Methods
How the index works
Nine measures, each scored against the other 68 towns, in two halves: what it takes to get into a home, and what it costs to stay.
The idea
The whole Upper Valley is expensive, so the index does not ask whether a town is affordable in absolute terms. It asks how each town compares with the other 68 on the costs a household actually faces, from the down payment to the drive to work.
Getting in covers the cost of getting into a home: what homes sell for, what rent runs, and whether the town is adding homes. Staying covers what it costs to live there each year once you have: property tax, the commute, the upkeep of older houses and private wells and septic systems, and the two budgets that drive future tax bills, schools and roads. The two halves count equally.
The nine measures
Getting in
- Typical home value 30% of the score
The price of the typical home sets the down payment and the mortgage.
- Median rent 13% of the score
Renting is how most people get a start in a town.
- New homes permitted 8% of the score
Towns that build more have more homes for newcomers to buy or rent.
Staying
- Property tax on a $400,000 home 18% of the score
The same house pays very different taxes from town to town, every year.
- Cost of the commute 15% of the score
Every mile to work is fuel, wear and time; far-out towns trade cheaper homes for driving.
- Age of the homes 5% of the score
Older houses cost more to heat, insulate and repair.
- Public water and sewer 3% of the score
Wells and septic systems are the owner’s to maintain and replace.
- School spending per pupil 5% of the score
School budgets drive the largest part of the tax rate.
- Town roads to maintain 5% of the score
Every mile of town road is plowed, graded and rebuilt from the town budget.
How towns are scored
For each measure, the middle town scores 50. A town as affordable as the one at the 5th percentile of all 69 scores 100, and one as costly as the town at the 95th scores 0, with straight lines in between on each side. Towns beyond either end are held at 100 or 0, so one extreme town cannot squash the scale for everyone else. School spending is scored only against towns in the same state, because Vermont and New Hampshire measure it differently. Rent estimates whose margin of error is more than half the estimate are left out.
Each half's score is the weighted average of its measures, and the overall score is the average of the two halves. When a town has no figure for a measure (a few small towns have too few rentals for a rent estimate, and towns that issue no building permits are not counted for new homes), that measure is left out of its score and the others count proportionally more. The weights are ours; on the rankings you can set your own.
Notes on the numbers
- Home values are the Zillow Home Value Index for August 2026: the value of a typical home, the middle third of the market. Zillow reports Bellows Falls and Saxtons River apart from Rockingham, and Wells River apart from Newbury; those towns combine the pieces by their number of homes. Zillow does not report Baltimore, Vermont or Orange, New Hampshire, whose values are estimated from the Census home value. In the 25 Vermont towns with enough sales to compare, Zillow's values track the state's recorded sale prices closely.
- Property tax is compared on the same $400,000 home in every town, because towns assess property at different shares of its market value and a rate alone misleads. Vermont's are the 2026 (fiscal 2027) rates; the state sets the school rate for every town, and 2 towns' own rates had not been published where we could find them, so theirs are estimated from last year's. New Hampshire sets 2026 rates town by town from October to December; 19 of our 34 towns had theirs by November 19, and the rest use 2025's. The state has not yet published the equalization figures that turn either year into tax per $1,000 of market value, so those are estimated from the change in what each town levies and what its homes are worth. Tested on 2024, the estimate missed by 6% for the middle town. The Vermont figures are before the income-based credit that lowers many residents' school taxes.
- The commute is where residents' main jobs actually are, from the Census Bureau's 2023 job records, not the distance to one employer. The straight line from home to job is lengthened by how much longer roads run here (42 percent, measured on routes to Dartmouth Hitchcock), and jobs more than 60 miles away, mostly remote work, are left out. It is priced at the IRS rate of 76 cents a mile, which counts fuel, wear, insurance and depreciation.
- Rent, home age and income figures are Census survey estimates for 2020 to 2024, with margins of error that are wide in small towns.
- Income needed to buy, shown on the homes page, assumes 10 percent down and this week's average 30-year rate (7.03% on 2026-09-24), with the mortgage and property tax taking 30 percent of income. Insurance, mortgage insurance and upkeep would add more.
What the index leaves out
Incomes: the index compares what towns cost, not what their residents earn. Insurance, electricity, broadband and child care, for which no town-level figures exist. Quality: a town with no school of its own or a long drive to a grocery store may be cheap for a reason, and the index does not say whether the schools are good. And the individual house: a new home on public water in an expensive town can cost less to keep than an old farmhouse in a cheap one.
Town spending per resident is on the taxes page but not in the score. In towns with many second homes or businesses, much of the town budget is paid by people who don't live there, so it says little about what residents pay. The tax on the same $400,000 home measures that directly.