The Restoration Capacity Index: Disaster Load per Firm by State

What this index measures, and what it does not

There is no official count of water and fire restoration firms. That work is classified across construction and services codes including NAICS 236, 238990, 561790 and 561720, mixed with firms doing unrelated work. The only federal establishment count tied to remediation is NAICS 562910, Remediation Services, which covers asbestos and lead abatement, contaminated buildings and site cleanup. It is environmental remediation, not water and fire restoration. Our remediation establishment census sets out the code in full.

So the restoration capacity index sets declared disaster load against the closest available official firm count. It is a comparison of states on a common, reproducible basis, not a count of every crew that could respond. A state with few NAICS 562910 firms may still have many water restoration firms filed under other codes.

Disaster load per establishment

The numerator is the mean number of county-level FEMA declarations per year for restoration-relevant incidents from 2015 to 2025, the same series behind the restoration demand index. The denominator is private NAICS 562910 establishments in BLS QCEW for 2026Q1. Nationally that is 1,394 declared events a year against 9,365 establishments, or 0.15 per establishment, and the median state sits at 0.09.

The leaders pair a heavy FEMA declaration record with a modest base of remediation firms. In FEMA data since 2015, the leading incident type for each of the top five is: Kentucky, severe storm (59% of county-level events); Oklahoma, severe storm (44% of county-level events); Kansas, severe storm (81% of county-level events); Nebraska, severe storm (53% of county-level events); North Dakota, flood (44% of county-level events). At the other end, BLS QCEW counts 55 establishments in Delaware, 297 establishments in Massachusetts, 241 establishments in Michigan, while those states average between 0.4 and 3.9 declared events a year in FEMA data. 2 jurisdictions with fewer than 25 establishments (South Dakota, District of Columbia) stay in the full table but are left out of the chart and the headline ranking, because one firm opening or closing would move their ratio sharply.

Disaster load per 1,000 workers

Establishment counts treat a two-person firm the same as a two-hundred-person one. The second index divides the same FEMA load by QCEW employment, per 1,000 workers. BLS suppresses employment where publishing it would identify a firm, and 4 jurisdictions show no disclosed employment for 2026Q1: South Dakota, Vermont, Wyoming, District of Columbia. They carry no worker figure. A third index uses BLS OEWS employment for SOC 47-4041, hazardous materials removal workers, 2025 estimates, which counts an occupation rather than an industry and is available for every state except Delaware.

Do the measures agree?

Comparison	States compared	Spearman rank correlation
Per establishment vs per 1,000 QCEW workers	47	0.97
Per establishment vs per 1,000 OEWS 47-4041 workers	50	0.93
Per 1,000 QCEW workers vs per 1,000 OEWS 47-4041 workers	46	0.93

All three denominators produce similar orderings. That matters because each has its own weakness: establishment counts ignore firm size, QCEW employment is suppressed in small states, and OEWS figures are survey estimates rounded to the nearest ten. When three imperfect measures agree at a Spearman correlation of 0.93 or better, the broad ranking is more trustworthy than any single figure in it.

Capacity pressure tiers

Each state is placed in a tercile on the establishment measure and on the worker measure. The composite tier averages each state’s percentile position on the two measures, or uses the establishment measure alone where employment is suppressed, and cuts that average into thirds. The second table lists the states in the top third on both measures, which is the stricter test.

Composite tier	States	Median events per year	Median establishments	Median events per establishment
High	17	39.5	86	0.42
Medium	17	15.7	139	0.09
Low	17	4.6	183	0.03

State	Events per year	Establishments	Events per establishment	Events per 1,000 workers
Kentucky	104.7	134	0.78	100.0
Oklahoma	53.4	86	0.62	58.6
Kansas	43.0	70	0.61	65.4
Nebraska	31.6	53	0.60	108.0
North Dakota	17.6	34	0.52	106.9
Iowa	27.1	54	0.50	56.9
Mississippi	51.4	108	0.48	70.7
Louisiana	102.4	242	0.42	33.0
Georgia	95.3	241	0.40	54.7
Arkansas	36.0	92	0.39	58.5
South Carolina	62.0	179	0.35	64.4
Alabama	39.5	133	0.30	35.0
North Carolina	82.8	296	0.28	42.8
West Virginia	17.5	71	0.25	49.3
Missouri	47.8	212	0.23	24.7

The full state table

Rank	State	Events per year	Establishments	Per establishment	Per 1,000 QCEW workers	Per 1,000 OEWS workers	Tier	Note
1	South Dakota	23.1	19	1.215	n/a	121.5	High	small base; employment suppressed
2	Kentucky	104.7	134	0.782	100.0	227.7	High
3	Oklahoma	53.4	86	0.621	58.6	82.1	High
4	Kansas	43.0	70	0.614	65.4	89.6	High
5	Nebraska	31.6	53	0.597	108.0	150.6	High
6	North Dakota	17.6	34	0.519	106.9	103.7	High
7	Iowa	27.1	54	0.502	56.9	100.3	High
8	Mississippi	51.4	108	0.476	70.7	205.5	High
9	Louisiana	102.4	242	0.423	33.0	116.3	High
10	Georgia	95.3	241	0.395	54.7	207.1	High
11	Arkansas	36.0	92	0.391	58.5	81.8	High
12	South Carolina	62.0	179	0.346	64.4	71.3	High
13	Vermont	12.6	37	0.342	n/a	180.5	High	employment suppressed
14	Alabama	39.5	133	0.297	35.0	113.0	High
15	North Carolina	82.8	296	0.280	42.8	66.3	High
16	West Virginia	17.5	71	0.247	49.3	135.0	High
17	Missouri	47.8	212	0.226	24.7	129.2	Medium
18	Montana	9.8	55	0.179	25.2	39.3	High
19	Virginia	36.5	205	0.178	16.8	39.2	Medium
20	Tennessee	45.0	255	0.176	20.7	78.9	Medium
21	Texas	110.8	649	0.171	13.5	24.9	Medium
22	Florida	107.5	734	0.147	22.7	60.4	Medium
23	Minnesota	20.1	147	0.137	17.2	27.5	Medium
24	Maine	6.6	49	0.135	16.3	34.9	Medium
25	Oregon	15.7	124	0.127	10.8	13.9	Medium
26	Washington	26.2	295	0.089	4.2	8.1	Medium
27	Alaska	6.1	74	0.082	8.9	43.5	Medium
28	Wyoming	2.2	27	0.081	n/a	72.7	Medium	employment suppressed
29	New Hampshire	4.1	61	0.067	10.5	29.2	Medium
30	California	50.5	758	0.067	4.1	6.9	Medium
31	Idaho	6.2	96	0.064	8.8	10.3	Medium
32	Hawaii	3.1	58	0.053	5.9	9.1	Medium
33	Wisconsin	6.9	139	0.050	5.7	11.7	Medium
34	Rhode Island	2.4	52	0.045	9.5	7.2	Medium
35	New Mexico	4.6	107	0.043	3.2	6.8	Low
36	Nevada	4.5	114	0.040	5.1	11.4	Low
37	New York	16.6	492	0.034	2.5	4.1	Low
38	Utah	3.0	90	0.033	4.0	5.7	Low
39	Indiana	6.0	183	0.033	3.1	6.1	Low
40	Illinois	6.6	215	0.031	4.1	3.4	Low
41	Connecticut	5.1	169	0.030	3.1	6.8	Low
42	Arizona	4.0	146	0.027	3.8	5.0	Low
43	New Jersey	7.9	296	0.027	2.8	3.5	Low
44	Ohio	6.0	293	0.020	1.4	7.2	Low
45	Maryland	2.5	145	0.017	1.0	3.7	Low
46	Pennsylvania	5.8	350	0.017	1.6	3.7	Low
47	Colorado	5.0	314	0.016	2.0	3.6	Low
48	Michigan	3.8	241	0.016	1.9	2.4	Low
49	Massachusetts	3.9	297	0.013	1.3	1.4	Low
50	Delaware	0.4	55	0.007	1.3	n/a	Low
51	District of Columbia	0.1	19	0.005	n/a	1.8	Low	small base; employment suppressed

Territories, reported separately

Territory	Events per year	Establishments	Per establishment	Per 1,000 QCEW workers
Puerto Rico	66.1	22	3.00	154.1

Puerto Rico’s ratio of 3.00 declared events per establishment is 3.8 times Kentucky’s. BLS QCEW counts 22 NAICS 562910 establishments there, and hurricane declarations make up 71% of its county-level FEMA events. A plausible reading is that island logistics also make cross-border support harder to compare with a mainland state. It is not ranked with the states. The other territories in the FEMA file (AS, GU, MP, VI) have no QCEW row for NAICS 562910 and cannot be indexed.

Why it matters

What follows is our reading, not FEMA or BLS data. A high ratio means declared disaster load is large relative to the official remediation firm base. For an owner, that can mean longer waits for specialist work after a big event. For a firm, it can point to a market where demand outruns the local base, at least in declaration years.

Two cautions apply. Declarations are not jobs, and firms travel. Catastrophe crews cross state lines after large events, so local capacity is never the only capacity. The restoration demand volatility index shows how unevenly the load arrives, which matters as much as its average: a state with a high average but one surge year behaves very differently from a steady one.

Method

Reproduce this

What this does not show

Questions this report answers

Which state has the most disaster work per remediation firm?

Among states with at least 25 NAICS 562910 establishments, Kentucky, at 0.78 FEMA county-level declarations per year per establishment.

Does this count water damage restoration companies?

No. No federal code isolates water and fire restoration. The index uses NAICS 562910, environmental remediation, as the closest official firm count.

Why is Puerto Rico not in the ranking?

Its ratio of 3.00 is an extreme outlier driven by a small firm base and island-wide hurricane declarations, so it is reported separately.

Which states have the least disaster load per firm?

Delaware, Massachusetts, Michigan, per FEMA declarations and BLS QCEW establishment counts.

Why are some worker figures missing?

BLS suppresses QCEW employment where publishing it could identify an individual firm. Those states carry an establishment figure only.

Does a high ratio mean firms there are overwhelmed?

Not necessarily. Declarations are not jobs, many restoration firms sit outside NAICS 562910, and crews travel to catastrophe work. The ratio compares states; it does not measure backlog.

Can I reuse this data?

Yes. The tables and derived figures are released under CC BY 4.0 with attribution to Remediation Insider. The underlying FEMA and BLS data is public.