Ultimatum and dictator in four-year-olds

Lucas, M., Wagner, L., & Chow, C. (2008). Fair game: The intuitive economics of resource exchange in four-year olds. Journal of Social, Evolutionary, And, 2(3), 74–88. Retrieved from

This is an interesting paper in an area that needs more research, namely, how children perform in economic games. The typical questions go as follows: Are children altruistic? Are they selfish?

The authors review and criticize some of the relevant previous research. Murnighan & Saxon (1998) found that kindergartners made larger ultimatum offers and accepted smaller ultimatum offers of candy than did third or sixth graders. However, Lucas et. al argue that Murnighan & Saxon’s results are probably not valid since they use a kind of simulated game, in which the child is asked to imagine that another child offers such and such amount of candy (instead of actually playing the ultimatum game). They also quote Harbaugh, Krause, & Liday (2003), who found that seven-year olds made and accepted smaller ultimatum proposals than adults; and Hill and Sally (2006), who found that 6-year-olds already make offers as fair as those of adults after repeated rounds of play. Benenson, Pascoe and Radmore (2007), on the other hand, found that very young children can be altruistic when playing dictator (4-year-olds donated, on average, 25% of a stake of ten stickers to another classmate).

The authors also mention a problem with some of the previous research: often, researchers make children play with tokens that have no inherent value; children are explained that they will be able to trade the tokens for other stuff later. Lucas et al. argue that researchers should use items with tangible, real value for children, such as candy, stickers or toys, since the use of “symbolic” tokens poses additional cognitive demands on children and might affect experimental results.

Lucas et al. found, in their own study, that children made, on average, offers of 4.7 stickers in the ultimatum game and 3.99 stickers in the dictator game. That is, children seem to be making quite fair offers.

It is well established that adults give an average of 40% of the money at stake in ultimatum. But there is a big difference between the 4-year-olds’ 47% and the adults’ 40%. In adults, the mode (i.e., the most frequent answer) is 50%, while some adults give 40%, 30% or less, and almost no one offers more than 50%. However, some of Lucas el al.’s 4-year-olds offer more than 50%.  They call this phenomenon hyperfair offers. There is a qualitative difference between adults and children: adults oscillate between the fair (“half and half”) and the strategic (“less than half”). Many children, by way of contrast, offer more than is fair, more than half the stake.

“(…) the percentage of hyperfair offers (…) increased from 18% in the dictator game to 33% in the ultimatum game. Adults, in contrast, almost never make hyperfair offers (only 3.5% of offers in Lucas, Et al., (2007) were hyperfair).”

Lucas et al. also report that, in their study, children did not seem to take into account the behavior of the other player in their responses, even if it was unfair. In the dictator game, children did not change the amount of their offer in response to receiving either a low or fair offer from the friend. Children’s offers for the second game were also not affected by whether the friend had accepted or rejected the child’s first offer.

The results of the dictator game suggest that children in the sample are quite altruistic: “The adult average offer of 20% of the stake in the dictator game (Camerer, 2003) is usually interpreted as evidence that individuals have preferences for altruism, since proposers could offer less in a dictator game without fear of rejection. With an average offer in the dictator game of 40% of the stake, our sample of children made more altruistic offers than adults.”

Children’s average offer of 4 stickers (or 40% of the stickers at stake) doubles adults’ typical offer of 20% of the money at stake in dictator, and does not seem to far away from the 47% children offered in the ultimatum game. How do Lucas et al. explain these data? “(Children’s) ability to perform a cost/benefit analysis was limited. They did not seem to appreciate the degree to which they could “shade” their offers without penalty.” Thus, Lucas et al. are assuming that children have the desire or goal of keeping as many stickers as possible but they that their strategic thinking is deficient. “Children were more generous than they needed to be and were limited in their ability to act strategically in bargaining games in order to maximize their own benefits while avoiding the costs of rejection.”

Lucas et al.’s conclusion: “children are quite altruistic”, “they may have an innate sense of fairness and altruism.” This result is, in my opinion, over-simplistic. Previous research in economic psychology has established that adults are altruistic and fair to some degree, but also a little bit selfish and strategic. Lucas et al. assume, therefore, that either children are born altruistic, fair and strategic or they learn these behaviors along the way. If research shows that children are completely selfish, then altruism is learned. If it shows they are altruistic from the start, then it must be innate.

I quote them: “We predicted that children would perform similarly to adults in showing preferences for fairness and altruism. Alternatively, and as some others have found, children could be less fair, indicating that fairness must be learned over the course of development.”

They start with a binary opposition between selfishness and altruism. In this approach, learning is seen as lineal, cumulative, unidirectional. They start with some innate concepts and, while learning, children simply absorb information from their milieu or copy adult models, until they reach the end-point.

There are other possibilities, however. For example: young children might be neither selfish nor altruistic. They might be following other types of reciprocity not related to the fair, contract-like, 50/50 reciprocity of adults. They might use an associative reciprocity of the kind “I give a lot of stickers to the other kid because I like to make friends” (see Faigenbaum, 2005), that are typical of children’s peer cultures and of exchanges within the family. Such an approach dispels the apparent inconsistencies of previous research: it’s not just that children are not yet able to think “strategically”. They are not even interested in this kind of reasoning.

Associative reciprocity might explain why kindergartners make large ultimatum offers and accepted small ultimatum offers of candy (Murnighan & Saxon, 1998) or why 4-year-olds can be so altruistic when playing dictator (Benenson, Pascoe and Radmore, 2007). It might also explain the results of Lucas et al.’s own research, for example, why children give 40% of the stake in the dictator game.

Some references

Faigenbaum, G. (2005). Children’s Economic Experience: Exchange. Buenos Aires: LibrosEnRed.

Harbaugh, W. T., Krause, K. S., & Liday, S. J. (2003). Bargaining by Children. Social Science Research Network, 1–40. https://doi.org/10.2139/ssrn.436504

Lucas, M., Wagner, L., & Chow, C. (2008). Fair game: The intuitive economics of resource exchange in four-year olds. Journal of Social, Evolutionary, And, 2(3), 74–88. Retrieved from

Murnighan, J. K., & Saxon, M. S. (1998). Ultimatum bargaining by children and adults. Journal of Economic Psychology, 19(4), 415–445. https://doi.org/10.1016/S0167-4870(98)00017-8


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s