Is willpower a battery that runs out?

Probably not, at least not in the simple way the famous theory claimed. For about fifteen years psychology's best-known account of willpower, called ego depletion, said self-control runs on a limited resource: use it up resisting one temptation and you'll have less left for the next. But large replication efforts in the 2010s failed to find the effect, and today it is one of the best-known casualties of psychology's replication crisis.

The idea came from the social psychologist Roy Baumeister, who compared self-control to a muscle that tires with use. In a 1998 experiment, people who had resisted tempting chocolates gave up sooner on a frustrating puzzle. Many studies followed, and a 2010 meta-analysis of 198 tests reported a moderate effect.

Then the cracks appeared. In 2015, researchers showed that the published literature was skewed by publication bias, and once they corrected for it, the effect shrank to almost nothing. In 2016, two dozen labs ran the same protocol on 2,141 people and found no effect at all. Baumeister objected that the protocol differed from his original, but a second project led by his collaborator Kathleen Vohs, with 36 labs and 3,531 people, found an effect of just 0.06, about a tenth of the 2010 estimate.

Effort can still feel tiring, of course. One alternative explanation is that after a hard task our motivation shifts toward reward and away from control, rather than a fuel tank running dry.