Every prequalification system in this industry runs on the same three numbers. Total Recordable Incident Rate, Days Away Restricted or Transferred, and the Experience Modification Rate that the insurer sets. They are easy to collect, easy to compare, and they arrive in a spreadsheet before anyone has met the crew.
They are also all lagging indicators. Each one describes what has already happened to a contractor, aggregated across every job they ran, in conditions that may look nothing like yours. A firm with a clean three-year record and a crew that has turned over twice since those years were earned is not the firm the numbers describe.
What the numbers actually measure
TRIR counts recordable injuries per two hundred thousand hours worked. It says nothing about severity, and nothing about the near misses that preceded the recordables. A contractor can hold a low TRIR and still be one bad decision away from a fatality, because the metric does not distinguish between a sprained ankle and a crew that stopped working under an energized line only because someone happened to look up.
EMR is an insurance instrument. It reflects claims history and payroll, and it lags the behaviour that produced it by years. DART is more useful than TRIR because it weights lost time, but it shares the same blindness to what the crew will do on Tuesday.
None of this makes the numbers useless. It makes them a floor rather than an evaluation. A contractor who cannot clear the floor should not be on the bid list. Clearing it tells you very little about who should win.
What low-bid procurement actually buys
When the award goes to the lowest number per acre, the pressure does not disappear. It moves. It lands on crew size, on how much supervision a job carries, on whether training happens on paid time or does not happen, and on how quickly a vacancy gets filled and by whom.
None of that is visible in the bid. All of it is visible on the right of way about six weeks in.
This is the mechanism behind a pattern most program managers will recognise: the cheap bid that comes in on price and then generates rework, access disputes, regulatory attention, and a schedule that slips past the season it was supposed to fit inside. The saving was real at award. It was spent later, out of a different budget line, and usually nobody reconciles the two.
Weak signals worth asking about
The things that predict field behaviour are mostly not measurable from a spreadsheet, but they are observable if you ask.
Will a crew member stop the job? Not whether the policy says they can — every policy says they can. Ask when it last happened, what the work was, and what happened to the person who called it. A contractor who cannot produce a recent example either does not have the culture or does not track it. Both are worth knowing.
What does a toolbox talk look like on day forty? Day one is always good. Ask what the fortieth one covered, and whether it addressed anything specific to the ground the crew was standing on.
Where does turnover sit, and where does it concentrate? Aggregate turnover is noise. Turnover concentrated among first-line supervisors is a signal, because that is the layer that decides whether a corner gets cut when the light is going.
Who is on the job when nobody senior is watching? Every contractor performs well on an audit day. The question is what the ratio of experienced to green looks like on an ordinary Thursday in week nine.
Familiarity with the ground is a safety input
A crew that has worked a stretch of right of way before knows where the access is soft, which landowner needs a call first, where the terrain does something the map does not show. That knowledge is not transferable through a bid document, and it is destroyed every time a program re-tenders to a new low bidder.
Continuity is usually argued as a cost benefit. It is at least as much a safety one, and the two are not separable — most of what makes a program expensive over ten years arrives through the same door as most of what makes it dangerous.
Where the evidence is thin
This is a case built on mechanism and field observation, not on a controlled study. Several parts of it deserve to be stated more carefully than they usually are.
There is reasonable published work on the limitations of TRIR as a predictor of serious injuries and fatalities, and on the difference between leading and lagging indicators generally — including Culture matters in contractor selection process, which this author wrote with Shawn Galloway of ProAct Safety for the October 2025 issue of Pipeline and Gas Journal, and which argues the case at more length. There is much less that isolates procurement structure as a causal variable in right-of-way vegetation work specifically. Nobody, as far as this author is aware, has run the study that would settle it: matched programs, same terrain, same species pressure, differing only in how the contract was awarded, tracked across a full ten-year cycle.
So the honest position is this. The direction of the effect is well supported by mechanism and widely recognised by people who run these programs. The magnitude is not established. Anyone quoting a specific percentage for what low-bid procurement costs a program over a decade is extrapolating, and should say so.
That study is the kind of thing a body like this one exists to make possible, because no single operator or contractor holds enough of the ground to run it alone.
Sources
Brent Oberlink and Shawn Galloway, "Culture matters in contractor selection process," Pipeline and Gas Journal, October 2025.