A year or so ago, you might have encountered advice on the news sites to trust others more in the morning, because as the day wears on people become more inclined to lie, cheat, or steal. (Forbes was one place that peddled this particular tidbit ). This idea may not be back in the news now, but it is back in researchers’ minds thanks to a follow-up duet of articles debating what really might be behind any influence of time of day on our moral decisions. So this will be the first part of a three-part series analyzing each article on the topic: today the original article, next week the alternative explanation, and the week after that the original authors’ response.
The original idea that we might be more moral in the mornings was founded on the fact that making moral decisions requires self-control (as I wrote about last year), and self-control is like a muscle that can wear out (which is described well by the New York Times). If you think of self-control as a limited resource that is going to get used up as you go through your day – opting for a healthy breakfast, keeping your hands on the wheel not your phone during morning rush hour, holding your tongue when you encounter that irksome co-worker in the hallway – then naturally things which require self-control would be more difficult as the day wears on. This means your instinctual self-interest may win out over honesty.
To test this idea, the researchers conducted four experiments. The first two were very similar, involving a basic task of indicating whether more dots appeared on the left or right side of the screen. They were paid based on what side they picked, not whether they were correct (5 cents when more dots were on the right, and only half a cent when they were on the left). This makes it rather tempting to just pick the side that pays more, not the side that actually had more dots. Do people resist that temptation better in the morning? Well, the college student volunteers who signed up for “morning” sessions (between 8 a.m. and 12 noon) picked the side that paid more about 4 fewer times, out of a hundred trials, than students who volunteered in the “afternoon” (between 12 noon and 6 p.m.).
Of course, neither side was outrageously unethical. The side that paid more had an obvious majority of dots on 16 trials, and 50 more were ambiguous, so if people were being perfectly fair I’d expect them to pick that side on about 41 trials. Instead, the afternoon group picked that side 24 times (versus 20 in the morning group) in the first experiment, and 20 times (versus 15.5 times) in the second experiment. Even if that were based on a more conservative estimate from just the ambiguous trials, that’s only guessing at chance – and perhaps leaning toward the side of the smaller payout. So while the afternoon group might have been a little less self-sacrificing in their guesses, and more wiling to err on the side of the bigger payout when the correct answer was in doubt, it’s a bit of an exaggeration to declare their behavior “clear cheating” (p. 98).
Aside from the rather small shift in immoral behavior, those experiments had several major issues in splitting volunteers into “morning” and “afternoon”. The distinction between those two seemed very blurry; a student could, as far as we can tell, sign up for 11:30 and be “morning”, but sign up for 12 noon and be “afternoon”. I’m not convinced that this particular half hour is going to involve major losses in self-control. And, of course, students selected which sessions they would volunteer for. It’s probably a bit of a stretch to suggest that people who volunteer for afternoon slots are less moral….but perhaps they do lack the self-control to crawl out of bed before noon (these are college students, remember) and this plays out in their moral decision making, regardless of time of day.
The next two experiments addressed these specific concerns by conducting the experiments online, through Amazon, with morning and afternoon more clearly separated as 8-11 a.m. and 3-6 p.m., and volunteers randomly assigned to complete the key part of the experiment at different times of day.
In the first of these experiments (the third experiment overall), volunteers had a choice between sending a true message, and being paid a quarter for it, and sending a blatant lie for twice the payday. In the afternoon, 65% opted for the lie and the double payday, while in the morning, only 43% decided to lie for the extra 25 cents. This was certainly a more compelling argument for actual lying. This experiment also tried to tie this willingness to lie to lower levels of self-control, with mixed results. A survey on cognitive fatigue showed that the afternoon participants were more mentally drained (by about .75 points on a scale of 1 to 5), suggesting that lower self-control resources could be behind the lying.
However, a second measure of self-control was the volunteers’ willingness to read The New York Review of Books or People magazine. About 60% of the afternoon participants opted for the magazine, versus only 40% of the morning volunteers, suggesting that by the afternoon people were less willing to pick the more cerebral reading material they “should” choose – but this was not quite statistically significant (p = .07, for the methodologically inclined), and they didn’t acknowledge that. I don’t mind researchers using significance levels that close – I’ve done so myself, in fact – as long as the difference from the usual cutoff for significance is discussed. Failing to even add an adjective such as “marginal” or “approaching” significance is flirting with bad science.
Finally, the second online (fourth overall) experiment used yet another task, of looking at a group of numbers (with up to 2 decimal points) and trying to find two that would add to an even 10.0. Doing so would earn a nickel…but half of the sets didn’t have any such pair. In the afternoon, volunteers claimed to have solved about 45% of the unsolvable sets, while morning volunteers claimed to solve only 25% of them. And, people who had indicated the week before that they were more concerned with morality (e.g., saying that falsifying your resume is not okay, no matter who else does it) only cheated on about 15% of the problems in the morning, but about 50% of the problems in the afternoon.
Although each of the four studies has more methodological concerns or exaggerations of findings than I would like, the collection of all four studies together does help shore up some of the weaker links in the story. Overall, I would say there is some intriguing potential for the idea that we find it easier to exert self-control, and therefore resist the temptation to lie or cheat, earlier in the day. Or, as my father put it, that “lack of sleep is the root of all evil”.
The good news (or bad news, if you were hoping for an excuse to take afternoons off) is that there are many options out there to help restore our morality in the afternoon. This research could be a call for re-introducing the mid-afternoon nap into our lives, or perhaps taking a walk in nature or treating ourselves to a taste of sugar, before we make any tough moral calls. Or, as we will see next week, it could be that the afternoons are only risky for the morning people, not for the night owls like myself.
Kouchaki, M., & Smith, I.H. (2014). The morning morality effect: the influence of time of day on unethical behavior. Psychological Science, 25 (1), 95-102 PMID: 24166855