REVIEW 3 major objections 4 minor 2 references
Preferences for efficiency, rather than preferences for morality, drive cooperation in the one-shot Stag-Hunt Game
T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read In the one-shot Stag-Hunt Game, cooperation is driven by preferences for efficiency, not by moral preferences for doing the right thing.
desk verdict New TOG-to-Stag-Hunt result is worth a referee, but the efficiency conclusion rests mainly on the 'unfair' frame and needs a neutral condition and order controls. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The Trade-Off Game (TOG), a dictator-style allocation task with two between-subjects framing conditions: in the efficient frame, the efficient allocation is labeled "be generous" and the equitable allocation "be ungenerous"; in the equitable frame, the equitable allocation is labeled "be fair" and the efficient allocation "be unfair". The design makes the same efficient choice morally praised in one condition and morally condemned in the other, so a choice that persists despite negative framing reveals a strong preference for efficiency. The correlation of that choice with Stag-Hunt cooperation is the central measure that separates efficiency motives from moral motives.
What would settle it
A direct test would play the Trade-Off Game before the Stag-Hunt Game, or use a Trade-Off Game with neutral labels, and check whether the correlation between efficient choices and Stag-Hunt cooperation survives; if it disappears, the result is an artifact of the order or of the moral framing rather than evidence for a stable efficiency preference.
Extended reading notes
Core claim
The paper reports that, in a one-shot anonymous Stag-Hunt Game, participants who later chose the efficient allocation in the Trade-Off Game were significantly more likely to cooperate, while those who chose the allocation framed as morally appropriate were not. This correlation was driven by participants in the equitable frame, where the efficient allocation was labeled "unfair": people who chose efficiency despite the negative moral label cooperated at substantially higher rates. The authors conclude that efficiency preferences, rather than moral preferences, drive Stag-Hunt cooperation, in contrast to the Prisoner's Dilemma, where moral preferences have been shown to be primary.
Load-bearing premise
The conclusion rests on the assumption that a choice in the Trade-Off Game, made after the Stag-Hunt Game and without risk, reveals a stable, context-independent preference for efficiency that also motivates the risky cooperative choice, and that the framing labels identify what participants actually regard as morally right.
Editorial extensions
If this is right
- If the finding is correct, efforts to encourage cooperation in risky coordination settings should emphasize the collective gains from cooperation rather than moral duty or doing the right thing.
- The motive behind cooperation depends on the strategic structure of the game: costly cooperation (Prisoner's Dilemma) may be moral, while risky cooperation (Stag-Hunt) is efficient.
- Models of social preferences should separate efficiency from morality as distinct channels, because they operate differently depending on whether cooperation is costly or risky.
- In the Stag-Hunt, an efficiency preference strong enough to overcome negative moral framing is the best predictor of the cooperative choice, implying that moral suasion alone may not raise cooperation in such games.
- The same methodology could be extended to other coordination games to map where moral versus efficiency preferences dominate.
Reading between the lines
- The pattern suggests that people who choose the efficient allocation under the "unfair" label have an unusually strong efficiency preference, whereas in the efficient frame the efficient choice is a default that carries no diagnostic power; this asymmetry is an insight the authors offer but do not fully formalize.
- A natural extension would be to measure risk preferences directly and test whether the efficiency effect on Stag-Hunt cooperation is moderated by risk aversion, since the cooperative choice is risky.
- If efficiency preferences are the true driver, then providing information about total surplus or the payoff-dominant equilibrium should raise cooperation even without moral framing.
- Because the Trade-Off Game was played after the Stag-Hunt Game, a short testable extension is to reverse the order; if the correlation disappears, the measured preference may be partly constructed by the preceding game.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports a pre-registered online experiment (N=436) in which participants first play a one-shot Stag-Hunt Game (SHG) and then a Trade-Off Game (TOG) that pits an equitable allocation against an efficient one, with the two options labeled either 'generous/ungenerous' or 'fair/unfair.' A logistic regression predicting SHG cooperation from TOG choice and frame condition shows a positive coefficient on choosing the efficient allocation, which the authors interpret as evidence that efficiency preferences, rather than moral preferences, drive SHG cooperation. The authors conclude that this contrasts with prior findings for the Prisoner's Dilemma, where moral preferences appear to be the primary driver.
Significance. If the central claim were cleanly identified, this would be an important contribution: it would show that the moral-preference account of prosociality in the Prisoner's Dilemma does not automatically extend to games in which cooperation is risky rather than costly. The study is pre-registered, uses a large sample, and reports robust standard errors, which are genuine strengths. However, as analyzed below, the current evidence does not support the strong conclusion in the title and abstract, because the only significant effect appears in the frame where the efficient option is labeled 'be unfair,' leaving a confound between efficiency preference and willingness to violate a moral label.
major comments (3)
- [Section 3, Table 4] The positive coefficient on TOG Efficient Choice (b2 = 0.612, p < 0.05) is significant only in the equitable frame; the interaction term (b3 = -0.704) is not significant, and the test b2 + b3 = 0 shows no effect in the efficient frame. Thus, the data do not show that efficiency preference per se predicts SHG cooperation; they show that choosing an allocation labeled 'be unfair' predicts it. This result is equally compatible with a disposition to disregard moral labels, or with risk-seeking/norm-independence, and the post hoc explanation in Section 4—that these participants have 'stronger preferences for efficiency' because they overcome the moral framing—is unfalsifiable without a neutral frame. The central claim is therefore underdetermined by the reported analysis.
- [Section 2.1] The Trade-Off Game is always administered after the Stag-Hunt Game, with no order counterbalancing. If prior SHG choices or outcomes influence TOG choices, the observed correlation does not identify a stable, context-independent efficiency preference. The authors should either test for order effects or substantially temper the causal language in the conclusion.
- [Section 2.2 and footnote 4] The moral-preferences hypothesis is tested by assuming that the labels 'be generous' and 'be fair' are perceived as morally right, based on Capraro and Rand (2018). No manipulation check is reported for this sample. Without such a check, the null result for the positively framed option could reflect failed framing validation rather than the absence of moral preferences.
minor comments (4)
- [Table 4] The chi-square tests are reported at the bottom of the table but are not described in the text; please state what null hypotheses they test and why they are relevant to the pre-registered hypotheses.
- [Section 2.1, footnote 3] The note says the payoffs were divided by two to produce roughly 50% efficient choices in the SHG, but the unconditional SHG cooperation rate is not reported in the main text before the conditional rates in Table 3; consider adding it.
- [Abstract and Section 4] The phrase 'primarily driven by efficiency motives' is too strong relative to the statistical evidence, which rests on a single significant coefficient in one frame; the conclusion should be scaled to the strength of the finding.
- [Section 2.1] The between-subjects random assignment to the two TOG frames is not accompanied by a demographic balance check; reporting one would strengthen confidence in the framing comparison.
Circularity Check
No significant circularity: the Trade-Off Game choice is an independent behavioral measure, and the framing validation is external, falsifiable prior work.
full rationale
The paper does not derive SHG cooperation from the Trade-Off Game (TOG) by construction. The TOG efficient choice is a separate behavioral measure, elicited after the Stag-Hunt Game, and the regression analysis merely tests preregistered correlations rather than fitting parameters that are then relabeled as predictions. The central finding is a contingent empirical pattern: TOG efficient choice predicts SHG cooperation mainly in the equitable frame, while the interaction test shows no effect in the efficient frame. This pattern is not an identity or an equation that reduces to its own inputs. The moral-label premise is imported from Capraro & Rand (2018) and Tappin & Capraro (2018), which are self-citations by the first author, but they are externally falsifiable empirical claims rather than uniqueness theorems or unverified assertions. Moreover, the present paper replicates the framing effect (82.73% vs 37.50%), so the moral-label premise does not rest solely on self-citation. The skeptic's concern about norm-independence or risk-seeking as an alternative explanation of the equitable-frame result is a validity threat, not a circularity: it challenges the interpretation of the correlation, not the independence of the measured variables. No circular step can be exhibited from the paper's own equations or definitions, so the appropriate score is 0.
Assumptions & free parameters
assumptions (3)
- domain assumption The Trade-Off Game framing labels from Capraro and Rand (2018) identify which allocation is morally appropriate, and this transfers to the current MTurk sample.
- domain assumption The Stag-Hunt Game with R > T = P > S captures risky cooperation without temptation, so a player's B choice in Table 2 reflects willingness to cooperate under strategic risk rather than other motives.
- domain assumption Choices made in the Trade-Off Game after the Stag-Hunt Game are unaffected by the preceding Stag-Hunt interaction, or any such effect is ignorable for the comparison.
Cite this review
Pith. "Pith review of Preferences for efficiency, rather than preferences for morality, drive cooperation in the one-shot Stag-Hunt Game." pith.science (2026). https://pith.science/paper/X5ROPBBU
@misc{pith2026190803091,
author = {Pith},
title = {Pith review of: Preferences for efficiency, rather than preferences for morality, drive cooperation in the one-shot Stag-Hunt Game},
year = {2026},
howpublished = {\url{https://pith.science/paper/X5ROPBBU}},
note = {Machine review of arXiv:1908.03091}
}
read the original abstract
Recent work highlights that cooperation in the one-shot Prisoner's dilemma (PD) is primarily driven by moral preferences for doing the right thing, rather than social preferences for equity or efficiency. By contrast, little is known on what motivates cooperation in the Stag-Hunt Game (SHG). Cooperation in the SHG fundamentally differs from cooperation in the PD in that it is not costly, but risky: players have no temptation to deviate from the cooperative outcome, but cooperation only pays off if the other player cooperates. Here, we provide data from a large (N=436), pre-registered, experiment. Contrary to what has been observed for the PD, we find that SHG cooperation is primarily driven by preferences for efficiency, rather than preferences for doing the right thing.
Reference graph
Works this paper leans on
-
[1]
safe” option A (T = P). Cooperation (i.e., playing B) is “risky
Introduction There is wide consensus among scholars that people’s capacity to cooperate is what has made human societies extremely successful, compared to other animal societies (Trivers, 1971; Ostrom, 2000; Skyrms, 2004; Rand & Nowak, 2013). Psychologists even argue that the psychological basis of cooperation, shared intentionality, is what makes humans ...
work page 2018
-
[4]
Conclusion Recent work highlights that (one-shot, anonymous) PD cooperation is primarily driven by moral preferences for doing the right thing, rather than by distributional preferences for equity or efficiency (Capraro & Rand, 2018). The current paper adapts the research methodology in Capraro & Rand (2018) to investigate behavior in the SHG, where coope...
work page 2018
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.