New research challenges the idea that willpower is a "limited resource"

When we use willpower to concentrate or to resist temptation, does it leave us depleted so that we have less self-control left over to tackle new challenges? This is a question fundamental to our understanding of human nature and yet a newly published investigation reveals that psychologists are in open disagreement as to the answer.

The idea that willpower is a limited resource, much like the fuel in your car, is popular in academic psychology and supported by many studies. In their recent report What You Need To Know About Willpower: The Psychological Science of Self-control, the American Psychological Association states "A growing body of research shows that resisting repeated temptations takes a mental toll. Some experts liken willpower to a muscle that can get fatigued from overuse."

This view was backed by an influential meta-analysis published in 2010 [pdf] that looked at the results from nearly 200 published experiments. But now a team led by Evan Carter at the University of Miami has argued that the 2010 study was seriously flawed and they've published their own series of meta-analyses, the findings of which undermine the limited resource theory (also known as the theory of ego depletion).

Many psychology studies on willpower follow a similar format: one group of participants is first asked to perform an initial challenging task designed to drain their willpower, before completing a second "outcome" task that also requires willpower. For comparison, a control group of participants performs the outcome task without the first challenge. Superior performance by the control participants (on the outcome task) is taken as evidence that the willpower of the first group was left depleted by the initial challenge, thus supporting the theory that willpower is a limited resource.

The new meta-analyses and the 2010 effort both consider the combined results from many studies following this format, but the new analyses are far stricter in that they only consider studies that used tasks well-established in the literature as ways to challenge willpower, including suppressing emotional reactions to videos and resisting tempting food, and that also used established tasks as outcome measures, including persistence on impossible anagrams, food consumption and standardised academic tests (such as the graduate record exam). The 2010 analysis, by contrast, included a far wider range of studies including those that stretch the definition of a willpower challenge to its limits, including darts playing and purely hypothetical temptations.

Another key difference between the 2010 study and the new analyses is that Carter and his team trawled conference reports to find unpublished studies on willpower. This is important because in this scientific field, as with most others, it's likely there has been a bias in the literature towards publishing positive results (in this case, those consistent with the popular idea that willpower becomes depleted with repeated use).

When Carter's team analysed the evidence from the 68 relevant published and 48 relevant unpublished studies that they identified, they found very little overall support for the idea that willpower is a limited resource. The one exception was when the outcome measure involved a standardised test – here performance did appear to be diminished by a prior self-control challenge.

But for other outcome tasks such as resisting food, the combined data from published and unpublished experiments either pointed to no effect of a prior self-control challenge, or there was worrying evidence of a publication bias for positive results, as was the case, for example, when the outcome challenge involved impossible anagrams or tests of working memory. The new meta-analyses even found some support for the idea that self-control improves through successive challenges, a result that's consistent with rival theories such as "learned industriousness".

This new series of meta-analyses should be not be taken as the end of the theory of willpower as a limited resource. Proponents of that theory will likely respond with their own counter-arguments, including questioning the use of unpublished work by the new study. However, the results certainly give pause. "We encourage scientists and non-scientists alike to seriously consider other theories of when and why self-control might fail," Carter and his team conclude. It's worth noting too that this message comes after the recent doubts raised about a related idea in willpower research – specifically, the notion that depleted self-control is caused by a lack of sugar in the body.

Further reading

Carter, E., Kofler, L., Forster, D., & McCullough, M. (2015). A Series of Meta-Analytic Tests of the Depletion Effect: Self-Control Does Not Seem to Rely on a Limited Resource. Journal of Experimental Psychology: General DOI: 10.1037/xge0000083