Commonwealth · Voters · Question 7
Question 7/Housing Divide: Renters vs. Homeowners, Urban vs. Rural
Across 350 cities and towns, the share of households that rent tracked the yes vote more closely than anything else we measured. Then density. Town income, property values and recent home building showed no link.
Published November 12, 2026
The numbers
Four numbers
23 of 23
cities and towns above 5,000 residents per square mile voted yes
57.8%
yes in the 70 cities and towns where the most households rent
−0.05
rank correlation of yes share with median family income, 348 cities and towns (0 means no consistent link)
32,685
more no than yes votes statewide (1,290,405 to 1,257,720)
The vote
What was on the ballot
Question 7 asked whether to allow single-family homes in residentially zoned areas on lots of at least 5,000 square feet, with at least 50 feet of frontage and access to public sewer and water, subject to reasonable local regulation. It failed, 1,257,720 yes to 1,290,405 no (49.4% yes). Counts are the Associated Press’s unofficial counts as published by WBUR, as of November 12, 2026; one town has no votes reported yet (New Ashford).
Renters
Only the fifth of towns with the most renters produced a yes majority on Question 7.
The 350 cities and towns with votes, sorted by the share of households that rent and cut into five groups of 70. In each group, amber bars to the left show no votes and slate bars to the right show yes votes, on one scale. The top fifth cast 43% of all votes: 635,373 yes to 463,806 no, net +171,567. The other four fifths voted no, by 204,252 votes combined; 47 of the 70 cities and towns in the top fifth voted yes, against 28 of 280 elsewhere.
Renters and income
Renters sort the towns. Income does not.
Each circle is one city or town. Up is the yes share; the heavier line is 50%. Both charts share that axis. Hover over or tap a circle for the city or town, its yes share and votes cast.
Above each chart is a rank correlation, which measures how closely two lists line up when the cities and towns are put in order on each. It runs from −1 to +1. At +1, the order would match exactly: more renters, higher yes share, every time. At 0, there is no consistent pattern either way. Below 0, the order runs the other way.
- Town voted yes (above the 50% line)
- Town voted no (below it)
- Circle area is proportional to votes cast
Yes share by renter share of households
Rank correlation with yes share: +0.63 across 350 cities and towns
Yes share by three measures of wealth and growth
Median family income
Rank correlation with yes share: −0.05 across 348 cities and towns
17 towns are at the Census top code, shown as $250K+ (the Census top code, $250,000 or more; ACS 2020–24); 2 towns (Gosnold, Monroe) have no estimate and are not plotted.
Property value per resident
Rank correlation with yes share: +0.08 across 350 cities and towns
New housing value per resident, FY2017–26
Rank correlation with yes share: −0.03 across 350 cities and towns
Residential new-growth value reported to the Department of Revenue, divided by 2020–24 population.
Yes share is clipped to the 20–80% axis; the lowest city or town is at 25.1% and the highest at 75.9%. These are town-level associations. They do not say how individual renters or owners voted.
Density
Every city and town above 5,000 residents per square mile voted yes.
Towns grouped by residents per square mile. Boston is one of the 23 cities and towns; press reports say the measure would not have applied there (see the size of the piles below). Below 2,500, 42 of 304 voted yes. Bars show net votes: left of zero is net no, right is net yes. Rank correlation of yes share with density: +0.45.
Density and renting go together: the denser a city or town, the larger the share of households that rent (rank correlation +0.62). So we compared like with like. Among cities and towns with similar shares of renters, the denser ones were barely more likely to vote yes (+0.09, close to no link). Among places of similar density, those with more renters still voted yes more often (+0.51). Most of what density seems to show, renter share already explains.
The exception is the fifth with the most renters, where 31% to 74% of households rent. There, denser places voted yes more often (+0.60 across those 70 cities and towns), and the link holds with renter share accounted for (+0.41). In the other four fifths it runs from −0.04 to +0.15.
Counties
Suffolk and Middlesex had the only large yes margins; Plymouth, Bristol and Worcester had the largest no margins.
Net votes by county, same form as above. The counties are listed from the largest net yes to the largest net no.
The size of the piles
The size of the piles
The 75 cities and towns that voted yes had a combined margin of +205,409 votes. The 275 that voted no had a combined margin of −238,094. Statewide, the result was −32,685.
Boston voted 119,998 yes to 53,286 no, net +66,712. Press reports say the measure would not have applied in Boston, whose zoning is outside the state zoning act the measure amends (Boston.com, September 22, 2026; WBUR, October 7, 2026; WGBH, October 22, 2026). Without Boston, the other 349 cities and towns voted 47.9% yes, 1,137,722 to 1,237,119.
The governor’s race
Beside the governor’s race
Healey received 64.1% of the two-candidate vote statewide, 1,628,580 to 912,149. Minogue carried 80 cities and towns; all of them voted no on Question 7. Healey carried 269 cities and towns, of which 195 voted no on Question 7. The governor’s vote is itself a vote, not a town characteristic, and it tracks Question 7 more closely than any measure above: rank correlation +0.74 between Healey’s share and the yes share. One town split its governor vote evenly (Monroe).
Analysis
What the results likely mean
These are town results, not a poll of voters. Several reads might be drawn from them.
The first is ownership. Where most households own, Question 7 asked owners to accept new houses on smaller lots next door. Their likely concerns are the street, the schools and the tax rate. Where more households rent, the same question may have read as a way into a first home. Town income showed no link (−0.05). The divide would seem to be about where people stand in the housing market, not what they earn.
The second is landscape. Rural places are mostly owner places, so this read rides on the first. Below 2,500 residents per square mile, 42 of 304 cities and towns voted yes. In a small town, lot size is also a landscape rule: it decides how much field and forest survives between houses. Towns with more forest leaned no (−0.32). But among those 304 lower-density cities and towns, the forest link nearly disappears (−0.11) while renter share holds (+0.53). Ownership would seem to track the rural vote more closely than landscape does.
The third is governing. The state split close to even: 49.4% yes. The more a town voted for Governor Healey, the more likely it was to vote yes (+0.74 across 350 cities and towns). Yet of the 269 cities and towns she carried, 195 voted no on Question 7; none of the 80 that voted for Minogue voted yes. The question split the governor’s own coalition.
Any housing policy from here has to hold both: the cities that see the shortage every day and the towns that weigh what new building would change. Can one statewide rule do that?
Who holds the lever
Who holds the lever
The Information for Voters says a no vote would make no change to the law relative to building single-family homes. Question 7 failed, so the law is unchanged, and minimum lot sizes continue to be set in each city’s and town’s own zoning.
Sources and limits
Sources and limits
Vote counts: Associated Press town results as published by WBUR (press source), Question 7 and governor, file last modified November 12, 2026 at 19:55 GMT. Question text: Secretary of the Commonwealth, Information for Voters 2026. Boston and the state zoning act: Boston.com, WBUR and WGBH (press). Renters and households, property value per resident and land area: ACS 2020–24 and Department of Revenue figures in the study’s town tables. Family income: ACS 2020–24 table B19113 via Census Reporter. New housing value: Department of Revenue new-growth reports, FY2017–26.
Forest (analysis section): the study’s Land and Water Index table, forest acres from MassGIS Land Cover/Land Use 2016 (deciduous and evergreen forest) as a share of the town’s land. The analysis section is our reading of the town pattern; it is labeled as such.
These are unofficial counts as of November 12, 2026 and will be replaced by the Secretary of the Commonwealth’s certified results when they are published. 2,383 of 2,384 precincts are reported statewide; New Ashford has no votes yet and is left out of every statistic, so 350 of 351 cities and towns are counted. Results are associations between town totals and town characteristics. They do not show how individual renters or owners voted, and nothing here is causal. Fifths: towns sorted by renter share and cut into five equal groups. Correlation: Spearman rank correlation across towns between the yes share and the variable; "with density held the same way" and "among cities and towns with similar shares of renters" are partial rank correlations.
Massanomics Weekly Every Tuesday
Your city or town, in the numbers that matter.
- This week’s story, with where your city or town stands
- Economics, housing and the state money that reaches you
- Land, climate, history and politics
- Side-by-side findings that show how it all connects