Playing fair: Good Manners or Compassion

Have you heard of the dictator game? If you haven’t, you play it like this: There are two players, player one has 10 euros (or dollars) and the second player has nothing. Player one has to make an offer on how to divide the money, and they can do anything from keeping all the money for themselves to giving the other person everything. Player two can only decide whether they will take the offer or refuse it, but when they refuse neither of the players will get anything. In conclusion, this does not make for a good parlor game.

So it is not a game that you take out when you are at a party, but psychological researchers have used it to study prosocial behavior (e.g. any behavior that people do to help others). In the past, economists proposed that people make choices based on pure self-interest. This would mean that people in the role of player one in the dictator game should make the lowest offer possible that is not nothing, as this allows them to have the largest amount of money possible for themselves. Subsequently, player twos should take any offer that is not nothing, as a little money is better than no money.

Of course, this is not how reality works (and don’t worry, economists also don’t think all people are heartless and cold calculators anymore). We have a sense of justice and fairness, so player twos refused offers in the dictator game that were not high enough to be called fair. But, on the other hand, player ones usually made offers that were more fair than expected (about 28% of the original amount). Even though in most cases people would not know who the other player was at the end of the game, thus these results can’t be explained by the fact that people want to increase their social relations with others. Seemingly, people do make choices that are selfless.

Situational factors do play a role in this decision, for example whether you play with real money, real people, or funnily enough, whether there is a mirror in the room in which you can see yourself.

In addition, there are of course big individual differences in how much people offer the second player. There are people that keep everything to themselves, but there are also people that give out an amount larger than half the amount. The latter may seem surprising, but think about having to share a cookie or a piece of cake with someone. You will not get two perfectly even halves, so you will have to decide whether to give away the largest or the smallest part. I would feel really uncomfortable keeping the larger half. The comparison is not perfect, but I do not normally decide what happens to money distributions in my daily life.

It has often been assumed that the amount of money people offer in the dictator game results from people’s tendency to be altruïstic. Altruïsm in psychological terms is any behavior that may cost you something, but helps another, done out of a feeling of empathy or compassion for the other. In other words, it is not real altruism if someone performs a selfless act with an ulterior motive in mind.

But just to be a little cynical, if you think about it, it is questionable that people feel any concern for the welfare of the other player in the dictator game. In most cases, the dictator game is done in a hypothetical manner, asking participants to imagine what they would do if they were making distributions in the game. It is hard to feel any empathy for someone that does not even exist. But still, with real people there is the problem that concern usually is felt when some harm is done to another. In the dictator game people can only gain some extra amount of money, they cannot lose any, and thus no harm is done. Nor does player one know whether this other player really needs the two extra bucks.

So what really causes these individual differences between money distributions? Zhoa, Ferguson, and Smillie (2017) propose that it is not compassion, but your tendency to be compliant to social norms that predicts your offer as player one. The title of their research kind of spoilers the outcome: “Individual differences in good manners rather than compassion predict fair allocations of wealth in the dictator game”.

As is true for any headline, you can’t make conclusions based only on the title. They did four studies, each building on the other. To distinguish between the two variables, compassion and good manners, they used two separate models.

The first model they used is the big five, which has come up in past blog posts and will definitely come up again in the future. In particular, they look at the agreeableness factor. As one might expect, being highly agreeable is related to making fairer distributions. However, agreeableness is a broad term that holds both facets reflecting good manners and compassion, thus the whole factor in itself is not useable in making a distinction. You can take it apart though in two smaller factors: Compassion and politeness. Where compassion reflects the tendency people have to feel empathic concern for others, politeness reflects the tendency to follow social rules.  

The second model is the HEXACO model, which is very similar to the big five, but it includes six personality variables. In the HEXACO model pro-social behavior is reflected into two variables instead of one: The agreeableness factor (to make things extra confusing), which reflects the tendency to be tolerant and forgiving towards others (e.g. reactive cooperation), and the Honesty-Humility factor which reflects the tendency to be fair even if you have the opportunity to exploit others (e.g. active cooperation). The latter is more reflective of adherence to social norms than the HEXACO agreeableness.

How these personality variables influenced allocations in the dictator game was studied using four samples. In the first and second study participants did a hypothetical version of the dictator game, where in the first study psychology students were used, and in the second study a sample from site (an amazon based site on which you can do small tasks for money). The second sample is thus more representative of a bigger group of people compared to the first one. The third study also used participants from MTurk, but the sample was bigger and they played a simpler version of the dictator game, they could not freely make their offers but could only choose between four options. Finally, in the last study they wanted to look at what happened if people would play with real money and people (they were told that the other player would not learn who they were at the end of the study though).

On a side note. To me it was really interesting that when real money was involved people would keep more money. In the last study people only offered 2.59 dollar on average, while in the other three studies an average amount of 4.03 to 4.57 dollar was offered to the other player.

But still in all studies, and especially the fourth one, Politeness and Honesty-Humility would be predictive of the amount offered to the other player, while Compassion and agreeableness would not be.

This means that the dictator game is not a good measure for pure altruïsm and that research using it as such, should be revisited.

What this study does not suggest however, is that compassion does not play a role in prosocial behavior at all. For one, politeness does not rule out compassion, in fact there is a large overlap between the two. For example, whom and how much you tip is dependent on social norms, but that does not mean there are no feelings of compassion when you are tipping. Imagine if you have been hogging one large table for the entire evening with your very loud, large group of friends. You know you have annoyed the waiter, so a little empathy is in order. This can partly motivate your tipping behavior.  

As a single study, it still needs a friend to build upon and confirm the results, but that doesn’t mean we can’t make some practical conclusions. For me it emphasizes that not all helping behavior would be increased by more compassion. I think it is commonly assumed that if people would just learn to feel some empathy for other groups, a lot of problems could be solved. However, this study suggests that there are situations in which more compassion would not make people more prosocial, but that actually social norms would make a difference. One of those situations is probably whether you play fair in whatever other parlor game you decide to play, instead of the dictator game.


If you want to read the full article, it can be found on google scholar using this reference:

Zhao, K., Ferguson, E., & Smillie, L. D. (2017). Individual differences in good manners rather than compassion predict fair allocations of wealth in the dictator game. Journal of personality, 85(2), 244-256

Or this link.

If you (or your university) don’t have access to this article, but you still really want to read it you can always e-mail the authors to request a copy!


Views All Time
Views All Time
Views Today
Views Today