California rider safety data: what it shows and what it does not

Motorcycle crash statistics get quoted at riders constantly, usually by someone arguing that riding is the problem. It is worth understanding what the data actually is, where it comes from, and what it cannot tell you — particularly about your own case. This page is descriptive. It states no legal rule and no figures. This is general information, not legal advice.

Where the data comes from

Californian motorcycle crash data is built from collision reports completed by law enforcement. Those reports are coded and compiled into the Statewide Integrated Traffic Records System, which is the source behind most state and local analysis. National datasets are assembled separately from state reporting and use their own definitions and thresholds.

That origin matters, because everything downstream inherits the properties of a police report — including its limits, which are set out at how to read a California traffic collision report.

What the data is reasonably good at

• Showing where crashes cluster. Location and roadway type are recorded reliably enough to identify intersections and corridors that produce repeat collisions.
• Showing the recurring crash configurations. The pattern that dominates serious motorcycle collisions is a vehicle turning or crossing into a rider's path at an intersection.
• Showing trends over time, provided the definitions have not changed in the period being compared.

The first two are genuinely useful to riders. A roadway that produces repeated collisions can support a claim about the roadway itself — see road hazards. And the dominant configuration is the subject of left-turn collisions.

What it is not good at

• Fault. The coded contributing factor is an officer's assessment recorded at the scene. In motorcycle crashes it is frequently formed without the rider's account, because the rider has been taken to hospital. The most seriously injured participant is the one least likely to have contributed to the record.
• Under-reporting. Collisions without significant injury are inconsistently reported, which distorts any severity ratio built from the data.
• Exposure. Comparing risk per mile requires knowing how many miles were ridden, which is estimated rather than measured. Rates built on estimated denominators carry that uncertainty and usually do not display it.
• Anything about an individual. This is the important one.

The aggregate-to-individual mistake

Here is the move to watch for. A defence or an adjuster observes that motorcyclists are over-represented in serious injury statistics, and treats that as though it said something about whether this rider, in this collision, did anything wrong.

It does not, and the reason is straightforward. A population-level rate describes a group. Fault is a finding about specific conduct by specific people at a specific place and time. The two are different kinds of statement, and one cannot substitute for the other.

The over-representation also has an obvious non-fault explanation: a rider has no structural protection, so the same collision that dents a car hospitalises a motorcyclist. Higher injury severity for the same crash is about physics, not about conduct. Reading a severity statistic as a conduct statistic is a category error, and it is one that gets made in claim files constantly.

How that assumption operates in practice: bias against riders.

What decides your case instead

The evidence of what happened: the physical reconstruction, the data from the vehicles, independent witnesses, camera footage, and the traffic rules that applied at that location. None of it is statistical, and all of it is specific to you.

How fault is decided · gear and evidence.

Common questions

Where does California motorcycle crash data come from?

Primarily from collision reports collected by law enforcement and compiled into the Statewide Integrated Traffic Records System, with national data assembled separately from state reporting. It is report-derived, which is the source of most of its limits.

Can an insurer use statistics to argue I was at fault?

They try, and it is not evidence of anything about you. Aggregate risk describes a population. Fault in your case is decided on what happened at that intersection, on that day, between those two road users. A group statistic is not a fact about a person.

What are the known limits of the data?

It reflects the officer’s view recorded at the scene, which in motorcycle crashes is often formed without the rider’s account because the rider has been taken to hospital. Minor collisions are under-reported. And exposure is poorly measured, so per-mile comparisons rest on estimates.

Is any of it useful to a rider?

Yes, in one respect. Data on where and how these crashes cluster — intersections, left turns, and drivers violating right of way — supports the point that the recurring failure is a driver not seeing a motorcycle, which is exactly what the left-turn cases are about.