Crunching thousands of counties…
Crunching thousands of counties…
The data, interrogated
We pointed our whole database at one question: what actually explains why one place costs more than another, and what moves the market as a whole? Below are the real correlations, groupings and trends — every number computed live from thousands of counties, every chart drawn from the same data that powers the rest of the site. Some confirm what agents feel in their bones. A few are genuinely surprising. And one popular story turns out to be mostly a myth.
Method: each scatter is one dot per U.S. county, at its latest reading; the red line is a least-squares fit and r is the Pearson correlation (−1 to +1). Correlation is not causation — these are patterns, not proofs.
Finding 01 · Income
Nothing else in the entire dataset predicts a county's home values as tightly as the income of the people who live there. It is the single strongest relationship we found — richer counties are pricier counties, almost in lockstep.
That sounds obvious, but the strength is the story: it means "is this an expensive place?" is very nearly the same question as "do the people here earn a lot?" The famous exceptions — resort towns where second-home money outbids local wages — are exactly the dots that sit above the line.
One dot per county · r computed across n=3062 counties · Data: U.S. Census ACS (income), Zillow ZHVI (value).
Finding 02 · Education
The share of adults holding a four-year degree predicts home value almost as powerfully as income itself. Where diplomas cluster, so does price.
For an agent choosing a farm area, this is a quietly useful signal: education levels are published years ahead of the price pressure they foreshadow, and they move slowly. A county quietly gaining degree-holders is a county whose prices tend to follow.
Finding 03 · Investors
Here is the tension every rental investor lives with, drawn straight from the data: the more expensive the home, the lower the rent it returns as a share of its price. Cheap markets throw off fat yields; trophy markets barely cash-flow.
It is a strong inverse relationship, and it explains why out-of-state investors pour into affordable metros while the priciest coastal counties are dominated by owner-occupiers and appreciation bets. Cash flow and prestige sit at opposite ends of this line.
Finding 04 · Migration
Follow the tax returns. Counties that import the most income — not just the most people, but the most dollars of adjusted gross income moving in — tend to be the expensive ones. Wealthy movers flow toward already-costly places and, at the margin, push them costlier still.
This is the mechanism behind the headlines about California money reshaping Boise or Austin: it is not the raw headcount that reprices a market, it is the size of the wallets arriving.
Finding 05 · Climate
You would expect danger to be a discount. It isn't — at least not yet. Across the country, higher wildfire risk lines up with higher home values, not lower.
The reason is geography, not recklessness: the places with the most wildfire exposure are the desirable, dry, wooded, view-rich counties of California and the Mountain West — the same places people pay a premium to live. Risk and desirability are tangled together, and for now desirability is winning. It is a clean example of why correlation must never be read as causation.
Finding 06 · Interest rates
Zoom out from individual places to the country as a whole and one force towers over the rest: the mortgage rate. When Freddie Mac's 30-year rate spiked in 2022, national home-value growth rolled over within months — the red line up, the blue line down.
Rates don't change what a house is; they change what a monthly payment buys. That is why a single number set in Washington ripples into every county on every other chart on this page.
National, monthly since 2016 · Data: Freddie Mac PMMS (30-yr rate), Zillow ZHVI (year-over-year growth).
Finding 07 · The 2020 surge
This is every state at once — a tangle of faint gray threads — with the U.S. average burning through the middle in red. Two things jump out. First, the 2008 dip and the long climb after it were shared: the threads move together. Second, the 2020–2022 surge was not a coastal story — it lifted the entire bundle, cheap states and expensive states alike, in the steepest synchronized climb in the record.
The spread between the threads also matters: it has fanned wider over twenty years, which is the geography of affordability pulling apart — the top states pulling away from the bottom faster than the average alone reveals.
Finding 08 · Bidding wars
Where homes cost the most, buyers are also the most likely to pay over asking. The share of sales closing above list price rises with the price of the market — competition concentrates where the trophies are.
The link is real but looser than the income relationship, which is the honest reading: expensive markets tend toward bidding wars, but plenty of affordable, fast-growing metros run just as hot. Price level tilts the odds; it doesn't decide the game.
Finding 09 · Supply
New construction skews toward the more valuable counties. That is builders being rational — you break ground where the finished homes fetch the most — but it complicates the tidy "more supply, lower prices" intuition. At the county scale, permits and prices rise together, because demand is pulling both.
The affordability effect of new supply is real, but it shows up later and locally; in a single cross-section it's swamped by the simple fact that hot, expensive places are also where the cranes are.
Finding 10 · The myth
Raw headcount migration — how many more people moved in than out — barely relates to how expensive a county is. The dots are a cloud; the line is nearly flat. Plenty of cheap counties are gaining people fast, and plenty of expensive ones are quietly losing them.
Contrast this with Finding 04: it was the income of the movers, not their number, that tracked price. That is the real lesson buried in the migration story — count the dollars, not the U-Hauls.
Every relationship above is rebuildable on your own markets. Pull any two metrics into a chart and see the pattern for your county.
n=3063 counties · Data: ACS (share with a bachelor's degree), Zillow ZHVI.
n=1352 counties with both rent and value data · Gross yield = annual ZORI rent ÷ ZHVI value.
n=3044 counties · Data: IRS county-to-county migration (adjusted gross income arriving), Zillow ZHVI.
n=3063 counties · Data: FEMA National Risk Index (wildfire), Zillow ZHVI.
Each faint line is one state's ZHVI, quarterly since 2005; the red line is the U.S. average. Data: Zillow Research.
n=2973 counties · Data: Redfin (share of homes sold above list), Zillow ZHVI.
n=2971 counties · Data: Census Building Permits Survey, Zillow ZHVI.
n=3037 counties · Data: IRS net migration (people), Zillow ZHVI.