Why car insurance premiums rose from 2022 to 2024

What the index recorded

The motor vehicle insurance index reached its steepest twelve-month rise of the episode in April 2024, at +22.6%. Nothing that fast had appeared in the series since November 1976. A rise of that size is not a normal pricing cycle; it is a correction, and corrections happen when prices have been wrong for a while.

Prices had been wrong in a specific direction. Insurers entered the period pricing on loss experience gathered during a stretch when people drove far less than usual, and the assumptions built into those filings stopped describing the road almost immediately after they took effect.

Severity, not frequency

The rise was driven mainly by what each claim cost rather than by how many claims arrived. Vehicles built in the last decade carry sensors, cameras and radar in the bumpers and windscreens, so a low-speed impact that once meant a panel and a coat of paint now means calibrated replacement of parts that did not previously exist. Parts availability lengthened repair times, and longer repairs mean more days of rental car per claim.

Labour rates in collision repair rose alongside every other skilled trade, and medical costs behind injury claims rose with them. Litigated bodily injury claims settled higher. Each of those effects lands on the same line of an insurer’s accounts, and they compound rather than offset.

Used vehicle values did something unusual and important. A total loss is settled at the vehicle’s actual cash value, so when used vehicle prices spiked, the cost of every write-off spiked with them, immediately and without any change in driver behaviour. Insurers were paying more for the same crashes.

Rate adequacy and the regulatory lag

A rate is adequate when it covers the losses and expenses it is expected to meet. For most of this period auto rates were not adequate, and the industry recorded underwriting losses that made the point plainly. Insurers responded by filing increases, and the increases had to travel through a regulatory pipeline built for stability rather than speed.

Every stage adds delay. Loss data is already months old when a filing is prepared. Review takes weeks or months, longer where prior approval applies and longer still where a filing is politically contested. Once approved, the new rate only reaches a policy at its next renewal, so a full policy term passes before the change is entirely in force.

The result is that the index kept climbing well after the underlying cost pressure had begun to ease, because what the index measured was the arrival of filings written against conditions that had already changed. Understanding that lag is the whole of understanding why premiums and costs move out of step.

Why the published surveys still show the older market

State premium surveys sit even further down the pipeline than the filings themselves, because a survey reports rates on file at a moment and is then published on the department’s own schedule. The oldest edition in this site’s committed data carries rates effective March 2023, in Arizona, while the newest carries August 2026, in Maryland.

Both are genuine regulator publications and both are the most recent thing their department has issued. Each figure on this site is therefore dated to the rates behind it rather than to the day the page was built, and the monthly price index carries the current direction separately. Reading an older survey as though it described this month is the error the dates exist to prevent.

Questions

How fast did prices rise at the peak?

The motor vehicle insurance index recorded its steepest twelve-month rise of the episode in April 2024, at +22.6%, which was the fastest reading in the series since November 1976.

Were there simply more crashes?

Claim severity did more of the work than claim frequency. Repair complexity, parts and labour costs, medical costs and used vehicle values all raised what an individual claim cost to settle.

Why did increases keep arriving after costs stopped rising?

Because filings are prepared on loss data that is already months old, reviewed over weeks or months, and then applied only at each policy’s renewal. A full policy term can pass before an approved change is entirely in force.