Paying for the Good | Dave Wiseman

The short version

Why it matters

When a safety program works, the money it brings in falls. That's the sign it's working. The value is the harm avoided, not the fines collected. The thing to watch is anyone who needs the fines to keep coming.

How to read the model

The first chart shows what the camera company earns under three kinds of contract, for the same program and the same drivers. Paid a share of fines, it earns less as passing falls. Paid a flat fee, it earns the same. Paid for results, it earns more. Only the last two reward success.

What it means for you

If you fund or run a program: pay a flat fee or pay for results counted by someone independent. Keep fine money out of operating budgets.
If you live with one: judge it by harm avoided, not tickets written, and ask to see the data.
If you build one: measure the harm it prevents. That's the value it adds to a community, and the part nobody counts yet.

The honest scale

Passing drivers kill at most about 3 school-age children a year across the whole US (NHTSA, A). One 1,000-bus program prevents a small fraction of one death a year. The case for these cameras rests on what isn't counted yet: injuries, near misses and the habit of stopping.

How the pieces connect

Pick a step to see the evidence behind it.

The model

A program you can tune. Every starting number comes from a source, shown under each slider and in the table below. Slider ranges stay inside what the data supports.

The program

Buses with cameras

Illegal passes per bus per school day 0.55 in the 2026 one-day count by 101,334 drivers (B). 180 school days.

Share of passes that end in a paid ticket Not published. In Florida, 70% of notices sent were paid (A).

Fine Florida $200, Montgomery County MD $250, Colorado's cap $300 (all A).

How much passing falls

Drop in passing once it's working

Years to get there New York's speed camera effects took about six months to show (A).

Switch the cameras off after year 5. When cities turned off red-light cameras, fatal red-light crashes ran 30% above expected (A). Here, half the gain is lost.

Three ways to pay the vendor

Share of fines Rhode Island's law sets 30% (B). The 50% starting point is our assumption.

Flat fee per bus per year We found no official published price for a flat-fee contract, so this starting point is our assumption.

Paid for results: base per bus per year

Paid for results: per pass prevented Our assumption. We found no school bus contract paid this way. Results would be counted by someone independent.

Lives

US school-age deaths a year from passing drivers 1.8 is the 2020-2024 average in the national loading-zone survey (B). 3.3 is the federal upper bound: all school-age pedestrians killed by vehicles other than the bus (A).

What the camera company earns each year

Same program, same drivers. Only the contract changes.

Illegal passes each year

Counting lives saved

Counting deaths alone, each illegal pass carries well under a dollar of expected harm, while fines run $200 to $300. That gap says the fine is doing a different job: it deters, and most of what it prevents (injuries, near misses, a culture of not stopping) isn't counted anywhere. For comparison, about 1.16 million people die in road crashes worldwide each year, and road injuries are the leading cause of death for ages 5 to 29 (WHO, A).

Where the starting numbers come from

Valuing what markets don't pay for

A stay-at-home parent who invests in their kid earns nothing, and the value to society is very high. Federal health economists put a standard value on unpaid time.

Hours a week of unpaid care At $20.27 an hour (HHS, 2026, A).

a year, at the federal standard rate. Unpaid household work in the US was worth about $5.7 trillion in 2024 (BEA, A).

Where the two views meet

Success shrinks revenue

Falling fines are the sign it's working.
A company that works itself out of a job has done its job.
We should value good that markets don't pay for.

Watch the dependence

Whoever needs the fines has a reason to keep violations high.
Good habits fade when enforcement stops.
A small flat fine can read as a price, and it weighs more on some families than others.

Designs that fit both

A flat fee with service standards, or pay for results counted independently.
Fund public services at a steady baseline, separate from fines.
Keep the chance of being caught high and fines fair.
Open the data so someone without a stake can measure the good.

Picture cameras on school buses that catch drivers who pass a stopped bus. Say 90% of drivers stop today, and after a few years it's 95% or 99%. That's a whole community changing how it behaves, and it's a good change.

The obvious objection: if the camera company is paid from the fines, it earns less as drivers improve. Doesn't that put the company against its own goal?

I don't think so. Falling revenue is what success looks like. The value is in the harm avoided, not the fines collected. A company that works itself out of a job has done its job.

The research mostly backs that up. It also turned up a few things worth getting right.

A fine is supposed to shrink

A charge for harm people push onto others exists to change behavior. When it works, it collects less. Pigou, The Economics of Welfare, 1932 ed. (A) Three charges show what success looks like:

Ireland's plastic bag levy (2002): revenue started falling in 2008, as expected, and bags dropped from about 5% of litter to 0.13%. IEEP for the European Commission, 2016 (B)
London's congestion charge: it cut traffic more than forecast, so it raised less than forecast. Fullerton, Leicester and Smith, Mirrlees Review, 2010 (B)
Federal tobacco taxes: revenue fell more than 30% in ten years, partly because fewer people smoke. GAO, 2025 (A)

A fine moves money from one pocket to another. It isn't a measure of value. Becker, Journal of Political Economy, 1968 (A) Road safety groups now tell cities to expect falling camera revenue and to treat zero tickets as the goal. AAA, Advocates, GHSA, IIHS and NSC, 2021 (B)

The danger is depending on the money

Revenue going down isn't the risk. The risk is someone who needs it to stay up.

Ferguson, Missouri: fines and fees grew from about 12% to about 23% of the city's general fund, and police were pushed to write more tickets. US Department of Justice, 2015 (A)
North Carolina: counties issued more tickets in the year after their revenue fell. Garrett and Wagner, 2009 (A)
Dallas: a camera company's CEO admitted paying local officials more than $3.5 million for school bus camera contracts. The school transportation agency closed in 2017 with about $103 million in debt. US Attorney, Northern District of Texas, 2019 (A)

The more a number drives decisions, the more pressure there is to bend it. Campbell, 1979 (A)

A community that comes to depend on fine money is pointing at a bigger question: which services do we want, and how should we pay for them? We still want police and other public servants funded at a steady baseline. If most people follow the rules, officers can spend less time writing tickets and more time on the work we'd want from them anyway: educating people, helping, looking after the community. Keeping fine revenue separate from operating budgets is already the recommended practice. US Commission on Civil Rights, 2017 (A)

Good habits need upkeep

When large US cities switched off their red-light cameras, fatal red-light-running crashes ran 30% higher than expected. Hu and Cicchino, 2017 (A; paid for by the Insurance Institute for Highway Safety, which auto insurers fund) Irish bag use crept back up until the levy was raised (B, IEEP).

So there's probably a steady state worth reaching and holding, and holding it takes ongoing tuning. Fines aren't the only lever:

The chance of being caught matters more than the size of the fine. National Institute of Justice, 2016 (A)
Telling people what their neighbors do works. Adding social-norm messages to UK tax letters raised payment, at no extra cost. Hallsworth, List, Metcalfe and Vlaev, 2017 (A)
Street design does work cameras can't. In Santiago, Chile, cameras on ordinary bus lanes didn't speed up buses, while bus-only streets did. González and Silva, Journal of Urban Economics, 2025 (A)

A small fine can become a price

Israeli daycare centers added a small fine for late pickups. Late pickups nearly doubled and stayed high after the fine was removed. Gneezy and Rustichini, 2000 (A) Parents treated it as the price of a little extra childcare.

That also means a flat fine lands differently on different people: some parents could easily afford it, and others couldn't. Fines that scale with income spread the weight more fairly. In Finland, larger income-based speeding fines cut repeat offenses by 15% over six months, mostly among higher earners. Kaila, Review of Economic Studies, 2026 (A)

The message has to stay clear: passing a school bus is wrong, not just costly.

How to pay for the good

If falling revenue means success, the contract should pay for success:

A flat fee with service standards. A 2023 federal guide calls this the model most often used for US speed camera programs. FHWA and NHTSA guide, 2023 (A) Colorado's 2025 school bus camera law also bars paying the vendor from fines or by quota. Colorado HB25-1230 (A)
Pay for verified results. The first social impact bond paid out only when reoffending fell at least 7.5% against a comparison group. It fell 9%. UK Ministry of Justice, 2017 (A) At Rikers Island the target was missed, and the city paid nothing. Vera Institute, 2015 (A) Someone independent has to do the counting.
Keep fine money away from whoever decides how much to enforce (US Commission on Civil Rights, above).
Open the data. Two federal studies couldn't show whether school bus cameras reduce passing, partly because the camera operator shared only totals. NHTSA, 2024 (B)

Valuing what markets don't pay for

A stay-at-home parent who invests in their kid earns nothing, and the value to society is very high. We already know how to count some of that:

Unpaid household work in the US: about $5.7 trillion in 2024. BEA, 2026 (A)
A standard rate for unpaid time: federal health economists use $20.27 an hour, about $42,000 a year for 40 hours a week (our arithmetic). HHS, 2026 (A)
A life saved: about $14.1 million (HHS, A).

We rarely reward that value. Paying a safety company for keeping kids safe, instead of for catching drivers, would be one small step toward recognizing it.

The honest scale

School bus cameras won't save thousands of lives. From 2015 to 2024, vehicles other than the bus killed 33 school-age pedestrians in school-transportation crashes nationally, about 3 a year at most. NHTSA, 2026 (A) For comparison, about 1.16 million people die in road crashes worldwide each year. WHO, 2026 (A)

Most of the value is in injuries, near misses and the habit of stopping, and nobody measures those well yet. A safer bus stop, like a well-raised kid, is worth far more than any revenue line shows. Measuring that good well is how we start to value it properly.

Who to read

The books most relevant to these ideas, including people who did the practical work. Start with the three marked in gold. Each title links to its library record; the notes are our summaries.

Sources

Every source used on these pages, with its grade and how much of it we read. A official statistics, laws or peer-reviewed research. B large studies with published methods. C news reports of contracts or vendor data, used only for slider starting points where nothing better exists. Every link was checked on 2026-11-09.