Research · Published 2026-08-01

Where America's oldest housing meets the worst hail

28.8 million Americans live in high-hail neighbourhoods where the typical home was built before 1980. The overlap is concentrated in a belt running from Iowa through Ohio.

Free to cite
Method published
Limitations stated
01Key findings

The numbers.

28.8M

people in high-hail tracts where the median home predates 1980

14.9M

of those live where the median home predates 1970

21.3%

of the US population lives in a high-hail tract (risk score above 80)

1959

median year built in Woodbury County, Iowa — the oldest housing in the high-hail set

02Findings

What the data shows.

Roof age is the single biggest driver of whether a hailstorm becomes a claim. An asphalt shingle roof has a service life of roughly 20 to 30 years, and a roof near the end of that window fails at hail sizes a newer roof shrugs off. So the question worth asking is not simply where hail falls — it is where hail falls on the oldest housing.

Joining FEMA's National Risk Index to Census American Community Survey housing data across 83,200 census tracts shows the overlap is not evenly spread. 28.8M people — 8.8% of the US population — live in a tract with a hail risk score above 80 where the median home was built before 1980. 14.9M live where the median home predates 1970.

The concentration is regional. Of the 25 largest counties where high hail risk meets pre-1970 housing, the majority sit in a belt running from eastern Iowa through Illinois, Indiana and Ohio — the older industrial Midwest, where housing stock is old for reasons that have nothing to do with weather, and hail exposure is high for reasons that have nothing to do with housing. The two simply coincide.

The exception at the top of the list is Potter County, Texas — Amarillo — which carries a hail risk score of 97.6, among the highest in the country, on housing whose median build year is 1963.

Counties over 100,000 people with a hail risk score above 80, ranked by oldest median housing
CountyHail riskMedian year builtMedian incomeOwner-occupiedPopulation
Woodbury, IA81.61959$73,21366.3%105,853
Bay, MI80.41960$62,44776.6%103,848
St. Louis, MO84.61960$80,54761.3%1,305,148
Macon, IL89.21961$61,97963.9%103,930
Allen, OH82.91961$65,11464.4%102,191
Columbiana, OH87.51962$58,87071.5%101,872
Madison, IN90.11962$61,34967.3%127,539
Stark, OH90.41963$68,48366.7%374,812
Black Hawk, IA84.81963$66,49363.7%131,044
Summit, OH87.01963$74,44163.8%540,333
Montgomery, OH86.91963$68,94561.9%537,193
Potter, TX97.61963$56,78355.2%118,498
Delaware, IN90.91965$54,93662.3%111,870
Sheboygan, WI90.11966$77,39073.2%117,195
Tazewell, IL93.61967$80,57878.0%131,288
Winnebago, IL93.01967$65,35863.2%285,160
LaPorte, IN82.31967$71,88573.5%112,304
Rock, WI81.81967$74,54769.4%163,608
Winnebago, WI88.91969$74,94964.6%171,623
Sangamon, IL83.61969$75,00268.2%196,182
03Method

How this was calculated.

FEMA National Risk Index hail data was joined to Census ACS housing data on census tract FIPS code, producing 83,200 tracts with both a hail risk score and a median year built. Tracts were aggregated to counties, with hail risk and housing measures averaged across a county's tracts and population summed.

The county table is limited to counties above 100,000 residents with an average hail risk score above 80, ranked by oldest median year built. The national figures use all tracts, unaggregated.

"High hail" throughout means a FEMA hail risk score above 80. The score is a composite of expected annual loss, social vulnerability and community resilience, expressed on a 0–100 scale.

04Limitations

What this does not show.

  • Median year built describes when a structure was built, not when its roof was last replaced. This is the most important caveat in the study. In high-hail areas insurance-funded replacement is common, so an old house may well carry a relatively new roof — the overlap identifies exposure, not confirmed roof age.

  • We do not have roof-age measurements at anything like this scale, so we cannot yet resolve that question. Testing it properly would require measured roof ages across a representative national sample, which does not currently exist in our data or anyone else's that we know of.

  • Census tract averages within a county can conceal wide variation. A county average of 1963 does not mean most homes were built near 1963.

  • ACS housing figures are survey estimates with their own margins of error, which are not carried through this aggregation.

05Citation

Using this study.

Free to cite with attribution. Underlying county and tract breakdowns are available on request, including cuts for a specific market.

RoofTap (2026). “Where America's oldest housing meets the worst hail.” https://www.rooftap.app/research/old-roofs-high-hail
Request data or an interview →
06Data

Download the full dataset.

The table above is truncated. This is the complete county-level set behind the study — 3,129 rows, covering 5 measures. Released under CC BY 4.0: use it anywhere, including commercially, with attribution.

Download CSV →Download JSON →
  • · Median year structure built
  • · FEMA hail risk score
  • · FEMA wind risk score
  • · Population
  • · Census tracts per county
07Check your own roof

Curious what this means for your address?

The studies above are national. This is your specific roof — size, pitch, and the storm history recorded at your address. Free, about six seconds.

08More research

Other studies.

The households most exposed to hail are the least able to pay the deductible

National census-tract analysis showing that median household income falls as modelled hail risk rises, and identifying the counties where high exposure, low income and high home-ownership coincide.

READ →

FEMA's hail risk score and actual hail events agree less than you would expect

Comparing FEMA National Risk Index hail scores against 161,403 NOAA storm events from 2021 to 2026, and identifying counties rated low-risk that report hail repeatedly.

READ →

DATA REFRESHED 2026-08-01