Contrary to popular psychological theory, believers in free will were no more generous or honest


By Christian Jarrett

There’s a popular idea in psychology that among the important factors shaping our honesty and generosity is our belief in the concept of free will. Believe more strongly in free will, so the theory goes, and you will be more inclined to prosocial behavior. Supporting this, studies that have momentarily undermined people’s belief in free will – for instance, by giving them a text to read about genetic determinism, or about how neuroscience shows our decisions are out of conscious control – have found that this increases people’s propensity for cheating and selfishness.

Such an effect seems understandable – after all, the notion that humans can choose whether to behave well or badly is fundamental to how we think about moral responsibility. It’s plausible that if you portray free will as an illusion then you provide people with a ready-made excuse for bad, selfish behavior, thus increasing the temptation for them to act that way.

As ever, however, reality is refusing to conform to a simple, intuitively appealing story. Recent attempts to replicate the influence of changing people’s free will beliefs on their subsequent moral behavior have failed, or have applied only to specific groups of people, but not others.

Now a series of four large studies conducted on Amazon’s survey website, each involving hundreds of people, has failed to find a correlation between people’s beliefs about free will and either their generosity toward charities or their inclination to cheat. Writing up their findings in Social Psychological and Personality Science, Damien Crone and Neil Levy at the University of Melbourne and Macquarie University said “… we believe there is good reason to doubt that free will beliefs have any substantial implications for everyday moral behaviors.”

The researchers asked their participants to complete two different comprehensive measures of their beliefs in free will, then gave them the opportunity to share some of their profits from a financial game with charities of their choice. The studies also included a measure of how central the participants saw virtuous traits as being to their identity. And in an online dice game, there was a chance to cheat secretly to earn a larger financial bonus.

Across all four studies, people who believed more strongly in free will were no more generous (if anything there was a trend in the other direction, toward less generosity), and cheated just as often. On the other hand, seeing virtuous traits as more central to one’s identity was associated with more prosocial behavior – suggesting the outcome measures were sensitive enough to detect meaningful associations in the data.

The new studies aren’t perfect, of course – for example, there are lots of other ways that one could conceptualize moral behavior – and should be seen as adding more detail to our understanding of this interesting area of inquiry, rather than shutting it down.

But in the context of the recent failed attempts to replicate the moral consequences of momentarily undermining belief in free will, the new findings do give reason to question the claims of those psychologists and commentators who have already raised the alarm rather loudly about the potential adverse effects for society of disseminating scientific findings that challenge the concept of human free will. As Crone and Levy conclude, “more research is required before actively discouraging free-will skepticism out of fear of moral degeneration.”

Are Free Will Believers Nicer People? (Four Studies Suggest Not)

Christian Jarrett (@Psych_Writer) is Editor of BPS Research Digest

5 thoughts on “Contrary to popular psychological theory, believers in free will were no more generous or honest”

  1. This finding (of a failure to replicate) does not surprise me at all. The original studies that I read demonstrating that belief in free will was positively associated with moral behavior (in a cheating context) were flawed. I had good reason to believe the authors completely miscoded their data. At the time, I was using the exact same computer program they used to measure cheating, but my results were the opposite of theirs. Sometimes failures to replicate simply indicate carelessness among research assistants, not something nefarious. That’s still bad for science when it happens, of course, but the implications are different from when there is intentional research misconduct.

Comments are closed.