Thursday, 18 August 2011

Empathy breeds altruism, unless a person feels they have low status. A brain-scan study with a lesson for riot-hit England

In a defining image of the recent English riots, a man helped an injured youngster to his feet while an accomplice stole from the same victim's bag. This sheer lack of empathy on the part of the perpetrators has shaken observers to their core. How could humans display such a lack of altruism toward their fellow man?

A possible clue comes from a new brain imaging study that has examined links between the neural correlates of empathy, an act of altruism, and participants' subjective sense of their social status. Among people who feel they have low status, the study finds, increased neural markers of empathy are actually related to reduced altruism. The researchers surmised this is because any feelings of empathy are quashed by a grudging sense of low status. This could be a kind of defence mechanism whereby self-interest dominates over empathy for others. A possible lesson is that by reversing people's feelings of low status, through educational opportunities and other interventions, we all gain, by reinstating the usual link between empathy and altruism.

Yina Ma and her team at Peking University scanned the brains of 33 student participants while they watched numerous video clips of people being pricked painfully in the face or hand by a needle, or touched on those same parts by a cotton bud (referred to as a Q-tip in the US). Extra activity in the brain, in response to the needle clips versus cotton bud clips, was taken to be a neural marker for empathy (seeing someone else in pain is known to trigger activity in the pain matrix of one's own brain).

The participants also rated their own empathy levels and their subjective sense of their socio-economic status. They were shown a ladder with ten rungs, with the top rung representing people with the best jobs and education and most money; participants then indicated which rung they saw themselves as occupying. Although the participants were students at the same university they varied in their subjective sense of status. Finally, the participants were left alone in a room with an anonymous donation box, labelled as raising money to help impoverished patients with cataracts.

Among patients who considered themselves privileged in terms of socio-economic status, there was a positive relationship between empathy and altruism. The more neural signs of empathy they displayed in the scanner (based on extra activity in the left somatosensory cortex when viewing needle clips), the more empathy they said they had, and the more money they chose to donate to charity. By contrast, among participants who considered themselves lower in socio-economic status, the opposite pattern was observed. The greater their empathy-related brain activity in the scanner (based on extra right somatosensory cortex and inferior frontal cortex activity in response to needle clips), the less empathy they said they had, and the less money they chose to donate to charity. The researchers said the empathy-related inferior frontal cortex activity observed in these participants could be a sign of inhibitory processes quashing the emotional impact of seeing another person in pain.

Note, there was no absolute difference in the amount of money donated by participants who self-identified as low or high socio-economic status. The finding is more subtle and suggests empathy has a differential effect on our altruistic behaviour depending on how we see our standing in the world.

"Our findings have significant implications to the social domain," the researchers said, "in that, besides improving objective socio-economic status, raising subjective socio-economic status via education may possibly manifold altruistic behaviours in human society."

The findings add to a complex literature that suggests lower socio-economic status is sometimes associated with more empathy and altruism, but sometimes associated with reduced empathy.

ResearchBlogging.orgMa, Y., Wang, C., and Han, S. (2011). Neural responses to perceived pain in others predict real-life monetary donations in different socioeconomic contexts. NeuroImage, 57 (3), 1273-1280 DOI: 10.1016/j.neuroimage.2011.05.003

This post was written by Christian Jarrett for the BPS Research Digest.


  1. May I suggest a further resource to learn more about empathy and compassion.
    The Center for Building a Culture of Empathy
    The Culture of Empathy website is the largest internet portal for resources and information about the values of empathy and compassion. It contains articles, conferences, definitions, experts, history, interviews,  videos, science and much more about empathy and compassion.

    I posted a link to your article in our
    Empathy and Compassion Magazine
    The latest news about empathy and compassion from around the world

  2. This sheer lack of empathy on the part of the perpetrators has shaken observers to their core. How could humans display such a lack of altruism toward their fellow man?

    Curiously, the set of people asking this question about the rioters is almost completely disjoint from the set of people who used the same words in reference to Margaret Thatcher.

  3. By the very nature of this study we're going to see a lot of selection effects that could be problematic in determining the true link between empathy, status and altruism. Whilst the explanation offered for the link between low status and expressions of altruism makes sense, it's no more valid than arguing that the low altruism effects are due to specific aspects of status. For instance, it could well be that high altruism, high status participants express empathy as altruism because of better moral guidance or upbringing than low altruism, low status. Consider, for example, the positive correlation between facets of high status (marital satisfaction, maternal weight, maternal education) and maternal sensitivity (Developmental Psychology, Vol 33(5)). Extreme lows of maternal sensitivity can cause affectionless psychopathy etc, so this too is an explanation. Furthermore, differences in the educational experiences of the participant could factor in. This is extremely likely, as it has been found multiple times that low recidivism in criminals is linked to correctional education (Steurer et al, 2003). I fear that we're going down that rather familiar dodgy path where psychology often finds itself, wherein we correlate two facets of behaviour and then guesstimate the cause with our favourite pet theory, ignoring that other theories are equally explanatory.

    Either way, this study is certainly interesting. I just hope that our government doesn't pick this study up in the face of the riots and run wild with it before proper follow up research has been completed to show how best to achieve social altruism, but knowing the UK government I bet that's precisely what it'll do if it hears of this; governments tend to have an annoying tendency to pay selective attention to science when they think it can explain away something that's otherwise troublesome for their re-election campaign.

  4. Hi Ben - thanks for your interesting comments. In defence of the paper, a couple of points sprang to mind. For example, you mentioned the confounding influence of education, but in this study all participants were undergrads at the same uni, so this was effectively controlled for. That leads to the second point, which is that this study is novel in its focus on *subjective" SES. This is just an initial study, but it's possible that feeling low status, regardless of the other factors you mentioned (e.g education, upbringing) is enough to reverse the usual positive association between empathy and altruism. And the tantalising implication, therefore, is that raising people's subjective sense of their status, we help maintain the positive relationship between empathy and altruism.

  5. There are two really great program that is running out of San Francisco (Build and Summer search) that take high school age children from low Socio/Eco regions and offers them the opportunity to have real experiences with other children their age who are from severely lower socio/eco regions. The result is an elevation in their own perceived status. with this, the results seem to demonstrate, on face value, that the children respond by feeling less sorry for themselves and are quicker to offer empathy towards others. I'm taking this based on what the children do, start programs, fundraise etc, for the areas and people that they had experience with. I thought the direction of your article, in one dimension, showed a parallel to this.

    Regarding the potential for confounds, such a difficult issue, it is so as you say tantalizing to see the potential direction of the results and then to be forced to not overgeneralize. I certainly agree that as an initial study it does give you something to ponder further.

    Jim MOss

  6. Hi Christian!

    It's true that they're at the same university, but this doesn't represent their general educational background since participants from lower status backgrounds are probably likelier to have attended 'rough' primary and secondary schools too. I doubt many psychologists would argue that development begins at the age of 18-ish when one starts uni!

    It's certainly an interesting link, there just seems to be far too little control of variables to draw the conclusion properly; we don't know precisely what facet of social class produces this outcome and thus it'd be unwise for any scientifically minded person to say X is linked to Y when A-Z were also involved and left uncontrolled with their relations to Y.

    As ever, I criticise not the blog, nor your bringing this to our attention (because I love both the blog and the questions raised by the studies you find), but the method of the study itself.

  7. Hi Ben - I really appreciate your comments on the studies, please keep them coming. The point about the participants all being at the same uni is just that most studies in psychology and sociology measure educational background by years spent in education or education level achieved, so by those standard measures, the participants in this study would obviously count as having the same educational background. I take your point that they could have very different schooling. But unless a study is particularly focused on education, it's unlikely to delve into that kind of detail.

    You're right of course that there are all sorts of possible confounding variables that we don't know about here. Subjective social status appears to be mediating the link between empathy-related brain activity and altruism in an important and interesting way, but perhaps there is some other factor or factors correlated with subjective social status that are the genuine causal/mediating factors at play. The fact that all the participants are at the same uni does help reduce this possibility a little, but you're right that it's a far from water-tight set-up. In fact, because there are so many possible confounds that could be influencing subjective social status and the dependent variables of interest, a useful next step might be to find a way to directly affect participants' subjective social status and see what effect that has on the empathy-altruism link. If lifting people's sense of status were to restore the usual positive link between empathy and altruism, this would provide us with more robust evidence that subjective social status is playing a genuinely causal role in the processes under investigation.