A questionnaire that is not interested in you personally
It is easy to read a proposal form as an assessment of character, and almost nobody enjoys being told that their job title changes the price of insuring a car. But a rating model has no opinion about anybody. It has a large record of what happened to previous policyholders, and it searches that record for variables that predict claims well enough to be worth including.
Occupation and address survive that filter because they predict, not because anybody decided they should. If a variable stopped predicting, it would be dropped without ceremony. Understanding the questions this way makes them considerably less irritating, and it also makes clear why arguing that you personally are not typical of your group has no effect on the quotation.
The address question is mostly about where the car sleeps
A postcode or its equivalent carries an enormous amount of information in a very small field. It encodes traffic density, the mix of junctions and parked cars, local vehicle theft and vandalism rates, the cost of repair labour in that area, flood exposure, and how many claims arose from vehicles kept there. That is why it moves prices as much as it does, often more than any single fact about the driver.
The precise question is usually where the vehicle is kept overnight rather than where the policyholder lives, and the two are not always the same. A car garaged at a workplace, or parked at a partner’s address most nights, or kept at a student term-time address, belongs where it actually stands. Registering it at a quieter address than the one it occupies is a misrepresentation with the ordinary consequences attached.
Occupation is a coarse proxy, and the wording matters
A job description compresses several things at once: typical journey patterns, typical hours, whether the work involves carrying tools or equipment, and how the vehicle is likely to be used during a working day. It also correlates with things nobody would ask about directly, which is part of why the factor works and part of why it is contested.
What surprises people is how much the exact phrasing changes the answer. Occupations are matched against a fixed list, and two descriptions a person would consider interchangeable can sit in different categories with different claims histories behind them. The rule is simply to choose the entry that describes the work most accurately. Selecting a neighbouring title because it quotes better is a misdescription, and it is one an insurer can check against an employer.
Why a price moves when nothing about your driving has
Changing an employer, being promoted, retiring or moving from employment to self-employment can move a premium noticeably even though the same person drives the same car on the same roads. That feels arbitrary, and from the individual’s point of view it more or less is. From the model’s point of view the category has changed, and the category carries a different average.
Retirement is the clearest example. The journey pattern genuinely changes — the peak-hour commute disappears, discretionary trips replace obligatory ones — and the model reflects that. This is also why any of these changes should be reported when they happen rather than at renewal. A policy describes a set of circumstances, and when the circumstances change the description has to keep up.
The fairness objection, and what regulators have done with it
The complaint against proxy variables is serious and worth stating properly: they price an individual on the behaviour of people they merely resemble, and some proxies correlate with characteristics that it would be unlawful to price on directly. That is not a hypothetical concern, and it has produced real regulatory intervention in several markets, on particular factors and with mixed results.
The counterargument is that removing a predictive factor does not remove the risk, it redistributes the cost onto drivers who did not generate it, and it can make cover harder to obtain for the people the intervention was meant to help. Both positions have evidence behind them and the debate is genuinely unresolved. Which factors may be used where is a matter of local law, and it changes.
What a driver can honestly do with any of this
Very little of it is adjustable, and the parts that are adjustable are tied to decisions far larger than an insurance premium. Nobody sensibly moves house to change a motor quotation. What is worth doing is making sure the entries are accurate rather than approximate, because a job description chosen carelessly can misprice a policy in either direction and creates an avoidable weakness if a claim is ever examined.
It is also worth checking the entries at every renewal, since occupations drift and addresses change without anybody thinking of the motor policy. A quotation is only as good as the description it was built on, and an out-of-date description is not a saving. It is an unresolved question waiting for the least convenient moment to be asked.
