The Peltzman effect
The evidence is real and the argument about it is still running: how strong it is, how far it travels, or whether it repeats. Trust the direction, and hold the numbers loosely.
An economist studying the effect of new car safety regulations in the 1970s reached a controversial conclusion. Mandating safety features like seatbelts, he argued, did not reduce harm as much as expected, because people unconsciously adjusted their behaviour in response to feeling safer. Drivers who felt more protected drove a little faster, a little less carefully, offsetting some of the safety benefit, and shifting some of the risk onto pedestrians and cyclists who got no such protection. The general idea, that making something feel safer can lead people to take more risk, became known as the Peltzman effect or risk compensation. It shows up in various settings, from sports safety equipment to financial safeguards. It is genuinely disputed how strong the effect is, and whether it fully or only partly offsets the safety gain varies a lot by situation, so it is best held as a real tendency to watch for rather than a reliable law.
Making something safer can partly backfire, because people respond to feeling protected by taking more risk, quietly eating into the benefit the safety measure was meant to provide. This is worth keeping in mind whenever you add a safeguard, a protection, or a safety net, whether in systems, teams, or your own life, because behaviour adjusts to the new sense of security in ways that can undercut it. Safety measures still help on balance. The catch is that the human response to feeling safe is to push a little harder against the new limit, so the real effect of a safeguard is often smaller than its designers expected, and occasionally it moves the risk onto someone else.
Moral hazard is about who pays. Somebody else carries the loss, so the arithmetic genuinely changes and spending rises. The Peltzman effect is about how safe something feels while the same person still carries the full loss. The seatbelt driver has transferred nothing to anyone and speeds up anyway. They get confused constantly because both end in more risk taken, and they need different fixes. Moral hazard is fixed by moving the cost back to the person deciding. Peltzman is not, because the cost never left them. A company car with full insurance is both at once: the driver pays for neither the damage nor the feeling of protection, and no single change to the policy addresses both.
Source: Sam Peltzman, 1975, Journal of Political Economy. The magnitude of the effect is genuinely disputed and situation-dependent.
Dan Gardner, 2008
On how people actually perceive and respond to risk, often in ways that surprise the experts.
Draw your own card. It does not take long, and it rewards taking your time.