A note on framing. This is a descriptive analysis of public data, organized around housing cost and income, with race shown as one dimension that moves with them. It documents where communities sit and how they differ in health and school outcomes; it does not measure why. Reasonable observers read these patterns differently: some emphasize exclusionary zoning and a historical racial wealth gap, others emphasize income, individual choice, and market demand, and recent regional research finds racial segregation easing even as income segregation grows. Where causes are named here, they are attributed to the cited research, not asserted as fact. The analysis describes systems and outcomes, not the intentions or character of any community or group.
Each dot is one municipality. The horizontal axis shows median home value; the vertical axis shows the Black and Latino share of residents. Across all towns the straight-line correlation between the two is near zero (about minus 0.07), so home value alone is a weak predictor of racial composition. The distribution is uneven by quadrant: few municipalities are both high-cost and high in Black and Latino share, while lower-cost municipalities span a wide range. Hover for detail, or select a dot to load a full profile.
As a single municipality, Boston registers as one moderate point: approximately 39 percent Black and Latino at a median home value near $730,000. That average conceals a great deal. Disaggregated by ZIP code, the same city ranges from roughly 3 percent Black and Latino in the North End and 13 percent in Back Bay, where median home values approach or exceed $1.5 million, to about 75 percent in Roxbury and 89 percent in Mattapan, at close to half the price. The toggle on the chart below disaggregates Boston accordingly. Any municipal-level figure, in Boston and elsewhere, can obscure this kind of internal division.
Higher-cost municipalities tend to have lower Black and Latino shares. How much of that relationship reflects race versus income is debated, since the two are closely linked: the same higher-cost towns include sizable Asian populations, and regional studies report racial segregation easing while income segregation rises. Lower Black and Latino shares also occur in lower-cost areas, notably in much of western Massachusetts, which is largely white and lower-income. The pattern is most concentrated in the higher-cost municipalities west and south of Boston.
The pattern above is specific to Black and Latino residents and does not hold for residents of color as a whole. Asian residents are well represented in many of the highest-cost suburbs. Lexington, for example, is about 5 percent Black and Latino but 33 percent Asian, and 44 percent residents of color overall, among the higher shares in the state. Use the vertical-axis control on the chart to see this directly: switching to the Asian or all-residents-of-color view moves several high-cost suburbs to the top, where the Black and Latino view leaves them at the bottom. For that reason this analysis reports each group on its own terms rather than collapsing them into one figure. Folding Asian residents in with white residents would erase them from the calculus; folding them into a single people-of-color total would obscure a real and substantial difference in how housing cost relates to each group.
Search any of 348 Massachusetts cities and towns, or a Boston neighborhood by name. Each profile shows the racial breakdown, real health outcomes (CDC PLACES), who owns versus rents, and the income it takes to buy or rent there, set against what households actually earn.
Higher-cost suburbs and a set of mid-size cities differ markedly in both housing cost and racial composition. Many of the latter are designated Gateway Cities by MassINC. Select any community to load its profile above.
High-cost municipalities (home value above $750K), ordered from least to most diverse. Rank 1 is the least racially diverse.
Communities ordered by Black and Latino population share, highest first. Rank 1 is the most diverse. Many are MassINC Gateway Cities.
Housing cost is associated with a range of community outcomes. Municipalities that differ in cost and racial composition also differ, on average, in school resources, public investment, and life expectancy. The diagram below presents the cumulative cycle as housing researchers describe it; this explorer documents the cross-sectional differences, not the longitudinal mechanism.
Massachusetts is not a pure property tax state. Since the 1993 Education Reform Act, the Chapter 70 formula sets a "foundation budget" (the minimum the state considers adequate), measures how much each town can contribute from its own property wealth and income, and sends state aid to cover the gap. That genuinely equalizes the floor: lower wealth cities like Lynn or Springfield receive far more state aid per pupil than Newton does.
Chapter 70 sets a floor, not a ceiling. Higher-wealth towns raise more than the minimum from their property base and spend above the foundation amount on items such as smaller classes, enrichment, facilities, and support staff. One documented example:
The state closes the gap at the bottom; local property wealth opens a new one at the top. And money is only one input: family wealth, housing stability, and the racial wealth gap shape outcomes well beyond the school budget.
A 2023 Boston Public Health Commission report found a 23 year life expectancy gap between a census tract in Back Bay (91.6 years) and one near Nubian Square in Roxbury (68.8 years), about two miles apart (BPHC, Health of Boston). Subsequent commission data show the gap between Boston's Black residents and other groups has continued to widen. Measured health outcomes follow the same pattern:
Health figures are CDC PLACES model-based estimates. Public health research associates outcomes like these with social determinants of health, including housing, income, education, and food access; the Boston Foundation's Greater Boston Housing Report Card discusses how housing policy, historically including redlining, relates to these conditions. This explorer documents the association with income and place, not the causal pathway.
The horizontal axis is median household income; the vertical axis is the selected health measure. Where the housing map showed almost no linear relationship with race (a correlation near 0.07), this relationship is pronounced. Switch the measure, and enter a community to locate it on the chart and view its population by age cohort.
A note on age and health data. The CDC and the Boston Public Health Commission report adult health (ages 18 and older); youth, working-age, and elder health outcomes are not published at a comparable municipal level. Life expectancy at birth summarizes the full life course in one figure and is the most complete available indicator; it is drawn from CDC and National Center for Health Statistics small-area estimates (2010 to 2015), aggregated from census tracts to municipalities and population-weighted. The Boston neighborhood values shown when the city is exploded use the same tract estimates aggregated to ZIP codes; a few small downtown ZIPs lack a stable estimate and are omitted. These ZIP averages are still smoother than the sharpest tract-to-tract gaps reported by the Boston Public Health Commission, such as the 23 year Back Bay to Roxbury difference noted below. The age cohorts shown for each community are from the Census (ACS 2020 to 2024); "young adults" uses the 18 to 24 bracket, the closest standard Census age band to the 18 to 25 range.
District performance on the state's standardized tests against median household income. Switch the grade band between the elementary and middle grades and the high school years, and enter a district to locate it.
Standardized-test proficiency is the share of students meeting or exceeding expectations on the state assessments (MCAS), via the U.S. Department of Education's EDFacts collection (2018, the most recent clean pre-pandemic year). Results are reported at the school-district level, so Boston appears as a single point and most small towns are represented by their regional district. The high school band reflects the grade 9 to 12 assessment.
MassINC gives these communities a name and a definition: 26 mid-size former mill cities (Lawrence, Brockton, Springfield, Holyoke, Lynn, and others) with income and educational attainment below the state average. They hold roughly a quarter of the state's population but a a larger share of its poverty (MassINC). Research documents differences across several systems:
The clearest statewide test is the 2021 MBTA Communities Act, which requires the 177 cities and towns served by transit to zone for multifamily housing as of right (meaning allowed automatically where the rules are met, without a special permit or discretionary town approval) near stations. By early 2026, 133 had adopted compliant zoning. That is real movement, and it took years of pressure. But the law changes what is allowed, not what is built: it creates zoning capacity and removes special-permit hurdles, while requiring no developer to break ground. As of 2025 the state counted only about 4,000 units in the pipeline, a small fraction of the capacity those rezonings opened up.
Two gaps separate a vote from a home. First, some higher-cost communities met the letter of the law with modest districts placed away from their established single-family areas, so the zoning map can change without the neighborhoods changing. Second, roughly a dozen communities have not adopted compliant zoning and are in dispute with the state over enforcement; supporters of those towns cite local control and infrastructure concerns, while the state cites the statute. The point here is not to single out any town: it is that adopting a zoning change and producing housing are different things, so the more telling measure over time is how many homes are actually built, and where.
These differences can be self-reinforcing: research describes a cycle in which fewer school resources and poorer health outcomes are associated with lower adult income, which in turn relates to where the next generation can afford to live. This explorer shows the cross-sectional differences; it does not measure the cycle directly.
Data sources, all public: U.S. Census Bureau ACS 2020 to 2024, the most recent 5-year release (race, income, home value, rent, tenure, age); four reclassified municipalities (Amesbury, Easthampton, Methuen, Watertown) retain 2019 to 2023 values · CDC PLACES 2025 release (adult health) · U.S. Department of Education EDFacts via the Urban Institute Education Data Portal (district standardized-test proficiency, grades 3 to 8, 2018) · CDC and NCHS small-area life expectancy estimates, USALEEP (2010 to 2015) · Mass. DESE and MassBudget (Chapter 70). Mortgage rate from Freddie Mac and Bankrate, May 2026.
The straight-line correlation between price and Black and Latino share is near zero (r is about minus 0.07). The relationship is not linear: lower-cost towns vary widely in racial composition, while higher-cost towns are consistently lower in Black and Latino share. The triangular shape means home value sets an effective upper bound on that share at the top of the market rather than predicting it throughout. Researchers attribute that upper bound to housing-cost and zoning barriers; this is their interpretation, not a measurement made here.
Stated precisely: the pattern is most concentrated in the higher-cost Greater Boston suburban ring (Weston, Wellesley, Newton, Lexington, Needham, Dover), where high home values, alongside single-family zoning, coincide with lower Black and Latino shares, while cities such as Lawrence, Chelsea, Springfield, Holyoke, Lynn, and Brockton have larger Black and Latino populations. Whether the relevant factor is best understood as race or as income is contested: because the two are closely linked, these data cannot separate them. Several of those higher-cost suburbs are racially diverse once Asian residents are counted, which is why the relationship is more accurately described as cost-linked than as racial.
What the western Massachusetts data show. The western part of the state is not simply a gap in the data; it is a second, distinct pattern worth describing rather than setting aside. Much of Franklin, Hampshire, and Berkshire county, and rural Worcester County, is both largely white and lower-income. Low diversity there does not coincide with high housing costs; homes are comparatively inexpensive, so a cost barrier is not a plausible explanation. The likelier factors, as regional economic research describes them, are different: long-running post-industrial and rural economic decline, distance from the immigration gateways and job centers that draw newer residents, an older population, and limited in-migration. Hampden County is a useful contrast, holding the diverse, lower-income cities of Springfield, Holyoke, and Chicopee. The west thus contains both lower-income white towns and lower-income diverse cities, but in neither case does the cost-exclusion pattern of the eastern suburbs appear to apply. Treating the two as the same phenomenon would conflate distinct dynamics.
Why each group is measured separately. The headline measure is the Black and Latino share, because that is the group for which housing cost and diversity move together. Asian residents are reported as their own dimension on the housing chart, not folded into the white majority and not merged into a single people-of-color total. Doing either would obscure the picture: several of the highest-cost suburbs have large Asian populations (Lexington is 33 percent Asian), so an all-residents-of-color measure does not show the same gradient. Reporting the groups separately is what makes the difference visible, and keeps AANHPI residents present in the analysis rather than erased by aggregation.
A note on scope: this describes systems and outcomes, not people, and it documents patterns rather than proving their causes. The disparities shown are real in the data; the explanations for them, including the high cost of housing, restrictive zoning, and the racial wealth gap that earlier policy such as redlining helped produce, come from the external research cited here, not from this tool, which contains no lending, wealth, or zoning-history data. Many residents of lower-diversity communities support integration and inclusive housing. A useful forward test is not whether a town adopts inclusive zoning, but whether homes are actually built, and where (see Why It Matters).
Source: U.S. Census Bureau, American Community Survey 2020 to 2024 5-Year Estimates, released January 2026 (table-based summary file: B03002 race and ethnicity, B19013 median household income, B25077 median home value, B25064 median gross rent, B25003 tenure, B01001 age). Four municipalities reclassified after the 2020 Census (Amesbury, Easthampton, Methuen, Watertown) retain 2019 to 2023 values and are noted as such. Geography: county subdivisions (Massachusetts municipalities). 348 municipalities with population ≥ 200 and reported home value retained. Income values are top-coded by the Census at $250,001. Figures are 5-year averages and carry margins of error; verify specific values at data.census.gov before publication. Health outcomes are CDC PLACES model-based estimates for adults 18 and older; life expectancy is from CDC and NCHS small-area estimates (2010 to 2015) aggregated to municipalities; school proficiency is from federal EDFacts (2018). The demographic and economic layers reflect the newest available 5-year ACS (2020 to 2024). The life-expectancy and school layers are older because no more recent comparable municipal dataset has been published; they should be read as structural context rather than current-year measurement. This document is a research instrument intended to inform further analysis.
This short survey is optional and anonymous. It collects no names, emails, or exact addresses, and responses are reviewed only in aggregate to understand who the explorer reaches and which parts are useful. Every question can be skipped.