Regional analysis · Allegheny County (Pittsburgh), Pennsylvania

More recorded storms than almost anywhere. Insurance costs that don’t follow.

Across the 100 most populous US counties, Allegheny — the county that contains Pittsburgh — has more recorded severe weather than all but two, yet only 2.9% of its mortgaged households report paying more than $3,000 a year to insure the home. In Oklahoma County, with fewer recorded events, 24.0% do. That is an 8.3× difference. One likely reason is what a standard policy leaves out.

2.9%of Allegheny mortgaged households report paying $3,000+ a year — against 24.0% in Oklahoma County, which recorded fewer severe weather events

Download the complete 100-county dataset (CSV) · Methodology and sources

1,108severe weather events recorded 2015–2024, against a median of 285 across the 100 counties
58.8%of mortgaged households report paying under $1,000 a year to insure the home
2.9%pay more than $3,000 — the 6th-lowest share among the 100
Zone 1EPA radon designation, the highest of three

Look up any of the 100 counties

NOAA records reports of severe weather, not a census of it, so counts describe how often something damaging was written down. Insurance shares are what mortgaged households told the Census they pay, in bands — there is no average premium in this data and none should be quoted from it.

The finding

Storm frequency and insurance price have come apart here more sharply than anywhere else in the dataset.

Between 2015 and 2024 the National Oceanic and Atmospheric Administration recorded 1,108 severe weather events in Allegheny County: 528 thunderstorm wind events, 330 floods and flash floods, 115 hailstorms and 12 tornadoes. That is the 3rd-highest count among the 100 counties examined and close to four times the median of 285. On thunderstorm wind alone the county ranks 2nd; on flooding, 4th.

Insurance prices in Pittsburgh and the rest of the county do not read like that at all. 58.8% of Allegheny's 206,078 mortgaged households told the American Community Survey they pay less than $1,000 a year to insure the home — the 7th-highest share reporting that band in the dataset. Only 2.9% pay more than $3,000.

Oklahoma County, Oklahoma records 864 severe weather events over the same decade — fewer than Allegheny — and 24.0% of its mortgaged households pay over $3,000. Allegheny records more weather and charges a quarter as many households that much.

CountyStorm events
2015–2024
Pay under $1,000Pay over $3,000
Maricopa, AZ 1,282 50.0% 3.9%
El paso, CO 1,128 22.8% 15.9%
Allegheny, PA 1,108 58.8% 2.9%
Cook, IL 1,034 39.8% 6.2%
Tarrant, TX 931 13.9% 20.8%
Fairfax, VA 904 38.4% 9.2%
Oklahoma, OK 864 18.0% 24.0%
Montgomery, MD 779 36.5% 9.0%

The eight counties with the highest recorded storm counts. Insurance shares are of mortgaged households, from ACS table B25141.

What the county actually gets

Wind and water, not the perils that usually drive a premium.

The composition matters more than the total. Allegheny's weather is overwhelmingly thunderstorm wind and flooding rather than the hail and tornado activity that drives pricing in the southern plains.

Thunderstorm Wind 528
Flash Flood 279
Hail 115
Flood 51
Winter Weather 44
Heavy Rain 28

Allegheny County by event type, NOAA Storm Events Database, 2015–2024.

That distinction is the most likely explanation, though this data cannot prove it. A standard HO-3 homeowners policy covers wind. It does not cover flood — flood is written separately, through the National Flood Insurance Program or a private equivalent, and a household that has not bought it is uncovered for the second most common event type in the county. Allegheny logged 330 floods and flash floods in ten years, 4th of the 100 counties.

So the low premium is not evidence that little happens here. It is partly evidence that the policy is not carrying what happens here. A 2026 national analysis on this site found the same pattern at scale: roughly three quarters of the building loss FEMA models for the United States falls on perils a standard homeowners policy excludes, inland flooding chief among them.

The second uninsured exposure

The county sits in EPA radon Zone 1, and three quarters of its housing predates 1980.

The Environmental Protection Agency places Allegheny in Zone 1, its highest category, meaning the predicted average indoor screening level exceeds 4 picocuries per litre — the level at which the EPA recommends action. 24 of the 100 counties carry that designation. Among them, Allegheny has the 4th-largest share of pre-1950 housing: 36.7% of its 604,558 units, with 77.2% built before 1980 — the 11th-oldest stock of the hundred. The counties with the oldest housing stock nationally are ranked on the same measure.

Construction era matters for radon because passive radon-resistant construction is a modern technique. The methods now written into model codes as radon-resistant new construction post-date the great majority of Allegheny's housing by decades, so most homes here were not built to resist it and were not tested when they were sold. No homeowners policy covers radon mitigation, and no premium reflects it. It is a several-thousand-dollar remediation that a buyer either finds during the inspection period or inherits — and if a test comes back high mid-purchase, who pays for radon mitigation is negotiable on a contingency and a deadline.

Both of the county's largest hazards — inland flooding and radon — sit outside the homeowners policy entirely. A cheap premium in Allegheny County is an accurate price for the narrow set of risks the policy actually carries, and says nothing about the two that it does not.

For newsrooms

Free to reuse, including the underlying county data.

This analysis and its data are published under a Creative Commons Attribution 4.0 licence, so a newsroom can republish the figures, redraw the charts, or run its own cut without asking. The three source datasets are all US Government works and public domain in their own right: the NOAA Storm Events Database, the Census Bureau's American Community Survey (year structure built, and table B25141 on homeowners insurance costs), and the EPA Map of Radon Zones.

If you are looking for Pittsburgh storm statistics, Allegheny County flood data, or home insurance costs by county, the numbers on this page are the whole of what we hold and the files below are the whole of the working. Two things are worth knowing before you quote them: NOAA counts are reports rather than a census of weather, and the ACS publishes insurance costs in bands, so there is no average premium here for any county.

Happy to pull the same cut for a different county, walk through the method, or check a figure before it prints — hello@beforeregret.com. If you find an error, say so and it will be corrected on the page with a note.

Method

How this was built.

Storm counts are every county-tagged entry in NOAA's Storm Events Database for the ten calendar years 2015–2024, summed by event type. Housing age and totals are American Community Survey 5-year estimates for year structure built. Insurance figures are ACS table B25141, which records what mortgaged households report paying, in bands — not quoted rates, and not a modelled premium. The radon designation is the EPA Map of Radon Zones. Land area is ALAND from the Census Bureau's 2023 Gazetteer file. The comparison set is the 100 counties in 31 states for which all five sources are complete.

One methodological choice is worth stating because it changes the answer. Comparing counties by events per square mile looks more rigorous than comparing raw counts and is not: across these 100 counties, event density correlates with population density at a Spearman coefficient of 0.681. NOAA's database records reports, and a storm that damages nothing in an empty place is often never written down, so density substantially maps where people live. Raw counts relate only weakly to land area (0.156) and to population (0.113), so raw counts are used throughout. Density is published in the dataset so the choice can be checked.

Limitations

What this cannot tell you.

NOAA's counts are reports of events, not a census of weather, and reporting density varies with population and with local spotter networks. They should be read as how often something damaging was recorded here, not as a physical storm frequency.

ACS insurance figures are self-reported and banded, so no county has a published mean and none is calculated here. They cover mortgaged households only; owners without a mortgage face no lender insurance requirement and are excluded from the table entirely. The figures do not distinguish policy form, deductible or coverage limit, so a cheap premium and an expensive one may be buying materially different things — which is part of the point, but it means the comparison is of price paid, not of value received.

Across all 100 counties the relationship between recorded storm events and the share paying over $3,000 is Spearman -0.101, Pearson -0.043 — effectively no relationship nationally. Allegheny is an extreme case of a pattern that is general, not an exception to a rule that otherwise holds.

Nothing here describes an individual property. Every figure is a county aggregate, and a specific house may face far more or far less than its county's profile.

Appendix

All 100 counties, ranked by recorded storm events.

#CountyEventsFlood<$1k>$3kPre-1980Radon
1 Maricopa, AZ 1,282 438 50% 4% 29% 2
2 El paso, CO 1,128 87 23% 16% 36% 1
3 Allegheny, PA 1,108 330 59% 3% 77% 1
4 Cook, IL 1,034 175 40% 6% 74% 2
5 Tarrant, TX 931 90 14% 21% 34% 3
6 Fairfax, VA 904 337 38% 9% 44% 1
7 Oklahoma, OK 864 134 18% 24% 55% 3
8 Montgomery, MD 779 280 37% 9% 49% 1
9 Tulsa, OK 769 169 18% 14% 53% 3
10 Davidson, TN 695 57 33% 9% 42% 1
11 Duval, FL 692 93 27% 12% 41% 3
12 Denton, TX 672 59 12% 25% 13% 3
13 San bernardino, CA 655 443 40% 7% 44% 2
14 Wake, NC 650 156 35% 5% 19% 2
15 Bexar, TX 647 191 29% 9% 37% 3
16 Baltimore, MD 620 266 42% 6% 59% 1
17 Dallas, TX 593 114 18% 21% 46% 3
18 Pima, AZ 549 187 59% 3% 39% 2
19 Collin, TX 545 33 11% 26% 11% 3
20 Montgomery, PA 527 114 36% 7% 62% 1
21 Harris, TX 524 105 22% 23% 38% 3
22 Worcester, MA 511 140 29% 6% 65% 1
23 Middlesex, MA 495 138 27% 11% 69% 1
24 Jackson, MO 480 80 23% 9% 65% 1
25 Fresno, CA 468 263 44% 6% 49% 2
26 Jefferson, KY 461 120 31% 7% 63% 1
27 Kern, CA 445 308 51% 4% 44% 2
28 San diego, CA 445 242 35% 10% 51% 3
29 Shelby, TN 440 67 27% 11% 55% 3
30 Hamilton, OH 439 110 44% 4% 73% 1
31 Will, IL 438 58 31% 5% 37% 2
32 Travis, TX 431 101 25% 12% 25% 3
33 Franklin, OH 415 78 49% 3% 52% 1
34 Middlesex, NJ 402 86 40% 4% 58% 2
35 Dupage, IL 395 62 35% 6% 54% 2
36 Philadelphia, PA 383 92 39% 5% 84% 3
37 Riverside, CA 380 206 34% 7% 30% 2
38 Hennepin, MN 370 13 26% 14% 60% 1
39 Suffolk, NY 364 110 17% 18% 69% 3
40 Cuyahoga, OH 359 57 51% 4% 81% 2
41 Prince george's, MD 346 116 32% 8% 54% 2
42 Bergen, NJ 342 167 30% 10% 75% 2
43 Honolulu, HI 339 68 45% 14% 58% 3
44 Lake, IL 339 39 33% 9% 47% 2
45 Clark, NV 337 150 59% 3% 18% 3
46 Miami-dade, FL 333 131 23% 49% 51% 2
47 Milwaukee, WI 316 21 56% 2% 80% 2
48 Jefferson, AL 305 50 35% 7% 58% 2
49 Palm beach, FL 304 66 17% 46% 34% 3
50 Erie, NY 285 56 62% 2% 77% 1
51 Broward, FL 284 73 23% 48% 50% 3
52 Marion, IN 279 41 40% 4% 59% 1
53 St. louis, MO 270 32 23% 12% 69% 2
54 Suffolk, MA 269 51 32% 19% 73% 3
55 Monroe, NY 265 36 70% 1% 70% 2
56 Wayne, MI 248 32 45% 4% 81% 3
57 Westchester, NY 247 114 28% 21% 77% 3
58 Fulton, GA 246 50 30% 13% 33% 1
59 Norfolk, MA 237 77 25% 14% 68% 2
60 Essex, MA 232 79 25% 11% 69% 1
61 Oakland, MI 229 13 37% 6% 59% 2
62 Mecklenburg, NC 228 35 38% 6% 26% 3
63 Hidalgo, TX 225 57 39% 6% 18% 3
64 Gwinnett, GA 222 12 26% 8% 14% 1
65 Dekalb, GA 202 17 28% 9% 45% 1
66 Providence, RI 197 82 28% 10% 76% 2
67 Denver, CO 194 9 28% 16% 57% 1
68 Salt lake, UT 193 16 60% 4% 44% 2
69 San francisco, CA 191 74 28% 16% 77% 2
70 Santa clara, CA 189 115 38% 9% 60% 2
71 Pinellas, FL 185 8 20% 27% 59% 3
72 Nassau, NY 184 79 16% 20% 86% 3
73 Macomb, MI 181 23 48% 2% 58% 3
74 Alameda, CA 170 125 32% 11% 66% 2
75 Orange, FL 156 11 19% 20% 27% 3
76 Los angeles, CA 143 89 40% 11% 71% 2
77 Cobb, GA 141 13 27% 9% 28% 1
78 Bernalillo, NM 139 25 47% 4% 48% 1
79 Hudson, NJ 135 67 37% 10% 64% 2
80 Fort bend, TX 134 25 18% 29% 13% 3
81 Sacramento, CA 133 110 44% 5% 49% 3
82 El paso, TX 126 38 46% 4% 40% 3
83 Orange, CA 119 65 37% 10% 57% 3
84 Hillsborough, FL 118 20 20% 22% 31% 2
85 Essex, NJ 118 75 27% 12% 75% 2
86 Queens, NY 112 59 33% 12% 84% 3
87 Lee, FL 108 20 26% 22% 21% 3
88 Utah, UT 108 19 60% 4% 26% 2
89 Polk, FL 101 1 26% 13% 32% 2
90 San mateo, CA 85 42 31% 13% 73% 2
91 Bronx, NY 85 26 33% 15% 80% 3
92 Contra costa, CA 70 44 30% 12% 54% 2
93 Kings, NY 68 28 31% 21% 81% 3
94 Ventura, CA 57 44 36% 11% 56% 1
95 San joaquin, CA 48 33 41% 6% 45% 3
96 New york, NY 41 19 50% 16% 79% 3
97 Multnomah, OR 32 5 54% 5% 61% 2
98 King, WA 11 1 39% 9% 47% 3
99 Snohomish, WA 8 5 45% 6% 36% 3
100 Pierce, WA 5 1 42% 6% 43% 3
Cite this Before Regret, “Recorded storm frequency and reported home-insurance costs in Allegheny County.” Analysis of NOAA Storm Events, ACS B25141 and EPA radon zones across 100 US counties. Full county dataset: CSV · JSON. Questions and corrections: hello@beforeregret.com
Embed the county lookup Free to use, no tracking, no attribution required beyond the credit line it carries.
<iframe src="https://www.beforeregret.com/research/allegheny-storm-premium/embed/"
        width="100%" height="420" style="border:1px solid #e5e0d8"
        title="Storm frequency and insurance cost by county"
        loading="lazy"></iframe>
It posts its own height to the parent page as beforeRegretEmbedHeight if you want it to resize with the content; without that listener the fixed height above still works.