High housing costs and the steady flow of families leaving Greater Boston reflect a simple reality: we’re not building enough homes. Not only should we build enough to affordably house everyone who already lives here, but also to support the kind of steady population growth happening in other parts of the country. When housing supply falls short, we see increased residential overcrowding, delayed household formation, long commutes, rising prices, and out-migration. By contrast, abundant housing supports social mobility, economic growth, cultural vibrancy, and a stronger tax base.

In advance of the release of our annual Greater Boston Housing Report Card on November 12, we’re publishing this preview of recent trends in building permits alongside analysis of new dataset on total housing units, the Census Address Count (scroll down for data at the municipal level). Permits are a useful leading indicator of homes likely to be built in the next few years. They are not a perfect proxy for actual production, but they provide a time series that allows us to see how the region has performed over many decades.

That record shows a clear pattern. Statewide permitting remains well below the levels reached in the 1970s and 1980s. Even during the relative rebound of the 2010s, permit totals fell far short of those earlier peaks, and the past few years have brought another slowdown, with only a modest uptick in 2024.

The Census Bureau’s Building Permit Survey (BPS) is the most comprehensive publicly available data on residential permits. While useful, the survey has serious flaws. These include:

• Local governments sometimes underreport or misreport their permit activity, which leads to incomplete or inaccurate totals. The very nature of this work makes it difficult to provide precise undercount levels, but it appears likely that BPS estimates often undercount true permitting activity.
• Permit data are not a net measure of housing change. If, for example, 10 homes are torn down and replaced with 50 apartments, the BPS simply reports permits for 50 new units without subtracting the 10 that were lost.
• The data exclude adaptive reuse projects, so the thousands of units created in converted mills across Massachusetts do not appear.
• Permits do not guarantee actual construction, and this distinction has grown more important in recent years as high interest rates and rising construction costs have delayed or canceled many projects.

A New Data Source on Housing Completions: The Census Address Count

Fortunately, the Census Bureau has started administering a promising alternative data product for tracking net new housing completions: the Census Address Count file. This dataset is derived from the U.S. Census’s Master Address File, a comprehensive list of addresses that underpins the decennial census, drawing from the U.S. Postal Service’s Delivery Sequence File, among other sources. It provides total counts of housing units at the census block level and is now released every six months, beginning with 2023 data. Importantly, it allows comparisons back to the April 2020 decennial census baseline, making it possible to track changes over time.

Address counts give us a more complete picture of the housing stock as it actually exists, not just what has been permitted. Comparing these counts over time also allows for better estimation of net changes in the count of housing units, automatically accounting for demolitions that are not tracked when simply counting newly permitted units.

Here, too, there are some important caveats. Conceptually, address counts and building permits measure different things. Permits remain useful because they signal future production activity. This can be especially important for projecting multifamily construction as it often takes years from permitting to completion for these larger residential projects. Additionally, the very nature of the Census Address Count being a new product should lead us to analyze it with caution. Housing researchers are just beginning to explore it, and so far, the Census Bureau has released very little documentation on the methods behind the estimates. We may learn over time about shortcomings or quirks in how the numbers are produced.

Bearing these cautions in mind, the early data appear quite useful, showing steady increases in the number of homes added to the housing stock since 2020. Statewide, the housing stock increased by 97,656 units from April 2020 to July 2025. In Greater Boston, the five-year increase was 71,135.

Comparing New Addresses and Building Permits: Two Different Stories

Simply looking at the new Census Address Count data on recent housing completions risks giving the impression that we are finally producing homes at a good clip and starting to make a real dent in our long-term housing shortage. The problem is that even with these recent increases, vacancies remain low, and prices remain elevated across Greater Boston, as we show in later sections of this year’s Housing Report Card. Clearly, we need far more housing than we currently have. Further, although completions have gone up over the past couple of years, the best rough data we have on permits suggest a significant slowdown in future production.

The good news about recent completions is therefore paired with real warning signs for the rate of future production. New permits in Greater Boston hovered around 13,000 units annually during the 2010s. They jumped to 15,000 in 2021 but have dropped sharply since. By 2024, Boston proper had fallen from typically permitting 3,000 to 4,000 units annually between 2015 and 2022 to far less than 3,000 annually during the last two years, with 2,219 permitted in 2024. In Boston, years 2023 and 2024 had the lowest counts of new permits since 2012.ⁱ Nationally, permitting peaked in 2022 before turning downward.ⁱⁱ

Early data for 2025 show a continued slowdown. While the above graph shows annual totals, the graph below compares year-to-date permits in the Boston Metropolitan Statistical Area from January through July 2025 with permit counts during the same months in each year going back to 2015. Year-to-date permitting in 2025 is down sharply from the late 2010s and early 2020s, with 2025 levels 44 percent below 2021. Much of the growth in the 2010s had come from large multifamily projects, and it’s this type of permitting that has seen the steepest decline over the past three to four years.

Macro-level causes explain much of the slowdown. Elevated construction costs and higher interest rates have made financing projects more difficult. Nationwide, construction materials remain expensive compared to pre-COVID levels, with costs in May 2025 roughly 40 percent higher than in March 2020.ⁱⁱⁱ Recent upticks in material prices, layered on top of recent tariffs that still apply to key inputs like lumber and steel, add further pressure. Restrictions on immigration enacted in 2025 may also raise labor costs, since a large share of workers in several homebuilding trades are foreign-born. All these recent headwinds compound longstanding local barriers such as restrictive zoning, costly building code requirements, and lengthy approval processes.

Placing Completions and Permits in Context of the State’s 220,000 Unit Production Target

So, what does this all mean in terms of tracking progress toward official housing production goals? The Affordable Homes Act, passed in August 2024, was followed by a new state housing plan that set an official state construction target of 220,000 additional homes between 2025 and 2035. This represents about 7 percent of the state’s 2020 housing stock. For Greater Boston, EOHLC’s Housing Needs Assessment forecasts even higher need, calling for 7.5 to 10 percent growth, or roughly 140,000 to 180,000 new units over the decade.^iv

The Executive Office of Housing and Livable Communities (EOHLC) reports that “90,400 homes have been completed, started construction, or proposed since the beginning of the Healey-Driscoll administration” in 2023, based on a mix of Census Address Count data, reporting of new units funded by state grant programs, and recent permitting activity.^v On its face, this figure risks giving the impression that Massachusetts is already nearly halfway toward its goal. But the new goal is for homes constructed starting in 2025, and most of the 90,400 homes being claimed by EOHLC were built or permitted during prior years.

If the pace of completions seen in 2023–2025 could be sustained, Massachusetts would be on track to meet the 220,000 target by 2035. This is encouraging because it suggests that our local construction industry has the scale capable of building at pace. But there are strong reasons for concern. Much of the recent surge in completions reflected pent-up projects delayed by the pandemic, many of which were financed with historically low interest rates in the late 2010s and early 2020s. Since multifamily projects often take two or more years to move from permit to completion, the high completion counts in 2023–2025 likely trace back to permits issued before financing conditions tightened. Today’s higher borrowing costs, tariffs on key building materials, and persistent construction workforce challenges make it far harder for new projects to pencil out financially.

So, while the good news is that thousands of new homes have recently come online, Massachusetts is likely entering a period of far slower completions. At current permitting levels, we are unlikely to come close to building 220,000 new units statewide between 2025 and 2035.

Municipal Trends in New Housing Completions

For people interested in what these data on new housing completions look like at the local level, we end by presenting Census Address Count estimates by municipality. Because the first release after the 2020 Census did not arrive until mid-2023, we present changes in two ways: annualized growth for the April 2020 to July 2023 period, and direct year-to-year changes from July 2023 through July 2025. Click on any of the column headers to sort.

It is important to interpret these estimates cautiously. As noted earlier, this is a new data product, and we are still learning about its strengths and limitations. Some of the swings in the early years may reflect coding adjustments rather than true changes in the housing stock. For instance, college dorms may have initially been coded as individual housing units before being corrected to group quarters, which are not counted as housing units for production purposes.

The top producers in total units over the full five-year period were Boston, Cambridge, Quincy and Somerville, all bringing online an increase of around 5 percent of their 2020 housing stock. By comparison, regionwide, the net increase of new units over these five years was 3.8 percent of the total housing stock.

Because larger cities naturally generate more units, we also present the data as the share of pre-existing housing stock that was added between 2020 and 2025. This framing highlights smaller communities that grew quickly relative to their base. Millis stands out, expanding its housing stock by almost 14 percent over five years. Woburn, Hanover, Maynard, Bridgewater, Revere, and Saugus also grew by more than 7 percent during the same period.

Finally, to show how different types of communities have contributed, we summarize completions by community type using MAPC’s Community Types classification. Metro Core Communities have been producing substantially more housing than other types, even after adjusting for their pre-existing stock. Between 2020 and 2025, Metro Core Communities added about 26,382 net new units, compared to 44,724 units across all other community types combined. Metro Core Communities top the list whether measured by total new units or by growth as a share of their 2020 housing stock: They expanded by 5.5 percent under the latter measure, while no other community type exceeded 4 percent.

Taken together, the data point in two directions. Address counts show that completions since 2020 have been steady enough to add tens of thousands of homes statewide, suggesting real progress in recent years. At the same time, permits for new construction have fallen from early-2020s highs, especially for large multifamily projects, which suggests a slower pace of future completions. Whether permitting rebounds will depend on financing costs, materials prices, labor supply, and local policy reform efforts, and each of those is uncertain. Against the state’s goal of 220,000 additional homes by 2035, recent output is encouraging but probably not sufficient if current permitting levels persist.

i. Metropolitan Area Planning Council (MAPC), “DataCommon,” June 3, 2025, https://datacommon.mapc.org/browser/datasets/384.

ii. U.S. Census Bureau and U.S. Department of Housing and Urban Development, New Privately-Owned Housing Units Authorized in Permit-Issuing Places: Total Units [PERMIT], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/PERMIT, October 2, 2025.

iii. U.S. Bureau of Labor Statistics, Producer Price Index by Commodity: Special Indexes: Construction Materials [WPUSI012011], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/WPUSI012011, September 22, 2025.

iv. Maura Healy et al., A Home for Everyone. Statewide Housing Needs Assessment (Executive Office of Housing & Livable Communities, 2025).

v. Executive Office of Housing and Livable Communities, How EOHLC Is Tracking Housing Production (Executive Office of Housing & Livable Communities, 2025), https://bankerandtradesman.com/wp-content/uploads/2025/10/Housing-Production-Estimates-Methodology-19-Aug-2025.pdf.