An independent, fact-checkable exposé · Australia
How Australian "road safety" became a revenue and surveillance machine — and what the government's own audits, budget papers, crash data and legislation actually show.
Every figure is drawn from government audits, budget papers, legislation, peer-reviewed studies and reputable reporting, and is labelled verified, reported or contested. Verify against the primary source before relying on it — that is the whole point.
Part 1 · The Strategy
The National Road Safety Strategy 2021–30 promised to halve the road toll by 2030 on the way to "Vision Zero." Five years in, deaths have risen almost every year — and governments can't even measure most of what they promised to deliver.
The Strategy, agreed by all federal, state and territory transport ministers, sets hard targets against a 2018–20 baseline of 1,142 deaths a year:
Source: National Road Safety Strategy 2021–30, pp.9–10.
To halve the toll by 2030, deaths need to be falling by roughly 50–60 a year. They are doing the opposite.
"The rate of annual road deaths per 100,000 population… declined over the five years to 2020, but since that low, has increased at an average of 3% per year… total fatalities have also been increasing by approximately 4% per year." — BITRE, Road Trauma Australia 2024
"Road deaths are rising, and governments are reluctant to report the data needed to measure most of these targets. This means our current plan looks more like a wish-list than a credible strategy." — Australian Automobile Association, Benchmarking the Performance of the NRSS
The Strategy has five headline performance indicators. According to the AAA's benchmarking, only two can actually be measured, because the underlying national data is never published.
| Key performance indicator | Can it be measured? |
|---|---|
| Total national road deaths | yes |
| Deaths of children aged 7 and under | yes |
| Serious injuries (national) | no national data |
| Deaths in city CBD areas | not reported |
| Deaths on national highways / high-speed network | not reported |
"Only two of the Strategy's five key performance indicators — total national deaths and deaths of young children — can be accurately measured." — AAA, Benchmarking the Performance of the NRSS (Mar Qtr 2023)
A strategy you can't measure is a strategy you can't be held to. The AAA has repeatedly urged that federal transport funding to the states be made conditional on releasing the crash data — a call governments have not adopted.
"We follow a Safe System / Vision Zero approach: safe roads, safe speeds, safe vehicles, safe road users. Enforcement is one pillar among many."
The pillar that scaled fastest is automated enforcement — the one that also generates revenue and surveillance data. Deaths keep rising while camera fine income is budgeted to grow. The infrastructure pillar (median barriers, sealed shoulders, star-rated roads) is chronically underfunded relative to the fines collected in its name.
Be fair — and be precise The strongest, most defensible criticism here is not "enforcement never works." It's that (a) the toll is rising, (b) most targets can't be measured because the data is withheld, and (c) the interventions that scaled are the ones that raise money and gather data. That is an accountability failure, and it is documented in the government's own reporting.
The previous National Road Safety Strategy (2011–2020) set a target to cut deaths and serious injuries by at least 30% by 2020. It missed. A 2018 independent inquiry (Woolley & Crozier) found progress had "come to a virtual standstill," deaths rose in 2015 and 2016, serious injuries were increasing, and — a decade in — there was still no nationally agreed way to even measure serious injuries. The inquiry called for "dramatic change… given the inadequately acknowledged national road injury epidemic."
So the pattern is now two strategies deep: ambitious targets, missing data, missed results — while the enforcement-and-revenue machinery expands regardless.
Sources: Inquiry into the National Road Safety Strategy 2011–2020; NRSS 2011–20 progress.
The strategy's north star is Vision Zero (from Sweden, 1997): zero deaths and serious injuries by 2050. As an aspiration it's admirable. As a policy lever it's powerful in a dangerous way — because if the only acceptable number is zero, then any measure, however intrusive or however marginal, can be justified as a step toward it.
Scholars reviewing the critiques note Vision Zero has been "accused of being paternalistic and unjust," with some measures "threatening the freedom, autonomy, and privacy of road users," and argue an "ideology of absolutes is morally incongruent with the fallible nature of human beings." The road-safety economist Rune Elvik warned decades ago that an uncompromising prioritisation of safety "would divert economic resources from other societal objectives" without cost-benefit discipline. A goal that can never be disproved is a goal that can license anything.
Sources: Arguments Against Vision Zero: A Literature Review (Springer); Elvik, Accident Analysis & Prevention. Framed as documented critique, not the view of the strategy's authors.
Here's the part they never advertise: a slice of the fine money is spent advertising the fines. Governments recycle the revenue into emotionally-engineered campaigns whose job is to manufacture your consent — to make you believe every camera, limit and levy is love, so you'll keep paying and keep quiet.
And it's deliberately manipulative. The TAC celebrated 25 years of its ad campaigns under the banner "Mission: to upset, outrage and appal" — stating the emotional-manipulation objective out loud. Yet when auditors looked for proof it works, they couldn't find it:
"The lack of a complying cost-benefit analysis or evidence of its long-term impact meant that effectiveness could not be confidently demonstrated." — NSW Auditor-General, on a road-safety campaign running annually since 2013
The academic literature on fear-appeal advertising says the same: shock tactics grab attention but their effect on actual behaviour is "mixed and inconsistent," with a documented wear-out effect. Meanwhile the brand machine expands — "Towards Zero" logos, 28-year AFL and cricket sponsorships, a "Road to Zero" attraction embedded in school excursions — normalising enforcement to children before they can drive. It's a state-funded loop: fine the public, spend their money persuading them the fine was a gift, repeat.
And it isn't working on the public either Despite all of it, the Victorian Road Safety Camera Commissioner's own survey found the share of people who believe cameras exist mainly to raise revenue rose from 53% (2020) to 65% (2024). Note the tell in the branding: the deliberate shift from "accident" to "crash" (to imply every death was preventable and therefore someone must be fined) — pushed by the very agency Victoria still calls the Transport Accident Commission.
Sources: NSW Auditor-General, Government Advertising 2022–23; Mumbrella (TAC "upset, outrage and appal"); TAC Annual Report; Road Safety Camera Commissioner survey; fear-appeal review (Carey & McDermott meta-analysis). The ~9% ratio pairs same-year NSW figures; the TAC $103M is a bundle of research/education/marketing.
When NSW removed mobile speed-camera warning signs in 2020, it justified the move with Monash University modelling claiming the changes could save "up to 43 lives" a year. That modelling was never fully released to the public. The government used unpublished numbers to strip the warnings, watched fines explode — then, in 2023, reversed course and restored the signs anyway, against the logic of its own modelling.
Auditors keep making recommendations that go the other way. The 2011 Victorian Auditor-General urged independent testing of camera accuracy and evaluation of never-assessed point-to-point cameras. The 2018 NSW Auditor-General questioned deployment hours and network deterrence. The public's own view is hardening: the Victorian Road Safety Camera Commissioner's survey found 65% now believe the main purpose of cameras is to raise revenue — up from 53% in 2020.
Sources: ACRS (the "43 lives" modelling); VAGO (2011); Road Safety Camera Commissioner survey.
The Office of Road Safety was established on 1 July 2019 as a branch of the federal Department of Infrastructure, to coordinate the national strategy and run the National Road Safety Data Hub. The Hub is where the missing data should live. The Strategy explicitly commits "to regularly publish progress on implementation, targets and safety performance indicators, enabled by a national data hub." That commitment is, on the AAA's assessment, unmet for three of five targets.
See Sources → The Strategy for every document cited here.
Part 2 · The Money
When a program is really about safety, success means the numbers fall and the revenue dries up. When it's about revenue, the fines are forecast to grow — and treasuries quietly bank on them. Here is what the states publish about themselves.
Two states publish a clean camera-revenue figure. Both are near half a billion dollars a year — each.
| State | Camera fine revenue | Year | Note |
|---|---|---|---|
| Victoria | $473 million | 2023-24 | Official (vic.gov.au) |
| Queensland | $464.3 million | 2023-24 | ~$337m net after $127.3m costs |
| WA (Road Trauma Trust Acct) | ~$123.2 million | 2022-23 | Infringement revenue, audited |
Sources: Victorian Government — Revenue from fines; Queensland Dept of Transport & Main Roads Annual Report (CDOP); WA Road Safety Council Annual Report 2023-24.
The clearest tell that this is revenue, not safety: governments write rising fine income into their forward estimates years in advance. You can't forecast growing "safety failures" and call the program a success.
| Jurisdiction | Forward estimate |
|---|---|
| NSW — "Fines, regulatory fees & other revenues" | $3.59bn (2024-25) → $3.90bn → $4.04bn → $4.15bn (2027-28) |
| Queensland — camera program (CDOP) | Forecast around $466–534m for 2024-25 |
| SA — mobile-phone cameras alone | $7.6m (2023-24) rising to $21.1m/yr indexed from 2026-27 |
| Queensland — camera revenue growth | $274.5m (2021-22) → $465.8m (2022-23) = +70% |
Sources: NSW Budget 2024-25, Budget Paper 1 (Revenue); SA Budget 2023-24; QLD annual report / budget. The NSW line bundles more than camera fines — but it is explicitly forecast to rise every year.
Note the honest caveat on NSW: that budget line includes regulatory fees and "other," so it isn't pure camera revenue. What it unambiguously shows is a treasury planning for fine and fee income to climb — the opposite of what a safety program that's working would produce.
If cameras exist to slow you at a hazard, warning you is the whole point. In late 2020 NSW removed the warning signs from mobile speed cameras. What happened next is the cleanest natural experiment in the country.
~3,222 low-range (≤10 km/h over) fines that month.
~27,855 low-range fines that month — a jump of roughly 1,595%.
In March 2021 alone, an estimated 10,000 extra fines raised an additional $6.33 million. Widely reported figures put the annual low-range take rising from about $4 million to about $45 million reported. Then, after sustained public backlash, the incoming government restored the warning signs in April 2023:
"Fines issued from mobile speed cameras have fallen nearly 90 per cent since warning signs were returned to the roadside in late April… approximately 6,650 fines were issued… compared to 55,387 fines in the corresponding period in 2022 — an annual reduction of 88 per cent." — NSW Government media release, 16 July 2023
Read that again Restoring the warnings — the thing that actually slows drivers before the hazard — cut fines by ~90%. If the fines were a proxy for danger, danger fell 90% overnight. It didn't. The fines were the point.
Sources: Border Mail; NSW Parliament Staysafe Committee, Mobile speed camera enforcement programs in NSW; NSW Government (2023).
Every state now "ring-fences" (hypothecates) camera revenue into a road-safety fund. It's the standard rebuttal to the revenue accusation. There are two problems with it.
| State | Ring-fenced fund | The catch |
|---|---|---|
| NSW | Community Road Safety Fund | Underspent every year since 2019-20 — from 12% to 20% underspent |
| WA | Road Trauma Trust Account | ~$119.8m sitting unspent at end 2023-24; Auditor-General criticised its management |
| VIC | Better Roads Victoria Trust | "Every dollar" reinvested — but into general road projects, not just safety |
| QLD | CDOP reinvestment (TORUM Act 1995) | Only the surplus above program costs is reinvested |
| SA | Community Road Safety Fund | All except the Victims of Crime Levy |
Problem one: hypothecation doesn't mean the money is spent on safety. NSW and WA both sit on huge, persistent underspends — a road-safety fund that isn't spent on road safety is just a tax with a nice name.
Problem two: hypothecation actually strengthens the perverse incentive. Once a road-safety budget depends on fine income, the system needs the fines to keep flowing — the exact opposite of an agency working to make itself obsolete.
Sources: NSW Audit Office — Regional Road Safety; WA Auditor-General — Road Trauma Trust Account; state government fund pages.
The same "safety and amenity" cover, one rung down at the councils. Parking fines are a budget line — forecast, relied on, and enforced to quota.
And when it's checked, it collapses. A Victorian Ombudsman investigation forced councils to refund ~$24 million across ~250,000 fines — because for a decade they'd unlawfully outsourced fine "reviews" to a private contractor. The Ombudsman found a council officer had approved 107 reviews in one minute — "about half a second per review."
A private "fine" is not a fine On private land — shopping centres, hospital car parks — operators like Wilson Parking or Care Park can't issue government fines. Their "penalty" is a breach-of-contract invoice, and tribunals have repeatedly found the amounts to be unenforceable penalties (e.g. Vico v Care Park, VCAT 2014). Consumer Affairs Victoria confirms these debts can't appear on your credit report. Get advice before you pay.
Sources: NSW Government (ticketless parking); Victorian Ombudsman; Consumer Affairs Victoria; Consumer Action Law Centre.
NSW's mobile-phone detection cameras earned $7.1 million in their first month (March 2020) and roughly $56 million in their first nine months. Every other state has since rolled out the same technology — see Surveillance for how they work and what they photograph.
Full citations on the Sources page.
Part · The Fund
Speed-camera money doesn't disappear into "road safety." In NSW it flows into a fund that buys more cameras and pays for more police enforcement — administered by the same department that runs the cameras. The people who decide how hard to police you are funded by how hard they police you. Here is that loop, in the government's own numbers.
The simplest way to see the incentive: compare what the cameras take in fines against what it costs to run the entire state police force. It's not a rounding error — it's roughly one dollar in ten.
[interactive chart — see the live site]
In Victoria, road-safety cameras issued $473 million in fines in 2023-24 verified — about a tenth of Victoria Police's ~$4.5 billion in government funding. In NSW, speed-camera revenue of roughly $636 million reported sits at around 11–12% of the NSW Police budget (~$5.51 billion). When a slice that large depends on writing tickets, "how much should we enforce?" stops being a pure safety question.
Sources: Victorian Government — revenue from fines; Victoria Police Annual Report 2024-25; NSW camera revenue via a NSW Parliament question-on-notice (reported); NSW Police Annual Report 2024-25. Years don't perfectly align (fines 2023-24 vs police funding 2024-25) — read the ratios as "roughly a tenth," not to the dollar.
Set up in 2012 (operating from 2013), the NSW Community Road Safety Fund receives all camera-detected speeding fines, red-light fines and mobile-phone & seatbelt fines, topped up from consolidated revenue. It is administered by Transport for NSW — the agency that runs the camera program — while Revenue NSW issues the penalty notices. verified
| Year | Total fund spend | Of which from camera fines |
|---|---|---|
| 2022-23 | $454m | $283m |
| 2023-24 | $479m | $261m |
The clause that closes the loop Money in the fund that isn't required for road-safety spending may be paid into the Consolidated Fund — general government revenue. So a levy raised in the name of safety can, by design, become ordinary income. That is the legal hinge on which "safety" turns into "budget."
Source: Transport for NSW — NSW Road Safety Progress Report 2024 (pp. 53-54). Fund machinery sits mainly in the Transport Administration Act 1988; we quote the mechanism, not a section number.
Follow one dollar. It leaves your account as a fine and comes back as the budget to catch the next driver — more cameras, more police. This isn't an accusation; it's the fund's own spending line.
In 2023-24 the fund spent $118.4 million running the automated camera enforcement program itself, and $36.7 million directly funding police enforcement (it was $29.7m the year before). Money taken from drivers in fines is spent catching more drivers. The enforcement pays for the enforcement. verified
Source: Transport for NSW — Road Safety Progress Report 2024 (Community Road Safety Fund expenditure, pp. 53-54).
If the fines were really about saving lives, the money would be spent, fast, where people die. The NSW Auditor-General found the opposite.
In fairness, the Auditor-General attributed the underspend to COVID, disasters and skills shortages, not deliberate diversion. But that's the point: whether by intent or dysfunction, the drive to collect never wavers, while the duty to spend it on safety does.
Source: NSW Auditor-General — Regional Road Safety (Nov 2023). verified
A government serious about ending speeding would budget for fine revenue to fall toward zero. Instead, NSW's forward estimates pencil fines in to keep rising, out to the horizon.
[interactive chart — see the live site]
The NSW Budget's "Fines" line runs from $713m (2023-24 actual) to a forecast $793m by 2028-29 — never once projected to drop. verified Queensland tells the same story: its camera program took roughly $275m in 2021-22 and about $466m the next year — a ~70% jump — all, by law, "reinvested in road safety." reported The machine is planned to grow.
Sources: NSW Budget 2025-26, Budget Paper No.1, Table 5.7; Queensland CDOP figures via government statements (reported).
You'll often hear that police work to secret "ticket quotas" or revenue targets. We looked, and couldn't stand that up with a primary source — so we don't assert it. The honest argument is stronger and needs no conspiracy: when the body that decides how hard to police you is funded by fining you, the incentive is built into the structure. No memo required.
Know how the fund is really spent — or what the targets really are? We work with government whistleblowers. If you're inside a police force, a transport agency, a camera contractor or a treasury office and you can show what the public can't see, reach us confidentially at whistle@theradicalparty.com. Every figure on this page came from the public record; the next one could come from you.
Cross-references: The Money · By State (fund underspend jurisdiction by jurisdiction) · Contracts & Corruption · full citations on Sources.
Part · The Toll Scam
Different racket, same trick. You pay to build the road through fuel excise, rego and taxes — then pay again, forever, to a private company to drive on it. The public carries the risk; a monopoly keeps the money; and the meter only ever goes up.
Transurban owns or operates 11 of NSW's 13 toll roads — plus Melbourne's CityLink and most of Brisbane's network. It is, in the words of the NSW Government's own review, "the only non-government owner of motorways in the state."
| Transurban FY2024 | Figure |
|---|---|
| Proportional toll revenue | A$3,535 million |
| Proportional EBITDA (margin 73%) | A$2,632 million |
| Statutory profit after tax | A$376 million |
| Distribution to security holders | 62¢ (more than the profit) |
Note the gap: it pays out to investors more than it books as profit — a feature of its "stapled trust" structure, which is also why it pays little or no company tax in many years (one half-year: A$21m tax payable on ~A$2bn of revenue; a prior half-year recorded a A$97m tax benefit). That's legal — and that's the point. The structure is built to minimise tax on money extracted from a public necessity.
Sources: NSW Government (Fels toll review); Transurban FY24 results; Michael West Media (tax structure).
These aren't temporary tolls that switch off once the road is paid for. They're locked in for a generation, with escalation clauses that outrun inflation by design.
| Road | Toll escalation & end date |
|---|---|
| WestConnex (M4/M5E/M8) | Rises by the greater of CPI or 4% every year — to ~2060 |
| CityLink (Melbourne) | 4.25% p.a. from 2019; concession extended to 2045 |
| Hills M2 / Eastern Distributor | Adjusted quarterly under long-term contracts |
| Westlink M7 | Concession extended to 2051 "funded by increased toll revenue" |
The independent toll review led by former competition tsar Allan Fels found Sydney motorists now pay about $2.5 billion a year — an estimated $195 billion over 35 years — across a "poorly-functioning patchwork" of 13 roads and 10 contracts "designed with financial returns back to toll road operators in mind rather than managing traffic."
Sources: NSW Government / Fels review; CityLink concession. Exact escalation formulae are set in concession deeds; the "greater of CPI or 4%" and "4.25%" terms are widely reported — verify in the deed for precise wording.
WestConnex cost about A$16.8 billion to build (itself a blowout from ~A$10bn, with the Auditor-General finding creative accounting hid over A$4bn more). It is forecast to collect about A$64 billion in tolls to 2060 — roughly four times its build cost. The government then sold it to a Transurban-led consortium in two tranches: A$9.26bn for 51% (2018) and A$11.1bn for the rest (2021) — A$20.4bn total — handing decades of that toll stream to private hands.
Heads they win, tails you pay When toll roads fail, the private investors lose — but the tolling never stops; the assets are just bought cheap and keep charging. Brisbane's AirportLink cost ~A$5.6bn to build and was bought out of administration for A$1.87bn — "less than half the price it cost to build" — and kept tolling. Sydney's Lane Cove Tunnel: into receivership, bought by Transurban for A$630m, kept tolling. Cross City Tunnel and Brisbane's Clem7: same story. The public built them; private buyers scooped the toll rights at a discount.
Sources: NSW Auditor-General (WestConnex); NSW Treasury (sale); SBS (AirportLink); Lane Cove Tunnel.
The real scandal is in the contracts. Sydney's Cross City Tunnel deal reportedly required the State to keep surrounding roads closed to funnel traffic into the tunnel — and to compensate the operator up to A$100 million a year for 30 years if it didn't, plus compensation for any public-transport improvements that reduced tunnel traffic. Reopening the closed public roads was said to risk up to A$1 billion in taxpayer compensation.
This is the template: governments guarantee private profits, hand over monopoly pricing for decades, agree not to compete with their own roads, and lose oversight once the asset is sold (after the WestConnex sale the Auditor-General "no longer has the mandate" to scrutinise its costs). Transurban, for its part, donated over A$730,000 to the major parties between 2000 and 2015. Socialise the cost. Privatise the profit. Silence the scrutiny.
Sources: Cross City Tunnel; NSW Auditor-General; Australian Greens (donations). Some compensation-clause specifics are from contemporaneous reporting of confidential deeds — treat exact figures as reported.
Australians pay fuel excise, registration, stamp duty and GST — nominally to fund roads. Then a toll is layered on top, paid to a private monopoly, escalating faster than inflation, for decades after the concrete is paid off. It's the same philosophy as the speed camera: a public good repackaged as a private revenue stream, sold to you as something being done for you.
Part · The Goldmines
If cameras were sited for safety, they'd cluster where people die. They don't. Australians die on rural roads; the cameras — and the money — sit on busy urban arterials, at the bottom of hills and in daytime 40 zones.
In NSW in 2024, 234 of 340 road deaths (69%) were on regional roads — despite only about a third of the population living there. The Minister said it herself. Yet, as the government concedes, "most fixed speed cameras in NSW are installed in metropolitan areas." The 2024 plan to add thousands of regional mobile-camera sites is itself an admission that the network was pointed at the wrong places.
"More than two thirds of those people died on regional roads, despite only a third of our population living in the regional areas of NSW." — NSW Minister for Regional Transport & Roads, 2024
Sources: BITRE, Road Trauma Australia 2024; NRSS Regional & Remote fact sheets; TfNSW 2023 Enforcement Camera Review.
The biggest killers on Australian roads are crash types no urban speed or red-light camera can address.
| Crash type | Share of fatal crashes | Why cameras miss it |
|---|---|---|
| Single-vehicle run-off-road | 39% | 96% are single-vehicle; half in 100+ km/h rural zones. Fixed by barriers & shoulders, not cameras. |
| Head-on | 20% | Rural high-speed roads; needs median barriers. |
| Combined (run-off + head-on) | ~59% | Overwhelmingly rural & infrastructure-driven. |
Meanwhile hospitalised injuries are ~67% urban — so the cameras sit where the fender-benders are, not where the fatalities are. The fix for rural death is un-billable: median barriers, sealed shoulders, rumble strips, better star-rated roads.
Source: BITRE Information Sheet 112 — Run-off-road crashes 2016–2020.
A handful of individual cameras generate staggering sums — and the most lucrative ones sit at engineered "gotcha" spots, not documented blackspots.
| Camera | Take | The mechanics |
|---|---|---|
| Rosanna Rd / Darebin St, Heidelberg (VIC) | 53,000+ drivers, $11M+ in one year | Limit drops to 40 km/h daytime on a major through-road |
| Eastern Distributor, Darlinghurst (NSW) | $4.56M/yr | Fast urban motorway — "NSW's most notorious" |
| Hale St, Paddington (QLD) | ~11,000 fines/yr | Downhill slope + a lower limit than the rest of the bypass |
| Hannell St, Wickham (NSW) | 8,283 phone fines, ~$3.3M (2023-24) | Single mobile-phone camera site, many times any other |
Queensland's own description of Hale Street is telling: "the combination of a downhill slope and a lower speed limit than the rest of the [bypass] manages to catch out thousands of inattentive drivers." That's not a description of a hazard. It's a description of a trap.
Sources: CarExpert (VIC/QLD top earners); NSW top-earning cameras (Revenue NSW data); Newcastle Herald (Hannell St). Rankings rotate year to year; figures are pinned to the year reported. Crash-history at these sites is not published — a prime FOI target.
The whole justification is "more enforcement = fewer deaths." So watch the two lines. Camera revenue climbs; the national road toll fell for two decades on engineering and safer cars, bottomed in 2020, and has risen every year since — through the biggest camera-enforcement expansion in Australian history.
[interactive chart — see the live site]
Source: BITRE Australian Road Deaths Database (2000–2022) + BITRE monthly bulletins (2023–25). Deaths fell ~1,817 → ~1,097 to 2020, then rose to ~1,317 in 2025.
[interactive chart — see the live site]
Source: Revenue NSW open dataset DSF010 (all speeding & red-light camera offences). The 2020–22 spike lines up exactly with the removal of mobile-camera warning signs (Nov 2020); revenue is climbing again to a record ~$347M in 2025-26. Note: NSW camera revenue vs national deaths are different geographies — shown together to make the enforcement-vs-outcome point, not a like-for-like comparison.
The cleanest test of "revenue vs safety": when fine income soars and the death toll rises together, the fines are not tracking danger.
The fair counter-point To keep this honest: TfNSW's 2023 review found fixed speed cameras cut casualty crashes by about 17% at the sites where they operate. Cameras can work where they're genuinely needed. The scandal isn't that cameras function — it's that they're pointed at revenue, not at the rural roads where Australians actually die.
Sources: CarExpert (QLD/VIC revenue vs toll); WhichCar (80% automated); TfNSW 2023 review (17% figure).
Part · The Evidence
"The evidence shows cameras save lives." Does it? The highest-tier reviews have no randomised trials, rate the evidence "moderate quality at best," and a measurable chunk of the apparent "camera effect" is a statistical illusion called regression to the mean.
The Cochrane systematic review — the gold standard of evidence synthesis — looked at speed cameras across 35 studies and concluded the direction was positive but the size could not be trusted:
"…an overall magnitude of this effect is currently not deducible due to heterogeneity and lack of methodological rigour." The included studies were "of overall moderate quality at best." — Wilson, Willis et al., Cochrane Database of Systematic Reviews (2010)
Crucially, Cochrane found no randomised controlled trials — every study is an observational before-and-after, the design most vulnerable to statistical distortion. The first systematic review (Pilkington & Kinra, BMJ 2005) reviewed 14 studies, rated none high-quality, and concluded the "level of evidence is relatively poor."
Sources: Cochrane review (2010); Pilkington & Kinra, BMJ 2005. Note: "positive direction, untrustworthy magnitude" is the fair reading — not "cameras do nothing."
Cameras get placed where crashes recently spiked. Crashes then fall back toward normal on their own — and the fall gets credited to the camera. This is regression to the mean (RTM), and governments' own evaluations show it inflates the results.
Roads with a high number of crashes in one period "are likely to have fewer during the following period, even if no measures are taken… the effects of the intervention may be overestimated." — European Commission road-safety directorate, on RTM in speed enforcement
The meta-analyst Høye flagged that a headline −51% fatal-crash figure "could partly be explained by regression to the mean." And Erke's red-light-camera meta-analysis found results are "more favourable when there is a lack of control for regression to the mean" — i.e. the sloppier the study, the better cameras look.
Sources: UK 4-year camera evaluation (2005), Appendix H (Mountain & Maher); Høye, EU SafetyCube synopsis; Erke (2009), Accident Analysis & Prevention.
Red-light cameras reduce dangerous side-on ("T-bone") crashes — but reliably increase rear-end collisions as drivers slam on the brakes to avoid a fine.
Overall crashes up ~15%; rear-end collisions up ~40%; right-angle crashes down ~10% (effects non-significant). Conclusion: "on the whole RLCs do not seem to be a successful safety measure."
Right-angle crashes down ~25%, rear-end up ~15%, but a net +$18.5M economic benefit because the crashes prevented are more severe than the ones caused. Both sides agree on the trade-off — they disagree on the net.
A separate 2018 meta-analysis: overall crashes −12%, right-angle −24%, but rear-end +32%. The honest summary: red-light cameras shift the crash mix; whether that's a net win depends on the intersection — and the cleaner the statistics, the smaller the claimed benefit.
Sources: Erke 2009; FHWA red-light camera evaluation (2005); Accident Analysis & Prevention (2018).
Much of Australia's pro-camera evidence comes from the Monash University Accident Research Centre (MUARC), whose evaluation of Victoria's fixed cameras found casualty-crash reductions of up to 47% on the camera-monitored approach. MUARC argues its estimates were "not inflated by regression to the mean," and that's an empirical argument worth taking seriously.
But two things are fair to note: its design is comparison-site regression, not the full empirical-Bayes method used to strip out RTM in the UK study above; and MUARC is funded by the road-safety agencies whose programs it evaluates. When the body that certifies the policy is paid by the body that runs it, independence is a reasonable question — not an accusation.
The bottom line Cameras probably do reduce crashes where they're properly sited — average-speed ("section") control looks the strongest of all. The defensible criticism is not "enforcement is useless." It's that (1) the benefit's size is routinely overstated by weak, RTM-contaminated studies, (2) red-light cameras create new crashes they rarely account for, and (3) the money and the siting — see The Goldmines — follow revenue, not risk.
Source: MUARC Report 307.
Part 3 · Speed
Nobody sensible thinks 160 in a school zone is fine. The trick is how "speed kills" slides from dangerous speed in the wrong place to any speed over an arbitrary line — because only the second version can be automated and billed.
Authorities say speed is "a factor" in roughly one in four to one in three fatal crashes. Victoria's TAC puts it at "one-in-four (25%)." But read their own definition:
"Speed, both excessive speed and inappropriate speed for the conditions… was a likely contributing factor in one-in-four fatal crashes." — Transport Accident Commission (Victoria)
That single number is doing three different jobs at once:
The missing statistic There is no published Australian figure that isolates "crashes caused by a driver exceeding the posted limit on an otherwise safe road" — the precise behaviour that low-level speed cameras target. Governments simply don't report that cut. The absence of that number, while "speed is a factor" is quoted everywhere, is itself the finding.
Note: crash-causation factors are not mutually exclusive — one fatal crash can be coded speed + alcohol + fatigue — so the published shares do not sum to 100%. Sources on the Sources page.
For a century, traffic engineers set limits using the 85th-percentile speed — the speed at or below which 85% of drivers naturally travel on a road. Set the limit there and you target genuine outliers. Set it well below, and you criminalise the safest, most typical driver — and manufacture a permanent, predictable stream of "offenders."
"85th Percentile Speed means the speed at or below which 85% of all vehicles are observed to travel under free-flowing conditions past a nominated point." — Main Roads Western Australia, Speed Zoning policy
Under the "Safe System" doctrine, Australian road authorities are steadily moving away from anchoring limits to how people actually drive, toward limits based on survivable impact speeds — which in practice means setting many limits below the 85th percentile. That may or may not be good safety policy. What it undeniably does is widen the gap between the legal limit and normal driving — and every km/h of that gap is billable.
Austroads' Guide to Road Safety Part 3: Safe Speed (2024) reframes limit-setting around the Safe System rather than observed behaviour. The US FHWA has likewise moved (2023 MUTCD) to discourage rigid reliance on the 85th percentile. See Sources.
How far over the limit before you're fined? In most of Australia, that's a secret — and legally, there is no guaranteed buffer at all.
| Jurisdiction | Published tolerance |
|---|---|
| Victoria | 2 km/h fixed (tightened from 3 in 2019); 3 km/h or 3% mobile published |
| NSW | Not published confidential |
| Queensland | Not published confidential |
| SA, WA, Tas, ACT, NT | Not published confidential |
Only Victoria publishes a figure. Everywhere else, the exact tolerance is treated as operational secret and, in law, any amount over the limit is an offence. A system that fines you at 2 km/h over — inside the error margin of many car speedometers — is not targeting danger.
Source: RACV — Enforcement of road rules in Victoria. We deliberately do not quote specific km/h figures for the confidential states; nobody publishes them.
The official case for near-zero tolerance rests almost entirely on one body of work — the Kloeden Adelaide studies. Take it seriously, because it's real. Then read the fine print.
In a 60 km/h zone, the risk of involvement in a casualty crash roughly doubles for each 5 km/h above 60 — a car travelling at 65 in a 60 zone has about double the crash risk. — Kloeden et al., CR 207 (BITRE)
The fine print that rarely makes the press release:
The honest position Speed matters. The defensible criticism isn't "enforcement does nothing" — it's that the safety benefit of chasing very small margins rests on a narrow evidence base, while the revenue benefit is large, immediate and budgeted. When the evidence is thin and the money is thick, ask which one is driving policy.
The classic Solomon curve (1964) found crash risk is U-shaped against traffic flow: vehicles going much slower or much faster than the surrounding traffic both face elevated risk, with the lowest risk near the median speed. If that's true, then speed variance — not raw speed — is a key danger, and a camera that pushes a cluster of drivers to brake below the natural flow can increase variance.
In fairness: the low-speed limb of the Solomon curve is the most disputed part of that research (later work by West & Dunn, Kloeden, and Australian data from Fildes et al. challenged it). We flag it as supported for the high-speed side, contested for the low-speed side — because a site that only quotes the convenient half of the evidence is doing exactly what we're criticising.
Overview and critiques: Solomon curve (with primary citations). Full list on Sources.
Part 4 · Surveillance
A fine is a one-off. A movement record is forever. The road-safety brand is the wrapper on a mass automatic-number-plate-recognition and in-cabin-camera network that grew with almost no public debate — and it keeps expanding into things it was never sold as.
The key fact about ANPR is architectural: it uses mass-surveillance capture. It doesn't only photograph cars on a watchlist — it reads and logs every plate that passes, then checks each against police and registration databases. Non-offenders aren't excluded; they're recorded.
ANPR "is a mass surveillance technique and breaches the human right of liberty of movement." The resulting database is "a 'honeypot' that attracts attention from many organisations" and is "impossible to protect against unauthorised access." — Australian Privacy Foundation
A proposed national system (CrimTrac's NAVR) envisaged retaining all ANPR sightings for five years — on the order of 70 million sightings a day. Queensland's own parliamentary inquiry had to recommend that data on non-offending vehicles be "cleansed nightly" — which tells you the default was to keep it.
Sources: Roger Clarke — Australia's ANPR surveillance; Australian Privacy Foundation; Astor Legal — ANPR.
Mobile-phone detection cameras don't photograph offenders — they photograph everyone, then let an AI decide who to keep. NSW's system was designed to check up to 135 million vehicles a year.
The cameras "take multiple images of every vehicle passing… as well as images of the front seats" — not just suspected offenders. An AI reviews the lot.
NSW live 2020 · QLD Nov 2021 · ACT Feb 2023 · VIC Jul 2023 · SA late 2023. What started as one state's trial is now the national default.
The same in-cabin AI was extended to seatbelt enforcement — QLD 2021, VIC & Tas 2023, NSW added. One lens, ever-expanding list of things it judges.
NSW says non-offender images are deleted "typically within an hour" (AI stage) or "within 72 hours" (human review). The safeguard is a policy, not a law you can enforce.
Sources: NRSPP; Transport & Main Roads QLD; CarExpert (rollout dates).
The clearest proof this is surveillance infrastructure first and safety second: watch what happens to "average speed" (point-to-point) cameras in NSW.
| When | What happened |
|---|---|
| 2010 | Average-speed cameras introduced in NSW — for heavy trucks and buses only, on freight-safety grounds. |
| 2010–2024 | The infrastructure to time every vehicle between two points sits there, used only on trucks. |
| Oct 2024 | Government announces a trial extending average-speed enforcement to ordinary cars. |
| 1 Jul 2025 | Enforcement of light vehicles begins on the trial sections. The "trucks only" promise is gone. |
And the most direct example of repurposing: during the 2020 pandemic, Victoria Police used ANPR — normally for stolen and unregistered vehicles — to enforce lockdown movement restrictions, catching drivers outside their permitted postcodes. Data gathered "for road safety" became a tool for something else entirely, "announced unilaterally… with little to no public discussion."
Sources: NSW Government — average-speed trial; NRMA; Policing Insight (Vic Police ANPR / COVID).
A surprising amount of the "public safety" network is operated by private companies, several of them foreign-owned. The state issues the fine; the contractor captures and processes the images.
| Operator | Role |
|---|---|
| Acusensus | NSW mobile phone & seatbelt detection cameras (program reported ~$88m) |
| Redflex (now US-owned Verra Mobility) | NSW mobile speed cameras since 2010; acquired by Verra Mobility in 2021 |
| Serco | Victoria's traffic camera operation (contract reported ~$90m) |
| SenSen Networks | AI average-speed enforcement cameras, NSW |
| Jenoptik | 400+ traffic-safety systems supplied across Australia over 20+ years |
Follow the incentive Globally, some camera contracts historically paid the operator a share of every fine (reportedly 40–50% under older Redflex-style models), while others pay a flat fee explicitly "to remove incentives to over-issue violations." We can't confirm which model each Australian contract uses — the contracts aren't public. That opacity, on a system that fines millions of people, is itself the story. It's a prime candidate for a Freedom-of-Information request.
Sources: iTnews (Acusensus); Serco; Redflex/Verra Mobility. Payment-model figures are from international reporting, not confirmed Australian contract terms — see Sources.
The photo you gave for a driver's licence has quietly become biometric infrastructure. Under the National Driver Licence Facial Recognition Solution (NDLFRS), agreed by leaders in 2017, states upload licence images to a central system that builds facial-recognition templates — Victoria, Tasmania and South Australia had already loaded roughly 6 million licences by 2019.
The first attempt to legislate broad access, the Identity-matching Services Bill 2019, was rejected by a bipartisan parliamentary committee on privacy grounds. A scaled-back law took effect at the end of 2024, largely prohibiting "one-to-many" face matching. But the trajectory is clear — and in June 2026 WA Police deployed live facial recognition in public spaces, described as an Australian first, scanning the faces of passers-by from a van. Civil liberties group EFA called it "the cherry sitting on top of the surveillance state cake."
Sources: iTnews (NDLFRS uploads); iTnews (2024 law); Electronic Frontiers Australia (WA live facial recognition).
Toll roads are a second movement-tracking layer. Transurban/Linkt's privacy policy confirms it discloses personal information to police and law-enforcement agencies "as required or permitted by law." And it's routine, at scale: documents tabled in NSW Parliament show law-enforcement requests to NSW toll agencies where toll movement data formed all or part of the information rose from 157 (last four months of 2012) to 996 in 2013 — a roughly six-fold jump in a single year.
Sources: Transurban privacy policy; NSW Parliament (toll-data law-enforcement requests).
The newest frontier is the vehicle itself. In the EU, Intelligent Speed Assistance became mandatory on all new cars from July 2024 — technology that knows the limit and your speed at every moment. And connected cars are already a monetised surveillance feed: in the US, regulators found General Motors collected drivers' precise location "as often as every 3 seconds" plus hard-braking and speeding data, and sold it to data brokers who fed it to insurers. California imposed a $12.75 million penalty. The same data that enables "safety" features is the data that builds a permanent behavioural record — and it can be sold, subpoenaed, or leaked.
Sources: ETSC (EU ISA mandate); US FTC (GM/OnStar). GM/ISA examples are international, included as the trajectory Australia is following.
Every "trust us, it's secure" assurance runs into the record. The 2020 Service NSW breach exposed 3.8 million documents — including driver licences and other ID — affecting ~104,000 people. Separately, over 54,000 scanned NSW driver licences were found sitting in an open, misconfigured cloud bucket. And the databases that ingest ANPR and licence data are repeatedly abused from the inside: 178 Victoria Police officers faced complaints over misuse of the LEAP database in one five-year window, and NSW officers have been prosecuted in 2026 for illegally accessing the COPS system (which holds 40+ million records). A honeypot this large cannot be kept safe — from outsiders or insiders.
Sources: Service NSW cyber incident; OAIC (54,000 licences); IBAC (LEAP misuse); NSW COPS prosecutions (2026).
For in-cabin cameras, NSW publishes a retention policy (non-offender images deleted within ~1–72 hours). For ANPR there is no consistent, published rule across states — retention ranges from "cleanse nightly" (a Queensland recommendation) to a proposed five years nationally, with Victoria's regime described by civil-liberties analysts as effectively unregulated: "it is unclear what records Victoria Police generate, keep, and/or cross-reference, nor how long data is kept."
"Nothing to hide" was never the test for whether the state gets to keep a movement history on millions of people who were never accused of anything.
Full citations on the Sources page.
Part · Cabin Cameras
Speed cameras watch the road. This newer generation watches you — photographing the front cabin of every vehicle that passes and running the image through AI to look for a phone in your hand or a seatbelt off your shoulder. Australia was the world's first to deploy them. They are being bolted, one offence at a time, onto a network that scans everyone.
NSW switched on the world's first mobile-phone detection cameras in a trial from December 2019, with fines from 1 March 2020. Each camera photographs every vehicle — capturing the number plate and a high-definition image of the front seats — and an algorithm flags the ones it thinks show an offence. By the state's own account the NSW network is scaled to check on the order of 135 million vehicles a year.
"The cameras take multiple images of every vehicle that passes… Images where no offence is detected are automatically excluded and deleted." — Queensland Government, road-safety camera page
That is the reassurance: the machine looks at everyone, but only "keeps" the guilty. The catch is that you can't audit it — you never see the images that were taken and deleted, and no government publishes how often the AI gets it wrong (see below).
Sources: Transport for NSW; NRSPP; Queensland Government; fine and revenue figures via NRMA.
What began as a NSW experiment is now a near-national grid, mostly built on the same Australian technology (Acusensus). Note the pattern: almost every state added seatbelt detection to the same hardware within a couple of years — the infrastructure sold to stop texting quietly grew a second job.
| Jurisdiction | Fines from | Phone fine | Seatbelt AI? |
|---|---|---|---|
| NSW | Mar 2020 | $423 | Yes — Jul 2024 |
| QLD | Nov 2021 | $1,209* | Yes — from launch |
| VIC | Jul 2023 | ~$556 | Yes — from launch |
| WA | 2024 | up to $1,000 | Yes |
| SA | Sep 2024 | $556 + levy | Rolling out |
| ACT | Feb 2024 | $514–$674 | Yes — Nov 2025 |
| TAS | Trial phase | $410 | Not at scale |
Dates and amounts vary by financial year and use-type; several are indexed annually. QLD runs the country's harshest penalty (*fine amount rises each July). reported — Sources: CarExpert (state-by-state); QLD Government; Victorian Government. Tasmania details are thin and are flagged contested.
If these cameras were purely a deterrent, switching them on would be uneventful. Instead, every "go live" is a revenue event — and the clearest proof is what happens during the warning period, when the machine counts offences it isn't yet fining.
| Jurisdiction | Warning period | What it recorded |
|---|---|---|
| WA | 8 months, no fines | 380,000+ breaches |
| SA | ~3 months | 64,454 warnings |
| QLD | ~3 months | 21,599 warnings |
| NSW (seatbelt) | none | 11,400 fines / 21 days |
NSW gave seatbelt detection no grace period at all. In the first 21 days it issued over 11,400 penalties — about 75% for wearing a belt incorrectly (still the full $410 fine and 3 points) — and in three weeks detected more seatbelt offences than the entire NSW Police force issued in all of 2023. WA's cameras, meanwhile, quietly logged 380,000+ breaches during an 8-month grace window before a single fine was posted. The "safety warning" period is really a demonstration of the revenue pipeline about to be switched on.
Sources: Transport for NSW media release (4 Aug 2024, PDF); WA/SA/QLD warning-period figures via CarExpert and state media releases. verified (NSW) · reported (WA/SA/QLD).
A system that photographs the inside of every car, decides guilt by algorithm, and reverses the burden of proof onto you ought to be the most transparent machine in the state. It is one of the least.
"There is a lack of any clear proposals… as to how the captured data may be used, accessed, stored or when it will be deleted." — NSW Council for Civil Liberties, on the detection-camera bill (2019)
Sources: iTnews — Queensland Audit Office finding (26 Sep 2025); NSW Council for Civil Liberties (15 Nov 2019); ACT wrongful-fine case via Canberra reporting (2026). verified (QLD audit, NSWCCL).
The honest way to see this technology is as a general-purpose cabin and plate scanner that happens to have started with phones. The same Acusensus units already read number plates to catch unregistered and speeding vehicles; seatbelts were added by software; average-speed timing rides on the same poles. Each new offence is a small, reasonable-sounding extension — and each one normalises a camera pointed inside your car.
A narrow tool to stop people texting at the wheel — "if you've done nothing wrong, you've nothing to fear."
A multi-function scanner — phone, seatbelt, registration, speed — photographing every cabin, run by private operators on contracts worth tens of millions (NSW ~$88m; QLD signed a $27.4m extension in Dec 2024), with the images you never see.
What we're not claiming We have found no confirmed case of these cabin images being handed to police for an unrelated investigation — the function creep so far is offence-type expansion, not cross-agency data use. And there is no reliable public "total revenue since 2020" figure, so we quote only the per-year numbers governments have released. The point isn't a single scandal; it's that a watch-everyone camera has been built with almost none of the transparency that watching everyone should require. See the receipts on Sources.
Sources: iTnews — TfNSW/Acusensus contract; QLD extension via state procurement reporting (Dec 2024). Cross-references: Surveillance — the grid, Faults & Errors.
Part · Contracts & Corruption
"Public safety" cameras are largely operated by private companies on contracts worth hundreds of millions of dollars — contracts you can't read. One of them, the firm that runs NSW speed cameras, was at the centre of a bribery scandal that sent people to federal prison.
Redflex Traffic Systems — founded in Australia, headquartered in South Melbourne — has operated NSW mobile speed cameras since 2010. In the United States, Redflex was the vendor at the heart of one of the biggest municipal corruption cases in Chicago's history.
| Who | What happened |
|---|---|
| John Bills — Chicago transport official | Took up to $2 million in cash, a Mercedes and an Arizona condo to steer a red-light contract worth ~$124 million to Redflex. Convicted on all 20 counts; sentenced to 10 years in federal prison (2016). |
| Karen Finley — Redflex CEO | Pleaded guilty; sentenced to 30 months in prison and $2M restitution (plus a separate 14-month term for an Ohio scheme). |
| Redflex (company) | Chicago severed ties; the city settled its lawsuit for $20 million. Redflex's stock collapsed (to A$0.17 by 2015) and it posted years of losses. |
In 2021 Redflex was acquired by the US firm Verra Mobility for ~A$152 million. Verra's automated-enforcement ("Government Solutions") business earned roughly US$460 million in 2025. This is the corporate lineage sitting behind a chunk of Australia's "road safety" network.
Sources: US DOJ; ABC7 (Bills, 10 years); CBS (Finley); NBC (settlement); Verra Mobility. All US except the Australian corporate origin and NSW operation.
Australian states outsource camera operation to private companies on very large contracts. The dollar totals surface; the actual terms — including how the operators are paid — largely don't.
| Operator | Contract |
|---|---|
| Serco (Victoria) | Traffic camera services — $178M over six years (2019 rebid), after a $90M extension in 2014 |
| Acusensus (NSW) | Phone/seatbelt program ~$88M; mobile-speed extension ~$16M per six months |
| Redflex / Verra Mobility (NSW) | Mobile speed cameras since 2010 (reported ~$112M program) |
| Jenoptik (WA, VIC) | 81 speed + red-light systems in WA (~A$28M); Victorian distracted-driver cameras (~€10M) |
The incentive problem. Overseas, some camera contracts historically paid the operator a cut of every fine — reportedly 40–50% under older Redflex-style deals. Maryland banned per-ticket payments outright, with legislators calling them "a form of kickback" that creates "a financial incentive for contractors to cheat to maximise profits."
Credit where due — then the catch To be fair: NSW states its mobile-speed operators are paid per hour of enforcement, not per fine, specifically to remove that incentive. That's the right model. The catch is that the actual contract terms are not public, so nobody outside government can verify what the incentives really are. On a system that fines millions of people, that opacity is itself the story — and a prime candidate for a Freedom-of-Information request.
Sources: Serco; Acusensus (ASX); Jenoptik; CBS (Maryland per-ticket ban). Payment-model percentages are from international reporting, not confirmed Australian contract terms.
Red-light cameras earn most when more people run reds. The fastest way to make that happen isn't reckless drivers — it's a shorter amber. Overseas, that's exactly what happened.
And the engineering research is unambiguous: lengthening the amber cuts red-light running more than the camera does. A Texas study found a 1-second increase cut violations by at least ~50%. In Philadelphia, adding a second of amber cut violations 36% before any camera was installed. The cheap, un-billable fix works — which is precisely why it's rarely the first resort.
Sources: Slate (Florida); Chicago 2.9s; Georgia HB 77; Texas TTI study; IIHS (Philadelphia). All international — no equivalent state-sanctioned amber-shortening scandal is documented in Australia, but AU's combined red-light-and-speed cameras issue two fines from one pass.
Part · Faults & Errors
Camera fines arrive with the authority of a machine that "doesn't lie." But cameras have been hacked, mis-calibrated, and caught fining cars for speeds they can't physically reach — and the burden of proving the error is dumped on you.
In June 2017 the WannaCry ransomware infected Victoria's fixed road-safety camera network after a maintenance contractor plugged in an infected USB stick.
| Detail | Figure |
|---|---|
| Camera devices infected | 110 (43 Redflex + 67 Jenoptik) |
| Cameras potentially exposed | up to 280 |
| Fines suspended as a precaution | ~8,000 |
| Fines withdrawn | 1,643 |
| Fines embargoed pending review | 5,500 |
Authorities said there was no evidence the ransomware had altered any fine's accuracy — the tickets were cancelled "to preserve public confidence." Which rather concedes the point: public confidence in automated fines rests on the assumption the system is clean, and that assumption had just been broken.
Sources: iTnews; Global News; Victorian Road Safety Camera Commissioner report, WannaCry malware infection (2018).
Sources: NSW refunds (documents released by the then-Opposition); Victoria faulty cameras / Datsun / West Gate. Some single-outlet specifics are reported figures; treat exact totals as "as reported."
In RTA of NSW v Michell (2006), the NSW Supreme Court dismissed the roads authority's appeal against an acquittal. The camera image lacked a legally required security indicator (proof the image wasn't altered), and Justice Adams held this could ground reasonable doubt even with no contrary evidence from the driver:
Computer processes "can go awry," and a "tribunal of fact is still bound … to consider the quality of that evidence." — Adams J, RTA of NSW v Michell (2006)
In other words, a camera certificate is evidence — not automatic guilt. But most people never test that, because of how the system is built (see below, and The Law).
Source: RTA v Michell (2006) NSWSC.
Mobile-phone detection cameras use AI to flag drivers, with a human reviewing before a fine issues. Reported grounds for successful challenges include a phone sitting untouched, a reflection or object misread as a phone, adjusting the air vents — even a hand near the face. The vendor's "95% capture" figure is a design target, not an independently verified false-positive rate.
The transparency hole Here's what's genuinely damning: no Australian authority publishes a false-positive rate or the proportion of contested camera fines that get withdrawn. NSW even hosts a dataset on fine reviews and waivers — but the headline error rate isn't advertised. Combined with owner-onus (you're presumed liable and must actively contest), the rational move for most people is to just pay. A system that profits from people not checking its accuracy has little reason to publish it.
Sources: NSW review/waiver data — data.nsw.gov.au; AAP FactCheck (enhanced images admissible). No official false-positive rate exists in public sources — a prime FOI/GIPA target.
Part · Punishing the Poor
A flat fine is a mosquito bite for the wealthy and a catastrophe for everyone else. And when it can't be paid, the machine escalates — costs, licence loss, seizure, and, until recently, prison. This is where "road safety" collides with justice.
A NSW fine for going 10–19 km/h over the limit is a flat $361 — whether you're on a pension or a seven-figure salary.
"While they are basically a mosquito bite for the wealthy, they can be earth-shattering for the poor… Australia's regressive speeding fine system effectively criminalises poverty." — The Australia Institute, Refining Fines (2024)
For someone on $50,000 a year, $361 is more than a third of a week's pay; for someone on $200,000 it's covered "by lunchtime on Monday." Several European countries use income-based "day fines" so the sting is proportionate. Australia doesn't.
Source: The Australia Institute.
An unpaid fine doesn't stay a fine. It escalates, administratively, with new costs at each step — and it takes the things people need to earn a living.
| Step | What happens |
|---|---|
| 1. Overdue | Enforcement costs added — up to $65 each time an action is taken |
| 2. Sanction | Driver licence and/or vehicle registration suspended — even for non-driving fines (Vic) |
| 3. Recovery | Sheriff seizure of goods, wage/bank garnishee, charge over land |
| 4. (Historically) Prison | Warrant of commitment — imprisonment for fine default |
Losing a licence over an unpaid fine means losing the ability to get to work — especially in regional and remote areas with no public transport — which pushes people into driving while suspended, which is itself a crime. "What starts as a social justice issue," one government submission put it, "often becomes a criminal justice issue."
Sources: Legal Aid NSW; ALRC Report 133, ch.12; Redfern Legal Centre.
In 2014, Ms Dhu, a 22-year-old Yamatji woman, was locked up in South Hedland, WA, over $3,622 in unpaid fines. She died in custody less than 48 hours later. The Coroner found her treatment "unprofessional and inhumane," and that staff had been influenced by "preconceived ideas about Aboriginal people."
Her death was not an isolated glitch. Between 2006 and 2015, 7,462 people were imprisoned for fine default in WA — roughly 11 a day. The burden fell overwhelmingly on Aboriginal and Torres Strait Islander people:
WA abolished imprisonment for fine default on 20 June 2020, and introduced a Custody Notification Service — reforms that came only after Ms Dhu's death and years of campaigning.
Sources: WA Coroner — Inquest into the death of Ms Dhu; ALRC Report 133. Handled as a matter of public record.
NSW runs a Work and Development Order scheme letting vulnerable people clear fines through treatment, counselling, courses or volunteering instead of money. It has helped more than 87,000 people clear over $200 million in debt. Every hardship scheme like it is a tacit admission that the flat-fine model does real damage to people who can't pay — damage the "road safety" framing never mentions.
The honest ask None of this argues for lawless roads. It argues that if the purpose were safety rather than revenue, the penalty would be proportionate to the risk and to the person — not a flat charge that a millionaire ignores and a single parent goes to court over. A speed camera can't tell the difference between the two. The budget papers can.
Part · Drug Testing
Random breath testing measures whether you're impaired — there's a legal 0.05 line. Roadside drug testing measures only whether a substance is present, with no impairment threshold at all. You can be stone-cold sober and still be guilty.
NSW Mobile Drug Testing (MDT) checks saliva for THC (cannabis), methamphetamine, cocaine and MDMA. It does not, and cannot, measure impairment.
"Roadside drug testing focuses on detection rather than impairment. This means you can be charged when certain drugs are present in your system, even if you don't feel intoxicated." — Alcohol and Drug Foundation
Impairment-based. A measured legal threshold (0.05 BAC). You can legally drive after a drink. ~6 million tests a year.
Presence-based. No threshold — any detectable trace is an offence. The prosecution "does not need to prove that the person was driving impaired."
Two opposite philosophies, in the same road-safety statute.
Sources: Alcohol and Drug Foundation; NUAA; Harm Reduction Australia, Drive Change NSW (2025).
Cannabis impairment lasts a few hours. THC stays detectable far longer — the Alcohol and Drug Foundation says saliva tests can pick it up for up to 3 days, and urine for around a month. So the test routinely catches people long after any effect has passed.
"THC levels in saliva do not reliably reflect cannabis intake or intoxication." — Prof. Iain McGregor, University of Sydney
The emblematic case: Joseph Carrall tested positive nine days after last using cannabis — after a police officer had told him "a week clear was fine." A magistrate acquitted him on the defence of honest and reasonable mistake of fact. Then, in February 2025, the NSW Court of Criminal Appeal confirmed drug-driving is an absolute liability offence — abolishing that defence entirely.
"It doesn't matter how you came to have the drug detected… you are liable. It's made a bad situation even worse. At least there was a defence in some circumstances. Now, there is no defence." — David Heilpern, former NSW magistrate
Sources: University of Sydney; Sydney Criminal Lawyers (absolute liability ruling).
A University of Sydney study tested the two roadside devices Australian police use and found significant error rates in both:
| Device | False negatives | False positives |
|---|---|---|
| Securetec DrugWipe | 9% | 5% |
| Dräger DrugTest 5000 | 16% | 10% |
The lead researcher: results "often came back positive when they should have been negative, or… negative when they should have actually been positive." These devices "were never designed to measure impairment."
Source: University of Sydney, Lambert Initiative (published in Drug Testing and Analysis).
Legal, prescribed medicinal-cannabis patients commit the presence offence if they drive with any detectable THC — even wholly unimpaired, using their medicine exactly as directed. Tasmania is the only state with a medical defence; Victoria is running a $4.9M closed-track trial to test whether unimpaired patients can drive safely, and NSW is introducing a limited scheme.
Two more red flags Regional NSW is drug-tested at roughly twice the state rate. And the per-test cost to taxpayers is not publicly disclosed — the very transparency that harm-reduction groups are demanding. A presence-based regime, wrong up to 1-in-6 times, with no defence, expanding by targeting, at an undisclosed cost: that is a lot of machine for very little demonstrated impairment.
Sources: BOCSAR (charges & testing trends); Premier of Victoria (trial); NSW Government (reform). Some figures (per-test cost, "only jurisdiction in the world") are flagged by their sources as needing confirmation and are treated cautiously here.
Part · The World Fights Back
Australia treats camera enforcement as settled and expanding. Elsewhere, towns switched cameras off with no rise in deaths, voters banned them at the ballot box, and whole states outlawed them on due-process grounds. The evidence is more contested than the press releases admit.
In 2009, Swindon became the first UK council to switch off its fixed speed cameras. The predicted carnage didn't come:
| Swindon, first 9 months after switch-off | Before | After |
|---|---|---|
| Total road casualties | 327 | 315 |
| Fatalities | 4 | 2 |
| Serious injuries | 48 | 44 |
Casualties fell slightly. Somerset later reported a similar "no worse without them" result.
Source: road.cc (Swindon figures). International.
We're not going to cherry-pick. When Oxfordshire switched off all its cameras in 2010 (after a budget cut), deaths in the affected period rose from 12 to 18, and it switched them back on eight months later. In France, after "yellow vest" protesters destroyed a huge share of the camera network, the road-safety agency linked a 17% rise in road deaths in February 2019 (with the usual caveat that other factors were in play).
So what's the truth? The international evidence is genuinely mixed — Swindon and Somerset saw no rise; Oxfordshire and France did. That's exactly the point: "cameras unambiguously save lives" is not what the natural experiments show. The effect depends heavily on the site — which is why where you put them (see The Goldmines) matters more than how many.
Sources: BBC (Oxfordshire); CNN (France).
Where US voters got a direct say, they overwhelmingly rejected automated enforcement — often explicitly as a revenue-and-due-process grievance.
| Place | Vote |
|---|---|
| Houston, TX (2010) | Banned red-light cameras, 52.8% |
| Anaheim, CA (2010) | Banned, 73% |
| Cleveland, OH (2014) | Banned by roughly 3-to-1 |
| Nov 2011 ballot sweep | 7 of 8 cities rejected cameras |
And it went statewide: Texas banned red-light cameras in 2019, the bill's author citing due process — "the right to due process matters." Several states (Maine, Mississippi, New Hampshire, South Carolina, West Virginia and others) ban speed and/or red-light cameras outright. Arizona killed its statewide freeway speed-camera program in 2010 after it missed revenue targets and roughly two-thirds of recipients simply ignored the citations.
Sources: Texas Tribune; Ballotpedia; Arizona program. International; the US state-ban list should be checked against each state's statute.
Britain's cameras used to run on "netting off" — partnerships kept their operating costs directly out of the fines. The government ended it in 2005–2007 because it was "too politically sensitive": it too openly tied camera operation to the money the fines produced. The result was telling — once cameras had to be funded from general budgets, the 2010 budget cuts led councils to switch them off. Cameras that were about safety would have been protected; cameras that were about revenue lost their reason to exist the moment they stopped paying for themselves.
This is the cautionary tale for Australia's hypothecated "road safety funds" (see The Money): the moment a safety budget depends on fine income, the system needs the fines to keep flowing.
Source: Road Safety GB — National Safety Camera Programme history.
In Italy, a masked vigilante nicknamed "Fleximan" cut down 15+ speed cameras with an angle grinder, becoming a folk hero online. In France, the "yellow vest" movement damaged 60–75% of the country's speed cameras, at a cost to the state of €660 million. You don't get that level of public anger at a measure people believe is really about saving their lives.
Sources: BBC (Fleximan); The Local (France). International — reported as evidence of public sentiment, not endorsement of vandalism.
Part · By State
Eight governments, one pattern. Only one publishes its speed tolerance. Some ring-fence the money then don't spend it; others tip it straight into consolidated revenue. And the territory with the lightest camera footprint has the worst road toll in the country.
| Jurisdiction | Tolerance published? | Latest camera/fine revenue | Where the money goes |
|---|---|---|---|
| NSW | No — most secretive | ~$275M (hist.); fell after signs returned | Community Road Safety Fund — ~$104M underspent |
| VIC | Yes — 2 km/h fixed / 3 km/h mobile | $473M (2023-24) | Better Roads Victoria Trust — but cameras run covert/unsigned |
| QLD | No | $464M gross, ~$337M net (2023-24) | Hypothecated (TORUM Act); majority from lowest-level fines |
| SA | No | 608,794 speed fines (2023-24) | Community Road Safety Fund |
| WA | No | $127M into RTTA (2022-23) | Road Trauma Trust Account — ~$120M sat unspent |
| TAS | No | $9.5M mobile cameras (to Oct 2024) | Consolidated — no dedicated fund (its own advisers want one) |
| ACT | No | ~$51M (2023-24) → ~$104M (2025-26) | Consolidated — no fund at all |
| NT | No | Not separately published | Consolidated — lightest enforcement, worst toll |
Sources on the Sources page. Only Victoria officially publishes its tolerance; every other jurisdiction keeps it confidential, arguing disclosure would create a de-facto higher limit. Demerit suspension is 12 points (13 in NSW).
Its "Community Road Safety Fund" has been underspent every year since 2019-20 — a ~$104M shortfall. A safety fund that isn't spent on safety is just a tax with a nicer name.
The only state honest enough to publish its 2 km/h tolerance — and the one that runs its mobile cameras hidden and unsigned, "anywhere, anytime."
More than half its speeding revenue comes from the lowest-level speeding band — the smallest overages, not the dangerous outliers.
Ran a formal independent Speed Camera Audit of its own siting — an admission the question of "safety vs revenue" is live.
The Auditor-General found ~$120M sitting unspent in the Road Trauma Trust Account, weak ministerial-approval records, and decade-long unevaluated projects.
Camera revenue goes to consolidated revenue — and the chair of the state's own Road Safety Advisory Council has publicly demanded it be ring-fenced.
A single mobile-phone camera gantry earns ~$1 million a month; fine revenue is set to double by 2026 — all into consolidated revenue, no road-safety fund.
The outlier that breaks the spell: no phone cameras, no point-to-point cameras — and the highest road-death rate in the country. If cameras were the answer, the NT would be safest.
Sources: NSW & WA Auditor-General; QLD/VIC/SA/TAS/ACT/NT government & reporting — full list on Sources.
Read the table as a whole and the pattern is unmistakable. The jurisdictions "ring-fence" the money into road-safety funds — then leave tens or hundreds of millions unspent (NSW, WA). The ones that don't bother ring-fencing tip it into consolidated revenue and forecast it to grow (ACT). Only one publishes how it enforces (VIC) — and hides the cameras anyway. And the single place with almost no automated enforcement (NT) has the worst toll — which means the deaths are about roads, distances and infrastructure, not about whether a camera caught you 3 km/h over in a city.
Part · Right to Drive?
"Driving is a privilege, not a right" is repeated like a law of nature. It isn't. It's a 20th-century invention — a political choice made barely a lifetime ago. Here's the honest history, the actual law, and why the "privilege" framing is a choice we're allowed to question.
For all of human history until about 1906, people moved freely — on foot, by horse, by bicycle — with no licence, no registration, no test, no plate. Then the car arrived, and governments built an entirely new bureaucracy around it.
| When | What was invented |
|---|---|
| 1906 | South Australia issues Australia's first driver's licence and introduces car registration |
| 1909–10 | Victoria's Motor Car Act 1909 and NSW's Motor Traffic Act 1909 create the offences of driving without a licence and driving an unregistered vehicle; first number plates issued |
| early years | The first licences were just a fee and an application — no driving test. Competency testing was bolted on later, piecemeal, over following decades |
Horse riders were never licensed. Cyclists, by and large, never were. The licence was created specifically for the motor car and then steadily expanded into the fee-points-fines-and- suspension apparatus we have now. The state didn't discover that driving was a privilege — it manufactured the category.
Sources: SA 1906 (first licence & registration); Tranter, "The History of the Haste-Wagons: the Motor Car Act 1909 (Vic)" (Melb Univ Law Review); NSW Motor Traffic Act 1909 / first plates 1910.
Section 92 of the Australian Constitution declares that "trade, commerce, and intercourse among the States… shall be absolutely free." Australians do have a real, court-recognised freedom to move across State borders — it's why wartime and ex-prisoner travel bans were struck down.
Free movement of people and goods between states, without protectionist burdens. Confirmed in R v Smithers (1912) and Gratwick v Johnson (1945): people are "free to pass to and fro among the States without burden, hindrance or restriction."
It is not a general "right to travel," doesn't cover movement within a state, and says nothing about cars or licences. And in Gerner v Victoria (2020) the High Court unanimously held there is no implied constitutional freedom of movement at all.
So here's the honest bottom line, and we won't pretend otherwise: as the law currently stands, driving is a regulated privilege, and licensing laws are valid. The popular "I'm travelling, not driving, so I need no licence" argument has been run in Australian courts many times and loses every single time — the NSW Judicial Commission records no case of it ever succeeding. Anyone who tells you otherwise is selling a courtroom disaster.
A warning, because we're honest "Sovereign citizen" / "freeman" tactics — refusing to give your name, claiming a licence is a contract you can decline, invoking s92 to drive unregistered — are not valid law and have a 0% success rate in Australian courts. They get people convicted, fined and occasionally jailed for contempt. This page is a political and historical argument, not legal advice, and definitely not that.
Sources: Section 92; Gerner v Victoria [2020] HCA 48; NSW Judicial Commission (pseudolaw never succeeds).
Because a "privilege" can be conditioned on anything, revoked at will, and priced however the state likes. Calling mobility a privilege is what makes the whole machine in this exposé possible.
In much of the country a car is not a luxury — it's the only way to reach work, school, a doctor. Treating that as a revocable favour from the state, rather than a basic freedom to be regulated proportionately, is a political choice. It was made in 1909. It can be questioned in 2026.
Not the pseudo-legal fantasy — the real one. Australia is unusual: it has no bill of rights and, after Gerner, no constitutional protection of freedom of movement within the country at all. Even the United States, for all its faults, recognises a constitutional "right to travel" (while still upholding sensible licensing). Australians' freedom to move rests almost entirely on ordinary statutes that any parliament can tighten tomorrow.
The legitimate reform case follows from that:
"Privilege, not a right" isn't a fact of nature handed down through the ages. It's a sentence a government wrote — and a citizenry can decide to rewrite.
Sources: ALRC — Freedom of Movement; US "right to travel" vs driving; Gerner v Victoria (2020).
Part 5 · Delegations
The most important fact about the fine machine is where its rules come from. The dollar amounts, what counts as a "camera," which offences make you guilty-by-default, and how many points cost you your licence are almost all set by regulation and ministerial instrument — the executive — not debated and voted by Parliament.
Each state has an enabling Act. But the Act mostly grants power to a minister or agency; the operative detail is pushed down into delegated legislation and instruments.
| State | Enabling Act | Key camera / penalty provisions |
|---|---|---|
| NSW | Road Transport Act 2013 | s 195 penalty notices · s 183 "approved traffic enforcement device" · s 273 delegation |
| VIC | Road Safety Act 1986 | s 66 operator-onus · s 80 / 83A camera evidence · reg 32 "prescribed road safety cameras" |
| QLD | Transport Operations (Road Use Management) Act 1995 | s 114 — person in charge "is taken to have committed" the offence |
| SA | Road Traffic Act 1961 | s 79B — owner "is guilty… unless it is proved" otherwise |
| WA | Road Traffic (Administration) Act 2008 | s 86–87 — "responsible person presumed to be driver" |
Section numbers and titles verified via AustLII / state legislation registers; see Sources for the exact provisions. Always read the current consolidated section before relying on it.
Here's the accountability gap. The things that determine your fine are set by the executive, by instruments most people have never heard of and Parliament rarely debates.
Why this matters When the amount, the trigger, the device definition and the points are all set by regulation, the government can ratchet the whole system — higher fines, more offences, tighter tolerances — without ever passing a law. You get to disallow an instrument only within a short window most people never see. It is government by administrative fiat, on an apparatus that touches nearly every adult.
Ordinary criminal law says the state must prove you did it. Camera enforcement flips that. Under "owner-onus" / "operator-onus" provisions, the registered owner is liable by default and must actively act to escape it.
The state must identify the driver and prove the offence beyond reasonable doubt before you're penalised.
You're fined first. To escape it you must nominate the actual driver by statutory declaration (NSW s 186 — within a set period), or prove you weren't driving (QLD s 114, SA s 79B, WA s 87), or elect to fight it in court on your own time and cost.
The practical effect: the burden of identifying the driver is shifted onto you, and the default outcome if you do nothing is that you pay. Multiply that by millions of notices and you have a system engineered for compliance-by-inertia.
Sources: NSW Road Transport Act 2013 ss 186–187; Vic Road Safety Act 1986 s 66; QLD TORUM Act 1995 s 114; SA Road Traffic Act 1961 s 79B; WA Road Traffic (Administration) Act 2008 s 87. See Sources.
Demerit points and licence suspension are administrative penalties, imposed by the agency, on thresholds set by regulation, with no separate court hearing. In NSW, reaching the points threshold (13 for an unrestricted licence, 14 for professional drivers) triggers suspension by Transport for NSW under ss 33/40 — the points values themselves prescribed by regulation under s 43. A fine, points, and the loss of your livelihood can all flow from a camera flash, entirely within the executive branch.
The cameras themselves are frequently operated by private contractors (see Surveillance) — Serco in Victoria, Redflex/Acusensus in NSW. The legal structure keeps the fine "public": the contractor captures and processes the image, but the infringement notice is issued by a state agency (Revenue NSW, Victoria Police, the department) under the "approved device" and delegation framework. So a private company's equipment, operating under a commercial contract whose terms you can't see, sits at the front of a process that ends in a government fine, demerit points and possible suspension.
Sources: NSW Parliament Staysafe Committee, Mobile speed camera enforcement programs in NSW; NSW Government fines pages. See Sources.
This isn't a fringe complaint. Parliament's own reference works and scrutiny bodies describe the gap.
Under Odgers' Australian Senate Practice (Ch 15), delegated instruments take effect on registration and can only be disallowed within 15 sitting days; the scrutiny committee reviews for technical and rights issues but generally avoids the underlying policy, and the sheer volume means most instruments are never debated. Analysts at the Institute of Public Affairs and Australian Public Law have separately warned that delegated legislation lets the executive "bypass democracy."
Put it together: the power to surveil (approved devices, ANPR), to fine (amounts by regulation, indexed automatically), to presume you guilty (owner-onus regulations), and to suspend your licence (administrative thresholds) is exercised almost entirely by the executive and its delegates — under instruments that receive minimal parliamentary scrutiny.
The counterweight that proves the point Victoria created a Road Safety Camera Commissioner (Act 2011) to oversee camera integrity and handle complaints. Tellingly, the Commissioner can review and report — but cannot set penalties or issue fines. The oversight is advisory; the power is executive.
Sources: Odgers' Australian Senate Practice, Ch 15; IPA — Bypassing Democracy; Road Safety Camera Commissioner Act 2011 (Vic).
Part · The Court Trap
The system fines you automatically, presumes you liable, and then makes the only way to fight it a criminal court — where, to get any mercy, you must stand up and say the word "guilty" first. Then it blames you for clogging the courts.
Elect to challenge a camera fine in NSW and it doesn't go to some gentle administrative tribunal. It goes to the Local Court's criminal jurisdiction — the same stream that handles assaults and thefts. The court's own website spells out the stakes:
"For some traffic offences, you could receive a criminal conviction or a licence disqualification… The Court may issue you a fine at a higher amount than the penalty notice… You may have to pay legislated court costs." — Local Court of NSW, "Going to court for a penalty notice"
Legal Aid confirms it plainly: "Your criminal record will also include… traffic offences where you have taken a fine to court, and the Court convicted you." So the machine's offer is: pay the automatic fine quietly, or roll the dice on a criminal record and a bigger fine to prove your innocence.
Sources: Local Court NSW; Legal Aid NSW.
"You could go to jail over a camera fine" sounds like hyperbole. Trace the chain — every link is in the legislation:
| Step | What happens |
|---|---|
| 1 | Camera fine issued. Don't/can't pay → |
| 2 | Overdue fine + $65 fee → enforcement order → |
| 3 | Revenue NSW suspends your licence (an administrative act) → |
| 4 | You drive (to work, to a doctor) → driving while suspended, s54 Road Transport Act → |
| 5 | A criminal offence carrying up to 6 months' jail (first offence) or 12 months (subsequent) |
To be precise: the camera fine itself isn't imprisonable — but the pathway it starts is. A missed fine becomes a suspended licence becomes a criminal charge becomes, potentially, jail. That is an absurd amount of coercive power hanging off a photograph of a numberplate.
Source: Road Transport Act 2013 (NSW) s54 (max 6 months first / 12 months subsequent); Legal Aid NSW (enforcement pathway). See also Punishing the Poor.
Here is the part that gives the game away. The "good outcome" everyone chases is a Section 10 — a finding with no conviction recorded. But to get it, you must first be found guilty or plead guilty.
"Section 10 is a sentencing option that allows the court, after a person has pleaded guilty or been found guilty, to discharge the offender without recording a conviction." — on s10, Crimes (Sentencing Procedure) Act 1999 (NSW)
Read that again. Even if the magistrate privately thinks the whole revenue-camera apparatus is nonsense, the only lever they can pull to spare you requires you to assume guilt for an "offence" a machine decided you committed and the law presumes you liable for. You cannot say "this system is unjust" and get leniency. You must say "I did it, please be kind." The same trap exists in Victoria (s76 / s8 Sentencing Act) and Queensland (s12 Penalties and Sentences Act).
Sources: s10 (NSW); s76 Sentencing Act (VIC); s12 P&S Act (QLD).
The stock complaint is that self-represented "time-wasters" bog down the system. The data points the other way. In NSW, traffic and vehicle offences are the single most common category of criminal matter — 45,371 defendants found guilty of a traffic/vehicle principal offence in one year (2024–25). These are prosecutions the state brings, off the back of automated detection, into a criminal court.
Our view — and it follows from the facts If contesting an automatic fine is the only way to test it, then the people "clogging the courts" are not the drivers exercising their sole avenue of defence — they're the agencies that built a guilty-by-default machine whose only appeal route is a criminal courtroom. The state manufactured the caseload, then blames the defendant for turning up.
Source: BOCSAR NSW Criminal Court Statistics (traffic/vehicle = most common principal offence). The "who clogs the courts" argument is our characterisation, grounded in that data.
Most people fighting a fine can't afford a lawyer for a $300 ticket, so they self-represent — and self-represented litigants fare measurably worse, disadvantaged by procedure and paperwork (Productivity Commission; AIJA). Worse, courts are now primed to treat anyone self-representing on a "the system is illegitimate" theme as a "sovereign citizen" to be shut down fast: the NSW Judicial Commission issues bench guidance on identifying pseudolegal litigants, and Local Courts handled 300+ such matters in six months.
That guidance exists for good reason — pseudolaw is genuine nonsense that has never once succeeded in an Australian court, and this site says so plainly (see Right to Drive?). But there's a side effect: an ordinary person with a legitimate grievance about an unjust fine can get rounded down into the same bucket — pattern-matched to the crank and dismissed before they've been heard. When the only people challenging the system get pre-labelled as cranks, the system never has to answer for itself.
Sources: AIJA (self-represented litigants fare worse); The Conversation (300+ cases; pseudolaw never wins). The "genuine grievances get lumped in with cranks" point is our inference from these facts, not a cited finding.
When highway patrol pulls you over, the officer is carrying a Glock — the standard NSW Police sidearm — and broad statutory stop powers. For an alleged 3 km/h over the limit, the encounter carries the full apparatus of armed state force. That is, by design, intimidating.
And then there's the ritual gratitude. Get let off with a warning, or fined for one offence instead of three, and you're expected to feel lucky — grateful to the officer for "discretion," grateful the fine wasn't stacked higher, grateful the system didn't ruin your week. It's the logic of a shakedown dressed as a favour: manufacture a threat, then present not-quite-executing it as generosity. You are being asked to thank people for the restraint they showed while taking your money under a presumption of guilt.
Fact vs framing The facts: traffic police are armed as standard and hold broad stop powers; fines, demerits and costs can be layered or reduced at official discretion. The framing — "intimidation," "shakedown," "manufactured gratitude" — is our editorial view of what that dynamic feels like from the driver's side. We think it's the honest description. You can judge.
Sources: Australian Institute of Criminology (police firearms carriage); NSW police stop powers (LEPRA s202).
A justice system confident its fines were fair wouldn't need to presume you guilty, criminalise the appeal, demand a confession for mercy, and pre-label its critics as cranks. It does all four. Meanwhile it's your one lawful avenue to be heard — see Take Action for how to use it anyway.
Part · What Actually Works
Criticism is easy; here's the constructive half. The interventions that actually cut road deaths are well-evidenced — and they share one feature: they're boring, un-billable, and design-based. Nobody gets a fine. Which is precisely why they're not the priority.
The single biggest, most replicated win in road safety isn't a camera — it's a shape. Replace a signalised intersection with a roundabout and the severe crashes largely vanish.
32 conflict points — including 16 crossing points that produce high-speed T-bones. Relies on drivers obeying a light; permits full-speed crossing; breeds red-light running (>1,000 US deaths/yr).
8 conflict points, zero crossing conflicts. Forces every driver to slow to ~15–30 km/h and look. Injury crashes down 72–80%; fatal/serious down up to ~90%.
That's not a rounding error — it's the difference between a fender-bender and a funeral. The FHWA estimates converting just 10% of America's signalised intersections would have prevented tens of thousands of crashes in a single year. Traffic lights don't make drivers safe; they make drivers obedient — staring at a bulb instead of assessing the road.
Sources: IIHS — Roundabouts; FHWA Roundabouts Informational Guide (conflict points); IIHS — Red-light running. The "lights breed obedience not awareness" point is an engineering argument (aligned with the self-explaining-roads work below), not a single measured statistic.
The counter-intuitive frontier: sometimes removing signs, lights and markings makes roads safer. The Dutch engineer Hans Monderman pioneered "shared space" — strip out the false certainty of traffic control and drivers, no longer on autopilot, slow down and negotiate by eye contact.
At the Laweiplein in Drachten, after the lights were removed: collisions fell from roughly 30 (with several injuries) over seven years to about 4 collisions and no injuries in the two years after. — shared-space redesign, Drachten, Netherlands
The mechanism is deliberate: a driver told exactly what to do stops thinking; a driver made responsible starts. contested — this works best on low-speed, low-volume urban streets, raises real accessibility concerns for blind and disabled pedestrians, and isn't a drop-in replacement for busy highways. But it proves the deeper point: attention and good design beat surveillance and punishment.
Source: Shared space / Hans Monderman (with criticisms); Project for Public Spaces.
Remember: ~two-thirds of Australian deaths are on rural roads (see The Goldmines) — run-off-road and head-on crashes no urban camera can touch. The fixes are un-glamorous, passive, always-on engineering:
These are the FHWA's "Proven Safety Countermeasures" — quantified, replicated, and they don't rely on catching anyone. A dollar spent on a median barrier on a country highway saves more lives than a dollar spent on a camera at the bottom of a city hill. The budgets say which one governments prefer.
Sources: FHWA — Rumble strips; FHWA Proven Safety Countermeasures.
Here's where the intuition ("just teach people to drive properly") is right — but only if you do it the hard way. The evidence is blunt about the easy way.
Bolting a driver-ed class onto high school does NOT reduce crashes — and can make things worse. The landmark randomised DeKalb County study and the Cochrane systematic review both found that school driver-ed mainly gets teens licensed earlier, increasing their road exposure, so any skills gain is cancelled out. Cochrane's words: it provides "no evidence that driver education reduces road crash involvement" and "may… increase the proportion of teenagers involved in traffic crashes."
What does work is rigour and structure, applied early and seriously:
The lesson the exposé keeps hitting: spend the effort up front making genuinely competent, attentive drivers and building forgiving roads — instead of licensing people cheaply, then harvesting them with cameras and calling the fines "safety."
Sources: Cochrane — school driver education; DeKalb County study; IIHS — graduated licensing; Finland multi-phase licensing. Honesty note: generic driver-ed evidence is strongly negative; graduated licensing is strongly positive; hazard-perception training (vs testing) is mixed — we've labelled each.
Norway has the lowest road-death rate in the world — about 1.6 per 100,000 people (87 deaths in 2024). Sweden, birthplace of Vision Zero, cut its toll by more than half. They didn't get there by hiding cameras and indexing fines. They got there by engineering the system so a human mistake doesn't cost a life — safe speeds built into the road, separated traffic, forgiving roadsides, rigorous licensing, safe vehicles.
Yes, this is "Vision Zero" — the real one Elsewhere on this site we criticise "Vision Zero" — but as a marketing slogan used to justify ever-more enforcement and revenue (see The Strategy). The Nordic original is a different animal: an engineering and design philosophy that shifts responsibility onto the people who build the system, not the individual to be fined. Australia adopted the logo and the enforcement; it skipped the expensive design work. That's why its toll is rising while Norway's falls.
Sources: ITF/OECD — Norway (2025); WEF — Sweden's Vision Zero.
We know what saves lives — roundabouts, median barriers, rumble strips, graduated licensing, honest speed limits, and drivers taught properly. None of it issues a fine, fills a revenue fund, or logs your movements. The tragedy isn't that we don't know the answer. It's that the answer doesn't pay.
Ready to push back? The Take Action page has FOI templates and how to challenge a fine.
Reference · Myth-busting
You'll hear the same lines every time you question a camera. Here's the comeback to each — short, and backed by the government's own audits, budgets and crash data. Screenshot one; win the argument.
Speed kills — so any camera saves lives.
Answer Speed is "a factor" in roughly one in four to one in three fatal crashes — but that bundles illegal high-range speeding with driving too fast for the conditions (often under the limit) and crashes caused by something else entirely. There is no published Australian figure for crashes caused by exceeding the limit on an otherwise safe road — the exact thing low-level camera fines target. Unpack the numbers →
If you don't speed, you won't get a fine.
Answer Only Victoria publishes its enforcement tolerance (2 km/h); everywhere else it's secret and any amount over is an offence. Add owner-onus (you must prove you weren't driving), AI phone/seatbelt cameras that misread objects, and NSW alone refunded nearly 19,000 wrongful fines after one software error. "Just don't speed" assumes a machine that never errs and a limit you're allowed to know. See the errors →
Cameras go where the crashes are.
Answer Victoria's Auditor-General found 15 of 33 sampled AI camera sites had no crash record at all. In NSW, 80% of fixed and red-light cameras sit in greater Sydney while most people are killed on country roads. And when NSW hid its mobile cameras in 2020, fines jumped ninefold — hiding a camera raises money, not safety. See it on the map →
It's about safety, not revenue.
Answer In one year NSW's "road safety" fund spent $118m running the cameras and $36.7m funding police — the enforcement pays for the enforcement. Camera fines equal about a tenth of the entire state police budget, and budget papers forecast fine revenue to keep rising, never fall. A safety program would budget for fines to shrink toward zero. Follow the fund →
Speed cameras have cut the road toll.
Answer The national toll is up ~22% since the plan to halve it — during the biggest camera-enforcement expansion in the country's history. About a quarter of the measured "camera effect" is just regression to the mean (crashes spike randomly, then fall back), and the research is rated "moderate quality at best." Weigh the evidence →
If you've nothing to hide, you've nothing to fear.
Answer A single ANPR unit reads 16 number plates a second — every plate logged, guilty or not, and kept for years. Phone-detection cameras photograph the inside of every car. This gear, sold "for road safety," has already been repurposed to enforce COVID lockdowns. The fine is the excuse; the surveillance is the product. See the grid →
The speed limit is the safe speed.
Answer Limits used to be anchored to how people actually drive (the 85th-percentile rule) — that engineering principle is being quietly retired. Only one state will tell you its camera tolerance, and the limits, fine amounts and owner-onus rules are set by regulation and delegation — not a vote. See who really decides →
Fines make people drive better.
Answer NSW drug-driving charges rose +836% while the state's own crime bureau found "no evidence" behaviour actually changed — the surge is an enforcement artefact. A flat fine costs a millionaire minutes and a single parent a fortnight, and unpaid fines can spiral into a licence loss and even a cell. That's punishment by wallet, not safety. See who pays →
The private operators are just efficient contractors.
Answer The company that has run NSW mobile speed cameras was at the centre of one of the biggest municipal bribery scandals in Chicago's history — executives went to federal prison. The Australian contracts, worth hundreds of millions, are secret. Overseas, some deals historically paid operators a cut of every ticket. Read the contracts →
Speed cameras are world's best practice.
Answer Then why have towns switched them off, cities voted them out, and whole regions banned them — without the bloodbath the industry predicts? Plenty of places treat cameras as a discredited shortcut and invest in roads that are safe by design instead. See who fought back →
You agreed to all this when you got your licence.
Answer You never needed a licence to ride a horse down the same road. "Mobility is a privilege, not a right" is a political choice, not a law of nature — and none of the fines, the secret tolerances, the in-cabin cameras or the plate-tracking was ever put to the public. Right to drive? →
Every claim above traces to a government audit, budget paper, Act or study on this site — check any of them on Sources. Grab a shareable card to make the point in a feed, or hand someone the whole printable report.
Part · Take Action
This whole exposé keeps hitting the same wall: the damning numbers aren't published. So make them. Freedom-of-Information (and NSW's GIPA / Queensland's RTI) laws let any member of the public demand them. Below are ready-to-send requests — with the exact, verified addresses to send them to. Copy, add your name and the date, and file. In NSW you can ask for informal release first, which is free.
Tests the "cameras go where the crashes are" claim. Address it to your state road agency / police (NSW: Transport for NSW or Revenue NSW; VIC: Dept of Justice & Community Safety; QLD: Transport & Main Roads).
To the FOI/GIPA Officer,
Under the [Government Information (Public Access) Act 2009 (NSW) /
Freedom of Information Act 1982 (Vic) / Right to Information Act 2009 (Qld) /
relevant FOI Act], I request access to the following information:
1. For the fixed/mobile speed or red-light camera located at
[ADDRESS / INTERSECTION], for the most recent 5 financial years:
(a) the number of infringements issued, by year;
(b) the total revenue raised, by year;
(c) the recorded casualty and fatal crash history at that location
for the 5 years BEFORE the camera was installed, and for each
year SINCE; and
(d) the site-selection justification / business case used to approve
the camera at that location.
I ask that documents be provided in electronic form. If any part is
refused, please cite the specific exemption and provide the balance.
[Name] [Contact details] [Date]
Tests the incentive to over-issue fines and the secrecy of the deals.
To the FOI/GIPA Officer, Under the [relevant FOI/GIPA Act], I request: 1. The current contract(s) between [agency] and the private operator(s) of the state's speed / red-light / mobile-phone detection cameras (e.g. Acusensus, Redflex/Verra Mobility, Serco, Jenoptik). 2. In particular, the clauses setting out HOW the operator is remunerated — specifically whether payment is per hour of enforcement, a fixed fee, or linked in any way to the number or value of infringements detected. 3. The total value of each such contract and its start and end dates. Where commercial-in-confidence is claimed, please release the payment- structure and total-value clauses with only genuinely commercial pricing detail redacted, and state the public-interest reasoning for any redaction. [Name] [Contact details] [Date]
The number no one advertises: how often the "infallible" machine is wrong.
To the FOI/GIPA Officer, Under the [relevant FOI/GIPA Act], for each of the last 3 financial years, I request statistics showing, for camera-detected traffic infringements: 1. the total number of infringements issued; 2. the number for which a review was requested; 3. the number withdrawn, waived or refunded, and the total value; 4. for mobile-phone/seatbelt AI cameras: the number of images flagged by the system that were rejected at human review (i.e. the machine's false-positive count or rate); and 5. any internal report on camera accuracy, calibration failures or erroneously issued fines during the period. Electronic format preferred. [Name] [Contact details] [Date]
Tests "it all goes back into road safety."
To the FOI/GIPA Officer, Under the [relevant FOI/GIPA Act], for each of the last 5 financial years, I request: 1. the total revenue paid INTO the [Community Road Safety Fund / Better Roads Victoria Trust / Road Trauma Trust Account / relevant fund] from camera and traffic fines; 2. the total amount actually SPENT from that fund each year; 3. the closing balance / accumulated unspent amount each year; and 4. a breakdown of what the spent money was spent on. Electronic format preferred. [Name] [Contact details] [Date]
Prises open the self-funding loop on The Fund — how much comes in, how much is spent, and how much bankrolls more cameras and police. Send to accessapplications@transport.nsw.gov.au (Transport for NSW).
Send to: accessapplications@transport.nsw.gov.au (Transport for NSW) Subject: Informal access request (GIPA) — Community Road Safety Fund To the Access to Information / GIPA Officer, Under the Government Information (Public Access) Act 2009 (NSW), I request informal release of the following. If informal release is declined, please treat this as a formal access application and advise the $30 application fee. For each of the last 5 financial years: 1. Total revenue paid INTO the Community Road Safety Fund, broken down by source — fixed speed cameras, mobile speed cameras, red-light cameras, mobile-phone/seatbelt cameras, average/point-to-point cameras, and any allocation from the Consolidated Fund. 2. Total amount SPENT from the Fund each year, itemised by program — in particular the amounts spent on (a) police enforcement and (b) operating the automated camera enforcement program itself. 3. The closing balance / accumulated unspent amount at the end of each year. 4. Any amount transferred FROM the Fund back into the Consolidated Fund. Please provide documents in electronic form (a spreadsheet if available). [Your name] · [email or postal address] · [date]
Tests the surveillance grid: how long every plate is kept, who can see it, and whether police work to revenue targets. Lodge with NSW Police (online GIPA portal; enquiries iau@police.nsw.gov.au).
Send to: NSW Police Force — GIPA online portal (police.nsw.gov.au → "GIPA");
enquiries iau@police.nsw.gov.au ($30 formal application)
Subject: GIPA application — ANPR retention and camera enforcement
To the Information Access Unit,
Under the Government Information (Public Access) Act 2009 (NSW), I request:
1. For automatic number-plate recognition (ANPR): the number of plates
read/logged per year for the last 3 years; the retention period for
reads that do NOT match any alert ("non-hit" reads); the policy
governing who may access ANPR data; and any record of ANPR data being
accessed for a purpose other than vehicle-registration or traffic
enforcement (e.g. movement or location tracking).
2. For mobile speed cameras: the total contracted and actual HOURS of
enforcement per year for the last 3 years.
3. Any internal document that sets a revenue target, KPI, quota or
expected number of infringements for traffic or camera enforcement.
Electronic format preferred.
[Your name] · [contact] · [date]
Goes to the heart of Speed: the isolated causal contribution of illegal speeding. Send to the Commonwealth department that houses BITRE — foi@infrastructure.gov.au (no fee; a plain email is enough).
Send to: foi@infrastructure.gov.au (Dept of Infrastructure / BITRE — no fee) Subject: FOI request — isolated causal contribution of illegal speeding To the FOI Coordinator, Under the Freedom of Information Act 1982 (Cth), I request access to any document, dataset or analysis held by the Department or by BITRE that: 1. quantifies the number or proportion of fatal or serious-injury crashes attributable specifically to a driver EXCEEDING THE POSTED SPEED LIMIT on a road that was otherwise compliant — as distinct from "travelling too fast for the conditions" (often at or below the limit), and as distinct from crashes primarily caused by alcohol, drugs, fatigue or distraction; and 2. sets out the methodology and definition by which a crash is classified as "speed-related" or "speeding-related" in national crash statistics. If no such document exists, I ask the Department to confirm that in writing. A plain electronic copy is fine. I'm happy to narrow scope to stay within any processing-charge threshold. [Your name] · [contact] · [date]
The secret margin, and whether the top earners are blackspots or traps (see Speed, Goldmines). Send to accessapplications@transport.nsw.gov.au (Transport for NSW).
Send to: accessapplications@transport.nsw.gov.au (Transport for NSW) Subject: Informal access request (GIPA) — speed camera tolerance & site data To the Access to Information / GIPA Officer, Under the Government Information (Public Access) Act 2009 (NSW), I request (informal release preferred; if declined, treat as a formal application and advise the $30 fee): 1. The speed enforcement threshold/tolerance applied before a FIXED camera, and before a MOBILE camera, issues an infringement (the km/h or % over the posted limit). 2. For the 10 highest-revenue fixed speed and red-light camera locations in the most recent financial year: the infringements issued and revenue raised by year (last 5 years), and the recorded injury and fatal crash history at each site for the 5 years BEFORE the camera was installed and for each year SINCE. 3. The site-selection justification / business case used to approve each of those cameras. Electronic form (spreadsheet if available). [Your name] · [contact] · [date]
The false-positive rate no one publishes, plus how the operator is paid (see Cabin Cameras, Contracts). To accessapplications@transport.nsw.gov.au.
Send to: accessapplications@transport.nsw.gov.au (Transport for NSW)
Subject: Informal access request (GIPA) — phone/seatbelt camera accuracy & contract
To the Access to Information / GIPA Officer,
Under the Government Information (Public Access) Act 2009 (NSW), for the last
3 financial years for mobile-phone and seatbelt detection cameras:
1. The number of vehicle images captured; the number AI-flagged as possible
offences; and the number REJECTED at human review (i.e. the machine's
false-positive count or rate).
2. The number of penalty notices issued; the number challenged/reviewed;
and the number WITHDRAWN or refunded, with the total value.
3. The retention/deletion policy for non-offence ("innocent") images.
4. The current contract(s) with the camera operator (Acusensus), in
particular the clauses setting out HOW the operator is paid — per hour,
fixed fee, or linked to the number/value of detections — and the total
contract value.
Where commercial-in-confidence is claimed over the payment structure, please
release those clauses with only genuine pricing detail redacted, and state
the public-interest reasoning.
[Your name] · [contact] · [date]
The full revenue picture and how often fines are quietly reversed (see The Money, Errors). To Revenue NSW via gipa@customerservice.nsw.gov.au.
Send to: gipa@customerservice.nsw.gov.au (Revenue NSW, via Dept of Customer Service) Subject: Informal access request (GIPA) — camera fine revenue & reviews To the GIPA Team, Under the Government Information (Public Access) Act 2009 (NSW), for each of the last 5 financial years: 1. Total penalty-notice revenue from CAMERA-detected offences, split by type (fixed speed, mobile speed, red-light, mobile-phone, seatbelt, average / point-to-point speed). 2. The number of penalty notices issued, by type. 3. The number for which a review was requested; the number withdrawn, waived or refunded; and the total dollar value withdrawn. 4. The number that escalated to enforcement action (e.g. licence or registration sanction). Electronic form (spreadsheet). [Your name] · [contact] · [date]
The Victorian versions of the same questions, including the 2026 VAGO "15 of 33" siting gap. To foi@transport.vic.gov.au (Dept of Transport & Planning); cc foi@police.vic.gov.au for enforcement/siting.
Send to: foi@transport.vic.gov.au (Dept of Transport & Planning) Cc: foi@police.vic.gov.au (Victoria Police — enforcement & siting) Subject: FOI request — road-safety camera revenue, tolerance and site selection To the FOI Officer, Under the Freedom of Information Act 1982 (Vic), for the last 3 financial years: 1. Road-safety-camera fine revenue by camera type. 2. The enforcement tolerance applied before a camera issues an infringement. 3. For distracted-driver and seatbelt cameras: images captured, AI-flagged, rejected at human review (false positives), fines issued, reviewed and withdrawn. 4. The crash history and site-selection basis for the AI camera sites examined in the 2026 VAGO audit (the finding that 15 of 33 sites had no crash record). I request a fee waiver on public-interest grounds. Electronic copy. [Your name] · [contact] · [date]
Where Queensland's ~$466m camera haul goes, and whether its cameras sit at crashes. Lodge via the e-RTI portal at rti.qld.gov.au; enquiries contactrti@tmr.qld.gov.au — ask for free "administrative access" first.
Send to: Qld Transport & Main Roads — e-RTI portal (rti.qld.gov.au, approved form); enquiries contactrti@tmr.qld.gov.au (ask for free "administrative access" first) Subject: RTI / administrative access — Camera Detected Offence Program To the RTI Team, Under the Right to Information Act 2009 (Qld) — or by administrative access if available — for the last 5 financial years: 1. Camera Detected Offence Program (CDOP) revenue by camera type. 2. The road-safety expenditure that revenue funded, itemised by program. 3. The speed enforcement tolerance applied before a camera issues a fine. 4. For the 10 highest-revenue camera sites, the recorded crash history for the 5 years before and after installation. Electronic form (spreadsheet if available). [Your name] · [contact] · [date]
The addresses below were checked against each agency's own page. Where an agency prefers a portal or a mandatory form, that's noted — using it is the safest route.
| Agency (what they hold) | Send to | Fee |
|---|---|---|
| Transport for NSW — cameras, the road-safety fund GIPA | accessapplications@transport.nsw.gov.au | free informal / $30 formal |
| Revenue NSW — fine revenue, penalty notices GIPA | gipa@customerservice.nsw.gov.au (via Dept of Customer Service) | free informal / $30 formal |
| NSW Police Force — mobile cameras, ANPR GIPA | Online GIPA portal at police.nsw.gov.au; enquiries iau@police.nsw.gov.au | $30 formal |
| Dept of Transport & Planning (VIC) — VicRoads / cameras FOI | foi@transport.vic.gov.au · portal online.foi.vic.gov.au | ~$34.50 (waivable) |
| Victoria Police — enforcement FOI | foi@police.vic.gov.au | ~$34.50 (waivable) |
| Transport & Main Roads (QLD) — camera program RTI | e-RTI portal rti.qld.gov.au (approved form); enquiries contactrti@tmr.qld.gov.au | ask for free "administrative access"; ~$59.60 formal |
| Queensland Police RTI | e-RTI portal rti.qld.gov.au; rti@police.qld.gov.au | as above |
| Dept of Infrastructure / BITRE — national crash data FOI (Cth) | foi@infrastructure.gov.au | no fee — a plain email is enough |
Fees change on 1 July each year (VIC/QLD) — confirm the current figure on the agency page. Always ask for a fee waiver on public-interest grounds. Got something back? Send it in via whistle@theradicalparty.com or GitHub — we'll add it, with credit or in confidence.
General information only — not legal advice. Rules and deadlines differ by state and change over time; check your official state authority and consider free community legal help before acting.
Sources
That's the entire point of this site. Below is the public record behind every claim — audits, budget papers, legislation, crash data and reporting. Where a figure is a government claim, contested, or illustrative, we say so on the page it appears.
Note: no Australian jurisdiction publishes an AI false-positive or appeal-withdrawal rate, and no primary "total revenue since 2020" figure exists — only per-year figures are quoted. The garbled "$34m to one driver" claim circulating online is unverified and is not used here.
Every claim on this site should be traceable to a document above. If you find an error, or you have a primary-source document (an FOI response, a contract, a budget line) that strengthens or corrects a point, that makes this stronger. The site is open-source:
github.com/kingomarwashere/safety-theatre
A note on integrity: several primary PDFs (some auditor-general and legislation pages) blocked automated retrieval during research, so a handful of figures were cross-checked against reputable secondary reporting and are labelled accordingly on each page. Section numbers and report titles are verified; before quoting an exact statutory wording, read the current consolidated provision at the linked register.
Almost none of it — the fines, the tolerances, the cameras that photograph inside your car, the fund that recycles your money into more of it — was debated in Parliament, let alone put to the public. The Radical Party exists to change who decides: direct democracy, where you vote on the policies, not just the politicians. Join at theradicalparty.com/join.
Work inside this system? We work with government whistleblowers — whistle@theradicalparty.com.