3.4 Benchmarking, targeting, and what a rate cannot tell you
“Is 2.1 good?” is the question every safety manager is asked and the one with no answer in the abstract. A rate means something against your industry, your establishment size, your state and your own previous year — in that order. Everything else is a scale check at best and a false reassurance at worst.
1 There is no OSHA threshold
No standard names a rate. No rate makes you compliant and no rate makes you non-compliant. Where a number appears in an OSHA program it is a targeting device, not a compliance line: the Site-Specific Targeting directive uses DART rates to build inspection lists, and a rate above the threshold means an inspection is likelier, not that a violation exists.
Customers set targets, and that is their right. A prequalification system that excludes contractors above a TRIR of 3.0 is a commercial rule, not a regulatory one. Knowing the difference matters when you are asked to explain a number: you are explaining a measurement, not defending an allegation.
2 The national picture, with its years attached
The Bureau of Labor Statistics measures the same three rates through its Survey of Occupational Injuries and Illnesses. These are the published figures for private industry, extracted from the BLS database in September 2026.
| Year | Total recordable cases | DART | Days away from work |
|---|---|---|---|
| 2024 | 2.3 | 1.4 | 0.8 |
| 2023 | 2.4 | 1.5 | 0.9 |
| 2022 | 2.7 | 1.7 | 1.2 |
| 2019 | 2.8 | 1.5 | 0.9 |
| 2014 | 3.2 | 1.7 | 1.0 |
Two things to take from it. The long trend is down — about 28% over eleven years. And the 2020 to 2022 block is distorted: the days-away rate jumped from 0.9 to 1.2 while the recordable rate barely moved, which is the signature of COVID-19 illness cases, recordable and almost always involving days away. Benchmarking a 2024 rate against 2021 flatters you for the wrong reason.
3 Four things that have to match
| Match | What goes wrong if you do not |
|---|---|
| Case type | A TRIR compared against a sector days-away rate makes you look about three times worse than you are |
| Year | BLS publishes roughly a year behind. Your current-year rate has no published benchmark yet |
| Universe | Private industry, all ownerships and public sector are three different populations |
| Level | The 300A is an establishment document and BLS measures establishments; a company-wide rate is a different figure |
Add establishment size to that list wherever you can. One case moves a small establishment's rate far more than a large one's, which is exactly why BLS breaks its own published rates into size categories and why OSHA's 2016 letter makes the point explicitly.
4 Finding the number for your own industry
BLS publishes industry detail in its annual incidence rate tables rather than as a time series, which is why it takes four steps rather than one.
- 1Find your six-digit NAICS code — it is on the 300A, on your workers compensation policy and on your business license.
- 2Open the BLS table of incidence rates by industry and case type, for the same year as your own rate.
- 3Read across to total recordable cases for the TRIR benchmark, and to cases with days away, restriction or transfer for DART.
- 4Check the establishment-size table in the same release if your establishment has fewer than 250 employees.
One sector comparison is worth carrying in your head, because it surprises people. In 2023, CPWR's Construction Chart Book — reproducing BLS tables — put the construction days-away rate at 1.0 per 100 full-time equivalents against 0.9 for private industry as a whole, tied fifth among the sectors listed. Construction stands out for its fatality rate, not for its nonfatal injury rate, and the same source records a 41.2% fall in construction's days-away rate between 2008 and 2023.
5 Four things a rate cannot tell you
- Whether the work is dangerous. It tells you how often harm was recorded. A low-rate establishment with no near miss reporting and no inspection findings is not demonstrating safety; it is demonstrating silence.
- Whether the program works. Rates lag by a year and move on small numbers. A change you made in March will not show up honestly until the year after next.
- Whether the next event will be serious. Case counts and fatality risk are different distributions. Industries with modest recordable rates and high fatality rates exist, and construction is one of them.
- Whether you are recording correctly. The rate is downstream of every classification decision in Part 1. An unusually clean number is a reason to audit the log, not to relax.
6 Reporting it to somebody else
Prequalification questionnaires rarely specify what they want, so answer the question they should have asked. Give the rate, the case count, the hours, the level (establishment or company), the year, and whether contractor and temporary worker hours are included. Six data points, one line each.
Two things not to do. Do not attach the 300 Log — a customer is not on the list of people entitled to your employees' names under 1904.29(b)(10), and the 300A carries everything they legitimately need. And do not quote a company rate as an establishment rate to clear a threshold. It is checkable, and it is the kind of thing that ends a relationship rather than a conversation.
- No OSHA standard sets an acceptable rate; thresholds in OSHA programs are targeting devices
- Private industry 2024: TRIR 2.3, DART 1.4, days away 0.8 — all ownerships 2.6, public sector 4.4
- Case type, year, universe and level all have to match before a comparison means anything
- Get your own number from the BLS table for your NAICS code and establishment size
- Construction’s days-away rate is close to the private-industry average; its fatality rate is not
- Report the rate with its case count, hours, level, year and population — never with the 300 Log attached
Free educational content — not OSHA-authorized training, no certificate or card issued. Follow your employer's program and the standards cited.