Every year the Internal Revenue Service does something quietly remarkable: it matches tax returns from one year to the next and publishes, county by county, how many households moved, where they went, and how much adjusted gross income moved with them. It is the closest thing anyone has to a census of relocation, and for Western Massachusetts in 2021–2022 it says something that upends the usual story.
Two of these four counties lost people and gained money. Two lost both. The difference is not subtle and it should change how you read the region.
The four counties, side by side
- Berkshire County: 3,384 people in, 3,651 out, net -267. Income balance $30.8 million. Arrivals reported $122,669 on average, leavers $96,826.
- Franklin County: 2,666 people in, 2,736 out, net -70. Income balance -$4.9 million. Arrivals reported $63,196 on average, leavers $63,534.
- Hampden County: 11,489 people in, 13,240 out, net -1,751. Income balance -$138.6 million. Arrivals reported $61,357 on average, leavers $70,192.
- Hampshire County: 6,282 people in, 6,773 out, net -491. Income balance -$4.4 million. Arrivals reported $76,078 on average, leavers $67,943.
What a positive income balance on a negative headcount means
Berkshire County shed 267 people on net and took in $30.8 million regardless. The arriving households reported an average adjusted gross income of $122,669 against $96,826 for the departing ones, a gap of about 27%.
In plain terms: the county is swapping households, and the incoming ones earn substantially more. If you are buying there, you are not entering a market being abandoned. You are entering a market being repriced by people with more money than the people leaving, and the assessment rolls back that up. It is the county with the region's highest median home value.
Hampden County is the mirror image. It lost 1,751 people and $138.6 million with them, and here the arrivals earn less than the leavers, $61,357 against $70,192. That is a different market with different pressures, and it is the one containing most of the region's population and most of its rental stock.
Why one region contains both
The pattern is a second-home market meeting a post-industrial one inside the same two-hour drive.
The southern Berkshire towns have been within weekend reach of New York for a century, and remote work extended that reach to full-time. Those buyers price against metropolitan incomes. Several of those towns now count the majority of their vacant housing as seasonal rather than empty, and one of them carries a median assessment above a million dollars.
The Connecticut valley cities are a different economy entirely: Springfield, Holyoke, Chicopee and Westfield built their housing stock for manufacturing employment that peaked around 1960. Springfield alone has 155,929 people against 174,463 then. That housing did not disappear, and a large supply meeting a smaller population is why the same $400,000 that buys a third of a house in the Berkshire hills buys a substantial one here.
What this tells a buyer
- Do not treat "Western Mass" as one market. On these numbers it is at least two, moving in opposite directions on the measure that matters most for future prices.
- Income inflow shows up in prices before it shows up in listings. A county taking in higher-earning households is a county where competition for the good properties is rising whether or not the population is.
- Population decline is not the signal you think. Ninety of the 101 towns here are larger than they were in 1960. The regional decline story is really a five-city story, and those five cities are also where the housing bargains are.
Two honest caveats
This data counts tax returns and exemptions, not people. Households that do not file, and people who move without a filing address changing, are invisible to it. It is also county-to-county, so a move from Springfield to Northampton crosses a county line and appears here, while a move from Springfield to Chicopee does not.
And it is one year. 2021–2022 sat in the tail of the pandemic relocation wave, which was not a normal period for American migration. A single year of this series describes that year. The direction has been consistent for longer, but the magnitudes will not hold.
The full migration picture, including which counties elsewhere the arrivals come from, is at who is moving in. If you have a workplace in mind rather than a county, live near work maps the towns inside a real commute of 106 places people actually work.
County totals also hide what individual towns are doing, and the town-level picture is stranger than the county one: which towns here are actually growing goes through it. If the move you are weighing is specifically to Berkshire County, the numbers behind the reputation are worth reading first in what the Berkshires data actually says. And if the move depends on working from home, check what the broadband record shows town by town before anything else, because it is the constraint people discover last.
Sources
- Internal Revenue Service. Statistics of Income county-to-county migration: matched tax returns, exemptions claimed and adjusted gross income moving between counties. Figures through 2021–2022.
- U.S. Census Bureau. Decennial population counts back to 1960, and the seasonal share of vacant housing.
"Figures through" is the period the publisher's data describes; "checked" is only when we last asked. The full sourcing: sources and method.