The finding
Inflation-adjusted flood damage shows no clean upward trend. Across 477,160 National Flood Insurance Program building claims with a loss year from 2015 to 2024, restated in 2024 dollars with the BLS CPI-U, the median claim swings between $17,179 (2020) and $66,458 (2024), a factor of 3.9. The swings track which large flood dominated each year’s claims, not a steady rise. Headlines that say flood damage is rising usually mix inflation and event mix together.
Headline figures
In this report
Key findings
- In 2024 dollars, the median NFIP building damage claim ranged from $17,179 in 2020 to $66,458 in 2024, a 3.9 times swing over loss years 2015 to 2024 (FEMA NFIP, BLS CPI-U).
- The real median rose in 6 of the 9 year to year steps and fell in 3. That is a sawtooth, not a trend.
- In nominal dollars the median grew 4.5 times from 2015 to 2024; after the CPI-U adjustment the same comparison is 3.4 times. Inflation accounts for the gap between those two multiples; what remains compares two different flood years, not a trend.
- The three loss years where one state supplied the largest share of claims (2017, 2022, 2024) are also the three years with the highest real median.
- Holding depth fixed at one foot of water, the real median still swings 4.3 times between 2015 and 2024, so composition goes beyond depth.
- Within Florida, the real median ranged from $13,957 (2015) to $80,921 (2024). Within Texas it ranged from $17,819 (2020) to $89,498 (2017). The two largest states peak in different years (FEMA NFIP, BLS CPI-U).
- The 99 foot depth sentinel appears 3 times in 2015 claims and 2,723 times in 2024 claims in this sample (FEMA NFIP), so depth cleaning matters more for recent years.
Real median damage by loss year
The chart restates the median building damage amount of every loss year in 2024 dollars. If flood damage per claim were rising steadily in real terms, the bars would climb. They do not. Large years and small years alternate: the highest year, 2024 at $66,458, sits beside 2023 at $32,338. Source: FEMA OpenFEMA NFIP claims and BLS CPI-U.
| Loss year | Claims | Median, nominal | Median, 2024 dollars |
|---|---|---|---|
| 2015 | 27,127 | $14,632 | $19,365 |
| 2016 | 72,017 | $39,322 | $51,394 |
| 2017 | 113,000 | $51,571 | $65,997 |
| 2018 | 34,586 | $16,500 | $20,612 |
| 2019 | 25,142 | $27,144 | $33,305 |
| 2020 | 24,729 | $14,174 | $17,179 |
| 2021 | 35,431 | $23,932 | $27,705 |
| 2022 | 43,766 | $54,191 | $58,086 |
| 2023 | 16,830 | $31,412 | $32,338 |
| 2024 | 84,532 | $66,458 | $66,458 |
How much of the change is inflation
The BLS CPI-U annual average rose 32.3% between 2015 and 2024. A claim assessed at $10,000 in 2015 carries the purchasing power of $13,235 in 2024, according to the BLS index. The table gives the multiplier for every year, so any figure from this report, or from the companion report on the water depth damage curve, can be restated on the same basis.
BLS had not published a 2025 annual average when the series was retrieved on 4 October 2026. The real series therefore stops at 2024 and is stated in 2024 dollars. We did not estimate or extrapolate a 2025 deflator. Claims with a loss year after 2024 (19,300 rows in the FEMA retrieval) are left out of every real figure; those years are also incomplete because claims are filed and settled after the loss.
| Loss year | CPI-U annual average | Multiplier to 2024 dollars |
|---|---|---|
| 2015 | 237.017 | 1.323 |
| 2016 | 240.007 | 1.307 |
| 2017 | 245.120 | 1.280 |
| 2018 | 251.107 | 1.249 |
| 2019 | 255.657 | 1.227 |
| 2020 | 258.811 | 1.212 |
| 2021 | 270.970 | 1.158 |
| 2022 | 292.655 | 1.072 |
| 2023 | 304.702 | 1.029 |
| 2024 | 313.689 | 1.000 |
Why the median swings: one state, one flood
Each loss year’s claims are not a stable cross-section of American flooding. They are dominated by whichever large flood happened that year, in whichever state it hit. The table shows the share of each year’s claims that came from the single largest state, next to the costliest flooding or tropical cyclone event in the NOAA billion-dollar disaster file for that year.
| Loss year | Largest state | Share of year’s claims | Real median (2024 $) | Costliest NOAA flood or cyclone event |
|---|---|---|---|---|
| 2015 | Texas | 38.1% | $19,365 | Texas and Oklahoma Flooding and Severe Weather (May 2015) |
| 2016 | Louisiana | 45.4% | $51,394 | Louisiana Flooding (August 2016) |
| 2017 | Texas | 67.5% | $65,997 | Hurricane Harvey (August 2017) |
| 2018 | North Carolina | 41.5% | $20,612 | Hurricane Michael (October 2018) |
| 2019 | Texas | 39.8% | $33,305 | Missouri River and North Central Flooding (March 2019) |
| 2020 | Florida | 30.2% | $17,179 | Hurricane Laura (August 2020) |
| 2021 | Louisiana | 39.7% | $27,705 | Hurricane Ida (August 2021) |
| 2022 | Florida | 83.3% | $58,086 | Hurricane Ian (September 2022) |
| 2023 | Florida | 51.1% | $32,338 | Typhoon Mawar (May 2023) |
| 2024 | Florida | 80.3% | $66,458 | Hurricane Helene (September 2024) |
The pattern is plain in the claims data: when one state supplies most of a year’s claims, that state’s flood sets the year’s median. A coastal surge that pushes water through thousands of homes in one state produces a different damage distribution from a year of scattered river and rainfall flooding across many states. Comparing the medians of two such years compares two different events, not the same risk at two points in time. That is why the question “is flood damage rising?” cannot be answered from yearly medians alone. Source: FEMA NFIP claims.
Like for like: one foot of water
One way to hold the claim mix steadier is to compare only claims with the same water depth. This control keeps claims recorded at one foot (from 1 foot up to, but not including, 2 feet), using the cleaned 0 to 10 foot depth range described below. It removes depth as a source of difference; it does not remove differences in building type, flood duration, contamination or state.
| Loss year | Claims at one foot | Real median (2024 $) | Depth recorded as 99 ft |
|---|---|---|---|
| 2015 | 9,570 | $19,175 | 3 |
| 2016 | 20,957 | $37,332 | 6 |
| 2017 | 47,289 | $60,973 | 38 |
| 2018 | 9,166 | $27,259 | 27 |
| 2019 | 6,694 | $44,923 | 474 |
| 2020 | 6,251 | $36,010 | 637 |
| 2021 | 9,593 | $49,291 | 800 |
| 2022 | 9,827 | $74,850 | 2,108 |
| 2023 | 4,333 | $61,588 | 929 |
| 2024 | 20,100 | $83,038 | 2,723 |
Holding depth fixed does not produce a flat line. The one foot real median runs from $19,175 in 2015 to $83,038 in 2024, a 4.3 times range, and 2 of its three highest years are among the three most concentrated years (2017, 2022, 2024). Like for like on depth is still not like for like on place. Source: FEMA NFIP claims.
Like for like: within the largest states
The second control compares each year only with itself inside one state. The four states with the most claims in the 2015 to 2024 sample are shown. Cells with fewer than 200 claims are marked rather than reported.
| Loss year | Florida | Texas | Louisiana | North Carolina |
|---|---|---|---|---|
| 2015 | $13,957 | $34,258 | $15,416 | $10,494 |
| 2016 | $32,203 | $45,879 | $84,013 | $18,120 |
| 2017 | $25,531 | $89,498 | $27,229 | under 200 claims |
| 2018 | $36,504 | $32,404 | $25,463 | $20,421 |
| 2019 | $24,975 | $60,130 | $31,817 | $17,050 |
| 2020 | $21,642 | $17,819 | $28,187 | $9,878 |
| 2021 | $19,458 | $19,033 | $50,315 | $19,521 |
| 2022 | $73,728 | $24,274 | under 200 claims | $7,904 |
| 2023 | $61,261 | $25,365 | under 200 claims | $8,517 |
| 2024 | $80,921 | $23,154 | $31,694 | $34,196 |
The two largest states do not share a pattern. Florida’s real median peaked at $80,921 in 2024; Texas’s peaked at $89,498 in 2017 and fell as low as $17,819 in 2020. Each state’s peak lines up with its own large flood year in the table above. Our reading of the within-state view is that damage per claim moves with the event, and that neither control shows a steady real increase that holds across states. State-level counts for the full period are in our flood claims by state report. Source: FEMA NFIP claims, BLS CPI-U.
Data quality treatment
The depth rules below are the same as in the water depth report, recomputed on the claims used here. The $1,000 damage floor and the loss year cut-off are added for this report.
| Rule | Claims affected | Treatment |
|---|---|---|
| Building damage under $1,000 or missing, loss years 2015 to 2024 | 130,426 | Excluded from every figure. Very small amounts are mostly closed or nominal claims, not repairs. |
| Loss year after 2024 | 19,300 | Excluded. No CPI-U annual average after 2024 was published at retrieval, and those years are incomplete. |
| Depth recorded as exactly 99 feet | 7,745 | Excluded from the depth control. Sentinel value. |
| Negative depth | 14,495 | Excluded from the depth control. Water below the first floor. |
| Depth above 20 feet (other than 99) | 16,175 | Excluded from the depth control. Not plausible at this frequency. |
| Depth above 10 and up to 20 feet | 11,798 | Outside the clean 0 to 10 foot band used for the control. |
| No depth recorded | 43,551 | Kept for the yearly and state medians, excluded from the depth control. |
After the damage floor the analysis sample is 477,160 claims. 383,396 of them carry a depth in the clean 0 to 10 foot band, and 143,780 sit in the one foot bucket used for the depth control. Source: FEMA NFIP claims.
Method
How we counted. We retrieved 626,886 NFIP redacted claim records with a loss year of 2015 or later from the FEMA OpenFEMA API on 4 October 2026, and the BLS CPI-U annual averages for 2015 to 2024 (series CUUR0000SA0) from the BLS Public Data API on 4 October 2026. We kept claims with a loss year from 2015 to 2024 and building damage of at least $1,000, which leaves 477,160 claims. Each claim’s building damage amount was multiplied by CPI-U 2024 divided by CPI-U of its loss year. We report medians, because the damage distribution is right skewed, and we report a cell only when it holds at least 200 claims. The depth control uses whole-foot bucket 1 from the clean 0 to 10 foot range. The state control uses the 4 states with the most claims in the sample. NOAA event names are taken from the NCEI billion-dollar file as yearly context only.
Reproduce this
Reproduce this
- Endpoint
https://www.fema.gov/api/open/v3/NfipClaims- Query
$filter=yearOfLoss ge 2015&$select=state,yearOfLoss,waterDepth,causeOfDamage,buildingDamageAmount,occupancyType- Rows retrieved
- 626,886 claims
- Retrieved
- 4 October 2026
The API pages at 5,000 rows via $top and $skip. No API key is required. The retrieval is the complete set of claims with a loss year of 2015 or later at the retrieval date; claims are filed and settled after the loss, so a later pull will return more rows.
Reproduce this
- Endpoint
https://api.bls.gov/publicAPI/v2/timeseries/data/- Query
POST, JSON body: {"seriesid":["CUUR0000SA0"],"startyear":"2015","endyear":"2024","annualaverage":true}- Rows retrieved
- 10 annual averages (2015 to 2024)
- Retrieved
- 4 October 2026
The endpoint accepts POST only; a GET request returns HTTP 405. Annual averages carry the period code M13. The API response noted that the year range is limited to ten years per request, which this range fits.
What this does not show
What this does not show
- Insured flood losses only. NFIP covers its policyholders. Uninsured flood damage and non-flood water losses are absent.
- CPI-U is not a construction cost index. It measures consumer prices in general. Deflating by it restates purchasing power; it does not track what repair materials or labour cost in each year.
- Damage amount is not restoration cost. The field is the assessed building damage on the claim, not what a contractor billed.
- Event mix dominates. Yearly medians compare different floods. Neither control (one foot of water, or a single state) removes every source of difference.
- Recent years are still filling. The retrieval holds every claim with a loss year of 2015 or later as at 4 October 2026 (626,886 claims), but claims arrive after the loss, so the latest years will grow on a later pull.
- No real figures after 2024. No published CPI-U annual average existed for 2025 at retrieval.
Most quotable figures
- In 2024 dollars, the median NFIP building damage claim swung 3.9 times between 2020 ($17,179) and 2024 ($66,458) (FEMA OpenFEMA, BLS CPI-U, n = 477,160).
- The real median NFIP claim rose in 6 of 9 years and fell in 3, loss years 2015 to 2024 (FEMA, BLS).
- In 2022, 83% of NFIP building claims came from a single state, Florida (FEMA OpenFEMA).
- US consumer prices (CPI-U) rose 32.3% from 2015 to 2024, so a 2015 flood claim needs a 1.323 multiplier to compare with 2024 (BLS).
Questions this report answers
Is flood damage per claim rising after inflation?
Not in a clean line. In NFIP claims from 2015 to 2024, the median in 2024 dollars rose in 6 year to year steps and fell in 3, ranging from $17,179 to $66,458 (FEMA, BLS).
Why are some years so much higher than others?
Because each year’s claims are dominated by its largest flood. When one state supplies most of the claims, that event sets the median for the whole year.
What dollars are the figures in?
2024 dollars, using the BLS CPI-U annual average. BLS had not published a 2025 annual average at retrieval, so 2024 is the latest base year available.
Why are 2025 and 2026 claims left out?
No CPI-U annual average for 2025 existed at retrieval, and claims for recent loss years are still being filed and settled. We did not invent a deflator for them.
Does adjusting for inflation give the repair cost?
No. CPI-U is a general consumer price index. It restates purchasing power. A construction cost index would be needed to track repair prices, and the claim amount is assessed damage, not a contractor invoice.
Does holding water depth fixed remove the swings?
No. At one foot of water the real median still ranges 4.3 times across 2015 to 2024 (FEMA, BLS).
Can I reuse this data?
Yes. The tables and derived figures are released under CC BY 4.0 with attribution to Remediation Insider. The claims and price data are public from FEMA and BLS.
How to cite
Remediation Insider. Inflation-Adjusted Flood Damage: A Decade of NFIP Claims in 2024 Dollars. https://remediationinsider.com/inflation-adjusted-flood-damage/ (retrieved 4 October 2026).
Data and tables in this report are released under CC BY 4.0. Attribution: Remediation Insider.
Sources
- Federal Emergency Management Agency. FEMA OpenFEMA NFIP Redacted Claims, v3. Claims with loss year 2015 or later. 626,886 records retrieved 4 October 2026.
- US Bureau of Labor Statistics. BLS Consumer Price Index for All Urban Consumers (CPI-U), series CUUR0000SA0, annual averages 2015 to 2024, Public Data API v2. Retrieved 4 October 2026.
- NOAA NCEI, U.S. Billion-Dollar Weather and Climate Disasters, final release covering 1980 to 2024 (DOI 10.25921/stkw-7w73). Event names only. Retrieved 4 October 2026.
- FEMA OpenFEMA Terms and Conditions.
- No survey, interview or physical inspection was conducted for this report. Every figure is computed from the retrieved public files.
This product uses the Federal Emergency Management Agency’s OpenFEMA API, but is not endorsed by FEMA. The Federal Government or FEMA cannot vouch for the data or analyses derived from these data after the data have been retrieved from the Agency’s website(s).