All of us are stressed, every now and then. Acute stress can have a profound impact on the human body and mind: both physical and psychological stressors affect the autonomic nervous system and the hypothalamic–pituitary–adrenal axis, leading to changes in cardiovascular and neuroendocrine measures. Stress also is shown to affect cognitive functions like memory and attention. Just recently, however, research discovered that acute stress also can influence our moral behaviour.
In a fascinating study, published in ‘Psychological Science’, 34 male participants were put under acute stress by using a standardized psychological stressor, well-established in stress research. Compared to the not stressed control group, the stressed participants not only showed physiological responses like increased heart rate and a higher levels of the ‘stress hormone’ cortisol – they also behaved more pro-socially. When offered to either share some money they got with another participant or to keep all for themselves, the stressed participants shared more. They also acted in a more trustworthy manner and were willing to trust others more themselves.
Another study, published in ‘Psychoneuroendocrinology’ this month, suggests that this phenomenon cannot be transferred to altruistic behaviour in general: participants under acute stress did not donate more money to an anonymous charity than their unstressed counterparts. Apparently, being confronted with an actual other person is necessary for stress to increase pro-social behaviour. An explanation might be that people under acute stress behave more pro-socially towards their peers in order to seek their comfort and support (‘tend and befriend hypothesis’).
When thinking about stress, most of us have its negative effects in mind like unpleasant emotions and health problems. Who would have thought that acute stress can serve the moral good?
Interesting studies.
I wonder what the long term effects of stress might be. There is an immediate need to reduce the arousal caused by stress (as described by Piliavin et al’s “arousal:cost-reward model”) which participants may intuitively believe can be done by acting pro-socially and inducing feelings of belonging, warmth, perceived approval etc. But if stress becomes chronic (e.g. money or employment worries), or short episodes are repeated regularly (e.g. people that perform in public), this pro-social effect may disappear or be reversed. For example, as people begin to feel more isolated or less supported as a result of being stressed, their ability to enjoy a sense of belonging may decrease, and so they may not expect their arousal levels to be reduced by pro-social behaviour. (Other studies show that social isolation decreases pro-social behaviour.)
It also seems a bit contradicatory to make people stressed (not very pro-social!) in order to make them act pro-socially. Therefore the practical application of these studies seems unclear at present.
“Who would have thought that acute stress can serve the moral good?”
I would seriously question that interpretation. In real life, people under acute stress are essentially at the mercy of other people’s sympathy, and will turn for support even to those they recognise as their enemies. This is a distorted state of affairs and will not necessarily lead to beneficial outcomes for anyone.
One interesting thing is that they did measure risk taking, which would be one first obvious objection: it could be the case that stressed individuals are only willing to share more because they tend to be more prone to risk taking.
Long-term and short-term stress have a sub-set of opposing effects and are mediated by different neurohormonal pathways (catecholamines vs. cortisol). Among other things, short-term tends to increase performance and have other beneficial effects, while long-term decreases performance. It might be the case that cooperation was increased through an increase in performance. Strong or mild stressors also have opposing effects, I would say the TSST stressor they used is really mild.
I would really doubt that giving high-voltage electric shocks for 10 1-minute sessions randomly spread throughout the day over 1 year, could make anyone more pro-social. What I would even more highly doubt is if children with more stressed childhoods would tend to be more pro-social. Several studies show that high stress in childhood leads to many problems in adult life, including increased criminality if I am not mistaken.
I believe this study only proves short-term mild stress, among many other already knew beneficial effects, increases prosocial behavior. That’s what you get for paying more attention, you make smarter and better choices.
As a newcomer in psychology, I hope I am allowed to ask a naïve question: What makes an experiment on 34 people only and under controlled conditions count as an evidence of how stress makes us behave pro-socially? Are not 34 people a too small sample for this broad conclusion? And is the “standardised psychological stressor” (I have no access to the linked article and could only read its small abstract) akin enough to “worldly” stress? (I do not intend to start a polemic, and would just like to understand better the epistemological background of such statements.)
Hi Elisa, Many thanks for your question and your interest in my blog piece! As for the sample size: 34 participants were half of the sample, the ones in the “stressed” condition only. About the same number of participants was in the control condition, i.e. they did the exact same tasks, but have not been stressed before. Comparing over 30 people with about the same number is a reasonable sample size, big enough to draw the conclusion – as the authors did – that acute (mild) stress can cause pro-social behaviour in a certain context (i.e., sharing money with another participant). Of course, we can’t infer the general claim “stress makes us more pro-social” from these data. One reason (besides others) is that stress is a very complex phenomenon – and a very broad term, both in the laboratory and the real world. I think there is no such a thing as “worldly” stress in the sense of a unified phenomenon. For example, being upset and worried for several months because of the illness of someone you love is, both physiologically and psychologically, very different from getting spooked by a friend with a scary face mask. Both would be “stress reactions”, though. (I can’t post links here, but in case you are interested you could look up “stress (biology)” in Wikipedia. There you can see how broad this term is.) It’s very important to distinguish, as others have said above, between acute and chronic stress and between mild and severe stress. These findings are about acute mild stress only.
Thanks for this blog post Nadira!
A related study examined whether people when stressed tend to make more deontological or utilitarian judgments in explicit moral dilemmas (Youssef et al., 2010). In personal dilemmas, such as whether you would push down a man from a bridge to save five people, subjects under stress are more likely to choose the deontological answer (don’t push him) than unstressed subjects.
As stress activates some of the same brain regions that are responsible for emotional processing, we might hypothesize that stress increases emotional gut reactions/intuitions. This would be in line with Joshua Greene’s dual process theory of moral judgement, which assumes that deontological judgements stem from more automatic and emotionally driven intuitions, whereas utilitarian judgments usually involve more reflective reasoning (at least in cases as the one mentioned above).
Maybe this also explains why people cooperate more when they a) are stressed and b) feel that others are present. We might speculate that people have an automatic intuition to cooperate conditional on the fact that others are watching them (evolved because of direct reciprocity or reputation building reasons). Stress activates this intuition automatically without giving people the chance to reconsider and change their mind.
This contradicts my theory of why the other study increased cooperation, as they used the same stressor. If my explanation for the study on the post were right, this same stressor would increase utilitarian thinking through an increase in cognition (whereas a stronger stressor would decrease utilitarian thinking and cooperation). I find hard to make sense of both datas. If Josh hopes utilitarian thinking is better suited to solve meta-cooperational problems, I would gather that he wouldn’t be very happy if a same stimulus both decreased utilitarian thinking and increased cooperation. One hopes that while solving meta-cooperation we shouldn’t mess up cooperation on the way. Although, on the other hand we would expect that an increase in meta-cooperation would decrease “parochial cooperation” to some extent.
But I am not sure people always default to cooperation under pression, cooperation is not adaptive in really unstable (and thus stressing) environments as there are few guarantees of reciprocity and punishment. Also, why would stress bypass the condition of being watched? Worth noticing, the stressor itself was being watched by others. Not sure how well disentangled the effects are.
Interesting points. I guess we would have to test your hypothesis experimentally.
I’m not sure whether Greene assumes that the same stimulus must lead to equal ethical decisions. In different situations a stimulus can activate different automatic reactions, which can be utilitarian or not (such as in counterintuitive cases of instrumentalization).
My point regarding the conditionality of cooperation was a bit ambiguous. To clarify: I wasn’t referring to the Trier Social Stress Test but to studies that seem to show that cooperation increases when people feel observed by others and want to fulfil their expectations (e.g. when they know that the other player in the Dictator Game expects a fair share http://papers.ssrn.com/sol3/papers.cfm?abstract_id=494422). I agree with you that cooperation isn’t necessarily increasing when people are stressed.
Comments are closed.