has been decimated by COVID-19, but the effects have been acutely felt by workers in lower-wage sectors such as hospitality and food service. This has led to elevated unemployment rates for workers with lower levels of education and people of color, as well as the associated consequences on housing cost burdens, food insecurity and other household challenges.
Current municipal data from the Massachusetts Department of Public Health (DPH) reveal a clear connection between place and the COVID-19 outbreak. In particular, data on COVID-19 case rates show significant concentrations in urban areas around the state, particularly Gateway Cities. While not revealing something “new” about inequality in the state and region, the COVID-19 outbreak underscores the critical importance of place, and therefore housing affordability and accessibility, as a social determinant of health and a driving factor in quality of life and upward mobility.
The following brief outlines the history of place in Massachusetts and its role in quality of life, social mobility and public health, and highlights how those issues are currently presenting themselves during the COVID-19 crisis.
As the 2019 Greater Boston Housing Report Card showed, the state and region have a long legacy of residential segregation. Dating back to before the civil rights era, many housing policies directly and indirectly prohibited residents of color from purchasing homes or land in majority white communities. Federal and local laws, as well as mortgage lending practices such as redlining, racial covenants and blockbusting, effectively limited the upward mobility of non-white families and hastened both racial segregation and urban decay, and drastically curtailed the ability of Black and Brown families to accumulate generational wealth.
Since the 1990s, the country has inched toward improvements in residential integration. In Massachusetts and Greater Boston, dramatic growth in foreign-born communities has led to increases in diversity around the state.
Despite this increase in diversity, Greater Boston continues to be highly segregated, driven, in part, by high housing prices, inadequate supply and exclusionary zoning practices in a number of communities limiting multi-family production.
Figure 2 measures segregation over time, using the dissimilarity index as a measure of “evenness.” Values range from 0 (complete integration) to 100 (complete segregation) with the value indicating the percentage of the racial/ethnic group that would need to move to be equally distributed across the region. A value of 60 or greater is generally considered indicative of a very high level of segregation.
Despite marginal improvements in integration over time, the Greater Boston region remains one of the most segregated metropolitan areas in the nation. The American Communities Project at Brown University ranked 384 metropolitan areas based on their indices of dissimilarity. Greater Boston was the 17th highest on the dissimilarity index among the Black population, and the 7th highest among the Latinx population.Figure 2
Access to opportunity and other elements of social mobility and quality of life are deeply connected to where a person lives. The maps below present two “Opportunity Indices” from the United States Department of Housing and Urban Development (HUD) and demonstrate the connection between place and quality of life indicators in Massachusetts.
While racial and ethnic segregation have decreased a small degree in Massachusetts in recent years, Black and Latinx households continue to be more likely to live in neighborhoods with greater exposure to poverty, higher incidence of unemployment, lower levels of educational attainment, less labor market engagement, and poorer air quality than neighborhoods where white households are more likely to reside.
As part of UMDI’s work for the Donahue Data Dash, the Boston Indicators COVID Community Data Lab and the Greater Boston Housing Report Card, the research team is examining how the outbreak of COVID-19 is disproportionately impacting different communities around the Commonwealth. While both the available data and the conditions surrounding COVID-19 are evolving, UMDI’s analysis focused specifically on the characteristics of municipalities and the known incidence of COVID-19. For this brief, our analysis focuses on the concentration of race and ethnicity, crowded housing and frontline workers in a community and the known positive testing rate of COVID-19.
As the Massachusetts Department of Public Health (DPH) made various elements of infection and testing data available at the municipal level, it was striking to see the collection of communities in Massachusetts with the apparent highest instances of COVID-19. To examine the ways different populations are affected by the coronavirus, the UMass Donahue Institute (UMDI) examined the rates of COVID-19 infections (confirmed positive tests) per 10,000 people for the 100 largest cities and towns in Massachusetts. UMDI then compared the rates of COVID-19 cases by town against various demographic and socioeconomic indicators, including the percentage of people of color, crowded housing, frontline workers, population density, average household size, per capita income and poverty rate. Those demographic and socioeconomic data were drawn from the U.S. Census American Community Survey five-year estimates for 2018.
The data show elevated known infection rates in urban areas, particularly those with crowded housing, larger communities of color and immigrants, and concentrations of low-income residents. Many of these lower-income communities of color, for instance, have higher shares of frontline workers (e.g., nurses and grocery store clerks) who risk infection every day by going to jobs providing essential services.
Individuals from particular socioeconomic groups are not biologically more likely to succumb to the virus, but through a range of “social determinants of health,” physical health intersects with the broader social context in which people live. In this way, COVID-19’s impact in Massachusetts underscores a number of existing demographic, geographic and socioeconomic inequalities in the state.
The correlation between communities with a high concentration of people of color and the confirmed COVID-19 case rate is striking and clear (Figure 5). Chelsea and Lawrence have the highest percentage of people of color (Chelsea: 79% and Lawrence: 86%) and respectively the highest COVID-19 case rates in the state (736 per 10,000 in Chelsea, 452 in Lawrence). Brockton, Randolph, Revere, Everett, Lowell, Worcester, Holyoke and Lynn, which comprise the rest of the top ten municipalities by COVID-19 infection rate, are all municipalities with large populations of color, and all but Randolph are Gateway Cities.
While the city of Boston’s infection rate ranks 19th out of the 100 largest municipalities in the state, conditions vary considerably across neighborhoods. If Boston neighborhoods were included separately in our COVID-19 infection rankings, East Boston would rank 6th, Hyde Park 8th, Dorchester 11th and Mattapan 14th. Each of these four neighborhoods are majority people of color. Conversely, South Boston, Charlestown and Back Bay/Downtown (all majority white) would rank 63th, 73th and 86th, respectively.
The most indicative measure of COVID-19 spread that UMDI observed is the rate of crowded housing. Crowded housing is defined by HUD as households with more than one occupant per room, (i.e., where people outnumber rooms). Cities with the highest crowded housing rates, including Chelsea, Lynn, Lawrence, Everett and Revere, also have the highest rates of COVID-19 infection. Crowded housing was more indicative of COVID-19 spread than a city’s population density alone. While population density measures the number of people for each square mile, the rate of crowded households measures the number of households with 1.01 or more occupants per room as a percentage of total household units.
By way of example, Somerville and Cambridge have the two highest rates of population density in the state but lower rates of COVID-19 infections than Chelsea, Lawrence and Everett, or other cities with higher rates of crowded households. Nearly 10 percent of Chelsea residents live in homes where people outnumber rooms (compared to 2 percent in Somerville and Cambridge), and nearly 3,000 people have tested positive in the city. Chelsea’s rate of COVID-19 infection is 736 cases per 10,000 people, the highest in the state.
UMDI found that municipalities with high COVID-19 rates also tend to have high shares of people working in frontline positions. Frontline workers are defined as employees in the health professions, grocery stores, transportation, protective services and other occupations allowed to continue to work after Governor Baker ordered non-essential businesses to close. While this correlation is less strong than the indicators for overcrowding and share of people of color, it shows that there is a relationship between frontline workers and COVID-19 infection rates.
First and foremost, COVID-19 is a public health issue. Infectious disease experts and epidemiologists have reported extensively on the infection patterns, risks and mitigating activities to help slow the spread of the virus (e.g., social distancing, wearing masks, etc.). As a result, public policy has focused on “flattening the curve” and reducing the spread of the disease in the aggregate, particularly as it relates to protecting elderly individuals and those with preexisting conditions, or populations most at risk of serious health complications and death related to COVID-19.
In terms of housing, the COVID-19 crisis reveals a few key issues in our state and region:
Our analysis shows that “crowded” housing (or the percent of housing units with more than one person living per room) is much more closely associated with elevated COVID-19 rates in a community than simply density (or the number of people per square mile). Crowding is a direct outcome of the cost of housing in Massachusetts, and more specifically Greater Boston, and the constraints facing families in finding household situations that they can afford.
As COVID-19 is a highly contagious infectious disease, it is not surprising to see outbreaks concentrating in urban areas, both in terms of raw counts and infection rates. There is a growing discourse around COVID-19, questioning what the virus and the fear of infectious disease may mean for density and urban living in the future. While proximity to other people is integral to urban living and is a risk factor in the spread COVID-19, the reality is that financial decisions that lead to crowded housing appear to be a larger driving factor in the COVID-19 outbreak than mere population density.
The 2019 Greater Boston Housing Report Card focused a great deal of attention on the challenges of affordability and supply in the region. The report concluded that insufficient housing supply, lack of housing affordability, and inequity in access to housing are the overriding challenges facing the region’s residential real estate market—and have been for decades. The COVID-19 crisis underscores these issues clearly. In short, few people want to live in “crowded housing.” Choosing to “double up” in bedrooms or live in multi-generational households is generally an economic decision for households. The connection between crowded housing and COVID-19 underscores the need for increased housing production and more affordable housing options, even in dense urban neighborhoods.
Residential segregation by race, ethnicity and income did not happen by accident. It arose from discriminatory practices in which the private housing industry and federal, state and local governments were active participants. Examples of current conditions that perpetuate racial segregation include discrimination in the housing, mortgage and insurance markets; exclusionary zoning; government policies affecting the location of, and access to, subsidized housing; disparities in private and public investments; and displacement of residents due to economic pressures. Some of these practices, particularly around restrictive zoning, limit housing variety and affordability in a number of communities.
Access to opportunity, housing and social and economic mobility is deeply connected to place and drives social determinants of health, or conditions in which people are born, grow, live, work and age that impact health outcomes, such as socioeconomic status, education, neighborhood and physical environment, employment, social support networks, and access to health care. The current analysis considers demographic and socioeconomic factors in Massachusetts communities as “social determinants of health” related to COVID-19. In short, there is a clear correlation across communities in Massachusetts between COVID-19 case rates and socioeconomic or demographic factors, particularly around the percent of people of color living in a community, the percent of households living in crowded housing, and the proportion of frontline workers living in a community.
Despite the clear correlations between place and health outcomes, these data are imperfect. Testing statewide remains unreliable, and reporting on race and ethnicity remains inconsistent. The current data focus specifically on the positive test results for COVID-19 in a community, but limited testing implies that there are likely more undetected cases.
COVID-19 is a crisis unlike any other we have experienced in our lifetimes. While there is a longing to leave its disruptions behind, for many communities, a return to “normal” would mean a return to the inequities in housing affordability and accessibility that have plagued our state and region for decades. As the state recovers from the current crisis, it will be important for public policy makers to seriously consider how the place where people live directly impacts quality of life and social mobility, housing equity must be at the forefront of our recovery efforts in the long term across the state and region.
 As of April 2020, the unemployment rate for individuals with a college degree was 9.2%, 23.5% for individuals with some college, 25.1% with a high school diploma, and 39.4% less than high school.
 As of April 2020, the unemployment rate for non-Latinx white people was 13.7%, compared with 25.4% for people of color.
 Racially restrictive covenants were contractual obligations in property deeds, typically created by private developers preventing occupancy by or real estate sales to non-white residents. See How Prevalent Were Racially Restrictive Covenants in 20th Century Philadelphia? Federal Reserve Bank of Philadelphia.
Redlining refers to the discriminatory practice of lending institutions in denying home loans for individuals and residences located within specific geographical areas of a city which the lender has deemed “high risk areas,” largely done on the basis of race.
Blockbusting is the practice of real estate brokers convincing homeowners to sell their houses for low prices by deliberately leveraging fear that a neighborhood’s socioeconomic demographics are changing and will decrease home values.
 Edozie, Rita Kiki, Barbara Lewis, Shauna Lo, Trevor Mattos, Mark Melnik, Lorna Rivera, Luc Schuster, Paul Watanabe, Gail Waterhouse, and J. Cedric Woods. Changing Faces of Greater Boston: A Report from Boston Indicators, The Boston Foundation, UMass Boston and the UMass Donahue Institute. May 8, 2019. https://www.bostonindicators.org/reports/ report-website-pages/changing-faces-of-greater-boston
 To date, the outbreak of COVID-19 is still imprecisely understood. Inconsistent availability of testing makes it difficult to say the true outbreak within communities. DPH data speak specifically to the positive testing in communities and, now, the amount of testing in communities. Obviously, researchers cannot know about COVID-19 incidence in individuals who are not tested for whatever reason.
 Holyoke’s high COVID-19 case rate can be partially attributed to the outbreak at the Holyoke Soldiers’ Home—accounting for 160, or 17.8% of all confirmed positive tests in the city.
 The following analysis is made specifically from a social science perspective and not from an infectious disease perspective. The analytical team are not infectious disease experts, but rather social scientists highlighting the ways in which demographic and economic characteristics interact with public health indicators.
 There is a substantial body of literature that details the history of residential segregation in the United States and the roles played by the real estate and homebuilding industries; lending and insurance institutions; the federal, state and local governments; and others, most notably Richard Rothstein’s 2017 book, The Color of Law.