The town that lost at the Supreme Judicial Court on Friday had already done the thing it was suing to avoid. Marshfield town meeting adopted compliant multifamily zoning on October 20, 2025, by a vote of 363 to 209, on the third attempt after rejecting it twice. The SJC handed down its decision eleven months later, on September 18, 2026, and the opinion runs 22 pages without mentioning that vote once.

It did not have to. The court was reviewing a complaint filed in February 2025, and on a motion to dismiss the world outside the four corners of that complaint does not exist. So the practical effect on Marshfield is close to zero. The town already drew its district. The real audience for this opinion is the eight communities that are still refusing, and anyone who has been underwriting land or multifamily in one of them on the theory that the law might go away.

That theory is done. Not because a court declared the statute bulletproof, but because the SJC just published, in detail, what a town would have to prove to get out from under it, and nothing a rezoning actually costs comes close to clearing that bar.

Here is the part I think most people are going to miss. The lawsuit was never what set land values in these towns. The map is. The court decided whether a town has to draw a multifamily district. It said nothing at all about where that district goes, and where it goes is the entire investment question. That gets settled in a school gymnasium at a town meeting, and in the eight holdout towns most of those meetings have not happened yet.

What the court actually decided on September 18

The case is Town of Marshfield & another vs. Commonwealth & another, SJC-13840. Argued March 4, 2026, decided September 18, 2026, written by Justice Serge Georges Jr., with Chief Justice Budd and Justices Gaziano, Kafker, Wendlandt, Dewar and Wolohojian all on the panel. No dissent, no concurrence.

Marshfield brought four arguments and lost all four.

On the unfunded mandate claim under G. L. c. 29, section 27C, the court held the complaint did not plausibly allege one. Marshfield said it spent money evaluating and drafting zoning bylaws and presenting them at town meeting. The court’s answer was that the complaint “does not describe the nature, amount, or necessity of those costs in a way that plausibly suggests they were unavoidable expenditures required by the MBTA Act, rather than incidental local administration expenses or costs Marshfield chose to incur in pursuing its preferred path to compliance.” Marshfield’s claim about future costs fared worse. Describing future costs as direct, the court said, “states a legal conclusion, not a well-pleaded fact.”

On the voting rights claim, the court did not even reach the merits. Marshfield and its select board have no standing to assert the constitutional voting rights of Marshfield town meeting voters, because a town is a political subdivision of the Commonwealth and cannot raise constitutional challenges against the state that created it. Board members suing in their official capacities, with no personal injury alleged, were out for the same reason.

On home rule, the court reached a claim it found was waived, because the issue was important enough to settle. The Home Rule Amendment “does not prevent the Legislature, through a valid general law addressing a matter of Statewide concern, from imposing substantive requirements that municipalities must implement through existing local procedures.” Housing is a statewide concern. The act is a valid general law.

And on the Zoning Act argument, that requiring compliant zoning effectively forces town meeting to vote yes, the court gave the line that the eight remaining holdouts should read twice: “Town meeting voters retain the authority to debate, amend, approve, or reject any proposal placed before them under section 5. But that authority does not include the power to exempt the town from an obligation imposed by a valid State law.”

Rejecting the article is still allowed. It just produces noncompliance, not an exemption.

The one door the SJC deliberately left open

I want to be precise here, because the coverage has been sloppier than the opinion.

The SJC did not hold that MBTA Communities Act compliance can never be an unfunded mandate. It said the opposite in so many words: “We therefore need not decide whether costs associated with MBTA Act compliance are categorically excluded from section 27C.” What died on Friday was Marshfield’s complaint, not the legal theory in the abstract.

That distinction matters for about five minutes, and then it stops mattering. Read what the court required. A town would have to plead specific, itemized, municipality-specific costs that were mandatory rather than voluntary, and that exceeded incidental local administration expense. Then look at what a rezoning actually generates: consultant fees, a planning board process, warrant articles, a public hearing, a town meeting. The court just labeled that entire category either incidental or self-inflicted.

The strongest fact Marshfield had was not its own. In February 2025 the Division of Local Mandates, inside the State Auditor’s office, sent letters to Methuen, Middleborough and Wrentham concluding that the MBTA Act is an unfunded local mandate. A state agency agreed with the towns. The SJC dispatched it in a paragraph: those letters were not about Marshfield, did not identify municipality-specific mandatory costs, and “do not supply the facts missing from the complaint.”

If a favorable determination letter from the State Auditor’s own division does not move the needle, I do not know what a ninth town is supposed to file.

Who is actually left, and what each of them owes

Attorney General Andrea Campbell sued nine communities on January 29, 2026: Dracut, East Bridgewater, Halifax, Holden, Marblehead, Middleton, Tewksbury, Wilmington and Winthrop. All nine had missed the July 14, 2025 deadline set by the post-Milton regulations. At that point 165 of the 177 covered communities had complied.

Marblehead has since adopted zoning and reached interim compliance in May 2026, with its application still in review at the Executive Office of Housing and Livable Communities. That leaves eight towns genuinely holding the line, out of 177.

These towns are not all the same problem, and lumping them together is where most of the commentary goes wrong. Four are commuter rail or served communities with a real transit obligation. Others are adjacent communities and adjacent small towns with no station at all, which means no half mile transit siting requirement and a great deal of freedom about where the district lands.

Wilmington is the one I would stare at. It has two commuter rail stations, Wilmington station on the Lowell Line at Main Street in the town center and North Wilmington on the Haverhill Line off Middlesex Avenue. Its obligation is 1,248 units with half the district inside a half mile of a station. The district that voters killed on May 3, 2025 ran along Main Street through the train station, and the town’s own estimate was that it would produce fewer than 200 actual new units. They rejected it anyway.

What holding out has already cost, in cash

Noncompliance is not a free option. Section 3A(b) cuts a town off from the Housing Choice Initiative, the Local Capital Projects Fund, MassWorks and HousingWorks by statute, and the Healey administration has extended the freeze to discretionary grants well beyond that list. The bills are not theoretical.

Tewksbury’s losses are the ones that would get my attention if I lived there. Three Department of Elementary and Secondary Education grants went ineligible: a $50,000 Early College planning grant, $250,000 in Early College designation funding spread over five years, and a $50,000 implementation grant. That is school money, traded for a zoning district that the town’s own planning consultants say it can site in an office park.

The Lowell Line test: Woburn complied, Wilmington did not

This is where I can bring something to the argument that the news coverage cannot, because we pull directly from MLS PIN rather than from a national portal’s town page.

The standard holdout argument is that by-right multifamily zoning will damage the single family market. Woburn and Wilmington sit on the same Lowell Line, eight minutes apart. Woburn rezoned. Wilmington refused. Here is what actually closed in both towns over the identical six month window from March 1 through August 31, 2026.

Woburn has had compliant by-right multifamily zoning on the books since 2023. Its single family average is $860,843 on 57 days. Wilmington held the line and sits at $872,591 on 61 days. The compliant town and the defiant town are within a percent and a half of each other on price and four days on market, and the defiant one is marginally slower.

If rezoning were the value destroyer these town meetings keep describing, that gap would not look like noise. It would look like something. It looks like nothing.

The column that should actually worry a Wilmington homeowner is the last one. Zero. In six months, across 120 closed sales of every kind, Wilmington did not record a single multifamily transaction. It also did not record a single vacant land sale, and I widened that one to a full twelve months before I believed it. A town with two commuter rail stations produced no land trades and no income property trades in a year.

Winthrop is already the thing it is fighting

Here is the number that ended the debate for me.

Winthrop is suing to avoid zoning for 882 units of multifamily housing. Over that same March through August window, 22 of Winthrop’s 99 closed sales were multifamily income properties, and only 23 were single family. Multifamily was 22.2 percent of everything that traded. Winthrop’s three families averaged $1,202,250 and its four families closed between $1,240,000 and $1,331,000.

Winthrop Town Council President Jim Letterie told CommonWealth Beacon, “I’m a no at 800 units. I’m a no at one unit. Not because of the number of units, but because we are being told what to do with our town.” I understand the sentiment. It is also an argument about autonomy, not about housing form, and it is worth separating the two, because a Winthrop resident reading a doomsday flyer about multifamily housing is reading about the town they already live in. The two and three family is the native building type on that peninsula. It sells at over a million dollars.

The mandate is settled. The map is wide open.

Now the part that matters for money.

Every one of these towns is going to end up with a compliant district. What none of this litigation determined is where. And the state’s own record shows how much latitude that leaves.

Boston Indicators published an analysis by Amy Dain in January 2026 finding that roughly 7,000 units are in the permitting, construction or occupancy pipeline across more than 100 projects in 34 communities as a direct result of Section 3A rezonings. That is real. But only about 30 percent of those pipeline units sit within a half mile of a train station, and 19 projects of 100 units or more account for roughly three quarters of the total. Dain’s own framing is that the law’s “hallmark flexibility leaves a lot of leeway for communities to embrace or sidestep the law’s housing goals.”

Sidestep is the operative word. Marshfield complied by putting its district where redevelopment is unlikely. Tewksbury, as an adjacent community with no station and therefore no transit siting requirement, is bringing two subdistricts to its October 6 special town meeting: one on Clark Road and one at the Lodge at Ames Pond, in an office research district. Both were recommended unanimously by the select board and planning board in June and confirmed as compliant by the Northern Middlesex Council of Governments. Both are legitimate. Neither is downtown.

So the thesis “buy land in a holdout town because the upzoning is coming” is half right, and the missing half is the expensive half. The upzoning is coming. It is not coming to every parcel, and in a town with no station it may not come anywhere near the parcel you assumed.

Tewksbury’s land market makes the point. Two commercial land parcels have closed there in the last year. 937 North Street, 3.16 acres, sold for $2,200,000 in September 2025 at 100 percent of its original ask after 503 days on market, which works out to $696,203 an acre. 1325 Main Street, 2.09 acres, closed on September 10, 2026 for $1,500,000 after 867 days on market and a cut from $1,750,000, which works out to $717,703 an acre. Two trades a year apart at effectively the same per acre price, both after more than a year and a half of marketing. Neither is inside the district the town votes on October 6.

That is what a land market with no zoning visibility looks like. Long holds, wide bid ask spreads, and pricing that has nothing to do with density because nobody knows what the density will be.

The honest counterargument, which is political and not legal

I do not think the remaining holdouts have a legal path. I do think they have a political one, and it would be dishonest to write this without saying so.

Massachusetts votes for governor in November 2026. Governor Healey’s Republican opponent, Mike Minogue, has said he would end state enforcement of the MBTA Communities Act and provide legal support to cities and towns that oppose it. That is not a fringe position in the nine towns that got sued.

Weigh it honestly. A new administration cannot repeal a statute, and Section 3A’s grant ineligibility is written into the law itself rather than into an enforcement posture. The Attorney General is separately elected and her authority to bring these suits was settled in Milton in January 2025. What a governor could do is slow-walk EOHLC, soften the regulations, and change the political weather. That is a meaningful risk to the timeline. It is not a path back to single family only zoning.

If I were underwriting a five year hold, I would treat the election as schedule risk, not as an exit.

How I would underwrite these eight towns right now

For investors and land buyers:

1. Stop paying for optionality you do not have. If your model has a branch where the town gets an exemption, delete it. After Milton and now Marshfield, that branch is gone.

2. Underwrite the map, not the mandate. Before you bid on a parcel in Wilmington, Tewksbury, Holden or Middleton, find the town’s actual proposed district boundaries in the planning board packet. In Tewksbury that is Clark Road and the Lodge at Ames Pond. Anything else is speculation about a second district that may never be drawn.

3. Separate station towns from adjacent towns. Wilmington has to put half its district inside a half mile of a station, which is a hard geographic constraint you can underwrite. Tewksbury, Holden, Middleton and Halifax have no such constraint, which means transit adjacency is an amenity story in those towns and not a zoning one.

4. Price the delay. 867 days on market for a 2.09 acre Main Street parcel is your carry assumption, not 180 days. These are not liquid markets and a favorable court ruling did not make them liquid.

For move-up buyers looking at Holden, Tewksbury or Wilmington specifically because those towns have resisted density:

5. You are buying a house, not a zoning outcome. If low density is the only reason a town is on your list, take it off. The resistance was never going to hold, and now it has been told so twice by a unanimous court.

6. Look at what the compliance actually produced next door. Woburn rezoned in 2023 and its single family market is indistinguishable from Wilmington’s today. That is the best available evidence of what happens to your equity, and it is better evidence than anything said at a town meeting.

For owners in the holdout towns:

7. Price in the grant freeze, because it is a real municipal budget item. Middleton gave up a $2,000,000 MassWorks award. Tewksbury gave up $350,000 in school grants. Those are costs that land on a tax rate or on a service cut, and they are the one genuine financial consequence of the fight that the litigation never addressed.

8. If you own a two or three family in one of these towns, you own the scarcest product on the board. Wilmington recorded zero multifamily trades in six months. That is not a soft market. That is an absent one, and it is about to get a legal supply of competitors.

Where this leaves things

Five years in, the score on the MBTA Communities Act is 177 communities covered, 8 still refusing, and zero successful legal challenges. Milton lost on enforcement. Marshfield lost on funding, standing, home rule and the Zoning Act, and by the time it lost it had already rezoned. The amicus briefs on the state’s side in the Marshfield case came from the Greater Boston Real Estate Board, the Massachusetts Association of Realtors, NAIOP Massachusetts, the Home Builders and Remodelers Association, MAPC, CHAPA and MassHousing. When the realtors, the developers, the planners and the affordable housing advocates all file on the same side, the outcome was not close.

What is genuinely unresolved is the part nobody litigated. Eight zoning maps have not been drawn yet, one of them gets voted on in Tewksbury on October 6, and the difference between a district at a train station and a district in an office park is the difference between a land basis that works and one that does not.

If you are weighing a parcel or a multifamily in one of these towns, or trying to decide whether a town’s resistance should factor into where you buy, I am happy to walk through the actual district boundaries and the closed comps with you. Reach out anytime.

Sources

• Town of Marshfield & another vs. Commonwealth & another, SJC-13840, slip opinion, September 18, 2026 (Massachusetts Supreme Judicial Court)
• MBTA Communities Act upheld by state’s highest court (The Boston Globe, September 18, 2026)
• SJC Dismisses Marshfield’s MBTA Zoning Lawsuit (Banker & Tradesman, September 18, 2026)
• SJC again upholds MBTA Communities housing law (CommonWealth Beacon, September 18, 2026)
• State’s highest court rejects latest MBTA Communities Act challenge (Marblehead Current, September 18, 2026)
• AG Campbell Sues Nine Communities for Noncompliance with MBTA Communities Law (Office of the Massachusetts Attorney General, January 29, 2026)
• Attorney General v. Town of Milton, 495 Mass. 183 (2025)
• Researchers: MBTA housing law responsible for 7,000 new housing units so far (Worcester Business Journal on the Boston Indicators report by Amy Dain, January 28, 2026)
• MBTA Communities Act Holdouts Are Losing State Funding (Streetsblog Massachusetts, January 16, 2026)
• MBTA Communities Act: What’s next for your town (Boston.com, January 27, 2026)
• Winthrop is latest front in MBTA Communities fight (CommonWealth Beacon)
• MBTA Communities (Town of Tewksbury, including the October 6, 2026 special town meeting subdistricts)
• Section 3A MBTA Communities Zoning (Town of Wilmington)
• Holden rejects second MBTA Communities proposal, faces legal action (Boston.com, March 3, 2026)
• Closed sale figures throughout are BMN Boston’s own analysis of MLS PIN data for the period March 1 to August 31, 2026, with land sales widened to September 2025 through September 2026.

Thirteen homes changed hands last year in the ZIP code Anthropic just leased into. All thirteen were in two buildings. Both buildings opened in 1989.

That is 02142, Kendall Square, the densest concentration of lab and office square footage in the country, and over the twelve months ending August 31 it produced thirteen closed residential sales out of 719 across Cambridge. Not one of them was built in this century. Zero closings of anything built since 2023. I pulled those numbers myself out of MLS PIN because no public tracker slices that small, and they are the cleanest answer I have found to the question buyers keep asking me.

The question is some version of this: Cambridge lab vacancy is at a record, developers are giving space back, so is the glut finally going to loosen up housing near Kendall Square? It is a reasonable thing to assume. Every office-to-residential conversion story this year has trained people to read an empty commercial building as a housing unit in waiting. And on September 9, Anthropic signed a 24,000 square foot lease at One Kendall Square and showed why the assumption is only half right.

Thirteen Sales, Two Buildings, Both Finished in 1989

Start with what the Kendall Square housing market actually is, because most people arguing about it have never looked. Here is every residential closing recorded in 02142 between September 1, 2025 and August 31, 2026. Not a top ten. All of them.

Two buildings on the Lechmere Canal, both delivered a year before CambridgeSide opened across the water. The rest of Cambridge closed at 99.3 to 101.5 percent of the original asking price depending on the ZIP. Kendall Square closed at 92.1 percent and sat a median 106 days. That is not a hot market and it is not a cold one. It is barely a market at all, because there is almost nothing there to trade.

So when somebody tells you the lab glut is going to fix housing supply near Kendall, ask them which buildings. The honest answer right now is two, and they were finished when the first Bush was president.

What Anthropic Signed, and What It Did Not Ask For

On September 9 Anthropic announced a lease for 24,000 square feet across two floors at One Kendall Square, the office, lab and retail complex on Binney Street that Alexandria Real Estate Equities bought from DivcoWest in 2016 for $725 million. The company has been running a roughly 50 person local team out of a WeWork in Central Square. The new space holds 120 workstations. Move in is early 2027.

Two details matter more than the square footage. The first is that Anthropic told the Boston Globe it is not seeking any state or local tax break for the expansion. In a state where nearly every corporate headcount announcement comes bundled with an incentive package, a company more than doubling its local footprint and declining to ask is a signal about how badly it wants the address.

The second is the landlord’s own occupancy. Banker & Tradesman reported the Alexandria campus at One Kendall Square finished the second quarter 86 percent occupied. Hold that against a Greater Boston lab market that CBRE put at 28.7 percent vacant in Q2 2026, up 250 basis points year over year from 26.2 percent. The record everyone quotes, 28.0 percent, was set in Q4 2025 and has already been broken twice since. Meanwhile the trophy Kendall campus is running at 14 percent vacancy. Same city. Same quarter. Half the emptiness.

The Vacancy Rate Is an Average of Two Markets

This is the part that gets lost. A metro vacancy rate is a weighted average, and Cambridge’s lab inventory splits into three submarkets that behave nothing alike. Lincoln Property Company’s Q2 2026 lab report breaks them out.

Look at the leasing column rather than the vacancy column. East Cambridge did 947,110 square feet of leasing in a single quarter. West Cambridge did 10,402. That is a ninety to one ratio between submarkets whose inventories differ by six to one. Sanofi alone accounted for most of the East Cambridge number with a ten year, 900,000 square foot extension at 350 and 450 Water Street, which is itself the point: when a large tenant recommits in Greater Boston right now, it recommits in Kendall.

Two more things worth understanding before you read another vacancy headline. Vacancy counts space nobody occupies. Availability counts space you could lease, including space a tenant is still paying for and trying to hand off. Mid Cambridge is 15.5 percent vacant and 36.3 percent available, and the gap is almost entirely sublease inventory. That distinction is why the same market can look like a disaster or a soft patch depending on which number an article picked.

Takeda Is Not Giving Back Kendall Square

The biggest give back in the market is the one people cite most loosely. In February, Takeda put roughly 630,000 square feet on the Cambridge sublease market across 35 Lansdowne Street, 40 Lansdowne Street and 300 Massachusetts Avenue, plus about 125,000 square feet at 75-125 Binney Street, as Bisnow reported. Headlines called it a retreat.

It is a consolidation. Takeda signed a fifteen year lease in 2022 on 585 Third Street, a roughly 650,000 square foot BioMed Realty development in Kendall Square that delivers in the second half of this year, and it is pulling its scattered legacy space into that one campus. The company’s Cambridge footprint goes from about 1.56 million square feet across seven buildings to roughly 1.43 million across four. The net reduction is on the order of 130,000 square feet. The gross number that made the headlines is more than four times that.

And notice the addresses. Lansdowne Street and 300 Mass Ave are Mid Cambridge. The space Takeda is keeping is Kendall. This is the entire pattern in one tenant: legacy, off-core, older buildings go back to the market, and the best located square footage gets held or re-signed. AstraZeneca is doing a version of the same thing at 290 Binney Street, a 570,000 square foot anchored delivery landing this half. Neither of those buildings is ever going to be an apartment.

AI Leasing Is Already Bigger Than Cambridge’s Vacant Lab

Anthropic’s 24,000 square feet is small on its own. It is not small as a data point, because it lands inside a trend CBRE has been measuring for a while. In its 2026 Tech Gateway Office Markets report, CBRE totals cumulative AI company leasing from 2019 through the first quarter of this year by market.

Boston is fourth at 3.7 million square feet, behind only the two Bay Area markets and Manhattan. Tech’s share of all United States office leasing hit 22.7 percent in the first quarter of 2026, up from 15.3 percent a year earlier and from a cyclical low of 12.7 percent in 2023. Nationally that is a story about office. In Cambridge it is a story about lab, because in Kendall Square the two are the same buildings with different mechanical systems.

The local roster is filling in. AI Proteins took 40,000 square feet at Related Beal’s One Kenmore Square in January for its headquarters. Lovable took 6,000 downtown at One Lincoln. LabCentral turned vacant space at 700 Main Street into an AI BioHub for artificial intelligence focused biotech startups instead of waiting for a traditional lab tenant. In February the Healey administration announced a state partnership with OpenAI alongside the launch of the Massachusetts AI Coalition. None of those are conversions to housing. All of them are square footage coming off the table.

Nine Days in September

I do not usually get to make an argument with a calendar. This month I can.

One of those three things required a signature. The other two required a moratorium, a working group, a rezoning, a Planning Board, and a City Council that is still amending the rules nineteen months after passing them. That asymmetry is the whole argument, and it did not need a theory to demonstrate it. It happened in nine days.

The Housing Is Going Three Miles Away

Cambridge is building. It is just not building where the lab market is tightest.

Every meaningful lab to housing pivot in this city has landed in West Cambridge, the 39.5 percent vacancy submarket out by Alewife and Fresh Pond. Boylston Properties swapped lab plans at 745 Concord Avenue for a residential tower. Toll Brothers and PGIM opened The Laurent at 55 Wheeler Street in June 2024, 525 units a ten minute walk from Alewife station. And on September 17 the Planning Board approved Healthpeak’s Cambridge Point, 4.6 million square feet on 46 acres in the Alewife Quadrangle with more than 2,300 homes, a project that began when Healthpeak spent $625 million assembling land, got frozen by the city’s 2022 moratorium on Alewife lab and office development, waited out a rezoning the Council finalized in September 2023, and only now has an approval.

One Kendall Square to Alewife station is about 3.3 miles as the crow flies. In Cambridge terms that is a different city. It is also the correct outcome, which is the part I want to be fair about. The Quadrangle has large parcels, surface parking, a subway terminus and a zoning framework that now requires housing alongside large commercial development. Andrew Copelotti of Boylston Properties put it plainly to Banker & Tradesman in 2024: “Thank God the moratorium came along. We’d have a built or half-built life science building in a tough market.” He is right. That space should be housing.

But read the Cambridge Point approval conditions and you find the sequencing problem in black and white. The pedestrian bridge connecting the project to Alewife station is required to start construction when 50 percent of the commercial portion is complete. The housing’s transit connection is gated on the lab and office space getting built first.

Housing Is the Only Bidder That Has to Ask Permission

Here is the mechanism, stated as plainly as I can.

When a well located lab or office building empties out, two kinds of capital look at it. One is a tenant with a signed term sheet who can be in the building in six months and who is bidding a rent number. The other is a residential developer whose bid is contingent on a zoning change, a permit, a financing, a construction cycle and, in Cambridge, a Planning Board. The first bidder closes in weeks. The second closes in years, if at all.

Landlords are not choosing housing over lab or lab over housing on ideology. They are choosing certainty. And because a discounted trophy rent still underwrites better than a residential redevelopment pro forma with five years of entitlement risk in front of it, the well located square footage clears to the faster bidder every time. Cambridge asking rents have fallen hard, from roughly $113 per square foot triple net at the Q3 2021 peak to $85.34 in Q2 2026, about 24 percent. A 24 percent haircut on Kendall rent is still a number a housing developer cannot outbid.

That leaves conversion to the buildings nobody is bidding on. Obsolete floor plates. Bad transit. Wrong block. Those are genuine candidates, and Boston has been working that problem with its own downtown conversion program, which I wrote about when the lab glut mostly got skipped by the conversion pipeline. The uncomfortable part is that the buildings that clear the conversion test are usually the ones you would least want to live near a job in.

The Conversion Math, Honestly

Let me put real numbers on the hope, because the hope is not crazy, it is just smaller than people think.

And that top line assumes conversion is even physically possible, which for purpose built lab it usually is not, and assumes the 2.6 million vacant square feet outside West Cambridge would be available, which the leasing data says it will not be.

Set that against what the city has actually produced. Cambridge legalized multifamily housing citywide on February 10, 2025, ending single family zoning and allowing four stories by right with a two story bonus for inclusionary units on larger lots. It was a genuine landmark. In the twelve months that followed, the city permitted fifty smaller developments and zero large scale projects, up from 43 buildings in 2024. Vice Mayor Burhan Azeem’s assessment was blunt: “we’ve not actually seen any of these buildings actually open yet.” Then on September 16 two Council committees voted to add setbacks and open space requirements back on top.

What the Sales Data Says About Buying in Cambridge Right Now

None of the above has loosened anything, and the closing record shows it. Here is what Cambridge did over the last twelve months, by ZIP, straight out of MLS PIN.

Kendall Square carries the highest price per square foot in the city, $1,137, and 1.8 percent of the sales. Cambridge’s newest housing, the 34 closings of homes built since 2023, averaged $2,580,117. Whatever the lab market is doing, it is not putting downward pressure on any of that. There are 209 active listings in Cambridge as I write this, averaging $1.9 million, with 62 more pending.

What I Would Actually Do With This

If you are buying in Cambridge, stop underwriting a future supply wave that is not coming to your block. The vacancy headline is real and it is describing buildings you will never live in. Price the neighborhood you want on its own closing record, and in East Cambridge that record says 100 sales a year at $860 a foot and 97.5 percent of ask, which is a normal, competitive market with no glut discount in it.

If you are investing, the split is the opportunity. West Cambridge is where the basis is cheapest, the rezoning is done, the entitlement risk is now proven survivable, and the Cambridge Point approval just established what 46 acres of the Quadrangle is worth. That is a ten year story with real execution risk and it is the only part of Cambridge where the lab glut converts into housing at scale. East Cambridge is a different trade entirely. You are buying scarcity next to an employment base that is currently being rebuilt by tenants with more capital than the biotechs they are replacing.

If you own near Kendall already, the Anthropic lease is good news and it is not the kind of good news that shows up next quarter. A 120 workstation office does not move a housing market. A hundred of them would, and the direction of travel is the thing to watch, not the size of any one signature.

What I would watch next: whether any AI tenant takes a full building in Kendall rather than two floors, whether Takeda’s Lansdowne Street space finds a subtenant or sits, and what the full City Council does on September 28 with the setback amendments. If Cambridge tightens its own housing ordinance in the same month it approves 2.6 million square feet of new commercial space at Alewife, that tells you which bidder this city is actually set up to serve.

We track Cambridge closings by ZIP and by submarket every month, including the cuts in this article, because the published town-level numbers are too coarse to price a specific block. If you are weighing a purchase near Kendall, Central or Alewife and want the real closing record for that pocket rather than a citywide average, reach out and I will pull it for you. If you already own in Cambridge and want to know what the last twelve months did to your number, start with a home value estimate.

Sources

• Boston Globe, Anthropic signs lease for 24,000 square foot office near Kendall Square, September 9, 2026
• Bisnow, Anthropic Doubling Local Presence With Cambridge Office Lease
• Banker & Tradesman, Anthropic Inks 24K SF Kendall Square Lease, September 10, 2026
• CBRE, Boston Metro Life Science Figures Q2 2026
• Lincoln Property Company, Q2 2026 Lab Market Report, Boston, Cambridge, Suburbs
• CBRE, 2026 Tech Gateway Office Markets
• Bisnow, Takeda Looks To Put 630K SF Up For Sublease Across Three Cambridge Properties, February 13, 2026
• Banker & Tradesman, Cambridge Approves 4.6M SF Healthpeak Alewife Project, September 17, 2026
• Harvard Crimson, Cambridge Sees Early-Stage Developments One Year After Overhauling Its Zoning Code, March 26, 2026
• Cambridge Day, Zoning update may see major changes, September 16, 2026
• City of Cambridge CDD, Cambridge Zones for Multifamily Housing Citywide, February 2025
• Cambridge Day, Close to a quarter of Cambridge lab space sits vacant, October 22, 2025
• Banker & Tradesman, Will Lab Slump Translate into Housing Gains?, September 29, 2024
• New England Real Estate Journal, Related Beal completes 40,000 s/f lease to AI Proteins at One Kenmore Sq.
• MLS PIN closed sales, Cambridge MA, 719 transactions September 1, 2025 through August 31, 2026, compiled by BMN Boston

The building that broke a $322 million bet never changed. Its tenants did not leave.

When BioMed Realty bought 1000 Washington St. and the connected 321 Harrison Ave. in the South End in early 2021, the plan was to turn 490,000 square feet of office into lab space. Boston was paying lab rents then and nobody wanted to be late. But the Massachusetts Department of Developmental Services stayed in its offices at 1000 Washington. So did the Board of Registration of Social Workers. You cannot rip out a floor and install a 10 foot air handler around a state agency that is still using the elevators. The lab conversion at that address never got off the ground.

On June 3, 2026, the keys went to the lenders.

I have had four separate conversations this month with people who read that headline and drew the same conclusion: empty labs, housing shortage, therefore apartments. It is a reasonable guess. It is also mostly wrong, and the way it is wrong matters if you own a condo in the South End, are shopping in South Boston, or are underwriting anything downtown. A business plan can change in a quarter. A floor to floor height cannot.

What actually changed hands on June 3

This was not a foreclosure auction on the courthouse steps. BioMed Realty, a Blackstone company, bought the two connected buildings from CIM Group and Nordblom in February 2021 for $314 million and financed them with a $322 million floating rate loan from a partnership of KKR Real Estate Finance Trust and AllianceBernstein. In June the deed went to that lender partnership, which Banker & Tradesman reported as an assumption of the $322.6 million mortgage. Tishman Speyer was named property manager on June 17.

The distinction is worth holding onto, because it tells you what happens next. Nobody bought this campus at a distressed clearing price that resets the block. A lender took title to protect a position. Lenders are not developers. They stabilize, they lease, and they sell later. The 321 Harrison side is not empty either. Monte Rosa Therapeutics signed for 63,000 square feet in December 2021, and PepGen, Pyxis Oncology, Nido Biosciences, Magnet Biomedicine and HilleVax followed in 2022. That half of the campus worked. The office half is what did not convert.

I wrote about a similar mechanic in Back Bay earlier this year, where the Park Square Building went to a conversion plan through a foreclosure rather than through a policy incentive. The pattern repeats: the debt gets written down first, and the housing question gets asked second.

The vacancy number depends on who you ask

You will see a lab vacancy figure quoted with a lot of confidence. Be careful with it, because the trackers do not agree, and the spread between them is wider than most of the moves they are describing.

A range of 26.4% to 34.2% is not a rounding difference. It reflects real disagreement about what counts as lab inventory in a market where a lot of buildings are half converted and marketed both ways. The honest version is that roughly a third of Greater Boston lab space is empty and the absolute figure, 16.6 million square feet, is the one that actually means something. I have seen a 42% figure circulating. I could not source it to any named tracker, so I am not using it. The real number is bad enough without inflating it.

For scale: Greater Boston lab inventory more than doubled in roughly a decade, from 18.6 million square feet in 2012 to over 41 million by CBRE’s count. The Boston Globe put it closer to 48.4 million using a different definition. We built a second lab market on top of the first one and the tenants for it did not arrive.

Why a failed lab is a bad apartment building

This is the part that gets skipped, and it is the whole argument.

Purpose built lab space is close to the worst residential conversion candidate in the city. Labs carry roughly three times the mechanical equipment of an office building. They need 10 foot by 20 foot air ducts because lab air gets exhausted rather than recirculated. The floor to floor heights are taller, which sounds like a feature until you do the arithmetic: a 294 foot lab building yields about 15 rentable floors, where an apartment building of the same height gets 25 or more. Elevators, stairs and plumbing stacks all sit in the wrong places. Architect Troy Depeiza’s summary in the Globe was that it would take “a whole lot of rejiggering to make that work,” and that it “wouldn’t be efficient.”

Then there is basis. Lab construction runs about $1,200 per square foot, roughly 50% above office. A developer who paid lab land prices and built to lab specs is carrying a cost per foot that apartment rents in Boston cannot service. That owner does not convert. That owner waits, or re leases to an office tenant, or hands the keys to a lender.

Which brings the 1000 Washington story full circle. The failed lab conversion is the reason that building is still a plausible housing candidate at all. Because the state agencies never moved out and the lab buildout never happened, it is still an office building with office bones. The half of the campus that succeeded as lab, 321 Harrison, is the half that is now hardest to turn into anything else.

The South Boston conversion that adds 24 apartments and no supply

A few blocks away at 69 A Street, a five story building sits on a site that has been chasing a use since 2016. The BPDA approved adding stories to an old rivet factory that year. CIEE, a Portland, Maine nonprofit, bought it in 2018 and got approval in 2019 to make it a headquarters. Then the pandemic hit, CIEE never moved in, and in 2021 it signed a purchase and sale with a life sciences firm to turn the building into labs. The Zoning Board of Appeal rejected it. A mostly residential stretch of A Street was not going to get a lab.

The 2026 filing converts floors two through four into 24 residential units. That reads like 24 units of new South Boston supply, and it is not. Per the project filing with the Planning Department, the units are designated for participants in CIEE’s own international internship programs, on placements that run twelve to eighteen months. They are tied to one organization’s program. They will never be listed, never be leased on the open market, and never show up in any inventory count you or I can shop.

What is actually converting in the South End

The real South End conversion story is two old office buildings, and neither one is a lab.

At 615 Albany St., a former Boston University Medical Center administrative building of about 20,500 square feet, bought for $3.4 million in 2023, is becoming 24 apartments with 5 income restricted. The developer is adding a sixth story to get there.

The bigger one is 95 Berkeley St. at Chandler, an 87,250 square foot six story building that sold for $24 million and has a building permit valued north of $25 million. It becomes 92 apartments, 18 of them income restricted at 60% of area median income, with office and up to 1,700 square feet of retail staying on the ground floor.

Now look at the unit mix at 95 Berkeley, because this is the number that should change how a South End owner reads the news: 46 studios, 41 one bedrooms, and 5 two bedrooms. Eighty seven of 92 units are studios or one bedrooms. That is what an old office floor plate gives you when you cut it into apartments. Deep interior space, limited window line, small units.

If you own a two bedroom condo in the South End, 95 Berkeley is not your competition. If you own a studio or a one bedroom that you rent out, it absolutely is. Those are two completely different conclusions from the same press release, and the difference is a unit mix buried in a filing.

The pipeline loses an order of magnitude at every stage

Boston’s Office to Residential Conversion Program is a genuinely good policy. It offers a 75% property tax abatement for 29 years, compresses review to about six months, and as of the December 2025 extension had applications covering 1,517 homes across 27 buildings, including 284 income restricted units, targeting 1.2 million square feet of office. The application window now runs through December 31, 2026, with construction required to start by December 31, 2027. City figures since then have put the proposed total higher, around 31 buildings and 1,785 units.

Here is what those filings have produced.

The bottleneck is not zoning and it is not the abatement. It is construction cost and financing. At 31 Milk St., a 110 unit conversion, the general contractor’s estimate rose more than $10 million during pricing and the developer had to find a new contractor. That deal closed a $40 million construction loan only after stacking the 75% city abatement with a federal historic tax credit and a $4 million state grant. As one person close to it put it, they got all those subsidies and barely got the deal financed. Conversion costs frequently run north of $300 a square foot, construction costs are up roughly 20% since 2024, and about 75% of a conversion’s cost stays uncertain until permitting is done. One construction executive’s estimate is that roughly 75% of current proposals actually become housing.

Massachusetts just opened this tool to every city and town through H.5562, which I covered when the conversion zoning statute passed. More towns adopting it does not make any individual building cheaper to convert.

What my own closing data says about these submarkets

Rather than rely on a national tracker, I pulled every MLS PIN closed condo sale in these Boston submarkets for the twelve months ending September 17, 2026. This is the transaction record, not an estimate.

Read the bottom row twice. The Financial District produced six condo closings in a year. They averaged 166 days on market and sold at 92.6% of original list price. Not one of them closed at or above asking. That is the thinnest, slowest for sale condo market in downtown Boston, and it is precisely where the conversion pipeline is concentrated.

South End and South Boston look nothing like that. Both are moving in roughly 87 to 89 days with about a third of sales closing at or over ask. Those are functioning markets.

Now split the same closings by construction vintage, because this is where the supply question gets answered.

In the South End, post 2015 condos sold at $1,418 per square foot against $1,174 for pre 2000 buildings, and they took 140 days instead of 80. Midtown runs the other way, where the older stock is the slower half, so this is not a universal rule. But in the two submarkets this post is about, new product is not cheap and it is not fast. Anyone assuming that conversions will arrive as bargain inventory should sit with that for a minute.

What this means for rent, which is where conversions actually land

Almost all of this supply is rental, so the for sale comps only tell you half the story. Here are closed leases from the same MLS PIN pull, twelve months ending September 17, 2026.

Downtown rents sit roughly $700 to $1,200 a month above South Boston for the same bedroom count. A developer converting an office floor plate downtown is underwriting to those numbers, not to South Boston’s. The idea that conversions arrive as cheap supply does not survive contact with the rent roll. They arrive as small units at downtown pricing, with an income restricted slice carved out by policy.

That income restricted slice is the part that genuinely adds affordability. At 95 Berkeley it is 18 of 92 units at 60% of area median income. At 615 Albany it is 5 of 24. Program wide it was 284 of 1,517 as of the December extension. Real, and modest.

What I would actually do right now

Steve’s read on this, and mine, is that these conversions are leading indicators worth tracking rather than headlines to react to. The refinement the data forces is about where and when.

If you are buying in the South End or South Boston: the conversion pipeline should not change your offer. Nothing in it lands in your timeline, and the units that do arrive are mostly studios and one bedrooms that do not compete with a two bedroom condo. Buy the building and the block.

If you are buying or underwriting downtown: this is where it matters. Six closings in the Financial District in a year at 166 days and zero at or over ask is a market with very little price discovery. Add several hundred conversion apartments into that and you have a rental submarket resetting while the condo submarket has almost no comps. Underwrite a longer hold and a wider exit spread than the glossy pro forma assumes.

If you own a small rental in the South End: 95 Berkeley is 87 studios and one bedrooms with a permit in hand. That is your competition, and it is closer than the citywide numbers suggest. Look at your renewal strategy for 2027 and 2028 now.

If you are selling: price against the vintage data, not the headline. New construction in the South End is sitting 140 days. If you are selling a renovated pre 2000 unit, you are in the faster segment and you should not let a nervous conversion narrative talk you into a discount you do not need to take.

The honest summary: Boston’s lab bust is enormous and real, the conversion program is working slowly and well, and the two are only loosely connected. Failed labs are mostly not becoming apartments, because the physics and the basis do not allow it. Old offices are. That is a smaller, slower, more local story than the headline suggests, and it is the one worth underwriting.

If you are weighing a purchase, a sale, or a hold in the South End, South Boston or downtown and want the closing data for your specific building rather than the neighborhood average, reach out and I will pull it. If you are trying to price a unit, the home value tool is a reasonable starting point before we talk.

Thanks,
Steve

The auction was held on St. Patrick’s Day. Ten parcels on the stretch of South Boston where Dorchester Avenue splits off Old Colony, roughly six acres of warehouses, a self-storage building and old seafood plants. One party bid. That party already held the mortgage.

Six months later, on September 10, that same group filed to build 1,945 homes on the site. It is among the largest housing proposals South Boston has seen in years, and every story about it has quoted the same number: $75 million for six acres, about $12.5 million an acre, New York money moving into Southie. The question I keep getting from owners on that side of the neighborhood is some version of the same one. Does this mean my building is worth more?

Here is my problem with the $75 million. A price is a number two parties agree on. A credit bid is a number one party writes to itself. There was no second bidder at that auction, so nothing about that figure tells you what the dirt is worth. The good news is that South Boston produced plenty of real prices this year, and I can read those. They say something more useful, and more surprising, than the headline does.

What actually got filed on September 10

The Letter of Intent went to the Boston Planning Department for 314-420 Dorchester Avenue. The program is four residential buildings, each capped at 200 feet, which is roughly 18 stories, totaling about 1.8 million square feet with approximately 45,000 square feet of ground floor retail. The team is exploring a grocery store on the site. CBT is the architect.

The developers are J.T. Magen & Co. and Extell Development, both out of New York, with The HYM Investment Group as the local partner. Tom O’Brien of HYM is the name on the filing. His stated goal is to get through permitting by early 2027, and his quote to the Globe was “the idea is to build as fast as possible.”

One correction to the coverage, because it matters if you live there. Nearly every outlet described the site as midway between the Broadway and Andrew Red Line stops. Pull the MBTA’s own station coordinates and measure it. The site sits 0.36 miles from Andrew and 0.50 miles from Broadway. It is not midway. It is a seven or eight minute walk to Andrew and a solid twelve to Broadway, and the buildings will read as an Andrew Square project to anyone who actually walks it.

For scale: the city’s 2016 PLAN: South Boston Dorchester Avenue study imagined 6,000 to 8,000 new units across the whole corridor between those two stations. This one filing is somewhere between a quarter and a third of that entire vision, on a single six acre block.

The $75 million was not a price

Andrew Collins, a South Boston developer, started assembling these parcels in 2016. He financed the assembly with mezzanine debt from an entity called Dot Developments LLC, which is affiliated with Magen and Extell. He never filed development plans. The loan went bad. The lender moved to foreclose, Collins sued to stop the auction alleging a fraudulent foreclosure scheme, and the court declined to issue a restraining order. The auction went forward on March 17.

Banker & Tradesman put it plainly in the headline: the lender submitted the bid. Magen and Extell were the only bidders, and their $75 million is roughly half of the $150 million in debt they said they were owed, according to court filings reported by the Boston Globe. Connect CRE filed the whole thing under the heading “Return to Lender,” which is exactly right.

So no outside capital wired $75 million for South Boston dirt in March. A creditor took back its own collateral, wrote off about half its paper to do it, and cleared the title problem that had frozen the block for a decade. That is a workout, not a comp. And the single most telling fact in the whole episode is the one nobody printed: at a public auction, for six acres inside a ten minute walk of two Red Line stops, in the middle of a housing shortage, nobody else showed up to bid.

What South Boston land costs when somebody actually buys it

Two real arms length land trades closed in South Boston this year, and both printed well above the auction number.

The first is small and it is in our MLS. 364-368 Athens Street, a tenth of an acre, closed August 21, 2026 at $2,150,000 against a $2,200,000 original ask. That is $21.5 million an acre. Small parcels carry a premium per acre, so take it with that caveat, but it is a real buyer paying real money three weeks before the Dot Ave filing.

The second is the one people keep mentioning to me. In late July, Procter & Gamble’s Gillette bought 232 A Street in Fort Point from Breakthrough Properties for $99.3 million, 2.4 acres, as part of a roughly $1 billion campus overhaul. That is $41.4 million an acre. Breakthrough had paid $80 million for the same 2.4 acres in 2021, so the same parcel went from $33.3 million to $41.4 million an acre in five years, up 24 percent, with both ends of that round trip being genuine arms length sales.

Read the table the right way. I am not claiming Dot Ave dirt is secretly worth $41 million an acre. Fort Point next to the Seaport and an industrial block off Old Colony are different products with different zoning and different buyers. The point is narrower and firmer than that. Every South Boston land number this year that came from a willing buyer and a willing seller landed above $21 million an acre, and the one number that came from a lender bidding against nobody landed at $12.5 million. Treating the lowest and least real of the three as the signal for the corridor is backwards.

The strange part is that the closer you get, the cheaper it is

This is where I stopped reading the news and went into MLS PIN, because the question my clients are actually asking is about their own building, not about Extell’s.

I pulled every closed residential sale inside one mile of the site this year, then sorted them by straight line distance from 380 Dorchester Avenue. 387 closings, January 1 through September 15, 2026. The gradient runs the wrong way.

Housing within a quarter mile of the biggest residential filing this corridor has ever seen trades at $743 a foot. Housing a mile away trades at $981. That is a 24 percent discount for being closest to the project, and the nearest ring is also the only one where sellers took a real haircut off their opening number, closing at 94.7 percent of original ask against roughly 97 percent everywhere else.

None of that is mysterious once you have walked it. The quarter mile ring around 314 Dorchester Avenue is the industrial edge. It is self-storage, truck bays, the Old Colony split and the rail cut. The reason it is cheap is the exact thing the filing proposes to demolish. Whether you think that is an opportunity or a trap is the entire investment question, and I will give you my answer below.

Inside that ring, the triple-deckers are the real discount

Narrow it to a half mile and split it by product type, and a second gap opens that I think is more actionable than the first.

Eight multifamily sales is a small sample and I am not going to pretend otherwise. But it is not a sample, it is the census. That is every two or three family that closed inside a half mile of the site this year, so here is all of it.

Two things in that table are worth your attention. 8 Glover Court closed at 143.9 percent of its original ask, which is a bidding war on a 1900 two family in the shadow of the site. And 13 Gates Street traded twice inside four months, in February at $1,350,000 and again in June at $1,450,000. Every building in the census predates 1930. This is original Southie triple-decker stock, and somebody is already working it.

What the rent roll says

The sale price only matters next to what the building collects. Same half mile ring, same year, 443 closed leases in MLS PIN.

One bedrooms leased at an average of $3,156 across 107 signings. Two bedrooms at $4,014 across 237. Three bedrooms at $5,001 across 77, and four bedrooms at $5,984 across 22. Average market time on a two bedroom was 32 days.

Run the arithmetic on the average three family in that census at $1,190,000. Three two bedroom units at $4,014 is $144,504 a year gross. That is a gross rent multiple of 8.2 and a gross yield of about 12.1 percent before a dollar of expenses, taxes or vacancy. Now run the same rent through a condo. The average condo in that ring closed at $968,329 and rents as a two bedroom for the same $4,014, which is $48,168 a year, a gross yield of 5.0 percent.

The triple-decker collects roughly two and a half times the gross yield of the condo next door, and you buy the square footage for 53 cents on the dollar. That spread is not new and it is not unique to South Boston. What is unusual is finding it this close to a project of this size, this late in the cycle.

A permit is not a building, and HYM can prove it

Now the part I would want to hear if I were on the other side of this conversation.

The local partner on Dot Ave is HYM, and HYM has the most instructive track record in Boston on exactly this question. In September 2020 the BPDA approved HYM’s Suffolk Downs plan in East Boston and Revere: 10,000 units on 161 acres, an $8 billion project the developer called the largest single creation of housing in Boston history.

Six years later, one residential building has opened. Amaya delivered in 2024 with 475 units. Portico, at 473 units, and The Arden, at 243, are in construction or about to start. Call it 475 units delivered out of 10,000 approved in six years, which is under 5 percent.

I want to be fair to O’Brien here, because the comparison is not perfectly clean. Suffolk Downs is 161 acres with roads, utilities and a whole district to build from scratch. Dot Ave is six compact acres with existing street frontage on two Red Line stops, and four buildings is a far simpler problem than a new neighborhood. It should move faster. But “faster than Suffolk Downs” is a low bar, and the honest shape of the schedule is this. Filing September 2026. Permits targeted early 2027, which almost never happens on the first target. First shovel after that, then roughly three years of vertical construction on a 200 foot building. The earliest anyone is signing a lease in tower one is around 2030, and the fourth tower is a mid-2030s question that depends on a capital market nobody can forecast.

If you are waiting for 1,945 units to soften South Boston rents, you are waiting into the next decade. What arrives long before the supply does is the construction. Old Colony and Dorchester Avenue are going to be a staging area, and the owners closest to the site absorb five or more years of trucks, noise and torn up street before they see a lobby.

Two live variables that could still move the unit count

Neither of these is settled, and both are worth watching if you own nearby.

The first is the affordability requirement. Boston’s Inclusionary Zoning rules, which replaced the old IDP for filings after October 1, 2024, require 17 percent of units at income restricted rents plus another 3 percent set aside for voucher holders, so 20 percent effective. On 1,945 units that is roughly 389 income restricted homes. In February 2026 District 2 Councilor Ed Flynn, who represents South Boston, filed a resolution asking for a temporary rollback to 13 percent. It was referred to the Committee on Housing and Community Development on February 25 and has not moved since. If it ever does, the difference on this project alone is about 136 income restricted units.

The second is the zoning itself. The 2016 corridor plan was approved by the BPDA board in December 2016, and the rezoning meant to implement it has still not been finished ten years later. That is the quiet reason a six acre assembly could sit idle through the biggest housing shortage in the city’s modern history and then go to a foreclosure auction with one bidder. Entitlement risk on this corridor is real, and it is priced into everything around it, including your building.

If you already own near Andrew Square

Do not reprice your house off a press release. A Letter of Intent is the first sheet of paper in Article 80, not an approval, and the last owner of this site held it for a decade and never filed anything at all.

What I would actually expect is gradual. The corridor finally has an anchor project with capital behind it and a clean title, which is more than it has had since 2016, and that supports steady appreciation pressure over a long horizon rather than a step change this year. Against that, if you are in the closest quarter mile, price in years of construction disruption you will live through before any of the upside shows up. If your plan was to sell in the next two years anyway, the filing is a mild positive for your story and I would use it in the listing. If your plan is to sell in 2032, you will likely do better, and you will earn it.

If you want a number on your specific building rather than a neighborhood average, our home value tool is the fastest starting point, and I am happy to run the block level comps behind it.

If you are buying or investing

My read is that the actionable trade here is the existing stock, not the new stock, and the data above is why.

The closest ring to this project currently prices at a 24 percent per foot discount to housing a mile away, and the small multifamily inside it prices at 53 cents on the condo dollar while collecting two and a half times the gross yield. You are buying the discount that the warehouses and the self-storage building created, in the one scenario where somebody has filed plans and put real money behind removing them. That is a cleaner risk than paying a premium for a 2030s delivery, and it is available today at ordinary Southie pricing.

Four things I would underwrite before writing an offer. Buy on the rent roll you can verify, not on the pipeline, because the 12.1 percent gross yield is what pays you while you wait and the entitlement is a free option on top. Walk the specific block at 7 a.m. and understand which side of the construction staging you will be on for five years. Read the 2016 corridor plan and check what the proposed zoning does to your parcel, since the rezoning is unfinished and your lot may be in it. And assume nothing about 2031 rents, because 1,945 units landing at once is the one scenario that could genuinely soften the submarket you just bought into.

The Gillette purchase in Fort Point in July and this filing in September are the second and third large outside checks written into South Boston this year. Institutional money is converging on the neighborhood from more than one direction, which is usually the part people notice last and the part that matters most.

If you own within a half mile of 314 Dorchester Avenue, or you are looking at a two or three family anywhere on the corridor, send me the address. I will pull the closed comps and the lease history on your block and tell you honestly whether the filing changes your number. Reach out anytime at bmnboston.com/contact or 617-955-2224.

Sources

• Boston Globe, “South Boston housing proposal among largest new projects in years,” September 10, 2026
• Boston Globe, “$75 million deal clears path for development of stalled Southie real estate project,” March 17, 2026
• Banker & Tradesman, “Lender Submits $75M Bid for Dot Ave. Block”
• Universal Hub, “Suffolk Downs developer teams up with New York firms for a six-acre project along Dorchester Avenue”
• Bisnow, “Auction Clears Path For Development On Contested Dorchester Avenue Properties,” March 23, 2026
• Boston Planning Department, PLAN: South Boston Dorchester Avenue
• Boston Globe, “Gillette buys back parking lots for nearly $20 million more than its 2021 sale price,” August 18, 2026
• Bisnow, “Boston Approves $8B Suffolk Downs Development,” September 2020
• HYM Investment Group, Suffolk Downs project page
• City of Boston, Inclusionary Development Policy update
• Boston City Council, File 2026-0432, resolution on a temporary IDP rollback to 13 percent
• MBTA V3 API, Andrew and Broadway station coordinates
• MLS PIN closed sale and closed lease records, Suffolk County, January 1 to September 15, 2026. Distance, per acre and yield analysis mine.

$283,019 an acre. $285,326 an acre. $294,118 an acre. Three pieces of Concord land, three recorded sales, all within four percent of each other on a per-acre basis. Two of those sellers owned what they sold. The one in the middle did not.

That middle number is 22-B Mattison Drive, the 1.84-acre lot that federal prosecutors say was stolen from its owners and sold to a developer for $525,000 in May 2024. Nearly every account of the case has described the sale as happening at half of what the land was worth, which is a reasonable way to describe the loss. It is not a good description of what the transaction looked like from inside the deal. I pulled the MLS PIN record for every closed land sale in Concord going back to 2019, and against that record the fraudulent sale is unremarkable. It priced itself into the market, not under it.

That is the part of this case worth your time. Deed fraud gets written up as an internet scam that happens to careless people. This one was not. It ran for seven months through a licensed Massachusetts agent, a live MLS listing, two attorneys and a real closing, and the professionals in that chain were the actual target. If you own land, a rental you rarely visit, or anything you inherited and have not thought about in a while, the exposure is real. If you practice in this business, the checklist you are probably relying on would have caught about half of this.

What happened on Mattison Drive

Omar and Halla Jaraki bought a vacant lot in the Mattison Farms subdivision in Concord in 1991 and later moved to South Carolina. They kept paying the taxes. They never listed it.

According to the federal charging documents and the civil complaint their attorney filed in Middlesex Superior Court, someone set up an email account in the owner’s name in August 2023 and contacted a Massachusetts real estate agent about selling the land. A listing agreement was signed electronically in October. The lot went live on MLS PIN on October 13, 2023 at $699,900. It was reduced to $599,900. A local developer offered $525,000 and the sale closed in May 2024, with the seller side executed under a power of attorney the owners never granted.

The impersonators used a counterfeit South Carolina driver’s license and a counterfeit United States passport in the owners’ names, communicated through Gmail and Google Voice accounts, and, per court filings reported by Boston.com, connected to those accounts from Nigerian IP addresses. The proceeds check, a little over $496,000, was routed to a UPS Store in the Philadelphia area and deposited into a brokerage account. The money was gone within weeks.

The Jarakis found out in August 2024, when they called Concord Town Hall looking for a tax bill that had stopped arriving. By then the lot had been cleared, a foundation poured and a house framed, financed by a $1.8 million construction loan recorded against land the developer had bought in good faith from a person who did not exist. Construction is now paused under a court injunction while the case is litigated.

The federal charges, and what a complaint is not

On July 23, 2026, the U.S. Attorney’s Office for the District of Massachusetts announced charges against three men, arrested a week earlier on July 16:

• Moshe Levi, 57, of Carrollton, Texas, charged with wire fraud conspiracy and money laundering conspiracy.
• Kyon James, 44, of Middleboro, Massachusetts, charged with money laundering conspiracy.
• Bradley Beauge, 41, of Somerset, New Jersey, charged with money laundering conspiracy.

Prosecutors allege the scheme identified vacant, unencumbered parcels in Massachusetts, Georgia, Indiana and Tennessee whose owners lived out of state, then impersonated those owners well enough to get real estate professionals to list, negotiate and deed the properties to buyers who had no idea. Across the four states the group allegedly shared or attempted to share in roughly $1.5 million between June 2023 and June 2024. The other targets reportedly included land owned by two lawyers in Georgia and a lottery winner’s parcel in Indiana.

One distinction that most of the coverage has blurred, and it changes how you should read the case. These men were charged by criminal complaint, not by grand jury indictment. A complaint is a sworn application supported by probable cause, filed by prosecutors, and it is how a federal case starts when agents need to arrest quickly. An indictment from a grand jury normally follows. All three are presumed innocent, and Levi is the only one charged with the fraud itself. James and Beauge are charged with moving the money, which is a different allegation.

The price was not the tell

Here is where the public record and the MLS record part company.

The lot has been described as selling for about half its value, anchored to an estimated market value of roughly $1 million. That estimate is defensible as a builder’s residual. Finished houses on identical 1.84-acre Mattison Drive lots have closed at $1,665,625 in 2016, $2,067,500 in 2018, $2,950,000 in 2022, $2,400,000 in 2025 and $2,740,000 in May 2026. Back a $2.4 million to $2.9 million finished product into land and construction cost and you land near seven figures for the dirt.

But nobody at that closing table was looking at a residual. They were looking at land comps. I queried MLS PIN directly for every closed land sale in Concord from January 2019 forward. Seventeen usable sales, and here is where the fraudulent one falls.

The stolen lot ranks fifth of seventeen. It sits between a 6.36-acre parcel on Musterfield Road at $283,019 an acre and a 1.87-acre parcel on Annursnac Hill Road at $294,118 an acre.

Run the tighter comparison and it gets worse. Restrict to Concord land sales between 1.8 and 2.2 acres, the closest thing to a true comparable, and the most recent one before this listing went live was 53A Annursnac Hill Road: 1.87 acres, closed July 2023 at $550,000. Three months later the Mattison lot listed at $699,900, which was 27 percent above that comp. It closed ten months after it at $525,000, within five percent of it.

The fraud did not undercut the market. It listed high, sat, cut once, and closed slightly under the last comparable sale. That is what an ordinary land listing does.

Why land is the asset class this scheme picks

Land gets targeted because it is vacant and usually unmortgaged. That is the standard explanation and it is true. The part that gets left out is that land is also the only residential asset class where a steep discount reads as normal.

I ran every Massachusetts closed sale in MLS PIN from September 2024 through September 2026 and compared how often each property type closes far below its original asking price.

Take those two facts together. A house that closes at 75 percent of its original ask is in the bottom one percent of Massachusetts sales and somebody asks why. A piece of land that does the same thing is in a group of roughly 190 sales a year, and nobody asks anything. Sitting on the market for seven months is not a warning sign in land either. It is the average.

The people who ran this understood that. They did not need to price the lot to move fast. They needed to price it so that no professional in the chain would feel the need to make a phone call.

The industry’s own red-flag list caught half of it

The American Land Title Association publishes a one-page seller impersonation fraud alert that is genuinely good and widely circulated to agents and closing attorneys. It lists four red flags. Score the Concord transaction against it.

Two out of four. And the two it missed are exactly the two that make a transaction feel wrong. Vacant land with an out-of-state owner describes a large share of legitimate land listings in Middlesex County, and an agent who treated it as a fraud signal on its own would flag most of their pipeline. The signals with teeth, a bargain price and a rushed timeline, were both absent here because the people running it were patient.

This is my argument in one line. A red-flag checklist that depends on a deal feeling wrong will not stop a scheme designed by someone who has read the checklist.

The paperwork was the tell

What would have stopped it was sitting in the file the whole time. The civil complaint, as laid out by the owners’ attorney Richard Vetstein on the Massachusetts Real Estate Law Blog, catalogs what the documents actually looked like:

• A South Carolina driver’s license and a United States passport bearing the identical photograph. Two separate agencies issue those credentials from two separate photo captures. They cannot match.
• A driver’s license missing standard security features and holograms.
• A notary clause left blank in parts, with the Texas county misspelled as “Tourrant” rather than Tarrant.
• A notary signature and stamp that appear to have been lifted off another document and dropped in with a PDF editor.
• Instructions to mail the deed and power of attorney to an apartment in Dallas, Texas, while the seller’s identification said South Carolina.
• Instructions to send a roughly $500,000 proceeds check to a UPS Store in Philadelphia.
• The real owners contacting Concord’s Natural Resources Director to say they had not listed anything.

None of that requires a forensic document examiner. A paralegal with an hour could have called the Tarrant County clerk to confirm the notary commission, or noticed that the license and passport photos were the same file. A South Carolina seller directing original documents to a Dallas apartment and proceeds to a Philadelphia mailbox is three inconsistent states in one transaction.

The Jarakis are suing the buyers, the developer and the attorneys for quiet title, trespass, civil conspiracy and negligence. I am not going to name the agent or the brokerage. They have not been charged with anything, the listing office was a legitimate Massachusetts firm about thirty miles south of the property, and my read is that they were worked the same way the attorneys were. That is the point. Competent people got beaten by a file that looked ordinary.

What owners can actually do, and what it will not do

The free fix is a Registry of Deeds alert. Massachusetts registries run a consumer notification service that emails you whenever a document is recorded against your name, address, or book and page. It costs nothing and takes about five minutes. I want every client who owns land or a paid-off rental signed up for it.

Now the part nobody tells you. This is not one system. Eleven registries share the statewide portal at cns.masslandrecords.com, including Middlesex South, Middlesex North, Suffolk and Worcester. The rest run their own separate sign-ups on their own websites. Norfolk, Plymouth, Barnstable, Essex North, Essex South, Bristol and Fall River each have a different page and a different form. You have to find yours.

Two limits you should know before you treat this as solved. The statewide service caps you at three properties per account and explicitly excludes commercial property. If you own six rentals through an LLC, this tool does not cover you, and that gap falls hardest on exactly the small landlords and investors who are most exposed.

The bigger limit is what the alert is. It fires when a document is recorded, which happens at or just after the closing. The registry’s own disclaimer says plainly that it has a legal duty to record documents even when those documents later turn out to be fraudulent. It will record a forged deed and then email you about it.

Apply that to Concord. The deed recorded in May 2024. An alert would have reached the Jarakis that week instead of in August, when they called about a missing tax bill. Three months of foundation work and a $1.8 million construction loan would not have happened. But the $496,000 was already gone. A registry alert is a smoke alarm, not a lock. Sign up anyway, because finding out in seventy-two hours instead of three months is the difference between a recoverable mess and a framed house on your land.

Who in Greater Boston is exposed

The profile prosecutors describe is specific: vacant, unencumbered, owned by someone out of state. Around here that describes more property than people assume.

One hundred sixty-five land parcels closed in Middlesex County alone in the last two years, at an average price of $831,953. There are currently about 1,300 active land listings across Massachusetts. Every one of those has an owner, and behind them sit thousands of parcels nobody has listed. If you are in one of these situations, you are in the target set:

• You own land in Massachusetts and live somewhere else. This is the single strongest predictor.
• You inherited a parcel or a house and have not transferred, sold or visited it.
• You own a rental free and clear and use a property manager, so no lender and no occupant would notice a title change.
• You hold a second home on the Cape or the islands that sits empty most of the year.
• Your property is held in a trust or an LLC whose registered address is a lawyer’s office rather than a place you read mail.

The common thread is not carelessness. It is distance. Every one of these owners is a person who would not notice a stranger walking the property, and whose only monthly contact with the asset is a tax bill. When the tax bill stops coming, that is the signal, and by then you are already months behind. If you are unsure what your Massachusetts holdings currently show on record, our property value tool is a reasonable starting point, and you can always reach out and I will pull the registry record with you.

What I want from the professionals in the chain

Owners can do one useful thing here. The professionals can do five, and I think we have to. Here is what we run on our side and what I would ask any agent or closing attorney you hire to commit to.

1. Treat “here is my new contact information” as a red flag, not a housekeeping note. Mid-transaction, a changed phone number or email address gets verified by calling the last known good number, not the new one just provided. This costs nothing and it breaks the single mechanism every impersonation scheme depends on.
2. Mail a confirmation letter to the address on the assessor’s record. The title industry already requires this on vacant land in many shops, and it is the step that would have ended the Concord fraud on day one. Concord’s assessor had the Jarakis’ South Carolina address. A letter to it would have reached them in October 2023, seven months before the closing.
3. Verify the notary, not the notarization. Commission records are public. If a deed is notarized in a county you have never heard of, call that county. The Concord file misspelled the county name on the stamp.
4. Do a live video call and ask something the file does not contain. A photo matching an uploaded ID proves the person holds a picture. Ask what is across the street from the lot, when they bought it, what they paid. An impersonator working from public records will get the purchase price right and the view wrong.
5. Look at where the money and the originals are going. Proceeds to a retail mailbox, originals to a third state, and a seller who never asks about net sheet detail are three things worth pausing over together, even when any one of them has an innocent explanation.

On title insurance, get the coverage right. A standard ALTA Owner’s Policy protects a buyer against forgery that happened before they bought, which is what the Concord developer is relying on now. The ALTA Homeowner’s Policy adds post-policy forgery coverage, protecting you against someone fraudulently transferring your property after you own it, and there is an endorsement, ALTA 49.1, that extends similar protection to an existing owner who has paid off the mortgage. Read the fine print on the Homeowner’s Policy though. It is available only for improved one to four family residential property, which means the product with the best post-closing forgery protection does not cover vacant land. The asset class the scheme targets is the one the coverage excludes.

The honest summary

Massachusetts owners reported $46,269,818 in real estate fraud losses to the FBI between 2019 and 2023 across 1,576 victims, the largest total in the FBI’s Boston Division, which also covers Maine, New Hampshire and Rhode Island. Regionally that is 2,301 victims and more than $61.5 million.

One caveat on those numbers, because the coverage usually drops it. The FBI says outright that its complaint center does not keep statistics specifically for deed fraud. Those figures are the whole real estate fraud category, which includes wire fraud at closing and rental scams. Deed fraud is a subset. The Bureau also says reported losses are almost certainly low, because many owners do not know where to report it or have not discovered the theft yet. The real number is unknowable, and both of those adjustments point in opposite directions, so treat $46 million as the scale of the problem rather than a precise count of stolen deeds.

What is not ambiguous is how this one worked. It did not beat a homeowner. It beat a licensed agent, a live MLS listing, two attorneys and a closing, using a price that matched the comps and a timeline that matched the average. Reading your deed carefully was never going to catch it.

Sign up for your registry’s alert this week. It takes five minutes and it is free. Then, the next time you sell anything, ask your attorney and your agent what they do to confirm a seller is who they say they are when nobody in the deal has ever met them in person. If the answer is that they look at a copy of a driver’s license, you already know how that goes.

We work with sellers, landlords and land owners across Greater Boston, and title questions like this one come up more often than they used to. If you own something you have not looked at in a while and want a second set of eyes on the record, get in touch. No charge and no pitch.

Sources

• U.S. Attorney’s Office, District of Massachusetts: Three Men Charged in $1.5 Million Deed Fraud and Money Laundering Schemes, July 23, 2026
• FBI Boston: Quit Claim Deed Fraud is on the Rise, April 1, 2025
• American Land Title Association: FBI Boston Issues Quit Claim Deed Fraud Warning, with the state by state loss breakdown
• ALTA: Combating Seller Impersonation Fraud and Benefits of ALTA’s Homeowner’s Policy of Title Insurance
• Boston.com: New court docs reveal how scammers allegedly impersonated Concord landowners, July 24, 2026
• CBS Boston: 3 charged after Concord, Massachusetts land stolen, sold to unsuspecting buyer in 2024
• Massachusetts Real Estate Law Blog (Richard D. Vetstein): the Concord title theft case and the civil complaint
• Secretary of the Commonwealth: Protect Your Home, registry consumer alert sign-ups by county
• Massachusetts Consumer Notification Service (statewide portal, participating registries and terms)
• Fall River Reporter: full text of the Department of Justice announcement, July 23, 2026
• MLS PIN closed sale records for Concord land and Mattison Drive single family homes, queried directly September 15, 2026 (proprietary)

The largest data center in Massachusetts is the Markley Group building in Lowell. It runs about 352,000 square feet. Right now there are roughly 6.19 million square feet of available office space in the Route 128 West submarket alone, the stretch that runs through Waltham, Newton and Needham. You could drop seventeen copies of the biggest data center in the state into the empty space in one suburban office market and still have room left over.

I keep that arithmetic handy because of a question I have been getting since last winter, usually from owners in Waltham, Burlington and Marlborough, and usually some version of the same worry. Data centers are eating housing land in Virginia. Are they coming for ours?

The short answer is no, and the longer answer is more useful than the short one. The fear arrived here measured in the wrong unit. What data centers actually compete for is megawatts and raw acreage with power already at the fence line. What housing conversions compete for is square footage and a zoning vote. Those are two different auctions, and in Greater Boston the bidders almost never meet. Where they do meet is on one line, and it is the line at the bottom of your utility bill.

Governor Healey’s Executive Order 658, signed on September 8, mostly settles the land question. It does not settle the bill.

What Executive Order 658 actually does

The order is narrower than the headlines suggested, and the narrowness is the point. It bars state agencies from permitting a data center with peak electricity demand above 25 megawatts unless the developer has first signed a community benefits agreement with the host city or town. Healey put it about as plainly as a governor can: unless a community says yes to a data center, the state is saying no.

Three other pieces matter. Any project over that threshold has to procure enough new clean electricity to match what it consumes. If it cannot, it pays into a newly created Ratepayer Protection Fund, and MassDEP has until December 31 to stand up the alternative compliance payment mechanism that collects it. And non-disclosure agreements between state agencies and data center developers are now prohibited, which is the provision I would have bet against and am glad to have lost.

This did not come out of nowhere. In late June the administration halted new applications for the qualified data center sales tax exemption that Massachusetts created in 2024, the one that offered a 20 year break to projects meeting a 100 job threshold. So inside of ten weeks the state took away the tax incentive and added a local veto. A separate ballot initiative filed for November 2028 would go further and require two thirds approval from local voters, plus a utility cost shielding contract and a water supply certification.

One number tells you how selective that 25 megawatt line is. Massachusetts today has about 168 megawatts of data center capacity spread across 58 facilities, which works out to an average facility of roughly 2.9 megawatts. The permit gate sits nearly nine times above the size of the typical data center that already exists here. Everything Massachusetts has built to date would pass under the bar untouched. EO 658 is not a data center ban. It is a hyperscale filter.

Massachusetts is not Loudoun County, and the map proves it

The imported version of this story comes from Northern Virginia, and the details there are genuinely alarming. Land that can take a data center trades at multiples of what a homebuilder can pay, and the National Association of Home Builders has documented builders losing sites outright. Loudoun County’s own board chair has described affordable housing developers getting outbid, and roughly 700 residential lots there left the pipeline when a seller chose data center rezoning over finishing a deal with a builder.

That is a real problem. It is also a land market problem, and it requires cheap land, available power and permissive zoning in the same place at the same time. Massachusetts has none of those three in the inner suburbs.

Look at where the fights are actually happening here. Holyoke banned new data centers outright in June. Westfield passed a 365 day moratorium on July 6 after residents looked at a proposal that would draw 274 megawatts, more than one and a half times the entire existing capacity of the state, and concluded they were being asked to host a city sized power plant. Lowell, which already hosts the largest facility in Massachusetts, adopted a moratorium of its own. Shutesbury, Mansfield and Everett have moved too, and Plymouth’s Select Board has recommended a one year zoning moratorium to Fall Town Meeting that would run from October 17. Hunneman counts at least 16 Massachusetts communities with a proposed or approved moratorium or ban.

Read that list again. Western Massachusetts and gateway cities. Not one town on the Route 128 belt. The pressure is landing where land is cheap and transmission capacity exists, which is exactly not Waltham.

The empty space on 128 is square feet, not acres

Here is the part most of the national coverage misses. Greater Boston’s commercial distress is not vacant land. It is vacant buildings, and specifically 1980s and 1990s office and lab buildings sitting on parking lots inside built out suburbs.

Route 128 West posted the highest availability rate of any Route 128 submarket at 26.6 percent, with sublease space at 5.5 percent. Route 128 North gave back 424,907 square feet of net absorption in a single quarter. Route 495 West has shed 297,363 square feet over twelve months and is asking $25.89 a foot, which is under half of what Back Bay commands.

Add the three belts together and you get roughly 16.7 million square feet of available space across Waltham, Burlington, Marlborough and their neighbors. That is about 47 times the footprint of the largest data center in Massachusetts. Square footage was never the scarce input in this market. Nobody is going to outbid a housing developer for an empty Winter Street office building by promising to fill it with servers, because filling it with servers is not the constraint. Getting 25 megawatts to the property line is.

The buildings landlords were pitching both ways

For about two years, the owners of this inventory have been running a dual track pitch. Every half empty suburban office park was simultaneously a candidate for an apartment conversion and a candidate for some form of digital infrastructure tenancy. I sat in on enough of those conversations to know the data center leg was always the thinner one, and EO 658 just cut it thinner.

Stack up what a hyperscale developer weighs when siting in Massachusetts. Commercial electricity here runs 24.79 cents per kilowatt hour year to date through June, third highest in the country behind Hawaii and California and 1.79 times the national average of 13.86 cents, according to EIA Electric Power Monthly Table 5.6.B. The 2024 sales tax exemption is paused. Any project over 25 megawatts now needs a signed community benefits agreement in a state where 16 communities have already moved to block them. And it has to bring its own clean power or pay a fee.

Against that, the conversion path got easier. On July 8 the House passed H.5562, the economic development bill that extends Boston’s office to residential conversion playbook to all 351 cities and towns and puts $50 million behind adoption. Watertown showed what it looks like in practice when a vacant office became 285 apartments. And in Waltham the City Council overrode Mayor McCarthy’s veto by a 12 to 1 vote on August 3, clearing BXP to pursue up to 1,200 homes at Bay Colony off Winter Street. Add the 300 to 400 units proposed off Jones Road and 600 more at the former Polaroid campus and Waltham has roughly 2,100 units of housing entitled or in motion on land that was office park a decade ago.

I want to be careful about the causal claim here, because I think a lot of coverage will overstate it. EO 658 did not convert a single building. What it did was remove the alternative bid that landlords could point to in a negotiation. That matters more than it sounds like it should.

What 2,100 units means against what actually trades

This is where I can go somewhere the national write ups cannot. We pull closed sale data straight from MLS PIN, so instead of guessing what the conversion pipeline means for the inner suburbs, I ran the actual condo closings in five Route 128 and Mass Pike towns for the first eight months of 2026.

Source: MLS PIN closed sales, condominium property types, January 1 through August 31, 2026. BMN Boston analysis.

Two things jump out. The first is Marlborough. Seventy condos closed there in eight months and exactly one of them was built in 2015 or later. Across all five towns, only 71 of 434 condo sales, about 16 percent, involved anything built in the last decade. These are not markets with a new construction problem. They are markets with a new construction absence.

The second is scale. Waltham closed 145 condos in eight months, an annual run rate near 218. The 2,100 units now in motion there equal roughly nine and a half years of the town’s entire condo transaction volume. Bay Colony alone is about five and a half years of it. Even accounting for the fact that much of this will deliver as rental rather than for sale, and that these projects take years, that is a genuine supply event for a town that has been starved of new product.

If you are buying a 1985 vintage two bedroom in Waltham today with the intention of selling in 2032, this is the number I would want you looking at. Not data centers.

The real new line item is your electric bill

Now the part I actually want owners worrying about, because it is the one thing in this story that can reach your monthly budget.

Massachusetts utilities have fielded interconnection requests from proposed data centers totaling roughly 2 gigawatts. That is about 2,000 megawatts against an existing statewide fleet of 168 megawatts, so the queue is nearly twelve times everything built here to date. The Boston Globe reported that serving it would require on the order of $1 billion in grid upgrades, and that at full draw it would equal the annual consumption of about 2.4 million Massachusetts households.

Grid upgrades get recovered through rates. Who pays which share is decided by cost allocation rules, and Massachusetts has not finished writing its own. Twenty three states have approved at least one large load tariff. Massachusetts is not yet among them. The House version of the energy affordability bill would direct the DPU to create a data center specific tariff, and the two chambers are in conference. EO 658’s Ratepayer Protection Fund is a real mechanism, but MassDEP still has to build it by December 31.

Meanwhile, look at what rates are already doing here, because the pattern is not what most people assume.

Massachusetts is the odd one out. Nationally, residential power got 7.4 percent more expensive over the last year and commercial got 6.1 percent more expensive. Here, commercial jumped 7.7 percent while residential actually came down slightly. Massachusetts holds the third highest price in the country in both categories, behind Hawaii and California, but right now the increase is landing on the business side of the meter.

That divergence cuts two ways, and both are worth understanding. It is a large part of why hyperscale developers were never going to love Massachusetts. It is also the leading edge of exactly the pressure homeowners should watch, because a residential rate that is merely flat at 30.06 cents while the commercial rate climbs is not a durable condition. If 2 gigawatts of new load arrives before the cost allocation rules are finished, the question of who pays for those wires gets answered by default rather than by design.

So put it in the budget conversation. When I sit with a buyer and we walk through carrying costs, electricity has historically been the line nobody models. Given where Massachusetts sits nationally, I now treat it the way we treat property taxes and insurance. It is a real, variable, policy sensitive number, not a rounding error.

Waltham already showed you the second bill

There is a companion cost here that gets less attention, and Waltham is the cleanest example of it in Greater Boston.

Close to half of Waltham’s property tax base comes from commercial real estate, a share exceeded only by Boston and Cambridge. When Route 128 West office values fall, and they have fallen hard, that burden shifts. Waltham raised its residential tax rate 5 percent this year to cover lost commercial revenue. Two nearby trades show how far values have moved. Stony Brook Office Park in Waltham sold in June at $94 per square foot, about 68 percent below its prior sale. Northeastern bought the Burlington BioCenter at $301 per square foot, also down roughly 68 percent from 2022.

This is the honest argument for conversions that does not get made enough. Residential redevelopment does not replace the commercial tax revenue a full office park generated. But a vacant building generates less than either, and it keeps declining. Mayor McCarthy’s objections to Bay Colony about traffic, setbacks and parking precedent were not unreasonable on the merits. The Council still had to weigh them against a tax base that is actively eroding, and voted 12 to 1.

For an owner in Waltham, Burlington or Marlborough, the practical read is that your assessment and your rate are both in motion, in opposite directions, for reasons that have nothing to do with your house.

What I tell buyers and investors right now

Five things, in the order I would act on them.

1. Stop pricing data center land risk into inner suburb decisions. There is no active hyperscale proposal on the 128 belt, the economics point away from Massachusetts, and EO 658 added a local veto on top. If you have been hesitating on a Waltham or Burlington purchase over this, it is not the reason to hesitate.

2. Treat the conversion pipeline as real supply, with a long fuse. Waltham’s 2,100 units are entitled or moving, not built. Rezoning to occupancy on a project like Bay Colony realistically runs five to eight years. If your hold period ends inside that window you will feel the construction and not the competition. If it ends after, plan for both.

3. Underwrite electricity like a real line item. At 30.06 cents residential, Massachusetts is 66 percent above the national average. For a small multifamily owner with common area load, that is not trivial, and the cost allocation fight over data center interconnection is unresolved. Ask for twelve months of actual bills, not an estimate.

4. In Marlborough and Framingham, understand what you are buying. One post 2015 condo sale in Marlborough in eight months means you are almost certainly buying older product at $261 per square foot. That can be a good trade. It is a different trade than buying new.

5. Watch three dates. MassDEP’s December 31 deadline for the alternative compliance payment mechanism. The energy affordability bill coming out of conference before year end. And whether the 2028 ballot initiative clears signatures. Those decide whether the ratepayer protections in EO 658 have teeth.

The bottom line

The land grab story is real, it is just somebody else’s. Massachusetts responded to the data center boom by making the biggest projects harder to build and by taking away the tax break, and the towns most exposed had already moved on their own. The result is that the vacant office and lab inventory along Route 128 and the Mass Pike has one fewer competing use than it did in June, at exactly the moment the state handed all 351 cities and towns a conversion tool.

That is good news if you are counting on that pipeline to add homes in the inner suburbs. It just shows up slowly, in units delivered in the 2030s.

The fast moving number is the one on your utility bill, and it is moving for reasons that are still being decided at the DPU and on Beacon Hill rather than in any zoning hearing. I would rather you track that than a data center that is not coming.

If you own along the 128 belt and want to know what the conversion pipeline does to your specific block, or you are weighing a purchase in Waltham, Burlington or Marlborough and want the closed sale data rather than the headline, reach out. You can also check what your home is worth today as a starting point.

A client shopping in Newtonville forwarded me the state’s press release the morning it came out, with one line above it: “Does this mean I should wait?”

It is a reasonable question. The number on top of that release is enormous. On September 3, 2026, Governor Healey announced $278 million in state and federal financing and tax credits for 29 housing developments in 20 communities, which the administration called the largest single round of awards from this program in more than a decade. That is not spin. It is a genuinely big round, and it landed directly in the towns we work in every day rather than only in the Gateway Cities that usually absorb these awards.

So I ran the award list through the only filter that mattered to my client. Of the 578 homes funded across Newton, Cambridge, Brookline, Watertown, Needham and Lexington, how many could he write an offer on?

Zero. Not one.

That is the whole tension in this announcement, and it is why I think it deserves a real explanation instead of a press-release rewrite. This money is going to build actual buildings on named parcels a short walk from homes my clients are touring. Cranes, trucks, road plates, two-year build-outs. It is also going to do almost nothing to the price of the house they are competing for. Both things are true, and the people selling you a story about this round usually pick one and drop the other.

What the state actually funded on September 3

The $278 million is not one pot. It breaks into $154.7 million in direct subsidy financing from the Executive Office of Housing and Livable Communities, $61.8 million in state low-income housing tax credits, $47.9 million in federal 4% credits and $13.2 million in federal 9% credits. The credits are the part that does the heavy lifting. Developers sell them to investors, and the state expects the round to pull in close to $500 million in private equity on top of the public money.

That structure tells you what kind of housing comes out the other end. Low-income housing tax credits carry deed restrictions that run for decades. A building financed this way is not a building that quietly converts to market-rate condos in year seven. It is locked, by design, and that is the point of the program.

The round covers Athol, Boston, Brockton, Brookline, Cambridge, Chelmsford, Chicopee, Erving, Fitchburg, Ipswich, Lexington, Lynn, Needham, Newburyport, Newton, Northampton, Springfield, Swampscott, Watertown and Worcester. Six of those twenty are inner suburbs where we list and sell constantly.

Where the money landed in the six inner suburbs

Here is the part worth knowing by address, because these are real parcels with real neighbors.

In Newton, the award goes to 793 Washington Street in Newtonville, where Beacon Communities plans a five-story building with a set-back sixth floor and ground-floor retail. The site today holds a two-story commercial block. Building it means losing Rice Valley, Che! Empanada, Centre Pieces Design and Bikofsky Insurance from that stretch. It sits directly across from the Newtonville commuter rail stop, and the current plan carries no resident parking. City Councilor Tarik Lucas has already pushed on both the displaced businesses and the parking question. Construction is targeted for spring or summer of 2027.

Worth noting, because people conflate the two: this is not the West Newton Armory. That project, 43 units at 1135 Washington Street, was funded years ago and is finishing construction now, with its lottery already run. Newton has two affordable projects on the same road, at different stages, and only one of them is in this round.

In Cambridge, the city’s share is about $18.2 million across two sites. Homeowner’s Rehab, Inc. is building 95 apartments at 28-30 Wendell Street in the Baldwin neighborhood, on land it bought from Lesley University in late 2023, with 55 homes for families and 40 for older residents. The Cambridge Housing Authority is starting its Corcoran Park rebuild in Strawberry Hill, replacing the first 29 of the 1951 complex’s 153 units with 67 new ones in a 46-unit elevator building and a 21-unit townhomes-over-flats building. The full Corcoran Park plan eventually runs to roughly 291 units in all-electric Passive House buildings, and it is also fixing the flooding and soil subsidence that have plagued part of that site. Both Cambridge projects are slated to break ground in 2027.

In Brookline, the Housing Authority is taking down 4-24 Walnut Street, which is 24 one-bedroom units, plus 28-42 Walnut Street, which is eight four-bedroom townhouses, and putting up a 96-unit passive house building in their place. Watertown’s Willow Park is the biggest of the six: the Watertown Housing Authority and Preservation of Affordable Housing are replacing a 60-unit state-aided public housing site on 2.3 acres with 138 homes, 60 of them deeply affordable and 78 restricted at 80% of area median income. Lexington Woods puts 40 homes on a three-acre town-owned parcel at Lowell and North Streets, and LexHAB picked up funding for eight accessory units.

578 funded homes is not 578 added homes

This is where most coverage of a round like this gets sloppy. Four of the eight inner-suburb projects are public housing rebuilds, which means the unit count in the press release is a gross number, not a net one. Corcoran Park takes down 29 to build 67. Brookline takes down 32 to build 96. Watertown takes down 60 to build 138.

Those are still good trades. Replacing aging stock at more than double the density is exactly what should happen on those parcels, and the residents get new construction instead of buildings from 1951. But if you are counting homes added to the region, you have to subtract.

Across those three rebuilds, 301 funded homes net out to 180. Needham’s Linden Terrace is a rebuild too, and its replacement count is not cleanly broken out by sub-phase, so I will not put a false decimal on it. The honest version is this: the six-town figure is lower than the 578 headline by at least 121 homes and probably more. The genuinely additive projects are Wendell Street, 793 Washington, Lexington Woods and the eight LexHAB units. Still real. Just not the headline.

None of it will ever hit the MLS

Here is the fact that answers my client’s question directly. This was a rental round. With one homeownership exception in Worcester, the entire $278 million produces income-restricted apartments, not homes for sale.

For a buyer competing on a Newton colonial or a Brookline two-bedroom condo, that means the competitive set does not change by a single unit. You will not see these addresses in a search. You will not tour one. You cannot bid on one. The pool of houses you are fighting over in spring 2027 is exactly the pool you would have been fighting over if this announcement had never happened.

There is a second-order argument that new affordable rentals relieve pressure on the market by giving some households a place to go. I think that is directionally true and practically invisible at this scale in these towns. A few hundred restricted apartments delivered over three years, in a submarket that closes more than 2,000 homes a year, is not something you will detect in a comp set.

What the resale market in these towns actually looks like right now

We pull our numbers straight from MLS PIN rather than from a national portal’s town page, which lets us cut the data the way it actually matters. Here is every closed residential sale in these six towns from January 1 through September 13, 2026.

Six months of supply is roughly a balanced market. Every one of these towns is running between 2.3 and 3.9, which is a seller’s market by any standard definition, and Needham and Watertown are the tightest of the group. Median market time sits between 56 and 68 days, and in Watertown more than half of all sales closed at or above the original asking price.

Now put the announcement next to that market.

The subsidized production is about 28% the size of a single year of open-market turnover in these towns, it arrives over three or more years, and none of it is for sale. Statewide the proportions are similar. The 2,500 homes in this round are a little over 1% of the 222,000 homes Massachusetts says it needs by 2035.

The Momentum Fund is the program people think this was

I keep seeing this round described as Momentum Fund money. It is not, and the distinction matters more than it sounds.

The Momentum Fund is a separate MassHousing program created by the same Affordable Homes Act. It exists precisely because the deed-restricted tax credit model does not produce middle-income housing. Instead of granting subsidy, it invests state dollars as equity, up to 49% of a project’s total equity, to make mixed-income buildings pencil. That is the one tool in the Affordable Homes Act designed to produce units at rents and prices closer to what an ordinary household pays without a subsidy.

Its most recent round put $15 million into 192 mixed-income rental homes in Milton, Middleton and Newburyport. Not Newton. Not Brookline. Not Cambridge.

So if your read on September 3 was “the state is finally funding housing that middle-income buyers can access in my town,” the honest answer is that the program which would do that is a different, much smaller program, and it did not fund anything in the inner suburbs this round.

Who these apartments are for

None of the above is an argument against the projects. It is worth being precise about who actually gets to live in them, because the phrase “affordable housing” does a lot of imprecise work in neighborhood conversations.

Under the FY2026 HUD limits for the Boston-Cambridge-Quincy area, published by the City of Cambridge Housing Department and effective June 1, 2026, area median income for a four-person household is $164,600. The 60% tier that most tax credit units target is $102,840 for that same household, and $72,000 for a single person. The 80% tier, which covers the 78 workforce units at Willow Park, runs to $137,100 for a household of four.

Those are ordinary working incomes for this region, not an edge case, and that is the point. In Newton, where the 2026 median sale is $1,615,000, a household earning $102,840 is not a buyer at that price under any financing structure that exists. These apartments do work the resale market cannot do, which is the argument for building them. It is also the reason they neither compete with nor relieve the resale market.

If you are selling near one of these sites

This is where the announcement becomes genuinely operational rather than theoretical.

Every one of these projects is a multi-year construction job on a specific parcel. The Cambridge sites break ground in 2027. Newton’s is targeted for spring or summer 2027. If your home is within a block or two, you are looking at staging areas, truck traffic, street work and noise across a meaningful stretch of your ownership.

Three practical points I would make to a seller near one of these addresses:

Timing is a real lever. Listing before a site goes active is a different sale than listing in month fourteen of a build. If you were already planning to sell within two years and you are on Wendell Street, Walnut Street or that block of Washington Street, the front of that window is cleaner.

Disclose what you know, and do not editorialize. Massachusetts is a caveat emptor state for sellers on most conditions, but your agent has a duty not to misrepresent. A permitted project down the street is public record and buyers’ agents will find it. Answer the question factually. Steering a conversation toward who might live there is both wrong and a Fair Housing problem, and I have watched agents talk themselves into trouble on exactly this.

Do not assume the effect is negative. Newton’s 793 Washington replaces a tired two-story commercial strip across from a commuter rail stop with a new building carrying ground-floor retail. Brookline’s and Cambridge’s projects replace 1950s-era buildings with new passive house construction and better open space. I have not seen a well-built affordable development show up as a discount in the comps in these towns, and I would not price a listing down on the assumption that it will.

If you are buying near one of these sites

The due diligence is straightforward, and almost nobody does it.

Pull the project’s page on the town or city website before you write the offer, not after. Cambridge, Newton and Brookline all publish plan sets, hearing calendars and schedules. You want the construction start, the expected duration, and the site plan showing where the staging and access points go. A building going up behind you is a different experience than one going up across a four-lane road.

Watch parking specifically. The 793 Washington plan carries no resident parking and the developer is still working out an arrangement with a neighboring complex. On a Newtonville street where residents already compete for spaces, that is a real quality-of-life variable to price into what you offer.

The transit adjacency cuts the other way and is usually underrated. These sites were picked because they sit near stations and bus lines, which is the same reason the surrounding blocks hold value. Buying near a rail stop in Newtonville or a well-served corner of Cambridge is a durable position regardless of what gets built on the adjacent parcel.

What I tell clients

I am glad this money is going out, and I would rather see it land in Newton and Brookline than watch these towns keep exporting the obligation to Lynn and Brockton. Six inner suburbs taking 578 units in a single round is a real shift in where Massachusetts builds subsidized housing, and it should continue.

What I will not do is let a client mistake it for market relief. The buyer competing for a $1.2 million Newton colonial or a Brookline two-bedroom is competing in a market with 3.1 months of supply and 755 active listings across six towns, and not one of the 578 funded homes changes that number. If you are waiting for this round to loosen your search, you are waiting for something that is not coming.

The things that actually move your position are the ordinary ones. Getting your financing genuinely locked down instead of loosely pre-qualified. Being early on the right street. Knowing which blocks are about to be under construction before you tour them rather than after you close. If you are weighing a sale near one of these sites, start with an honest read on what your home is worth today, then we can talk about whether the front of the construction window matters for your timing. If you want to talk through a specific address, reach out and I will pull the project file and the block’s comps with you.

Sources

• Governor Healey Announces Major Statewide Plan to Build Over 2,500 New Homes, Commonwealth of Massachusetts, September 3, 2026
• Governor Healey Awards Financing to Two Affordable Housing Developments in Cambridge, City of Cambridge Community Development Department
• Two Cambridge Housing Developments Receive Millions in State Funds, The Harvard Crimson, September 9, 2026
• Corcoran Park Modernization, Cambridge Housing Authority
• 28-30 Wendell Street, City of Cambridge Community Development Department
• 100% affordable housing development is moving forward on Washington Street, Fig City News, March 2026
• 793 Washington Street Affordable Housing Development, City of Newton
• Willow Park, Preservation of Affordable Housing
• Willow Park Redevelopment Project, Watertown Housing Authority
• Walnut/High Street Redevelopment, Community Preservation Coalition
• Development, Brookline Housing Authority
• Developer presents 100% affordable housing development design, The Lexington Observer
• Creation of 192 New Mixed-Income Rental Homes through Momentum Fund, Commonwealth of Massachusetts
• A Home For Everyone: Progress toward 222,000, Commonwealth of Massachusetts
• Housing Program Income Limits, effective June 1, 2026, City of Cambridge Housing Department
• MLS PIN closed and active residential listing data for Newton, Cambridge, Brookline, Watertown, Needham and Lexington, pulled September 13, 2026. Analysis by BMN Boston.

On March 4, 2025, Boston 25 News ran a segment on an architect putting up three townhomes in Taunton. The units were penciled at the low $500,000s. The story noted they would hit the market that spring, and that rising material costs could push the number higher before they ever got there.

That was eighteen months ago. Forecasts like that get made constantly and almost never get graded, so I went back and pulled the actual closing record for Taunton new construction out of MLS PIN.

Twenty newly built attached homes closed in Taunton between January 1 and August 31 of this year. The median was $460,000. The two 2026-vintage units on Couch Street closed at $549,900 and $557,400, and one of them went above asking. The low $500,000s call was fair. Costs did push the top of the range higher.

But the price is not the interesting number. The calendar is. The 2025-vintage units in that set took a median of 99 days to sell. The 2026-vintage units took 195.5 days. Same city, same product, same buyers. Roughly double the market time in one year.

That gap is the whole story of Massachusetts housing in 2026, and it is the part almost nobody is pricing correctly.

Massachusetts fixed the permission problem. That was the easy half.

Over the last three years this state did something genuinely hard. The MBTA Communities law forced 177 municipalities to zone for multifamily housing by right near transit. The Affordable Homes Act made accessory dwelling units legal by right in every community that allows single-family homes, effective February 2, 2025. And the FY2027 budget rewrote the variance standard, which quietly reopened the path to the stacked three-family that Greater Boston was built on and then banned.

Those are real wins. I have written about the triple-decker recodification and the by-right ADU rules at length, and I am not walking any of it back.

Here is what all of it has in common. Every one of those laws changed who is allowed to say no. Not one of them changed what the building costs.

And the production numbers show it. Boston Indicators senior fellow Amy Dain published an early look at the MBTA Communities pipeline in January, and found roughly 7,000 units across more than 100 projects, in 34 of the 177 communities. She called the gains real but modest. Only about 30 percent of those units sit within a half mile of a train station. On the ADU side, the state reported 854 approvals in the first six months of 2026 and 2,084 total since the start of 2025.

Those are decent numbers for a policy this young. They are not a wave. And they are being counted in a state that permitted 3,031 single-family homes in the first seven months of this year, against 4,551 in the same stretch of 2017.

The cost side moved, and it moved in the opposite direction

While the permission ledger was being rewritten, the materials ledger was getting repriced by people who have never heard of the MBTA Communities Act.

The shorthand I hear from builders, and the one I had been using myself, lands on two numbers: a 50 percent steel tariff and a 34.5 percent duty on Canadian lumber. The steel number is right. The lumber number is not, and the direction of the error matters, because the real figure today is worse.

The 34.5 percent figure is not invented. It is just early. Commerce issued preliminary results in its seventh administrative review on April 13, 2026, signaling a cut in the antidumping and countervailing rate from 35.16 percent down to 24.83 percent. Add the 10 percent Section 232 duty and you land right around 34.8 percent.

The catch is that preliminary results do not change what anyone pays. Cash deposit rates hold until the final determination, which is expected in October. So 34.5 percent is not the current burden. It is the relief that might arrive next month. Right now the stack is about 45 percent, which is exactly how NAHB describes it.

Why this lands harder on spruce-pine-fir than on anything else

New England framing is built out of spruce-pine-fir. It carries the structural loads our snow and wind codes ask for, and the trees that produce it need a cold climate. David O’Sullivan, who represents the Home Builders and Remodelers Association of Massachusetts at NAHB, put it plainly to GBH in April 2025: “You’re not going to get it to grow in Colorado or Rhode Island.”

That is the part that makes this a Massachusetts story rather than a national one. A builder in Georgia can substitute southern yellow pine. A builder in Dedham cannot substitute a climate. The tariff lands on the specific species our stick framing depends on, and there is no domestic workaround sitting one state over.

It is showing up in the price. The NAHB framing lumber composite sat at $521.35 per thousand board feet on August 28, up 9.3 percent year over year even after three straight weekly declines. Building material costs overall rose 6.7 percent over the last twelve months.

And the number everyone quotes for the damage is already stale. The figure is NAHB’s estimate that tariffs add about $10,900 to a typical home. It is a good number and I use it too. It is also badly out of date in a way that understates the problem.

That figure comes from the April 2025 NAHB/Wells Fargo builder survey. Look at the timeline. The Section 232 tariffs on timber and lumber did not take effect until October 14, 2025. Cabinets went from 25 to 50 percent on January 1, 2026. The Canadian duty jumped from 14.5 percent to 35 percent in August 2025. Every one of those happened after builders were asked the question.

So $10,900 is what builders estimated before the wood tariffs existed. Treat it as the bottom of the range.

The small-builder penalty is the real Massachusetts number

Here is the finding I think matters most, and it got almost no coverage when NAHB published it on August 26.

Massachusetts is a small-builder state. We are not a market of national production builders dropping 300-home subdivisions. We are a market of infill: a three-unit conversion in Somerville, a pair of townhomes in Taunton, a teardown in Newton, an ADU behind an existing house in Arlington. Those are exactly the builders sitting in the 9.1 percent column.

Which means the tariff exposure in this state is concentrated in the same small operators the zoning reforms were designed to activate. We handed them permission and then handed them the worst version of the invoice.

What new construction actually closed for in Massachusetts this year

Enough theory. I pulled every Massachusetts residential closing in MLS PIN from January 1 through August 31 of this year with a known year built, then split it by whether the home was finished in 2025 or later.

That is 2,124 newly built homes against 35,944 existing ones. New construction was 5.6 percent of the market.

Read the top band and the bottom band together. New construction is more than twice as concentrated above $1.2 million as resale is, and it is almost absent below $400,000. Only 20.9 percent of newly built homes closed under $600,000 this year. For existing homes it was 43.0 percent.

That is the answer to whether the building boom will bring prices down. The new supply is not arriving at the price points that need supply. It is arriving at the top, where it competes with other expensive homes and does nothing for a buyer trying to get in at $550,000.

A builder’s price is a floor. A seller’s price is an opinion.

The median newly built home in Massachusetts closed at $900,000 this year against $650,000 for existing. Per square foot it was $433 against $372, a 16 percent premium.

Some of that premium is real. A 2026 house has a current-code envelope, new mechanicals, and no deferred maintenance, and that is worth paying for. But the premium is not purely a preference premium, and here is how you can tell.

Both groups closed at a median of 100 percent of original asking price. Identical. What was not identical was how long it took.

This is the mechanism, and it is the thing I would want a buyer to understand before waiting on new construction.

A resale seller’s price is an opinion. Their basis is a purchase they already made, often years ago at a lower number, so they can always go lower. When the market softens, they cut, and the house sells.

A builder’s price is a floor. It is the materials invoice plus labor plus land plus carry plus whatever thin margin is left, and a builder cannot sell below it without writing a check to the buyer. So when demand softens, a builder does not cut. A builder waits. Tariffs raise the floor, and a raised floor does not show up as a bigger number on the listing sheet. It shows up as time, and as the homes at the bottom of the market that simply never get built.

Those 200-day Johnson Street closings in Taunton are what a raised floor looks like from the outside.

Permits are up. That is not the same as relief.

I want to be fair to the other side of this, because there is a real one.

Massachusetts authorized 3,031 single-family permits in the first seven months of 2026, against 2,820 in the same period of 2025, according to Census data. That is a 7.5 percent increase, and June was the strongest single month since 2024. Something is working.

But look at the level, not just the change. The comparable figure was 4,551 in 2017 and 4,038 in 2021. Full-year 2025 came in at 4,999, the weakest year in the series. A 7.5 percent gain off the bottom of a decade is a bounce, not a boom, and it is nowhere near the scale of the shortfall. Reading a percentage change without checking the level it is changing from is how people talk themselves into waiting.

What I would actually do this fall, as a buyer

The existing market is where a buyer has room to negotiate right now, and I do not say that as a preference. I say it because the numbers are lopsided.

Banker & Tradesman reported on September 6 that Greater Boston active listings, single-family and condo combined, were up 15.2 percent year over year for the four weeks ending August 23, using Redfin data. Meanwhile only 22,045 single-family homes had sold statewide through July 31, a 0.9 percent decline. More listings, slightly fewer buyers closing.

Worth being precise about that figure, because I have seen it repeated as a statewide number. It is Greater Boston, and it counts condos alongside single-families.

Practically, here is the read:

• Negotiate on the existing home, not the new one. 45.9 percent of resale closings this year came in under the original asking price. The seller has room. The builder does not.
• Do not price a new build off the sticker alone. Ask what the escalation clause says. A 195-day marketing period is a builder carrying financing, and carry has to come out of somewhere.
• If you want new construction anyway, shop the finished spec. A completed unit that has been sitting is a builder whose carry clock is running. A to-be-built contract signed today is a builder passing you a materials risk that has an open-ended end date.
• Reconsider the older house you dismissed. The 16 percent per-square-foot spread between new and existing buys a lot of renovation, and renovation you control is cheaper than renovation priced into a builder’s floor.

None of this is an argument that new construction is a bad product. It is an argument that in 2026 you are paying a premium for it in both dollars and waiting, and the supply narrative that is supposed to justify the wait is not showing up in the closing data.

And if you are the one building

For the small builders and the ADU-curious owners reading this, the exposure is ongoing. It is not a spike that already happened.

There are two dates on the calendar that matter. Commerce is expected to issue final results in the seventh administrative review in October, which could take the Canadian duty from 35.16 percent to 24.83 percent. And the reduced 15 percent cap on residential HVAC equipment runs through December 31, 2027, then reverts.

Neither of those is a reason to sit still. Both are reasons to write contracts that survive whichever way they go:

• Lock supplier bids in writing with an expiration date you can actually hold. A verbal quote is not a lock, and in a 9.1 percent materials year it will not hold for the length of a build.
• Get the material cost pass-through language nailed down now. Define the trigger, the index, the cap, and who eats the first increment. Ambiguity in that clause always resolves against the smaller party, and on an infill job that is you.
• Buy the steel derivatives early and the HVAC before 2028. Hangers, fasteners and connectors sit at 25 percent, and residential HVAC has a dated window of relief. Those are the two places sequencing genuinely saves money.
• Price the ADU off today’s stack, not last year’s. If your feasibility number came from a 2025 estimate, it is stale by roughly a 6.7 percent materials move plus a wood tariff that did not exist when you ran it.

The honest version of the supply argument

I am pro-supply. I supported the MBTA Communities law, I think the ADU statute is the best housing policy this state has passed in twenty years, and I want the triple-decker back.

But zoning reform is a permission instrument, and permission is the cheap half of a house. The expensive half is arriving from Canada at 45 percent and from the mill at 50 percent, it lands hardest on the smallest builders, and it is currently moving faster than the permission side is being used.

Anyone telling you to wait for the new-construction wave to bring your price down is describing a mechanism that the 2026 closing record does not support. Three point six percent of new construction closed under $400,000. The wave, so far, is breaking at the top of the market.

If you are trying to buy in the next twelve months, the existing inventory that is sitting right now is the better trade. If you want me to run the actual numbers on a specific property against comparable new construction in that town, reach out and I will pull the MLS data and show you the spread.

In the twelve months through last week, 157 homes changed hands within a mile of New Bedford’s new commuter rail station. Within a mile of the East Taunton station, twelve did. Within a mile of Freetown, two.

The stop with almost nothing around it is the one getting 275 new apartments. The stop surrounded by 157 recent sales is the one whose own mayor says it is still too early to see any new housing development at all.

That inversion is the whole story of South Coast Rail’s first eighteen months, and it is the opposite of the pitch most buyers are hearing right now. The train is identical at every stop. It leaves at the same times, costs the same $12.25, and takes roughly the same hour and forty minutes to reach South Station. What differs from station to station is the dirt. Rail buys access, and access only turns into housing where there is vacant land inside the walkshed to turn it into. New Bedford has stations. It does not have empty acres next to them. East Taunton had almost nothing but empty acres, which is exactly why it is the one with cranes.

So before you underwrite a South Coast purchase on the theory that a rail line lifts a region, it is worth looking at what the closed sales actually did. I pulled them.

The ridership is real, and it is the part nobody should argue with

Start with what worked, because plenty did. Service opened March 24, 2025, the first passenger rail to New Bedford and Fall River since 1958. In the three months before it opened, the old Middleborough line averaged about 6,500 weekday riders. It now averages about 10,000. From April through December 2025, ridership on the line rose 37% against the same months in 2024, roughly 600,000 additional trips, according to CommonWealth Beacon.

The context matters more than the percentage. Over a comparable stretch, MBTA ridership systemwide fell 6% while this line climbed, and The New Bedford Light reported that weekday boardings rose 36% against a 13% weekend gain. That split tells you these are commuters, not day-trippers. The state had projected about 1,600 new daily trips. The line delivered closer to 2,000 to 3,000.

One precision note, because the shorthand circulating locally is slightly off. South Coast Rail is now the fourth-busiest commuter rail line by total ridership, but fifth by average weekday ridership, per the Boston Globe’s August 2026 reporting. Reliability has held up too: 92% on-time out of Fall River and 91% out of New Bedford, against a 91% network average. As a transportation project, this thing works. I am not going to pretend otherwise, and anyone telling you the line was a boondoggle is arguing with the boardings.

I measured the station premium. There isn’t one.

Here is where the story turns. The assumption baked into almost every South Coast listing conversation I have had this year is that proximity to a new station is worth something. So I tested it directly. Using MLS PIN closed sales from September 1, 2025 through September 8, 2026, I geocoded every transaction against the published coordinates of the five new southern stations and compared what sold within one mile of each station against everything else that sold in the same city over the same window.

Same city. Same twelve months. The only variable is distance to the platform.

Plus one and a half percent. Plus one half of one percent. Zero. Eighteen months after the trains started running, buyers in New Bedford and Fall River are not paying a measurable premium to live next to the platform. Those gaps are noise, not signal, and I would not defend any of them as a real number.

This is not a distressed-market story either, which is what makes it interesting. Over the same window, Fall River closed at a median 100.0% of original list price with 50.6% of sales at or above the original ask, and New Bedford at 99.6% and 49.6%. Boston closed at 98.0% with only 37.9% at or over. On the metric that measures whether sellers are getting their number, the South Coast is running tighter than Boston. These markets are not soft. They simply are not pricing the train.

Why East Taunton got 275 apartments and New Bedford got a shrug

New Bedford Mayor Jon Mitchell has been blunt about this, and he is right. It is too early to see any new housing development near the city’s two new stations, he told CommonWealth Beacon, because of where they sit. Church Street station is in a single-family residential neighborhood. The downtown station is on industrial waterfront. Neither, in his words, is a prime location for new housing development.

The closed sales confirm it in a way a quote cannot. Inside a half mile of Church Street station, 22 properties sold in twelve months and 18 of them, 82%, were single-family houses. That is not a development site. That is a built-out neighborhood of owner-occupied homes, and you do not assemble a 275-unit apartment project out of it at any price. The downtown New Bedford station is on Acushnet Avenue about seven tenths of a mile north of downtown, hemmed by Route 18 on one side and the Wamsutta rail layover yard on the other. There is no there to build on.

Now look at the same measurement at East Taunton. Twelve sales inside a mile in a full year, and eight of them were units in a single condominium complex built in 1985 to 1987. The other four were single-family houses that are technically in Berkley. Inside a half mile of that platform, exactly one property sold all year.

That emptiness is the asset. The Hanover Company and Honeycomb broke ground in September 2025 on 275 apartments across four buildings, 137 one-bedrooms and 138 two-bedrooms, plus a 4,980 square foot retail building at the station entrance on Route 140, as the Taunton Daily Gazette reported. Roughly 22 months of construction, with first move-ins about 16 months in. The developer has said plainly that without the train, the interest level would have been considerably lower.

Taunton Mayor Shaunna O’Connell put the demand side in one line: developers know about rail and the advantages it brings, and they have been coming to us. Worth noting that she is the mayor being approached, not Fall River’s. Taunton also did the unglamorous work early, approving transit-oriented zoning near the station back in December 2022, more than two years before a train ever ran.

New Bedford already rezoned. It didn’t matter.

The lazy version of this story is that New Bedford was asleep. It wasn’t, and that is what makes it a genuinely useful warning rather than a scolding.

New Bedford adopted the Clasky Common Transit-Oriented Development overlay on August 26, 2025, five months after the first train. The district runs along Purchase Street from the Octopus intersection up to the Car Barn, and it is not timid: 8 to 10 stories in the subarea closest to the waterfront, 4 to 6 stories on the east side of Purchase Street, 3 to 4 on the west, with the residential parking minimum cut to 0.75 spaces per unit in the densest subarea. The city followed with a broader zoning package that the City Council approved on June 8, 2026 and the mayor signed into the code that July.

So the permission exists. Ten stories are legal a short walk from the downtown platform. And eighteen months in, the mayor is still saying it is too early. That gap is the lesson: in New Bedford, zoning was never the binding constraint. Assembled, buildable, non-industrial land inside the walkshed is, and no city council vote conjures that. When someone tells you a town “upzoned near the station,” your next question should be how many contiguous vacant acres that upzoning actually landed on.

It is also why I would be careful with the seller expectations that Mitchell himself flagged: owners of New Bedford three-families asking, in his words, crazy money on the theory that values will soar once rail arrives. Rail arrived. Check the table above.

Fall River’s real catalyst is a 20-acre state land deal that hasn’t picked a winner yet

Fall River is the most interesting of the three, and it is the one being described least accurately in listing copy right now.

The genuine catalyst there is not organic developer interest. It is that the state manufactured land supply. The Route 79 and Davol Street corridor project, roughly $135 million of it already spent, converted an elevated highway into surface boulevard and freed up about 20 acres of Taunton River waterfront directly adjacent to the Fall River Depot station. On July 24, 2026, MassDOT issued the invitation to bid on it. The requirement is more than 1,400 homes plus mixed-use commercial and public space, with the state estimating roughly $1 billion of private investment and more than $200 million in increased property value. Proposals are due October 21, 2026, and MassDOT has said it may select between one and three developers.

That is a serious project and it deserves attention. It is also, as of today, an open solicitation with no selected developer, no financing, no permits and no shovel. If you are buying in Fall River this fall, you are buying next to a bid deadline. There is a real difference between that and buying next to a building under construction, and the price you pay should reflect which one you are actually getting. The bid result in late October is the single most concrete near-term catalyst on this entire line, and it is worth watching if you own or are considering anything in the Fall River market.

The commute is the ceiling, and here is what it costs

Everything above runs into one hard constraint. The ride is long.

Typical through-train running time between South Station and the Fall River and New Bedford terminals is 94 to 109 minutes. CommonWealth Beacon puts the realistic New Bedford trip at 100 to 110 minutes without delays, against roughly an hour by car in good traffic. Trains run about every 70 minutes on weekdays, 32 direct trips to South Station split between the two branches, and about every two hours on weekends.

Run the break-even before you assume the monthly pass. At $388 against a $24.50 round trip, the pass only pays for itself at about 16 travel days a month, which is four days a week. At the two or three days a week most riders actually do, single tickets are cheaper and the annual number lands closer to $2,000 than $4,656. Either way it compares favorably to a second car and a daily Route 24 grind.

The harder cost is the clock. Three to three and a half hours of round-trip transit is a five-day-a-week proposition very few people sustain, which is why nearly every rider quoted in the Globe’s coverage described going in two or three days a week, not five.

That is the honest shape of the market. South Coast Rail is a superb hybrid-schedule asset and a poor five-day one. Mitchell’s own version of this is that the line is not going to usher a lot of wealthy folks into the city, and on the daily-commute question I think he is right. Underwrite two or three days a week, because that is what the timetable actually supports.

The investor math: real arbitrage against Somerville, none against Dorchester

For investors the headline numbers look irresistible. Over the last twelve months the median New Bedford multifamily closed at $595,000 and $207 per square foot. Fall River closed at $650,000 and $200 per square foot. Somerville’s median multifamily closed at $1,300,000 and $452 per square foot. Less than half the price per foot for a building on a train line to the same South Station.

That framing is incomplete, and the incompleteness is where people lose money. Cheap buildings sit in cheap rent markets. The only comparison that means anything is price against the rent the building actually produces, so I ran three-family closed sales against closed two-bedroom leases in the same town and the same window.

The arbitrage against the inner ring is real. New Bedford at 9.9 times gross against Somerville at 14.1 and Malden at 12.3 is a meaningful yield advantage, and it is the strongest case anyone can make for buying down here.

The part nobody mentions is the Boston row. Boston three-families also closed at 9.9 times gross, dead even with New Bedford, on 246 sales against New Bedford’s 84. Those Boston three-families are largely Dorchester, Mattapan and Hyde Park triple-deckers, which is precisely the stock a South Coast investor is trying to trade out of. So on gross yield there is no South Coast arbitrage against Boston’s own multifamily belt. There is only an arbitrage against Somerville, Medford and Malden, which are different markets with different tenant demand and different exit liquidity.

Two more things belong in that underwriting. New Bedford two-bedrooms closed at a median $1,800 and three-bedrooms at $1,900, so the rent ladder is compressed and adding bedrooms buys you very little. And the older housing stock down here carries higher relative operating and capital costs than the multiple suggests. If you want to pressure-test a specific building rather than a town median, our investment property analyzer is the right place to run it.

Phase 2 is a hope, not a plan

The last thing to strip out of your model is the fast train.

The promised full build would route trains through Stoughton with electrification and new stops in Easton, Raynham and Taunton, cutting New Bedford to South Station to roughly 77 minutes. That would genuinely change the affordability math, because 77 minutes is a commute a person can do four or five days a week and 105 minutes is not. A 2017 estimate had it attracting more than double Phase 1’s ridership.

It is also not funded. Phase 2 does not appear in the MBTA’s current five-year Capital Investment Plan, the $9.8 billion FY2026 to FY2030 program covering more than 660 projects. MassDOT’s stated position is that available funding is going to maintenance and repair of existing infrastructure first. The 2017 and 2018 estimates put the full build at $3.2 to $3.42 billion including Phase 1’s roughly $1 billion, but former Transportation Committee chair William Straus has since put the realistic figure at $6 to $7 billion. The Stoughton alignment still has to clear permitting through the Hockomock Swamp, with unresolved questions on wetlands, endangered species and drinking water. Eighteen months after Phase 1 opened, there is no financing plan.

Taunton’s mayor said she hopes to see phase two happen in her lifetime. That is a candid description of the timeline from someone who wants it built. Treat any appreciation you are underwriting on the strength of a faster electrified line as unfunded, unpermitted and undated. It may well happen. It is not a schedule.

What I would actually do

South Coast Rail is a legitimate affordability move for the right buyer. It is not a regional appreciation trade, and the closed sales are unambiguous on that point. Here is how I would separate the two.

If you are buying to live there and commute two or three days a week: this is the strongest case on the line. A Fall River single-family closed at a median $485,000 so far in 2026 and New Bedford at $449,950, against $752,500 in Quincy and $877,500 in Boston. You are saving real money and paying for it in time, not in price. Just buy the house you want on its own merits, because the data says you are not paying extra for the platform and you should not accept a premium ask for one either.

If you are buying for appreciation: be honest that you are making a land bet on a specific parcel, not a bet on a line. The two locations where that bet has actual mechanics behind it are East Taunton, where 275 units and a retail building are physically going up, and the Fall River waterfront, where the state controls 20 acres and picks bidders after October 21. Everywhere else on this line, the appreciation thesis is a story.

If you are buying multifamily: run the gross multiple before the price per foot. The yield edge over Somerville and Malden is real and worth having. The yield edge over Dorchester does not exist right now, so if you already own Boston triple-deckers, moving capital down here on yield alone is not the trade it looks like. You would be buying a different set of risks at the same multiple.

If you already own down there and are hearing that rail will lift your value: it has not, over eighteen months, by any measure I can find in the closed data. Price your sale off comparable homes in your own city, not off the station.

The thing to hold onto is that a station is not a strategy. Six stops opened on the same day with the same trains, and the only one with cranes is the one that had nothing standing next to it. If you are weighing a South Coast purchase and want the numbers run against your actual address and your actual commute, reach out and we will look at it together. If you are earlier than that, our investment property coverage and buying guides are a good place to start.

Sources

• CommonWealth Beacon, “Gateway City leaders celebrate one year of South Coast Rail, but final phase remains in limbo,” Hallie Claflin, June 11, 2026
• The Boston Globe, “South Coast Rail is a godsend for commuters working in Boston,” Omar Mohammed, August 26, 2026
• The New Bedford Light, “Commuter rail ridership surges with South Coast Rail,” Grace Ferguson, November 19, 2025
• The New Bedford Light, “Will South Coast Rail’s ‘full build’ ever happen?,” Grace Ferguson, February 17, 2026
• Taunton Daily Gazette, “Apartment complex breaks ground near East Taunton Commuter Rail Station,” Beth McDermott, September 18, 2025
• MassDOT, “MassDOT Opens Bidding for Transformational Waterfront Development in Fall River,” July 24, 2026
• Fall River Reporter, “MassDOT Seeks Bids for $1 Billion Fall River Waterfront Transformation Project”
• MBTA, South Coast Rail project page (service levels, station list, full build status)
• MBTA, Commuter Rail Fares (Zone 8)
• MBTA, FY2026 to FY2030 Capital Investment Plan
• City of New Bedford, Clasky Common Transit-Oriented Development Overlay District primer
• The New Bedford Light, “New Bedford proposes commuter-friendly zoning around South Coast Rail,” March 4, 2025
• MLS PIN closed sales and closed leases, September 1, 2025 through September 8, 2026, compiled by BMN Boston. Station ring analysis computed against published MBTA station coordinates.

Nobody paid $43 million for the Edgartown estate.

Two people did. The registry recorded two deeds on the same compound: 61 Butler’s Cove Road and an adjoining parcel for $27 million, and 57 Butler’s Cove Road along with 15 and 16 Forever Wild Way for $16 million. Two separate realty trusts, set up by different attorneys, closed on two halves of one property. Banker & Tradesman reported both, and the Vineyard Gazette named the trustees. The $43 million everyone printed is a sum. It is not a price anyone agreed to.

That matters more than it sounds, and not because the headline is wrong. It matters because the way the record was set is the actual trend. Buyers at this level are assembling, not buying. And a version of that behavior is already showing up on the mainland, in Weston and Brookline and on Beacon Hill, where I work. What is not showing up on the mainland is the part every seller I talk to has taken from this story: that the top of the market is hot and moving fast.

It is not. I pulled the numbers.

Start with what the record actually broke

Michael Bronner, the Digitas founder, spent more than $19 million assembling the Herring Creek Farm property between 2019 and 2023, then listed it in June 2026 at $49 million through Maggie Gold Seelig of MGS Group and Brian Dougherty of Corcoran Property Advisors. It closed at $43 million, per the MV Times. The prior Vineyard record was a $37.5 million Katama sale last year.

The state record is messier. In September 2023 Dave Portnoy paid $42 million for 68 and 72 Monomoy Road on Nantucket, which Forbes and everyone else called the most expensive home ever sold in Massachusetts. That itself broke a $38.1 million Nantucket sale from earlier the same summer. So the honest scoreboard reads this way: the Edgartown compound is the largest dollar total ever transacted on one Massachusetts residential property, and Portnoy’s $42 million is still the largest sum a single buyer has ever paid for one.

Both things are true. Only one of them was in the headline.

Those “6 sales over $10 million” are island numbers

The statistic doing the most work in the coverage is this one: 12 Massachusetts sales over $5 million in the first half of 2026 against 7 a year earlier, and 6 over $10 million against just 1. I have now been sent that line three times this week by clients asking what it means for their house in Wellesley.

It is a Martha’s Vineyard statistic. It comes from the Tea Lane Associates 2026 mid-year report, which the Vineyard Gazette covered on August 25, and it counts sales on the island. Not the state. The Vineyard runs on the LINK multiple listing service. Nantucket has its own. The mainland runs on MLS PIN, and none of those three systems can see each other. When an island report says “sales over $10 million,” it means island sales over $10 million, and the mainland does not appear in the count at all.

The rest of that report is worth reading, because it cuts against the boom framing. Island transactions fell 5% year over year, from 155 to 148. Dollar volume rose 23%, from $304 million to $375 million. Properties over $3 million made up roughly a quarter of sales and more than half of every dollar spent. And 75% of everything that sold on Martha’s Vineyard in the first half of 2026 sold below its asking price.

That last number is the one nobody quoted.

The mainland keeps its own book, and I read it

So I ran the mainland’s book directly. What follows is my own cut of the MLS PIN closed sale records for Massachusetts, every residential closing from January 1 through August 31, 2026. That is 34,834 sales. No aggregator, no portal estimate, no rounded press figure.

Sixteen of them closed at $10 million or more. Here is every single one.

Read that table again with a seller’s eye. The fastest sale on it, 46 Chestnut Street on Beacon Hill, went in 25 days and still closed $3 million under ask. The two that got full price were the two that were priced at what they were worth on day one. Everything else negotiated.

To check my own work, I compared the pull against Banker & Tradesman’s registry based sales report. Their August 20 entry for 430 Stuart Street penthouse 35B at $14,250,000 matches my row to the dollar and the day. The book is clean.

Weston already broke a record in May, and nobody called it a state record

Look at the top row. On May 29, 2026, 9 Atlas Lane in Weston closed at $25,825,000. It is a 17,000 square foot house designed by Thomas Catalano, built in 2011, overlooking the Weston Reservoir, and Boston.com reported it as the highest priced home sale in Weston history. It was also reported as the most expensive non-waterfront single family home ever sold in Massachusetts.

It got a fraction of the Edgartown coverage. It was three months earlier, twelve miles from Boston, and it sold with an additional buildable lot included in the deal.

There is your mainland compound. Somebody paid $25.8 million for a house and the land next to it, in a town on Route 128, in the same year that a Vineyard estate had to be cut in half to find two buyers. The same instinct, priced in a commuter suburb.

And it took 326 days and a $3.175 million reduction off the $29 million original ask to get there. Almost eleven months. That is the part of the mainland record nobody wrote up.

The bifurcation is real, and it is a supply story first

Now the part I think is exactly right, and it holds up under the data completely. The Warren Group data, reported by Banker & Tradesman, has the Greater Boston luxury median at $2.8 million in May 2026, up 7.4% year over year, against a non-luxury median of $750,000 up 2.1%. Statewide the luxury median is $2.5 million and up 6.6% while the non-luxury median is flat at $665,000.

The price gap gets the attention. The listing counts are what actually explain it.

Nationally, luxury and non-luxury listings both crept up by about a point. In Greater Boston they moved in opposite directions by nearly eighteen points combined. Owners of ordinary houses are listing. Owners of expensive ones are staying put. When you shrink the numerator of a median and grow the denominator of the other one, you get a widening price gap that has very little to do with demand.

Your town medians did cross $2 million

Here is where the local read is simply correct, and I can show it from the closed records rather than from a portal estimate. These are single family medians, MLS PIN, closed January 1 through August 31, 2026.

Three quarters of every single family sale in Brookline this year traded at $2 million or more. In Weston it was the same share. If you own in one of these towns, you are not in the median Massachusetts market and you should stop reading about it.

Insulated from rates also means in no hurry

The standard explanation for the split is that wealthy buyers do not finance, so the Freddie Mac survey rate of 6.71% as of September 3, 2026 does not reach them. That is mostly right. On the Vineyard, broker Jim Feiner told the Gazette that lower end island buyers are staring at $7,000 to $8,000 monthly payments. A cash buyer at $12 million is not.

What gets left out is the second half of that sentence. A buyer with no loan also has no rate lock about to expire, no pre-approval clock, no lender pushing to close before the quarter ends. Nothing in a cash buyer’s file creates urgency. The exact condition that protects your price level is the condition that removes your buyer’s deadline.

Which is why the mainland book reads the way it does.

And the higher you go, the wider it opens. Between $2 million and $3 million the median sale closes at 98.0% of the original ask in 75 days, and 29% of sellers take a cut of more than 5%. From $3 million to $5 million it is 96.6% and 84 days, with 40.5% taking that cut. Above $5 million it is 93.7% and 112 days, and 57.6% of sellers give up more than 5%.

That is the sentence I would put on the wall. Above $5 million in Massachusetts this year, more than half of all sellers closed at least 5% below what they first asked, and it took them close to four months.

Compounds, not trophy houses

Tom Wallace, the principal broker at Wallace & Co. Sotheby’s, gave the MV Times the most useful line in the whole story: “This property reflects a movement toward a desire for not just a home, but a family compound. And it is an appetite for this kind of property that is ever increasing, and very appealing.”

He is describing why a $49 million listing found its buyers only after it stopped being one listing.

The mainland version is quieter but it is in the data. Look back at the table: 46 Beacon Street sold two separate penthouses this year, PH8 in July and PH7 in August. 107-109 Chestnut Street is two Beacon Hill townhouse addresses carrying one penthouse. Weston’s record sale carried an extra buildable lot. And MLS PIN recorded 43 land closings at $1 million or more across Massachusetts this year, including a $2,950,000 lot in Brookline in July, $2,800,000 in Wellesley in January, and $4,100,000 in Cambridge in January. Nobody pays three million dollars for a Brookline lot to flip it.

The behavior is the same everywhere. Buy the piece next door. Control the edges. It shows up as a family compound on Slough Cove and as a combined-floor penthouse on Beacon Hill, but it is one instinct, and it changes how a neighboring owner should think about their own exit.

What this means if you are listing at $2 million this fall

The broader market is loosening underneath all of this. Massachusetts MLS sales through July 31 were down 0.9% year over year at 22,045 homes, July new listings ran 9.4% above last year, and Greater Boston listings in the four weeks ending August 23 were up 15.2% for single families and condominiums combined, per Banker & Tradesman. More competition, not less.

Here is what I would actually do.

1. Price at your ceiling, because that is what a list price is up here. In the under $1 million market a list price is an opening bid and 55% of sellers beat it. At $2 million and up only 36% do, and above $5 million it is 14%. Your number is where the negotiation starts and then goes down. Set it where you would be genuinely happy, not where you hope a bidding war lands.

2. Budget four months, not four weeks. Seventy-eight days is the median at $2 million and up, and 112 above $5 million. Those are medians, so half of everything takes longer. 397 Commonwealth Avenue took 291 days. 18 Marlborough took 262. Weston’s record took 326. If your plan needs a closing by Thanksgiving, you needed to be live in July.

3. Never benchmark against a median-home comp. Your neighbor’s $850,000 colonial that drew four offers in nine days tells you nothing about your $2.4 million house. Different buyers, different money, different clock. Ask your agent for the closed $2 million-plus set in your town with days on market and percent of original ask attached, and if they cannot produce it, that is your answer about who is pricing your house.

4. Reprice early or not at all. Every property on that $10 million table that took a real haircut also sat. The market read the days on market before it read the price. A cut in week six reads as responsive. A cut in month seven reads as a problem, and buyers at this level have the patience to wait for month nine.

5. Ask who owns the parcel next door. If your property abuts land, a subdividable lot, or a unit on the same floor, the assembly buyer is a real and specific buyer for you, and that buyer is not found through a portal. That is a phone call, and it is often the difference between 91% of ask and 100%.

6. Run the tax math before you sign, not after. Massachusetts adds a 4% surtax on taxable income above $1,107,750 for the 2026 tax year, and a large gain lands in a single year. If a January closing changes the answer versus a November one, that belongs in the pricing conversation now.

The honest summary

A record was set. It was set by splitting an estate in two and selling the halves separately, after a $6 million reduction off the ask, in a market where three quarters of island sales close below asking. On the mainland, sixteen homes traded above $10 million in eight months and not one of them beat its original price.

The luxury tier is genuinely stronger than the middle. Prices are up 7.4% in Greater Boston against 2.1% for everything else, and the reason is that owners of expensive homes are not listing while everyone else is. That is a real and durable advantage if you own one.

It is not the same thing as a fast sale, and this fall the difference between those two ideas is going to cost some sellers a lot of money.

If you are weighing a $2 million-plus listing this fall, I will pull the closed comps for your street with days on market and percent of original ask attached, and tell you plainly what I think the number is. You can start here or reach me directly at 617.955.2224.

Sources

• Author’s query of MLS PIN closed sale records, Massachusetts residential closings, January 1 to August 31, 2026 (n = 34,834). MLS PIN does not cover Nantucket or Martha’s Vineyard.
• Banker & Tradesman, “Martha’s Vineyard Estate Sells for Record Price” (September 3, 2026)
• Vineyard Gazette, “Edgartown Properties Sell for $43 Million” (September 4, 2026)
• MV Times, “Edgartown property sale breaks state record at $43 million” (September 3, 2026)
• Vineyard Gazette, “Luxury Home Sales Push Island Real Estate Market to New Peaks” (August 25, 2026), reporting the Tea Lane Associates 2026 mid-year report
• Tea Lane Associates, Martha’s Vineyard market reports
• Banker & Tradesman, “Boston Luxury Market Pricing Outperforming Non-Luxury Homes” (July 1, 2026), citing The Warren Group and Redfin
• Banker & Tradesman, “Will a Looser Fall Housing Market Unlock More Sales?” (September 6, 2026), citing the Massachusetts Association of Realtors
• Freddie Mac Primary Mortgage Market Survey (30-year fixed at 6.71%, week of September 3, 2026)
• Boston.com, “A look at the $25.8 million Weston home that broke a town record” (June 10, 2026)
• Forbes, on the $42 million Nantucket sale (September 29, 2023)
• Redfin, all-cash home purchases, March 2026 (28.8% of purchases, May 27, 2026)
• Mass.gov, Massachusetts 4% surtax on taxable income ($1,107,750 threshold, 2026 tax year)