The finding
The restoration job size distribution is dominated by small losses. Across 7,974,241 owner-occupied housing inspections recorded by FEMA Individual Assistance between 2002 and 2026, 68.3% of inspections fall in the $1 to $10,000 inspected damage band, while only 2.6% exceed $30,000. Catastrophic losses carry the trade coverage. They are not where the volume sits.
Headline figures
In this report
Key findings
- 68.3% of the 7,974,241 owner-occupied inspections in the FEMA file fall in the $1 to $10,000 inspected damage band, covering 5,444,521 inspections.
- Only 2.6% of inspections, 207,682 of them, record inspected damage above $30,000.
- 21.1% of inspections, 1,679,659, found no inspected damage at all. An inspection is not the same as a loss.
- Restricting to inspections that found damage, the picture barely changes: 86.5% of 6,294,582 damaged properties sit under $10,000 and 3.3% sit above $30,000.
- Severity is not evenly spread. California records 16.9% of its inspections above $30,000 against 0.16% in Massachusetts, a 106 times spread across the 28 states that clear the reporting floor.
- Average inspected damage per inspection runs from $36,784 in California down to $1,893 in Rhode Island.
- The two middle bands together, $10,001 to $30,000, account for just 8.1% of inspections, which is less than one inspection in twelve.
- The file covers 630 declared disasters across 2,951 county records and 30,350 postcodes.
The distribution in one chart
The shape is the finding. One band holds 68.3% of all inspections and the three bands above it hold 10.7% between them. That is not a bell curve with a long tail. It is a wall at the bottom of the scale with very little above it.
The full band distribution
| Inspected damage band | Inspections | Share of all inspections | Share of damaged properties |
|---|---|---|---|
| No inspected damage | 1,679,659 | 21.1% | n/a |
| $1 to $10,000 | 5,444,521 | 68.3% | 86.5% |
| $10,001 to $20,000 | 464,598 | 5.8% | 7.4% |
| $20,001 to $30,000 | 177,781 | 2.2% | 2.8% |
| Above $30,000 | 207,682 | 2.6% | 3.3% |
The bands are the dataset’s own, so there is no discretion in where the cut points sit. That matters for a distribution claim, because a bucket boundary chosen after the fact can manufacture almost any shape. The $1 to $10,000 band is wide, spanning an order of magnitude, and the three bands above it are narrow by comparison. Even so, the wide bottom band holds 5,444,521 of 7,974,241 inspections.
Reading it as damaged properties only
21.1% of inspections found no inspected damage. Those are real inspections of real properties, but they are not restoration work, so they belong outside the distribution when the question is about job size. Removing them leaves 6,294,582 inspections that recorded damage.
The distribution hardly moves. 86.5% of damaged owner-occupied properties sit in the $1 to $10,000 band and 3.3% sit above $30,000. Whichever denominator is used, between two thirds and three quarters of recorded damage events are small-dollar events.
Severe damage share by state
| State | Inspections | Above $30,000 | Severe share | $1 to $10,000 share |
|---|---|---|---|---|
| California | 80,064 | 13,549 | 16.92% | 44.8% |
| Oklahoma | 29,065 | 2,052 | 7.06% | 66.0% |
| New York | 229,993 | 15,984 | 6.95% | 58.7% |
| Iowa | 50,795 | 2,715 | 5.35% | 76.0% |
| Kentucky | 68,268 | 3,368 | 4.93% | 67.1% |
| Arkansas | 25,010 | 1,215 | 4.86% | 67.1% |
| Louisiana | 1,002,538 | 47,846 | 4.77% | 59.0% |
| Missouri | 46,874 | 2,083 | 4.44% | 66.3% |
| Mississippi | 347,561 | 15,005 | 4.32% | 68.0% |
| New Jersey | 204,593 | 8,761 | 4.28% | 70.3% |
| Tennessee | 77,412 | 3,235 | 4.18% | 72.4% |
| West Virginia | 51,844 | 1,514 | 2.92% | 73.9% |
The spread is wide enough to be a capacity finding rather than noise. California runs at 16.9% severe against a national 2.6%, and Massachusetts runs at 0.16%. A firm serving one of the states near the top of this table will meet large losses several times more often than a firm of the same size near the bottom, for reasons that sit in the hazard mix rather than in anything either firm does.
The hazard mix behind these numbers differs by state. Our disaster demand volatility report sets out how unevenly declared events arrive year to year, which is the other half of the same capacity question.
Average inspected damage by state
| State | Inspections | Average inspected damage |
|---|---|---|
| California | 80,064 | $36,784 |
| New York | 229,993 | $8,877 |
| Oklahoma | 29,065 | $8,754 |
| Mississippi | 347,561 | $7,749 |
| Louisiana | 1,002,538 | $7,584 |
| Missouri | 46,874 | $7,115 |
| Arkansas | 25,010 | $6,918 |
| Kentucky | 68,268 | $6,775 |
| Iowa | 50,795 | $6,547 |
| New Jersey | 204,593 | $6,515 |
| Tennessee | 77,412 | $5,704 |
| West Virginia | 51,844 | $5,265 |
| Texas | 1,028,880 | $4,677 |
| Indiana | 79,624 | $4,223 |
| Florida | 1,799,754 | $3,742 |
This figure is computed from the summed dollar total and the summed inspection count, not by averaging the per-record averages the dataset already carries. Averaging an average would weight a three-inspection postcode the same as a thirty-thousand-inspection county, which would be wrong. The national figure on the same basis is $5,003 per inspection.
What this means for how a firm is built
What follows is our reading of the figures above, not something the FEMA data states.
A restoration business organised around catastrophic work is organised around the minority of events. 2.6% of owner-occupied inspections in this file record damage above $30,000. The capability to handle those losses has real value, and large losses carry disproportionate revenue, but the day-to-day arrival rate of work looks like the bottom band. Our restoration capacity index sets that disaster load against the firms available in each state.
Read against the band table, the operational implication is about throughput rather than scale. A pipeline that turns over 5,444,521 inspections’ worth of small jobs rewards fast scheduling, tight documentation and low fixed cost per job. A structure built for the 207,682 largest events rewards equipment depth and project management. Those are different businesses, and the distribution says which one meets more events. Our remediation establishment census gives the average number of employees per firm in each state.
The counterweight is value concentration. These bands count events, not dollars, and a single loss above $30,000 can be worth more than ten jobs at the bottom of the scale. The distribution describes frequency. It does not describe where revenue lands, and this dataset cannot answer that question because it records assistance assessments rather than contractor invoices.
Method
How we counted. We retrieved 160,390 rows from the FEMA OpenFEMA HousingAssistanceOwners v2 dataset on 3 October 2026. Each row is an aggregate for one declared disaster, state, county, city and postcode, so we sum the band counts rather than count rows. Summing the five inspected damage band fields gives 7,974,241 inspections, which matches the totalInspected field exactly, so no reconciliation was needed. Shares are that band count divided by the same total. Average inspected damage is summed totalDamage divided by summed totalInspected, computed per state and nationally. We did not take a mean of the averageFemaInspectedDamage field, because averaging pre-averaged values weights a three-inspection postcode equally with a county of thirty thousand. State tables are limited to the 28 states with at least 20,000 inspections. Territory codes (AS, FM, GU, MP, VI) are excluded from state rankings because population and reporting bases are not comparable, though they remain in the national totals. No inflation adjustment is applied, so dollar bands across 2002 to 2026 are nominal.
Reproduce this
Reproduce this
- Endpoint
https://www.fema.gov/api/open/v2/HousingAssistanceOwners- Query
$top=5000 paged with $skip- Rows retrieved
- 160,390 owner rows
- Retrieved
- 3 October 2026
The API pages at 5,000 rows via $top and $skip. No API key is required. Re-running this query will return more rows as later disasters are added to the file.
What this does not show
What this does not show
- FEMA-inspected damage is an assistance assessment, not a restoration cost. The figure records what an inspection assessed for the purpose of determining eligibility for assistance. It is not a contractor estimate, not an invoice and not an insured loss.
- Declared disasters only. Every row in this file belongs to a presidentially declared disaster. Undeclared events, which are the large majority of water losses, do not appear at all.
- Registrants only. A property enters this dataset when a resident registers with FEMA. Households that did not register, including those with insurance that covered the loss, are absent.
- Owner-occupied only. This report uses the owners file. Renter records use a different severity schema and are covered separately in our owner renter damage gap report.
- Events, not dollars. The bands count inspections. They say nothing about how revenue is distributed across job sizes.
- No inflation adjustment. The dollar band boundaries are fixed in nominal terms across the whole period, so a loss that would have cleared $30,000 in 2024 might not have in 2004.
- Aggregated geography. Rows are pre-aggregated by FEMA to disaster, state, county, city and postcode. No property-level record is available, so no distribution within a band can be recovered.
Most quotable figures
- 68.3% of owner-occupied FEMA housing inspections record inspected damage between $1 and $10,000, against 2.6% above $30,000 (FEMA OpenFEMA HousingAssistanceOwners v2, 2002 to 2026, n = 7,974,241 inspections).
- Excluding inspections that found no damage, 86.5% of 6,294,582 damaged owner-occupied properties still sit under $10,000 of inspected damage (FEMA OpenFEMA, retrieved 3 October 2026).
- Severe damage share ranges from 16.9% of inspections in California to 0.16% in Massachusetts, a 106 times spread (FEMA OpenFEMA, 28 states above 20,000 inspections).
- Average FEMA-inspected damage is $5,003 per owner-occupied inspection nationally, computed from summed dollars over summed inspections (FEMA OpenFEMA, retrieved 3 October 2026).
Questions this report answers
What is the typical restoration job size?
In FEMA owner-occupied inspection records covering 2002 to 2026, 68.3% of inspections fall between $1 and $10,000 of inspected damage. That is the modal band by a wide margin. The figure is an assistance assessment rather than a restoration invoice, so it indicates scale rather than price.
How common are large losses?
Uncommon. 2.6% of the 7,974,241 owner-occupied inspections in the FEMA file record damage above $30,000, which is 207,682 inspections across 630 declared disasters.
Why do so many inspections show no damage?
21.1% of inspections recorded no inspected damage. A FEMA inspection follows a registration, and a registration follows a disaster declaration covering the area, so properties are inspected that turn out to be undamaged.
Which states see the most severe damage?
Of states with at least 20,000 inspections, California records the highest share above $30,000 at 16.9%, followed by Oklahoma at 7.1% and New York at 6.9%.
Does this include burst pipes and appliance leaks?
No. The dataset covers presidentially declared disasters only. An ordinary plumbing failure or appliance leak never enters it, which means the true population of small water losses is larger than this distribution shows, not smaller.
Is FEMA-inspected damage the same as restoration cost?
No. It is an assessment made to determine eligibility for assistance. Restoration cost depends on labour rates, materials, contamination category and scope, none of which this field records.
Can I reuse these figures?
Yes. The tables and derived figures are released under CC BY 4.0 with attribution to Remediation Insider. The underlying FEMA data is public and reproducible from the endpoint listed above.
How to cite
Remediation Insider. The Restoration Job Size Distribution: Where the Work Actually Sits. https://remediationinsider.com/restoration-job-size-distribution/ (retrieved 2 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 Housing Assistance Owners, v2. 160,390 aggregated rows covering 630 declared disasters, retrieved 3 October 2026.
- FEMA OpenFEMA Terms and Conditions.
- Damage band definitions and field names are taken from the FEMA OpenFEMA data dictionary for the HousingAssistanceOwners dataset.
- No survey, interview or physical inspection was conducted for this report. Every figure is computed at build time from the retrieved FEMA file.
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).