For the first time in this stretch of daily snapshots, the mega-sample of bid results tracked across the region shows low bids landing BELOW the engineer's own cost estimate, not above it. Across 1,000 tracked results, the average low bid now sits 4.7% under the engineer's estimate.
That is a reversal, not a dip. The same dataset — same 1,000-result sample size — showed low bids running 31.7% OVER the engineer's estimate one cycle ago. A market that was blowing past its own cost projections by nearly a third is now, on paper, coming in under them.
“Bidders were blowing past the engineer's number by nearly a third. Now, on paper, they're coming in under it — same bidder count, same sample size.”
The average bidder count held steady at 8.3 firms per job, barely off the prior cycle's 9.1, so this isn't a story about fewer people chasing work. The competition didn't change. What changed is where the winning number landed relative to what the engineer thought the job would cost.
New Jersey Department of Transportation and New Jersey's Treasury Division of Property Management & Construction remain the two heaviest-volume agencies in the results sample, with 101 and 74 tracked results respectively this cycle — enough weight that a shift in how either agency's estimates are built, or how its bidders are pricing against them, can move the whole average. MTA Construction & Development (21 results) and NYC School Construction Authority (42 results) are the next-largest blocks in the sample.
On the open-bid side, the board itself shrank slightly, from 449 open packages to 440. New York State's Office of General Services, the Dormitory Authority, and the state's Parks agency remain the steadiest posters in the pipeline, each holding double digits of open work. The flip in pricing isn't coming from a flood of new, cheaper-to-build jobs — the available pipeline barely moved.
A swing this size in a sample this large deserves a plain caveat: the average blends everything from a six-figure HVAC swap to a nine-figure transit contract, and a handful of outsized results can tug a percentage like this hard in either direction. One cycle of a flipped sign is a flag to watch, not yet a trend to bank on.
What this means for you: If you're estimating against a published engineer's number right now, don't assume the old cushion for overruns still applies — price the job on its own merits, because the market-wide gap just closed and crossed to the other side. How we count: Figures come from The RFI Wire's tracked bid-results board, 1,000 results in both this cycle and the prior one, comparing each posted low bid to the agency's published engineer's estimate; the percentage is the average gap across the full sample, not any single job.