In a hub mostly devoted to dismantling overclaimed behavioral-science findings, this article is about one that earned its place in the canon. The dictator game has been run thousands of times, in dozens of countries, across more than four decades. The meta-analytic base is large, the paradigm is simple, the mechanism candidates are well-specified, and the replication record is one of the cleanest in experimental economics. If you are an executive trying to figure out which behavioral findings to actually build strategy on, the dictator game is one of the small set of safe bets.
What makes it interesting is not just that it replicates. What makes it interesting is what it strips out compared to its more famous cousin, the ultimatum game. In the ultimatum game, the responder can punish an unfair proposer by rejecting the offer, which means proposers have a strategic reason to be generous: getting nothing is worse than getting a smaller share. Disentangling “I am being fair because I value fairness” from “I am being fair because I am afraid you will reject me” is impossible in the ultimatum game by design. The dictator game removes the rejection threat entirely. The recipient has no decision to make. The dictator can keep everything. And yet, across the entire meta-analytic record, dictators consistently give a meaningful fraction away — most often around a fifth to a third of the stakes — for reasons that cannot be reduced to strategic self-protection.
That is what makes the dictator game the cleanest experimental test of pure other-regarding behavior in the behavioral-economics toolkit. And that is why this article belongs in a hub mostly devoted to takedowns. Calibration is the point. After watching power posing collapse, ego depletion implode, money priming evaporate, and the bystander effect get rewritten, a reader is owed the parts that held up. The dictator game is one of them.
Forsythe 1994 — The Paradigm Crystallizes
The clean paradigm we now call the dictator game was specified in 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: 10.1006/game.1994.1021.
The Forsythe team was responding to a specific interpretive problem in the ultimatum-game literature. By the early 1990s, it was well established in WEIRD samples that ultimatum-game proposers offered substantially more than narrow self-interest predicted — typically 40 to 50 percent of the pie. The question was: why? One reading was that proposers had genuine preferences for fair allocations and were acting on those preferences. The competing reading was that proposers were strategic: they understood that responders would reject low offers, and they offered higher amounts not because they cared about fairness but because they wanted to avoid getting nothing. These two interpretations make identical predictions in the ultimatum game and cannot be separated by ultimatum data alone.
Forsythe and colleagues designed the dictator game to break the tie. They paired anonymous subjects, gave one subject (the dictator) a sum of money, and instructed the dictator to choose how much to give to the other subject (the recipient). The recipient had no decision. Whatever the dictator allocated, the recipient received. There was no rejection mechanism, no possibility of retaliation, no repeated interaction. The recipient was a passive party whose role in the game was to receive whatever fraction the dictator chose to send.
Under the standard game-theoretic prediction, dictators should send zero. The dictator’s payoff is strictly maximized by keeping the entire endowment. There is no strategic reason to send anything. There is no rejection to fear, no reputation to protect, no future interaction to influence. A rational, self-interested dictator keeps it all.
In the Forsythe sample, dictators sent on average around 23 percent of the stakes. The modal allocation was substantially below the modal ultimatum-game offer, confirming that part of the ultimatum-game generosity was indeed strategic. But the dictator-game allocations were not zero. Substantial fractions of dictators sent meaningful amounts to anonymous strangers who had no power to do anything about a stingy allocation. The paradigm had successfully isolated a non-strategic component of generosity, and that component was clearly different from zero.
The 1994 paper was foundational in two ways. First, it cleanly separated the strategic and non-strategic components of ultimatum-game generosity. Second, it gave the field a paradigm — the dictator game — that has been used in thousands of subsequent studies and has become one of the workhorse instruments of experimental economics.
Hoffman 1996 — Social Distance and Anonymity
If the Forsythe 1994 paper established the paradigm, the paper that mapped its boundary conditions and stress-tested the mechanism is Hoffman, E., McCabe, K., & Smith, V. L. (1996). “Social distance and other-regarding behavior in dictator games.” American Economic Review, 86(3), 653—660. Vernon Smith would later win the Nobel Prize for experimental economics in 2002. The Hoffman team’s contribution was a series of dictator-game variants that systematically varied the social distance between dictator and experimenter and showed how the giving pattern shifted.
The standard Forsythe protocol had a particular feature: the experimenter knew which dictator had made which allocation. Even though the recipient was anonymous, the experimenter could in principle judge a stingy dictator. Hoffman and colleagues asked: how much of the observed generosity comes from this experimenter-observability, and how much would survive in a setting where the dictator’s allocation was completely private?
They constructed a “double-blind” dictator-game protocol in which neither the recipient nor the experimenter could connect a specific dictator to a specific allocation. The procedure used unmarked envelopes and a randomized identifier system that made it computationally impossible for the experimenter to reconstruct who had given what. Under this fully anonymous condition, mean offers dropped substantially — but not to zero. A meaningful fraction of dictators still sent positive amounts even when no one would ever know who they were.
The Hoffman paper is important for two reasons. First, it documented that the observed generosity in standard dictator games is partly driven by social-image concerns — the desire not to look stingy in front of the experimenter, or more generally not to look stingy in front of anyone whose opinion the dictator can imagine. Removing that audience effect reduces giving substantially. Second, it documented that the giving pattern does not collapse to zero even under fully private conditions. There is a non-trivial baseline of pure, audience-independent other-regarding behavior, sitting underneath a larger pile of social-image-driven giving. Both components are real. Both contribute to the aggregate dictator-game pattern. And both have policy implications, but they are different implications: interventions that operate on audience visibility (public recognition, social proof) work on the social-image component; interventions that operate on the baseline (intrinsic appeals, framing the recipient as deserving) work on the other-regarding component.
The Hoffman 1996 work has been replicated extensively. Subsequent meta-analyses confirm that the audience-effect component is large and that the baseline component is non-zero. This is a cleaner mechanistic decomposition than most behavioral-economics findings ever achieve, and it is part of why the dictator-game literature has aged well: the field has known for two decades that the phenomenon has multiple components, and the components themselves have been measured.
Engel 2011 — The 616-Study Meta-Analysis
The single most important paper for the strategist trying to decide whether to bet on dictator-game findings is Engel, C. (2011). “Dictator games: A meta study.” Experimental Economics, 14(4), 583—610. DOI: 10.1007/s10683-011-9283-7.
Christoph Engel of the Max Planck Institute for Research on Collective Goods in Bonn assembled what is, by a wide margin, the most comprehensive synthesis of the dictator-game literature ever conducted. The meta-analysis covered 616 studies and 20,813 dictators, drawn from publications spanning roughly two decades of experimental work. The geographic spread was substantial, the demographic spread was substantial, the variation in experimental protocols was substantial, and the resulting dataset gave Engel statistical power to estimate effects that no individual study could pin down.
The headline finding is the central tendency. Across the entire meta-analytic record, dictators on average gave away approximately 28.35 percent of the endowment. Roughly 36 percent of dictators gave nothing — the rational-actor prediction. Roughly 17 percent gave exactly half — the equal-split norm. Roughly 5 percent gave more than half. The remaining dictators distributed across the rest of the range, with a clear secondary cluster around the 20-to-30 percent range. The modal pattern is not equal splitting and is not zero; it is a moderate positive allocation, with substantial heterogeneity around that mean.
The meta-analytic mean of 28 percent is, in its own right, important. It places a quantitative anchor on what “the dictator-game effect” actually looks like when you average across two decades of studies. Behavioral-economics writers sometimes describe dictator-game giving as “substantial generosity” or “near-equal splits”; both descriptions are wrong. Dictators give meaningful amounts but, on average, retain about 70 percent of the endowment. The pattern is consistent with weak-to-moderate other-regarding preferences, not strong fairness norms.
The meta-analysis also documented systematic moderators. Stake size has a measurable but modest effect: larger stakes reduce the proportion given, but the absolute amount given still rises with stakes (giving 20 percent of $100 is more than giving 30 percent of $10). Anonymity has a substantial effect, in the Hoffman direction: more anonymous protocols produce less giving. The identity of the recipient matters: dictators give more when the recipient is identified as deserving (e.g., a charity, a needy individual) and less when the recipient is presented as a peer. Dictator demographics matter: women give somewhat more than men on average, older subjects give somewhat more than younger subjects, and there are reliable differences across student fields of study (economics students give somewhat less, consistent with a long-standing literature finding).
For the strategist’s purposes, the meta-analytic structure is what matters. Engel 2011 is not a single dramatic experiment that might or might not replicate. It is a synthesis of 616 separate dictator-game studies with consistent quantitative parameters, well-specified moderators, and a sample size large enough that even modest sub-effects can be estimated with confidence. This is the kind of evidence base that survives publication-bias correction, that does not depend on any one lab’s procedures, and that gives the field a calibrated quantitative anchor rather than a vague verbal claim. Most behavioral-economics findings do not have a meta-analysis of this scope. The dictator game does, and the meta-analysis supports the basic pattern.
Cross-Cultural Variation — Henrich 2001 and the WEIRD Question
The 15-society cross-cultural project led by Henrich and colleagues, which is the central topic of the ultimatum game anti-example article in this hub, also included dictator-game work in several societies. Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., & McElreath, R. (2001). “In search of Homo economicus: Behavioral experiments in 15 small-scale societies.” American Economic Review, 91(2), 73—78. DOI: 10.1257/aer.91.2.73, and the larger collaborative work that followed it, documented dictator-game patterns across small-scale societies that had never previously appeared in the behavioral-economics literature.
The cross-cultural pattern in dictator games is similar in structure to the cross-cultural pattern in ultimatum games but smaller in range. Dictator-game means across the studied societies clustered between roughly 20 and 40 percent of the stakes — narrower variation than the 26-to-58 percent range observed in ultimatum-game offers. This difference is itself informative. In the ultimatum game, the responder’s anticipated rejection threshold drives much of the cross-cultural variation in proposer offers: in societies where receiving an unsolicited gift creates an obligation (like the Au and Gnau of Papua New Guinea), responders sometimes reject hyper-fair offers, which suppresses ultimatum-game offer rates differently across cultures. Removing the rejection threat in the dictator game removes much of this cultural-variance amplifier and leaves a tighter cross-cultural distribution.
The qualitative pattern that survives across cultures is clear. In essentially every society studied, dictator-game means were substantially above zero. The strict rational-actor prediction failed everywhere — not by the same magnitude, but in the same direction. Other-regarding behavior in the dictator game is not a WEIRD artifact. It is a broadly human pattern, with cultural variation in magnitude but cross-cultural agreement on existence.
This is an important difference from the ultimatum game case. The ultimatum-game anti-example is partly a lesson in how WEIRD samples can mislead about magnitudes and even directions of behavioral patterns. The dictator-game anti-example is a lesson in how, on a more pared-down paradigm with the strategic component stripped out, the underlying phenomenon is more cross-culturally stable. Both findings are real; both have honest scope conditions. The dictator-game scope is wider than the ultimatum-game scope, which is part of why it earns a higher confidence rating for strategic use.
List 2007 — The Boundary Conditions
The single most important paper for understanding what the dictator-game pattern is not measuring is List, J. A. (2007). “On the interpretation of giving in dictator games.” Journal of Political Economy, 115(3), 482—493. DOI: 10.1086/519249.
John List is a Chicago experimental economist whose research program has, more than anyone else’s, stress-tested the external validity of laboratory experimental economics. The 2007 paper is a careful demonstration that the standard dictator-game result is sensitive to features of the experimental protocol that are easy to overlook, and that some of the apparent “generosity” measured by standard protocols disappears when those features are varied.
List ran several modified dictator-game protocols. The most-cited variant gave the dictator not only the option to send money to the recipient but also the option to take money from the recipient. In the take-allowed condition, mean dictator-allocated amounts dropped substantially — many dictators who would have given a positive amount in the standard send-only condition gave zero or negative amounts in the take-allowed condition. The List interpretation is that the standard dictator game presents a constrained choice set in which the dictator can only be neutral or generous; introducing the option to be selfish (by taking) reveals that some of the standard-protocol giving was reluctant compliance with an implicit experimenter expectation rather than active preference for the recipient’s welfare.
A second variant introduced an “earned endowment” — dictators earned the money through real effort before the allocation decision. In the earned-endowment condition, mean dictator-allocated amounts also dropped: when the money felt like the dictator’s own, the impulse to share it decreased. This finding has been replicated in subsequent work and is part of the basis for the broader literature on entitlement effects in economic decisions.
List 2007 is sometimes cited, incorrectly, as a refutation of the dictator-game finding. It is not. List himself does not claim that. What List established is that the standard dictator-game protocol overstates the magnitude of pure other-regarding preferences because it presents an artificial choice set and because the standard windfall endowment removes natural entitlement concerns. The honest reading is: the dictator-game effect exists, the meta-analytic estimate of approximately 28 percent of stakes is a real number, and that number is conditional on a particular protocol whose features (no take option, windfall endowment, experimenter visibility) inflate the effect relative to what would obtain in a fully natural decision setting. The corrected estimate of “real-world altruistic giving toward anonymous strangers” is smaller than the standard dictator-game pattern, but it is still positive.
This is what robust behavioral-economics findings look like. They have a paradigm, the paradigm produces a quantitative pattern, the pattern replicates across many settings, the moderators are identified, the boundary conditions are mapped, and the findings can be used responsibly by anyone willing to read the qualifications. This is the opposite of the power-posing pattern (one dramatic paper, weak replications, no mechanism, no boundary mapping) and the opposite of the ego-depletion pattern (decades of supportive citations, then a massive multi-lab replication that found no effect at all). Findings like the dictator game are why behavioral economics, on net, has produced real knowledge — they are just not as numerous as the popular literature implies.
Real-World Applications — Charity, Workplace, Defaults
The dictator-game pattern has direct implications for several real-world strategic domains. The mechanism evidence is strong enough that these implications are reasonable to act on, with calibration.
Charitable giving. The dictator-game finding that recipient identity affects giving — dictators give more to identified deserving recipients than to anonymous peers — corresponds to a robust real-world finding that charities raise more money when the recipient is identified (the “identifiable victim effect,” documented across many field-experimental studies). The dictator-game finding that anonymity reduces giving corresponds to the real-world finding that public donation lists, social-recognition naming conventions, and visible donor walls increase fundraising. The dictator-game finding that women give somewhat more than men on average corresponds to a robust real-world fundraising pattern. None of these field patterns are explained solely by the dictator-game evidence, but the laboratory evidence and the field evidence converge on the same direction, which is unusual in behavioral economics and reassuring when it occurs.
Workplace fairness norms. Compensation decisions in organizational settings have a partial dictator-game structure: senior leaders allocate bonus pools, executives allocate equity, managers allocate raises. The dictator-game evidence suggests that even without enforcement mechanisms, decision-makers will allocate non-trivial fractions to subordinates rather than maximizing their own share. The same evidence also predicts substantial individual heterogeneity — some dictators give zero, some give meaningful amounts — which corresponds to substantial real-world variation in how generous different managers and executives are with discretionary allocations. The audience-effect literature predicts that publicly observable compensation decisions will be more generous than privately observable ones, which is consistent with the much-documented effect of public disclosure on executive compensation (where the direction is sometimes the opposite — exposure to public scrutiny pulls compensation toward conformity with peer norms, which can mean either up or down).
Default-architecture interactions. The default-effect literature and the dictator-game literature interact in interesting ways. Choice architects who design default contribution rates for charitable giving (e.g., default round-up programs, opt-out tipping defaults) are operating in a domain where both effects apply: the default-effect literature predicts the default will be sticky, and the dictator-game literature predicts that the chosen default amount will be processed through the same other-regarding cognitive system that produces dictator-game giving. Empirically, default-tipping experiments show patterns consistent with both effects operating simultaneously. Strategists designing such systems can reasonably use the dictator-game mean (around 28 percent of stakes) as a rough anchor for what fraction people will accept as a default in a one-shot allocation context, with the understanding that the actual optimal default depends on the specific protocol and the implicit normative framing.
These applications are the kind of strategic decision-usefulness that justifies including the dictator game in the working toolkit. The findings replicate. The mechanism is identified. The boundary conditions are mapped. The applications follow predictably from the laboratory evidence. This is what good behavioral economics looks like, and the contrast with the much larger pile of fragile findings is exactly why this hub spends so much effort distinguishing the two categories.
What’s Honest To Say About Dictator-Game Findings
The honest summary, after working through the evidence, is this. The dictator-game pattern is robust. The 28 percent mean giving estimated by Engel 2011 across 616 studies and 20,813 dictators is a real number. The cross-cultural pattern, while showing meaningful variation in magnitude, supports the same qualitative finding across the small-scale societies studied. The mechanism candidates (other-regarding preferences, warm-glow giving, social-image concerns) are identified, and the audience-vs-baseline decomposition from Hoffman 1996 has been confirmed in subsequent work. The List 2007 boundary conditions are real and important — the standard protocol overstates pure altruism — but they refine the interpretation rather than refuting it.
What is wrong to say: that dictator-game findings prove humans are intrinsically generous, that the 28 percent mean is a fundamental constant of human nature, that the laboratory pattern translates one-for-one to real-world giving behavior, or that the cross-cultural pattern is identical across societies. None of these stronger claims survive the careful reading. The weaker claim — that other-regarding behavior exists, is quantitatively meaningful, is mechanistically multiply-determined, and is moderated by identifiable features of the situation — is well-supported and is enough to ground real strategic decisions.
This is the pattern of a mature behavioral-science finding. The dictator game has earned that maturity over four decades of cumulative work, and the strategist who wants to build on behavioral-economics evidence should weight findings of this quality much higher than findings of the power-posing or ego-depletion variety. The dictator-game evidence is not perfect, but it is substantially better than the field’s average, and it deserves the trust that the field’s average does not.
Strategist Takeaway — How To Use This Finding
For executives evaluating whether to incorporate dictator-game-derived insights into product, pricing, or organizational decisions:
Treat the 28 percent meta-analytic mean as an anchor, not a constant. The actual fraction depends heavily on the specific protocol, the perceived deservingness of the recipient, the visibility of the decision, and the source of the endowment. Use the meta-analytic mean as a starting hypothesis for what fraction people will allocate to others in a one-shot allocation context, and run local validation to calibrate the actual number for your domain.
Apply the audience-effect lever deliberately. When you want to increase prosocial behavior (charitable matching, peer donations, organ-donor sign-up at the DMV), make the decision visible — to colleagues, to a public list, to a social-recognition framing. When you want a clean measure of intrinsic preference (e.g., for surveying customer values), reduce the audience visibility as much as possible. The Hoffman 1996 evidence quantifies this effect clearly.
Apply the recipient-identification lever in fundraising and prosocial design. Identified-recipient framing reliably increases giving. Generic-peer framing reliably reduces it. This is a low-cost intervention with consistent evidence across both laboratory and field.
Respect the cross-cultural caveats but do not overuse them. Unlike the ultimatum-game case, the dictator-game pattern is more stable across cultures. You can use dictator-game-informed strategies in markets outside the WEIRD samples with somewhat more confidence than you can use ultimatum-game-informed strategies. Confirm the magnitude locally; do not assume the existence question needs to be re-asked.
Read the List 2007 caveats before extrapolating to natural settings. Standard dictator-game protocols overstate real-world altruistic giving because they remove the option to take, present windfall endowments, and create an implicit experimenter audience. When you build a real-world system that should be informed by dictator-game evidence, expect the real-world allocation pattern to be somewhat less generous than the laboratory mean. The direction is right; the magnitude needs adjustment.
Sources
- 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: 10.1006/game.1994.1021. ScienceDirect.
- Hoffman, E., McCabe, K., & Smith, V. L. (1996). Social distance and other-regarding behavior in dictator games. American Economic Review, 86(3), 653—660. JSTOR.
- Engel, C. (2011). Dictator games: A meta study. Experimental Economics, 14(4), 583—610. DOI: 10.1007/s10683-011-9283-7. SpringerLink.
- Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., & McElreath, R. (2001). In search of Homo economicus: Behavioral experiments in 15 small-scale societies. American Economic Review, 91(2), 73—78. DOI: 10.1257/aer.91.2.73. AEA.
- List, J. A. (2007). On the interpretation of giving in dictator games. Journal of Political Economy, 115(3), 482—493. DOI: 10.1086/519249. UChicago Press.
- Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2—3), 61—83. DOI: 10.1017/S0140525X0999152X.
- Camerer, C. F. (2003). Behavioral Game Theory: Experiments in Strategic Interaction. Princeton University Press. ISBN: 978-0691090399.
Related
- The Replication Crisis Hub --- full index of dismantled, contested, and surviving behavioral-science findings.
- The Ultimatum Game Across Cultures --- the companion anti-example demonstrating how WEIRD ultimatum-game evidence misled the field about cross-cultural generalizability, and how dictator-game evidence interacts with that critique.
- Prospect Theory --- another framework-level anti-example showing what robust behavioral economics looks like at scale.
- Defaults and Status Quo Bias --- the choice-architecture anti-example that interacts directly with dictator-game findings in default-allocation design.
- Axelrod’s Tit-for-Tat --- the cooperation-game framework whose mechanism overlaps with the strategic component of the ultimatum game and contrasts with the non-strategic dictator-game finding.
- Framing Effect --- the canonical demonstration that decision context matters; relevant to how dictator-game allocations shift under different recipient and endowment framings.
FAQ
Why is the dictator game considered a cleaner test of altruism than the ultimatum game?
The ultimatum game confounds two distinct motivations. A proposer who offers 40 percent of the stakes might do so because they value fairness, or because they fear rejection. There is no way to separate these in ultimatum data. The dictator game removes the rejection threat entirely. Whatever the dictator gives is given without strategic reason — the recipient cannot punish a stingy allocation. Any positive allocation in the dictator game is therefore evidence of non-strategic other-regarding behavior, which is conceptually distinct from the mixed motivations that drive ultimatum-game generosity.
What is the actual percentage that dictators give on average?
The Engel 2011 meta-analysis of 616 studies and 20,813 dictators puts the mean at approximately 28.35 percent of the stakes. That mean includes about 36 percent of dictators giving zero (the rational-actor prediction), about 17 percent giving exactly half, and the rest distributed across the range. The modal pattern is positive but moderate. Aggregate giving is substantially below 50 percent and substantially above zero.
Does dictator-game giving disappear under full anonymity?
It drops substantially but does not collapse to zero. Hoffman, McCabe, and Smith 1996 ran a double-blind protocol in which neither the experimenter nor the recipient could connect any specific dictator to any specific allocation. Mean giving dropped meaningfully under this condition, confirming that part of the standard pattern is driven by social-image concerns. But a positive baseline remained, demonstrating that some dictators give for reasons that do not depend on being observed at all.
Doesn’t List 2007 refute the dictator-game finding?
No. List 2007 demonstrates that the standard dictator-game protocol overstates the magnitude of pure altruism because it removes the option to take, presents windfall endowments, and creates an implicit experimenter audience. When these features are varied, the giving pattern reduces. But the giving pattern does not disappear, and List does not claim it does. The honest reading is that the dictator-game pattern is real, the standard-protocol mean is biased upward relative to fully natural settings, and the corrected estimate is smaller but still positive.
Does this finding apply outside WEIRD samples?
Yes, with caveats about magnitude. The Henrich 2001 cross-cultural project documented dictator-game patterns in several small-scale societies that had never previously appeared in the behavioral-economics literature. Means clustered between roughly 20 and 40 percent across the studied societies — narrower variation than the corresponding ultimatum-game range. The qualitative pattern (positive mean giving, well above zero) appeared in essentially every society studied. The cross-cultural robustness of the dictator-game finding is greater than the cross-cultural robustness of the ultimatum-game finding, because removing the rejection threat removes a key source of culturally-driven variation in proposer behavior.
How does this fit with the broader replication-crisis picture?
This article is an anti-example: a behavioral-economics finding that survives careful scrutiny. Most of the findings in this hub did not. The honest picture is that behavioral economics has produced a small number of robust, large, mechanism-grounded findings — defaults, prospect theory, dictator-game altruism, ultimatum-game generosity within scope conditions — and a much larger number of fragile findings that should be weighted accordingly. The dictator-game evidence is in the first category, and the strategist’s correct response is to take it seriously while remaining skeptical of the surrounding field average.
What is the single most actionable implication for an executive?
When designing systems that depend on people allocating something (money, time, recognition, opportunity) to others, expect a moderate baseline of positive allocation even without enforcement, expect that visibility amplifies that baseline substantially, and expect that identifying the recipient as a deserving individual amplifies it further. These three levers — baseline, visibility, recipient identification — are all supported by the dictator-game evidence and all have direct applications in fundraising design, organizational compensation, default-allocation architecture, and prosocial product features.