Your Injury Rate Is a Measure of Two Things, and One of Them Is How Your Sites Record
When sites classify recordable cases differently, corporate injury numbers compare recording practice as much as safety performance. Case-level submission makes that variation visible to a regulator.
Two of your plants report materially different injury rates. The obvious reading is that one is safer than the other. The other reading — that one records more conservatively than the other — is equally consistent with the data, and considerably more common than most organisations assume.
Until you can rule the second explanation out, every site comparison you make is partly a comparison of paperwork.
Where the judgement actually sits
29 CFR Part 1904 is more prescriptive than its reputation suggests. Whether a case is recordable turns on defined criteria — death, days away from work, restricted work or transfer, medical treatment beyond first aid, loss of consciousness, or diagnosis of a significant injury or illness by a licensed health care professional. First aid is itself a defined list rather than a judgement call, and the distinction between first aid and medical treatment beyond first aid is where a great deal of site-to-site variation originates.
Two more areas generate disagreement reliably:
Work-relatedness. The standard presumes work-relatedness where an event or exposure in the work environment caused or contributed to the condition, with specific exceptions. Sites vary in how readily they apply those exceptions, particularly for conditions with a plausible non-work explanation.
Restricted work. Whether a case involves restricted work or transfer depends on whether the employee is kept from performing routine functions or from working a full shift. Sites with flexible job assignments can often accommodate an injured employee without formally restricting them, which produces a genuinely different classification for a genuinely similar injury.
None of this requires anyone to act in bad faith. It requires only that two competent people, applying a standard with real judgement in it, land differently — repeatedly, over years, at different locations.
Why it goes unnoticed
Each site keeps its own log. For decades the OSHA 300 was an establishment-level document. Nobody read forty of them side by side, so systematic differences never surfaced.
Rates are compared, methods are not. Corporate reporting compares outputs. It rarely asks how each site arrived at its numbers, and the number that arrives at head office carries no indication of the judgement behind it.
Low numbers do not get investigated. A site reporting a rate well below the group average is treated as a success story and studied for practices to copy. A site reporting a rate well above is investigated. The asymmetry means conservative recorders get scrutiny and permissive recorders get praised.
Incentives point one way. Where site leadership is measured on injury rate, the pressure on marginal classification decisions is not neutral. This does not require anyone to falsify anything — it operates entirely within the space where the standard permits judgement.
Case-level submission removes the cover
For establishments subject to the case-level electronic submission requirements, the underlying 300 and 301 data goes to OSHA, not just the annual summary.
That changes the visibility of the problem completely. Variation that was previously distributed across forty separate logs, seen only by the people who kept them, arrives in one place, attributable to one employer, in a form that can be compared. A regulator looking at your submissions can see that two of your plants treat similar cases differently, and they do not have to visit either site to see it.
The response to this is emphatically not to reclassify historical cases to look consistent. Altering records to produce a tidier picture is a substantially worse problem than the inconsistency it conceals. The response is to make the current decision-making consistent, and to be able to explain the historical variation honestly if asked.
Getting to consistency
Find out where you actually are. Take eight to ten real anonymised scenarios — the marginal ones, not the obvious ones — and ask the person responsible for recordkeeping at each site to classify them independently. The spread in the answers is your baseline. Most organisations running this exercise for the first time are surprised by it.
Train on the boundaries, not the standard. Everyone knows a fatality is recordable. The value is entirely in the grey band: first aid versus medical treatment, the work-relatedness exceptions, what constitutes restricted work. Build the training from your own scenarios.
Give marginal cases somewhere to go. A named person or small group who can be consulted on a genuinely ambiguous case, whose decisions are recorded and become precedent. This is the single highest-leverage change, because it converts forty independent judgements into one line of reasoning applied forty times.
Record the reasoning on close calls. Not for every case — for the ones where a competent person could have gone either way. A short note on why produces both consistency over time and an explanation if the case is ever questioned.
Separate the measure from the incentive. Where site leadership is evaluated on recordable rate, consider whether leading indicators — hazards reported and closed, inspections completed, training currency — carry more of the weight. The organisations with the most trustworthy injury data tend to be the ones that do not punish sites for reporting.
The number you get afterwards
Consistent classification usually makes the group rate go up, at least initially. Sites that were under-recording start recording, and nothing about the underlying safety performance has changed.
That is worth preparing leadership for before it happens, because a rising number that reflects better measurement is easy to misread as deteriorating performance — and the reaction to that misreading tends to push the recording practice straight back where it came from.
General guidance only, not legal advice. Recordability determinations are fact-specific. Verify against 29 CFR Part 1904 and consult a qualified professional on particular cases.
OSHA standards cited
- 29 CFR Part 1904
Recording and Reporting Occupational Injuries and Illness
Always verify current OSHA standards at osha.gov. This article reflects standards in effect at the date of publication.
Not sure where you stand?
Take the 5-minute compliance assessment. Answer a few questions about your business and get a prioritized list of what OSHA expects, free.
Related Articles
Continue learning about OSHA compliance and workplace safety
Recordkeeping
When a Site Changes Ownership Mid-Year, the OSHA 300 Log Doesn't Reset — It Splits
29 CFR 1904.34 splits OSHA 300 log responsibility at the ownership transfer date. Multi-site operators need a recordkeeping handoff built into every deal.
Recordkeeping
Electronic Injury Reporting Is Per Establishment, Not Per Company — And That's Where Multi-Site Operators Get Caught
OSHA's electronic submission rules under 29 CFR 1904.41 apply establishment by establishment. Multi-site operators routinely submit for headquarters and miss the sites that actually trigger the requirement.
Construction Safety
Excavation's Competent Person Is Designated at the Site. Corporate's Org Chart Assigns the Role by Title.
OSHA's excavation standard, 29 CFR 1926.651, requires a competent person who evaluates that specific site's soil daily. A title alone cannot substitute for it.