The Towns Where Most Houses Are Empty on a Tuesday

Drive through Hancock on a Tuesday in February and most of the houses are dark. That is not a town in decline. It is a town where 63.6% of all housing is somebody's weekend place, and it is the most extreme example of a pattern that shapes property across a large part of this region.

At the last census, 11,619 housing units across the 101 towns were held for seasonal, recreational or occasional use. That is 3.2% of all housing here, and 36% of everything the census counted as vacant. In 49 of the 101 towns, most of the "vacant" housing is not empty at all. Somebody owns it, heats it occasionally, and pays tax on it.

The Vacancy Rate You Read Is Measuring Two Different Things

This is the practical reason to care. Every site that ranks towns publishes a vacancy rate, and a high one reads as a warning: empty houses, falling demand, a place people are leaving.

In this region that reading is wrong about half the time. A hill town with a twenty percent vacancy rate made up almost entirely of second homes is not emptying out. It is a market where people from elsewhere have bought a large share of the housing stock and use it thirty weekends a year. Those are opposite situations and the headline rate cannot tell them apart.

The column that separates them is the last one in the table below. Where seasonal use is the bulk of the vacancy, you are looking at a weekend market. Where it is a small share, the vacancy is genuinely empty housing, and that is worth understanding for entirely different reasons.

The Boom That Did Not Happen

Here is where the data contradicts what almost everyone assumes, including us before we looked properly. Seasonal housing in this region has not been growing. Across three censuses:

• 2000: 10,962 seasonal units, 42% of all vacant housing
• 2010: 12,303 seasonal units, 39% of all vacant housing
• 2020: 11,619 seasonal units, 36% of all vacant housing

The count peaked in 2010 and has fallen by 684 units since. As a share of vacant housing it has declined at every count, from 42% to 36%. The region added roughly 24,622 housing units over those twenty years and essentially none of the net addition was seasonal.

That runs against the story you hear locally, and the story is not simply wrong. What has happened is a conversion rather than an expansion. Second homes have been becoming primary homes, particularly since 2020, and a house that was somebody's weekend place and is now somebody's address moves out of this column and into the occupied one. The units did not disappear. They filled up.

Which is a different and more interesting pressure on the market than the one people describe. The complaint is usually that outsiders are buying houses and leaving them empty. The measured change is closer to the reverse: houses that stood empty most of the year now have somebody living in them full time, competing for the same schools, the same trades and the same broadband.

What This Does To Prices And To Tax

A weekend market prices against the incomes of the people buying into it, and those incomes are not local. That is the whole explanation for the fact that Berkshire County, the rural one, has the highest median assessment of the four counties here, which we go through in the town-by-town price picture.

The tax effect runs the other way and is genuinely favourable to residents. Second homes are taxable property whose owners consume very little in municipal services: no children in the schools, a house occupied a fraction of the year, and no vote at town meeting. They enlarge the taxable base without enlarging the budget, which pushes the rate per thousand down for everybody. Several of the lowest residential tax rates in the region sit in exactly these towns, and why the cheapest rates are where they are covers the mechanism.

The cost lands elsewhere. A town where half the houses are used part-time struggles to keep a shop, a school or a fire roster, because the population that supports those things is the year-round one and it is smaller than the housing stock suggests. If you are moving to a town high in this table, that is the thing to look at, not the tax rate.

Every Town

Sortable and filterable. 17 towns have more than a fifth of their entire housing stock in seasonal use. Otis has the most seasonal units in absolute terms at 766.

One Limit Worth Naming

The census counts a unit as seasonal when nobody lives there as a usual residence and it is kept for part-year use. That single category holds the family camp on a pond, the restored farmhouse used eight weekends a year, and the house bought purely to let out by the week. Those are very different things for a neighbourhood and the count cannot distinguish them.

It is also a snapshot taken on one day in April, which is a specific and not especially representative moment in a region whose second-home season runs from June to October.

If you are weighing a town where this column is large, what the Berkshire numbers actually say is the companion piece, since that county is where most of this housing sits. For a specific property rather than a town pattern, look the address up.