The average gap between the low bid and the next-closest competitor on New York-area public construction jobs sits at 89.4%, up from 22.9% in the prior reading of The RFI Wire's bid results data.

Both readings cover the same sample size: 1,000 recorded bid results apiece. The average number of bidders per job barely moved, 8.2 versus 8.3. The average low bid against the engineer's estimate held flat too, 5.2% under engineer's estimate in both readings. Only the spread between the top two bids jumped.

“The same crowd of roughly eight bidders per job is still showing up. What changed is how far apart they are from each other.”

That combination is the tell. If fewer firms were bidding, a wider spread would be easy to explain — thin competition lets outliers in. That is not what happened here. The same crowd of roughly eight bidders per job is still showing up, and the low bid is still landing in the same place relative to what the engineer expected the job to cost. What changed is how far apart the bidders are from each other once they are in the room.

A spread that wide usually means one of two things on the ground: either a handful of jobs with extreme outlier bids pulled the average up, or pricing uncertainty has genuinely widened across the board — material costs, insurance, labor availability — and firms are no longer converging on a number the way they were. With the same sample size on both sides of this reading, the shift is real movement in the numbers as recorded, not a trick of a bigger or smaller batch.

New Jersey's DOT and DPMC remain the two largest single sources of recorded results in this sample, together accounting for 175 of the 1,000 rows — more than the next ten agencies combined. A pricing shift of this size inside a sample that concentrated in two agencies deserves a second look at whether it is a market-wide move or a couple of New Jersey lettings doing the heavy lifting.

What this means for you: If you bid these jobs, check your number against the field more carefully right now — a wider spread means more risk of being a lonely outlier on either end, high or low. How we count: Spread, bidder count, and low-vs-estimate figures cover the same 1,000-result sample in each reading; both are averages across all recorded bid results, not a single job.