Choosing a higher deductible reduces the premium by more than the arithmetic of expected losses alone would suggest. Several distinct effects stack on top of each other.
Small claims are the most numerous
Loss distributions are heavily weighted towards small events. Minor damage happens far more often than total loss, so a deductible removes the bulk of claim occurrences.
Each of those removed claims carried an expected payout, and the sum of many small expected payouts is a meaningful share of the premium.
The insurer's exposure to catastrophic events barely changes, which is why the reduction in cover feels smaller to the policyholder than the reduction in price.
Handling a claim costs money regardless of size
Processing any claim involves notification, assessment, possibly an inspection, and payment. Much of that cost is fixed and does not scale down with the amount claimed.
For small claims, the administrative expense can be a large fraction of the payout itself, which makes them disproportionately expensive to cover.
Eliminating them through a deductible therefore saves the insurer both the payout and the handling cost, and part of that saving is passed into the premium.
Deductibles change behaviour as well as exposure
A policyholder bearing the first portion of any loss has a direct interest in preventing damage and in maintaining the insured property.
They are also less likely to submit a marginal claim, since the recoverable amount above the deductible may not justify the effort or the effect on future pricing.
Insurers price for this. A higher deductible signals a policyholder who will absorb minor losses, which lowers expected claims beyond the mechanical effect.
Percentage deductibles behave differently
Some policies, particularly for named catastrophe perils, set the deductible as a share of the insured value rather than a fixed sum.
The amount payable by the policyholder then scales with the value insured, and can be far larger than a flat deductible on the same policy.
These provisions often apply only to specific causes of loss, which means the same policy can carry two very different deductibles depending on what happened.
The limit of the trade-off
Raising a deductible only makes sense up to the amount that could be paid without difficulty, since the point of the cover is to prevent an unmanageable loss.
The premium saving also flattens out. Beyond a certain level, further increases remove very little additional expected cost for the insurer and return little to the policyholder.
The useful range therefore sits between the level where small claims are removed and the level where an uninsured loss would create real hardship.