Look at the gap between what agencies think a job costs and what contractors actually bid, and something shifted overnight. Low bids are now landing 7.5% under the engineer's estimate on average, across all 1,000 results on the board. A day ago, that gap was 4.3%.
That's not a rounding error. That's the estimate line moving from "pretty close" to "padded by real money" in the span of one cycle's worth of tabulations feeding into the average.
“Bidders are pricing jobs more aggressively than the agencies who wrote the estimates.”
Here's why you should care even if you didn't bid anything this week. The engineer's estimate is the number every GC, sub, and owner uses to sanity-check a bid before it goes in — too far under it and you start wondering what you missed; too far over and you wonder if the agency's cost model is stale. When the average low bid sits 7.5% under that number, it means bidders are pricing jobs more aggressively than the agencies who wrote the estimates.
A few things drive that kind of swing. More bidders on a given package usually means a harder floor — nine firms bidding the same scope will find the real number faster than three will, and the board's average bidder count sits at 9 per job right now. Material costs easing off their highs plays a part too; agencies often write estimates a year or more ahead of a bid date, so if steel or asphalt pricing softened since the estimate was drafted, the low bid comes in under it almost automatically. And there's a simpler explanation worth saying out loud: some agencies just estimate conservatively, building in a cushion that a hungry bid pool eats straight through.
The spread between low and second bid hasn't moved — it's holding at 17.3%, same as the prior cycle. That matters because it tells you this isn't a story about the field getting closer together. Bidders are still spreading out from each other by the same margin they always do. What changed is where the whole pack sits relative to the agency's number, not how tight the pack is internally.
If you're building an estimate for a public job right now, don't anchor to the posted engineer's estimate at face value — treat it as a ceiling that historically leaves 7 to 8 cents on the dollar of daylight between the top of the range and where jobs are actually clearing. If you're an agency watching your own program, a gap that wide two cycles running is worth asking your estimators about, not just banking as savings.
None of this singles out one job or one agency — this is an average across a thousand tabulated results, and it moves with whatever mix of jobs and bidders is landing on the board on a given day. But a swing from -4.3% to -7.5% in one cycle is large enough to be a real signal, not noise, and it's worth watching whether it holds, corrects, or keeps widening from here.