When your insurer scores you in real time

The scene is already ordinary. An app on your phone, or a small box under the dashboard, watches how you drive and turns your behavior into a score, and the score becomes a premium. The good driver pays less, the reckless one pays more, and each is handed back, month after month, the numbered reflection of their own conduct. The same logic is spreading to other fields, the smartwatch counting your steps for your health coverage, the domestic sensor watching your wiring for your home policy. This is called behavioral pricing, and it arrives wearing the friendly features of fairness, you no longer pay for others, you pay for yourself. It is precisely this promise of fairness that must be examined, because it conceals a reversal.

Here is the thesis. Behavioral pricing seems to make insurance fairer, and in a sense it does, you pay your own risk and not the group's. But the fairer insurance becomes at the level of the individual, the less it insures. For insurance was never founded on individual fairness, it is founded on pooling the lucky with the unlucky, and that silent transfer is the protection itself. To price everyone exactly is to dissolve the group. Perfect personalization is the quiet abolition of insurance.

What pooling does, what the score undoes

To grasp the reversal, one must see what pooling actually achieves. Insurance works because it does not bind the premium to individual risk, because part of what is paid by those who will have no accident funds those who will. No one knows in advance which side they will fall on, and it is this shared ignorance that makes the wager bearable. The premium is your share in a collective risk rather than the price of your own, a difference of nature and not of degree. In one case you buy protection, in the other you settle a forecast bill.

Behavioral pricing eats at this foundation detail by detail. By scoring you continuously, it brings your premium close to your own expected loss, it gradually separates the good risks from the bad, it hands each person their own cost. Pushed to its end, this logic yields a population of individuals each paying exactly their own expected loss, and at that limit insurance stops transferring risk. It becomes a mirror, handing you back your risk with a management markup. The right image is the umbrella. Pooling is a large shared umbrella under which people huddle. Behavioral pricing gives everyone an umbrella cut to their own weather forecast, and at that point no one shelters anyone, and the one whose storm is fiercest receives the smallest umbrella.

From a bet on the future to a verdict on the present

Behavioral pricing does not only shrink pooling, it changes the very nature of what you buy, and this shift is worth pausing on. Classic insurance is a bet on an unknown future. You do not know whether you will crash, so you join with others who do not know either, and the group absorbs whoever fate selects. Behavioral pricing turns this bet on the unknown into a continuous judgment of your present conduct. It stops being coverage against destiny and becomes a running verdict on your observed behavior. What you buy is no longer protection against chance, it is a reward for your compliance.

This slide carries a disciplinary logic that must be named. Once the insurer watches you, it no longer merely prices your risk, it begins to shape your conduct. The score that drops when you brake gently, that rises when you drive at night, that congratulates you for walking ten thousand steps, sets the price and governs behavior in the same stroke. The insurer becomes a quiet regulator of daily life, nudging you to drive, eat, sleep and move as it prefers, because it watches you and its gaze carries a cost. Insurance then leaves its historic role, indemnifying misfortune, for a new one, steering conduct. This is not necessarily bad, but it is no longer the same thing, and one should know it on signing.

To this power to steer is added an opacity that deepens its reach. The insured sees their score, rarely the formula that produces it. They do not know the exact weight of a brake, a trip, a late hour, and they discover their mark without being able to dispute it, still less correct it. The judgment is delivered by a box no one opens for them, and two insurers may score the same conduct differently with no way for the insured to arbitrate. The information relationship has even reversed. Classic insurance feared that the insured knew more than the insurer about their own risk, that was the source of moral hazard. Behavioral pricing turns the table, now it is the insurer who knows more about you than you do yourself, and this new, silent asymmetry shifts power to the side that holds the sensor and reads the data.

One sees better, at this stage, why the sense of fairness misleads. It rests on a moral intuition, that each should pay according to what they do, whereas insurance rests on the opposite intuition, that misfortune should be shared whoever it strikes. Both intuitions are legitimate, but they cannot coexist without limit, because the more one satisfies the first, the more one betrays the second. Behavioral pricing decides, without saying so, in favor of individual responsibility and against solidarity, and it presents this eminently political choice under the neutral mask of a better measurement of risk.

Today's reward, tomorrow's exclusion

These schemes are no longer a matter of anticipation. Usage-based motor insurance, which adjusts the premium to actual driving, has settled into many markets, health programs rewarding physical activity measured by a smartwatch are multiplying, and domestic sensors that watch water, electricity or intrusion are entering home policies. Each advances under the banner of prevention and discount, and each collects, along the way, a continuous stream of intimate data on how you drive, sleep, move, live. The question of price is therefore doubled by a question of privacy, because scoring a behavior first requires watching it, and insurance becomes one of the vastest instruments of observation of ordinary life ever consented to.

One must be honest with the opposing case, for it is strong. Behavioral pricing genuinely rewards care, and that reward has virtuous effects. A driver who knows their driving is watched brakes more gently, an insured who tracks their health moves more, so behavioral pricing can reduce accidents and illness, turning insurance into a tool of prevention rather than a mere fund for compensation. It fights, besides, two old ills of flat pricing, moral hazard, since the watched insured slackens less, and adverse selection, since the good risks stop fleeing a pool where they felt fleeced. For the careful majority, the bill falls and safety rises. That is not nothing, and to wave it away would be dishonest.

But the other side of the scale is heavy. Behavioral pricing punishes the unlucky by circumstance, the one who must drive at night because their work demands it, the one whose health is shaped by their genes, their hardship or their poverty before it is shaped by their choices. It erodes solidarity by making every risk visible and billable, it gradually isolates those whose profile is costly, and it drifts toward a world where those who most need protection are those who can least afford it. This is the risk spiral, already met in the face of climate, where insurance withdraws precisely from where it would be most needed, leaving behind a population of the uninsurable whom perfect pricing has singled out one by one.

The deepest knot lies in this observation, there is a threshold beyond which a fairer price becomes anti-social. A society needs certain risks to be shared, health, mobility, housing, precisely so that misfortune is pooled rather than borne alone. Pushed to its limit, behavioral pricing un-shares these risks, it hands each person their own burden and undoes the sharing that made the burden bearable. Where to set that threshold is a political decision, never an actuarial one, even when it hides beneath equations. And it is because this choice is political that it should not be left to the play of premiums alone.

The spiral no one chooses

The most troubling part is that this shift requires no one's will, it imposes itself by the sheer mechanics of the market. Let a single insurer start rewarding good risks with a discount, and those good risks migrate to it, leaving the others with an impoverished book, costlier to cover. Forced to raise their prices, those others in turn see their best profiles flee, until they adopt the score to survive. So even an insurer convinced of the value of pooling is pushed, by competition, to segment in order not to inherit the bad risks alone. It is a prisoner's dilemma applied to solidarity, each knows sharing would benefit all, but none can hold it alone if a rival defects. The market cannot stay at pooling once one player leaves it, and the spiral that dismantles sharing needs no architect, only a first defector.

One will object that nothing is imposed, that the insured remains free to refuse the app and keep a flat premium. But that freedom is a trompe-l'oeil. Once the careful obtain a discount by letting themselves be watched, refusing means paying more, and the flat rate becomes the tariff of those who have something to hide, or who simply value their privacy. Consent is then manufactured by the price gap, no one forces you, they merely make refusal costly, which, beyond a certain gap, comes to the same thing. Surveillance is made economically rational rather than imposed, and that is a far more effective way to install it than coercion, because it leaves the watched the illusion of having chosen it.

The proof from genes

There is a striking demonstration that society already knows where this threshold lies, and it comes from the genetic file. The genetic test is the perfect-pricing tool par excellence, it predicts part of health risk with a precision no observed conduct will ever match. Yet many countries have chosen to forbid insurers from using it, refusing to let a test result set a premium or justify a denial of coverage. This choice matters, because it shows that before ultimate precision, society did not say yes, it said no, in the name of solidarity. It judged that insuring someone against an illness written in their genes, which they did not choose, mattered more than the actuarial justice of billing them for their destiny.

Behavioral pricing is the same debate moved one notch along. The driving score, the step counter, the domestic sensor are softer, more consented, more gradual versions of the genetic test, and they advance where the ban has not yet been placed. The question they pose is identical, how far do we let the measurement of risk decide solidarity. The answer given for genes draws a clear moral boundary, it remains to be seen whether we will hold it when the measurement turns smiling, voluntary, rewarded with a discount, rather than cold and imposed. For a consented surrender is still a surrender, and a solidarity dismantled premium by premium vanishes as surely as one forbidden by law.

Your reflection, and nothing else to sell

There remains a more distant worry, but one already visible in outline. Nothing forbids, in time, that these scores cease to belong to each insurer and become a portable, aggregated risk reputation, following the individual from one contract to the next as a credit score follows a borrower. Yesterday's bad driver would then drag their mark everywhere, the person with a health profile judged unfavourable would be preceded by their file, and one insurer's refusal would become a signal picked up by the next. One would move from an internal segmentation, already heavy, to a durable marking of individuals by their risk, where error and misfortune would be paid for long after. This is the point where behavioral pricing stops being a matter of premium and becomes a matter of social status, and where the question of solidarity meets the question of liberty.

The score that rewards you today is the same instrument that may exclude you tomorrow, because what measures your virtue also measures your misfortune. Behavioral pricing forces insurance to face a question it had long side-stepped, do we want a system that hands each person back their own risk, or one that shares it. One cannot have both without limit, and to pretend otherwise is to let the second option dissolve in favor of the first, without ever having decided it openly.

The more precisely your insurer knows you, the less it needs anyone to carry you, and the day it knows you perfectly, it has nothing left to sell you but your own reflection. An insurance that has an answer to everything about you is no longer insurance, it is a forecast account managed by a third party, where the chance you thought you were transferring has been handed back to you, wrapped in a score and billed to the cent. The original promise of insurance was to lift you, for a time, out of the tyranny of your own risk. Behavioral pricing, if allowed to run its course, returns you to it, while persuading you it was fairer this way. It will then remain to decide, collectively, what we truly want to protect, the fairness of the price, or the sharing of fate. And it will have to be decided soon, for every finer premium makes turning back a little harder, every good risk that leaves weakens the common pot a little more, and a solidarity is not rebuilt as fast as it is dismantled. The choice is still open, it will not stay so indefinitely.