Fairfield's Speed Cameras: A Sourced Timeline

Before speed cameras go in, C.G.S. § 14-307e requires two things beyond passing the ordinance: a public hearing on the camera plan before it's submitted to the state, and a vote of the legislative body (in Fairfield, the RTM and Board of Selectmen) to submit it. § 14-307e(a)(1) · CT DOT ATESD FAQ

These aren't optional formalities. In Connecticut a town holds only the powers the state grants it, and the state grants this one on conditions — so the hearing and the vote are the guardrails that decide whether the authority to run the cameras exists at all. (It's the same principle as zoning: a change made without the required public hearing doesn't stand.)

Greenwich turned its school-zone cameras on, then paused them in spring 2026 — because the town realized its public meetings hadn't counted as the required public hearing, and the program had never gone before its RTM before launch. The same gaps at issue here. So Greenwich did it over, correctly: it held a public hearing, and its RTM voted on the program and its specifics. The cameras are now set to return — running only during school hours, with $50 fines. The First Selectman's word for the result: "It settled it."

The lesson isn't "cameras are bad." It's that doing the process properly is achievable, doesn't tank the program, and restores trust — which is all anyone here is asking for. CT Mirror ↗

You'll have about two minutes at the mic. You don't need all of these — pick the one or two that matter most to you and ask them in your own words:

1. The plan hearing. State law (§ 14-307e) required a public hearing on the camera plan — the locations, hours, and fines — before it was sent to the state. Which meeting was that hearing, and where is its public notice?
2. The vote to submit. The law also required a vote of the legislative body to submit the plan to the state. When did the RTM or the Board of Selectmen take that vote?
3. Recommit vs. revise. Item 5 asks you to "confirm" a plan the state approved a year ago — and on Aug. 17 this body was told, on the record, that voting no or tabling it "doesn't actually change the operational plan." If the vote changes nothing, what is it for — and will anything tonight actually amend the hours, the fines, or the data rules?
4. Who judges the appeals. On Aug. 17 the town confirmed the appeals hearing officer and hearing clerk are paid out of ticket revenue. Should the person deciding whether a ticket stands be funded by the tickets being upheld?
5. The fines. The $50 / $75 fines and the $15 fee are all state maximums, and the town chose the maximum on each. Will the RTM lower them, or explain why the max is right for Fairfield?
6. Data retention. The town's privacy policy promises our data is destroyed within 30 days of a paid fine — but the signed vendor contract says records are kept up to 90 days, citing Florida law. Which one actually governs, and how long is our plate data really kept?
7. Repeat-offender tracking. If our data is deleted at 30 days, how does the system charge the higher "second violation" fine months later? Is a repeat-offender history kept — for how long, and does it ever reset?
8. The hours. The safety case is about school hours, but the cameras run around the clock and the speed studies were done in summer with school out. Will enforcement be tied to actual school activity?
9. The evidence. Will the town publish the data that would let residents judge the program independently — violation and crash rates per vehicle (with traffic volumes), a consistent before-and-after, and a comparison road without cameras — rather than just raw violation counts?
10. The vendor's fee. Our own ordinance says vendor fees can't be contingent on citations issued or fines paid — yet Altumint is paid $15 for every paid ticket, $381,450 so far. How is that consistent with § 102-7?
11. The license-plate readers. The contract bundles in free Flock license-plate readers and Condor cameras — one of each per speed camera — and makes the town responsible for their data rules. The Governor just asked the state to review exactly this technology. Has any of it been installed, and where is the town's ALPR policy?
12. The exit clause. If anyone says we're locked into a five-year contract: the agreement allows either party to terminate for convenience after one year on 90 days' notice, without damages. The town has options — will the RTM use that leverage to renegotiate the terms?
13. The cost of losing an appeal. The ordinance (§ 102-12(F)) says a resident who contests a ticket and loses "shall" be assessed "fines, penalties, costs, and fees" — but nowhere defines or caps what those penalties, costs, and fees are. What can losing an appeal actually cost, in writing? Residents deciding whether to contest a ticket deserve to know before they ask for a hearing.

A few facts from the August RTM packet and the signed contract, each checkable against the source:

1. Every ticket payment goes first into a bank account controlled by the vendor ("lockbox"), which deducts its fees and sends the town its share monthly — the vendor pays itself first.
2. The contract has a confidentiality clause: both sides agree not to disclose information about the program's performance "except as required by law, or by mutual agreement."
3. Fairfield police rejected 6,239 of the 35,099 violations the vendor's system flagged — about one in six. Human review is doing real work.
4. The town's one-citation-per-month cap quietly spared residents 3,430 tickets (about $171,500).
5. Citations went out with the wrong fine amounts during rollout — roughly $104,200 worth. (Residents were undercharged; the vendor is compensating the town, not the other way around.)
6. Of appeals actually heard by July 22: 8 of 13 won. When residents finally get a hearing, the tickets lose 62% of the time. There are 948 more waiting, so hearings are moving from 24 every two weeks to 75 every Friday.
7. The town told residents the cameras "operate only during school hours" (town FAQ) — while the plan filed with the state says enforcement "will be 24/7." Both documents appear in the same RTM packet.
8. The vendor's own report shows the sites are wildly different: one Riverfield stretch measured 3–7% of drivers speeding; a Burr stretch measured 33–55% — yet the terms are identical everywhere.

   

9. Roughly 3 out of 4 tickets go to non-residents (76.9% in June, 78.5% in July, per the vendor's report) — and the top source is Bridgeport, whose ZIP codes fill five of the top ten. Whatever your view of the cameras, Fairfield's camera revenue is largely paid by commuters from a neighboring, less affluent city.
10. The town's ordinance (§ 102-19) requires the annual report to state how many vehicle owners received more than one violation. The report in the August packet omits that number — even though the system plainly tracks repeats (its one-ticket-per-month cap suppressed 3,430 citations). A fair question for Aug. 24: where is it?

The town cites figures like "87% fewer violations" and "114,000 violations in the first weeks." They sound dramatic — but they can't actually be evaluated, because they're missing what any fair assessment needs:

• Rates, not raw counts — violations per 1,000 vehicles, with traffic volumes. A big number can just mean a busy road.
• A consistent before-and-after — the same speeds measured the same way before the cameras, during the warning period, and each month after. (The plan's "before" speeds were measured in July, with school out.)
• The real safety outcome — crashes and injuries at each site, not just violation counts.
• Comparison roads without cameras — to separate the cameras' effect from town-wide trends like the education campaign or seasonal slowing.
• School-hours vs. off-hours — since the safety case is about school hours, but the cameras run around the clock.
• Unique vehicles, same methodology throughout — the town has cited a drop from 94,000+ warnings in May to about 8,000 in early June. But May counted warnings with no cap (one commuter could generate dozens), while June issuances were limited to one per vehicle — so part of that drop is the counting changing, not the driving. And the numbers don't reconcile: 94,000 May warnings, yet the CFO reports only 35,099 violations ever submitted for police review through June 30. Four different counters are in circulation — vendor "events," warnings, violations submitted, citations mailed — and officials quote whichever suits the moment. The fix is simple: publish unique vehicles detected per month, same method every month. That number can't be inflated by repeat-counting or deflated by caps. If it fell sharply, the cameras work, and the data will say so.

Ask the town to publish this before the RTM votes. If the program works, the data will show it.

And is the limit even set right? A 20 mph limit makes sense when children are actually crossing — but these cameras run around the clock, and the town's own studies clock normal traffic at 32–38 mph on these roads, with tickets triggering at 31. Outside school arrival and dismissal, that means ticketing ordinary drivers for going the speed the road was engineered for — not catching dangerous ones. A police officer uses judgment; a camera has none. A real safety program would tie enforcement to when kids are present and to genuinely unsafe speeds — the specifics the RTM should set in the open.