Dictator game
One player simply decides how much, if any, to give the other.
What it means
The dictator game is a simple experimental paradigm in which one participant, the 'dictator,' is given a sum of money and unilaterally decides how much, if any, to give to a second, passive participant who has no power to accept, reject, or respond in any way. Its design purpose is precise: by stripping away the recipient's ability to retaliate, it isolates pure other-regarding motivation — generosity, altruism, or fairness norms — from the strategic fear of rejection that contaminates the related ultimatum game. The robust finding is that dictators give nonzero amounts, commonly in the range of 20–30%, which is taken as evidence of genuine social preferences rather than purely self-interested behavior. A crucial nuance is that giving is highly sensitive to context and framing: anonymity from the experimenter, how the entitlement to the money was established, the social distance to the recipient, and subtle 'moral wiggle room' all sharply change allocations, suggesting that some apparent generosity reflects a desire to appear fair or to comply with norms rather than pure altruism. It matters as a foundational tool for measuring altruism and fairness, for testing theories of social preferences, and for understanding when and why people share resources absent any strategic incentive.
What the evidence shows
Engel's 2011 meta-study pooled 129 papers and more than 20,000 dictators, and the headline numbers are stable: roughly 63 percent transfer something, and the average gift is about 28 percent of the pie. But that average hides a split. Over a third of dictators keep the entire sum, and a distinct cluster divides the money exactly in half, so the modal choices are keep-all and split-evenly rather than anything in between. Giving climbs when the recipient looks deserving or identifiable and falls as the recipient becomes more abstract or the dictator more anonymous. The stylised fact 'people share' is therefore less a smooth tendency than a mixture of self-interested and fairness-minded types responding to the frame.
Altruism or artefact?
Whether these transfers reveal stable altruism is genuinely contested. List (2007) and Bardsley (2008) widened the menu so a dictator could also take money from the recipient; transfers collapsed, and much of the apparent generosity vanished, suggesting subjects were reading the narrow options an experimenter handed them rather than expressing a fixed preference. Cherry, Frykblom and Shogren (2002) found that when dictators earned the stake through a task and anonymity was doubly enforced, about 95 percent gave nothing at all. Dana, Weber and Kuang (2007), with related 'quiet exit' work, showed that many give only when the choice is unavoidable: offer plausible deniability and they quietly take the money. The generosity is real, but it is fragile and partly manufactured by the protocol.
What moves the number
Because the paradigm is bare, small design choices swing behaviour hard, and each is a thumb on the scale. Concealing the choice from the experimenter through double-blind procedures lowers giving, so being observed matters even when the recipient never learns who decided. Social distance works the same way: the more abstract and anonymous the recipient, the less arrives. Entitlement matters too, as money earned rather than granted is shared far less. Framing the identical choice as taking rather than giving suppresses transfers, while a named, identifiable recipient or a real charity raises them. The practical lesson for anyone deploying the game is that the protocol is never neutral, and numbers only compare across studies that hold these levers constant.
Using it as a measure
The dictator game is the workhorse for quantifying prosocial preference. Its data feed models like inequity aversion, let researchers compare children, cultures and professions on the same scale, and help calibrate charitable and development interventions. Its limits are equally important. External validity is debated: one-shot lab giving predicts real-world generosity only weakly, most samples are Western undergraduates, and hypothetical or trivial stakes tend to inflate transfers. Treat a dictator-game figure as a context-bound snapshot of norm-sensitive sharing under a specific set of rules, not a stable personality trait you can carry to the field. The tool has recently been turned on large language model agents as well, as a way to probe whether machine behaviour mimics human fairness norms.
Examples
Many 'dictators' hand over 20–30% of the money despite facing no consequences for keeping it all.
Dictators who earned their stake by working through a task first give noticeably less than those simply handed the money. Generosity depends on whose money it feels like it was.
Offer dictators the chance not to see what their choice costs the other person and many take it, then keep the lot — a preference for not knowing, rather than plain selfishness.
Anthropologists running the game across small-scale societies find allocations track local sharing norms, showing the dictator reads the situation for what fairness locally demands rather than acting on a fixed level of altruism.
Make the passive partner a named cause, a specific food bank rather than an anonymous participant, and dictators part with more, because a deserving, identifiable recipient is harder to walk past than a faceless stranger.
First described in Forsythe et al. (1994); Kahneman, Knetsch & Thaler (1986).
Key references
- Engel, C. (2011). Dictator games: a meta study. Experimental Economics, 14(4), 583-610. doi.org/10.1007/s10683-011-9283-7
- Bardsley, N. (2008). Dictator game giving: altruism or artefact? Experimental Economics, 11(2), 122-133. doi.org/10.1007/s10683-007-9172-2
- Dana, J., Weber, R. A., & Kuang, J. X. (2007). Exploiting moral wiggle room: experiments demonstrating an illusory preference for fairness. Economic Theory, 33(1), 67-80. doi.org/10.1007/s00199-006-0153-z
- List, J. A. (2007). On the interpretation of giving in dictator games. Journal of Political Economy, 115(3), 482-493. doi.org/10.1086/519249
- Cherry, T. L., Frykblom, P., & Shogren, J. F. (2002). Hardnose the dictator. American Economic Review, 92(4), 1218-1221. doi.org/10.1257/00028280260344740
- Forsythe, R., Horowitz, J. L., Savin, N. E., & Sefton, M. (1994). Fairness in simple bargaining experiments. Games and Economic Behavior, 6(3), 347-369. doi.org/10.1006/game.1994.1021