3.1 One formula and the 200,000 constant

1 What the 200,000 is

The same OSHA letter explains the constant: 200,000 is “the number of hours 100 employees working 40 hours per week, 50 weeks per year would work”. Multiplying by it converts your own hours into a rate per 100 full-time workers, which is what makes a twelve-person crew and a four-thousand-person plant comparable.

Two consequences follow, and both get argued about:

• The constant never changes. It does not become 240,000 because your crews work 48 weeks of overtime. The hours you divide by already carry that.
• It does not shrink for a part-time workforce. Same reason. A rate is per 100 full-time equivalents, and the denominator produces the equivalence.

2 The arithmetic, three times

Because the denominator is identical, the three rates always fall in the same order and the gaps between them are informative. A TRIR of 5.6 with a DART of 5.4 describes a very different year from a TRIR of 5.6 with a DART of 1.1.

Report to one decimal place. Carrying three implies a precision the case count does not have: one case either way moves a small establishment's rate by whole points, which is exactly why BLS publishes its own rates broken out by establishment size.

3 Where each input comes from

Both numbers are already on the 300A summary you certified in February. The case count is the sum of the case columns; the hours are the figure you entered for total hours worked by all employees covered by the log.

That last row is the odd one out, and it is worth pausing on. A severity rate counts days, not cases, so it answers a different question: not how often people are hurt, but how badly. The next chapter takes the four apart properly.

4 Small numbers move violently

A company with 60,000 hours worked — roughly thirty people — and no recordable cases has a TRIR of zero. One stitched laceration next year puts it at 3.3, above the national average, with nothing about the workplace having changed.

OSHA's 2016 letter makes the point directly: a single case has a much greater effect on a small establishment's rate than on a large one's, which is why BLS breaks its published rates into establishment size categories. The practical response is to stop reading a single year at small hour counts.

1. 1
   Add three years of cases and three years of hours together.
2. 2
   Apply the same formula, with the same 200,000 constant, to the combined totals.
3. 3
   Report it as a three-year rate and say so.
4. 4
   Keep watching the single-year number for direction, not for level.

5 Rates that are not incidence rates

Two numbers get quoted alongside TRIR and DART in prequalification packets and are built on entirely different machinery. Know the difference so you do not try to reconcile them.

The experience modification rate — EMR or mod — comes from workers' compensation, not from OSHA. It compares your claim experience to the expected experience for your classification, over a multi-year window, using a rating bureau's formula. It is about cost, it lags, and it is calculated by someone else. Nothing in Part 1904 produces it and you cannot derive it from your log.

The severity rate is an incidence rate in form but a day count in substance, and different organizations define it differently — days away only, or days away plus restricted days. Whenever you quote one, say which.

6 Publish the inputs, not just the output

Most arguments about rates are not arguments about arithmetic. They are arguments about whose hours were counted and which cases were included, conducted by two people who each think the other is quoting the same thing.

There is a one-line cure. Whenever you report a rate, report the case count and the hours worked next to it: “TRIR 2.1 — 5 recordable cases, 476,000 hours, establishment level, 2025”. Anyone can check it, nobody can misread it, and the discipline forces you to know where your own hours came from.