The comparison, in full
Filing costs you the deductible now plus the extra premium for as long as the claim stays rated. Paying yourself costs the repair, once. So filing is cheaper when the repair exceeds the deductible plus the total surcharge, and that sum is the break-even figure this page reports. A $1,000 deductible and $1,300 of extra premium over three years puts the crossover at $2,300: below it, paying yourself is cheaper, above it, filing is, and the further above it you go the more clearly filing wins.
The consequence people miss is that the deductible is not the threshold. A repair a few hundred dollars above the deductible pays out a few hundred dollars and can cost more than a thousand in premium, which is a bad trade made to feel like a good one by the fact that the payout arrives immediately and the cost arrives in twelve-month instalments.
The two numbers this depends on, which nobody publishes
How much a premium rises after an at-fault claim, and how long it stays elevated, are set by each insurer, filed with each state's regulator, and applied against your particular record. They differ between companies for identical drivers and identical accidents. They differ between states, some of which restrict what may be surcharged and for how long. A first claim on a long clean record is treated differently from a second claim in three years. There is no industry schedule, and any page that gives you one is inventing it.
So both are inputs here, and the way to fill them in is to ask rather than to guess. Your insurer can often tell you what a claim of a given type would do to your renewal, and a renewal notice or a fresh quote that reflects the claim is better than any estimate. If you cannot get either, run the calculation twice, at a pessimistic and an optimistic figure, and see whether the decision changes. If it does, the honest conclusion is that you do not know enough to decide yet, and finding out is worth a phone call.
The fade, and what it does to the total
A surcharge is often largest at the first renewal and smaller afterwards as the incident ages, until it drops off entirely. The taper option models that as a straight-line decline across the years you set, so a three-year surcharge starting at $400 costs $400, then $267, then $133, for $800 in total rather than $1,200. Turn it off and the increase is flat for the whole period, which is the more conservative assumption and the one to use if you have no information.
The difference between the two is not small: over three years the flat assumption is 50 percent higher than the tapered one, and that is enough to move the break-even by several hundred dollars. Which pattern your insurer uses is one of the things worth asking about at the same time as the size of the increase.
Things the arithmetic does not decide
Money is one input to this decision and it is not the only one. Anyone was injured, or might later claim to have been, means the exposure has no ceiling and the insurer is who handles it; paying cash for a repair does not extinguish an injury claim brought months later. Fault is genuinely disputed, or another vehicle or property was involved, means the same thing. The repair estimate is a first estimate, and body shops find things once panels come off, so a figure just below the break-even is not really below it.
There is also the policy itself. Many policies require you to report an accident regardless of whether you intend to claim, and quietly not reporting one can affect coverage later. Check the reporting obligation before you treat "handle it privately" as an option, because that is a term of the contract rather than a preference.
And if the break-even is uncomfortably close either way, notice what that is telling you: the deductible may be set wrong for how you actually use the policy. A higher deductible with a lower premium suits someone who was never going to file for a small repair anyway, and the saving is collected every year rather than only in the years something happens.
Questions people ask
How much does insurance go up after an at-fault accident?
It depends on the insurer, the state, the type and size of the claim, and your record before it, and there is no figure that holds across the industry. That is why this calculator takes the increase as an input instead of supplying one. The way to get a real number is to ask your insurer what a claim of this type would do to your renewal, or to obtain a quote that reflects the incident. Comparing that figure against a competitor is also worth doing, since the same claim is not weighted identically everywhere.
Does a claim I do not file still affect my rates?
A repair you pay for yourself and never report does not create a claim record with your insurer. Reporting an incident without claiming is a different matter: some insurers record such notifications and some rate on them, and practice varies. Accidents can also reach an insurer through other channels, including a report from the other party or a state accident report. The safe assumption is that anything involving another vehicle or a police report is known about whether or not you claim on it.
Should I raise my deductible?
If the arithmetic on this page keeps telling you not to file, then you are paying for coverage you do not intend to use. A higher deductible lowers the premium every year, whereas the lower deductible only pays off in the years you claim and file. The constraint is liquidity, not arithmetic: the deductible is a number you must be able to produce at short notice, and setting it above what you could comfortably pay converts a manageable event into a crisis. Set it at the largest figure you could cover from savings without difficulty.
What about comprehensive claims, like hail or a broken windshield?
They are usually treated differently from at-fault collisions, often with a smaller effect or none, and glass claims are handled separately again in several states, sometimes with no deductible at all. The comparison on this page still works, but the increase you enter should reflect that type of claim rather than an at-fault collision figure, and for glass in particular it may genuinely be zero. As with everything else here, that is a question for your own policy rather than for a general rule.