The less likely an unwanted behaviour is to change voluntarily, the more likely a programme designed to change it is to cause net harm.
The third law in Losing Ground (1984). Murray reasoned that behaviour which people will not change on their own needs strong incentives to change, and strong incentives also reach, and reward, people who would have behaved well anyway.
In practice
- A programme paying bonuses to former offenders who stay out of trouble makes its largest payments to those who would have gone straight without it.
- A school rewards previously disruptive pupils for good behaviour, and the pupils who were always well behaved notice.
- Possible explanations
- One explanation is that inducements large enough to move resistant behaviour distort the choices of everyone else who can qualify. Another is selection: programmes attract those who gain most from them, not those who most need to change.
- In its favour
- Consistent with principal-agent reasoning and with evaluations in which generous inducements produced perverse responses.
- Against it
- Many programmes aimed at hard-to-change behaviour, such as intensive early-childhood or addiction treatment, show net benefits in controlled evaluations. The law names a risk without showing how often it is realised.
- Standing
- Stated as a law but not tested as one; the evaluations cut both ways. Compare Murray's law of unintended rewards and the cobra effect.
- Subject
- Economics & Incentives
- Essay
- The theories behind every argument about money