Crunching the numbers. Responsibly.
wealth visibility facilitates the downstream consequences of initial inequality—in initially more unequal situations, wealth visibility leads to greater inequality than when wealth is invisible
randomized factorial comparisons across 80 sessions show the interaction and corresponding simple effects
Replication outlook: favorable
making wealth visible has adverse welfare consequences, yielding lower levels of overall cooperation, inter-connectedness, and wealth
randomized visibility conditions show convergent reductions in wealth, cooperation, and network degree
Replication outlook: favorable
High initial levels of economic inequality alone, however, have relatively few deleterious welfare effects
multiple welfare outcomes support limited effects, but parts of the conclusion rely on nonsignificant contrasts
Replication outlook: uncertain
in the no initial inequality condition, visibility does not affect inequality
the single condition comparison is nonsignificant without an equivalence test
Derived from the full evaluation — not a separate score.
Strengths
Initial inequality and local wealth visibility are independently manipulated while mean starting wealth is held constant. This directly identifies the visibility-by-inequality comparison central to the paper.
↳ Experimental design; Extended Data Table 1; Fig. 2
Visible-wealth conditions show lower average wealth, cooperation, and network degree, while Fig. 2 reports the inequality interaction. The outcomes form a coherent account of welfare effects within the experiment.
↳ Results, Figs. 2–3
The Discussion names untested structural and institutional conditions and explicitly states that laboratory results do not translate directly into the real world.
↳ Discussion, limitations and final paragraphs
Limitations
The Results and Extended Data report numerous coefficients, subgroup interactions, and outcome tests without an explicit multiple-comparison procedure. This primarily tempers confidence in secondary and heterogeneity findings.
↳ Results, Figs. 2–4; Extended Data Figs. 5–6
Fig. 4 links relative wealth and prior behavior to cooperation within treatment conditions, but these individual-level relationships are not independently randomized. They explain the aggregate pattern without fully identifying mediation.
↳ Results, Fig. 4; Extended Data Fig. 6
The study uses one short network game with fixed payoff, topology, and rewiring parameters. The Discussion acknowledges that these features, institutions, and sanctions may alter outcomes.
↳ Experimental design; Discussion, limitations paragraph
The randomized factorial design and large multi-session sample support a Quality score near 4.0, with Fig. 2 directly identifying the visibility-by-inequality interaction. Figs. 3–4 and the simulations provide a coherent account of welfare outcomes and behavior, although the individual-level mechanism is less directly identified than the treatment effect. Numerous secondary contrasts lack an explicit multiplicity adjustment, which keeps Methodological Rigour below the upper end of its band. Impact Potential is lower because the Discussion proposes broad relevance to wealth concealment and pay secrecy without field validation or a concrete decision-maker pathway.
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Strong4.0
The factorial manipulation separates initial inequality from wealth visibility and addresses mixed prior findings with a distinct interactional result. The advance is meaningful but demonstrated within one laboratory network game.
wealth visibility facilitates the downstream consequences of initial inequality
Random assignment, equal mean initial wealth, 80 independent sessions, and six experimental cells directly identify the principal treatment comparisons. Confidence is moderated by numerous uncorrected secondary tests and the pooled low-inequality conditions.
Subjects were randomly assigned to be rich or poor
The presentation moves coherently from aggregate inequality to welfare outcomes, individual behavior, and simulation. Real-world implications are explicitly qualified, although the mechanism account is less directly identified than the randomized treatment effects.
the results of laboratory experiments do not translate directly into the real world
The discussion engages mixed public-goods findings and connects the study to network cooperation, social comparison, and pay secrecy. It names several untested design features and appropriately limits societal interpretation, though alternative explanations receive less detailed treatment.
Prior work regarding the role of inequality in contributions to public goods has reported mixed results
Lower confidence on Methodological Rigour — domain match limited.
Caveats3 of 4 checks
The design and reported numerical results are internally coherent, with significant and null findings reported consistently. Numerous secondary tests lack an explicit multiple-comparison procedure, reducing confidence in some heterogeneity and secondary findings rather than the principal randomized comparison.
Competing interests and data availability are disclosed, but no explicit ethics-approval statement appears in the supplied text for this compensated human-subjects experiment. The absence is a conduct-reporting gap rather than evidence of unethical conduct.
Flags: 2 declared / 5 total
Reference integrity check could not run: no references section detected in extracted text.
0 references in manuscript
No retraction notice found in Retraction Watch.
Sources: Retraction Watch ✓
Low2.8
Wealth disclosure and pay secrecy provide recognizable policy and organizational contexts, but the paper does not identify a specific user or decision process. Its practical connection therefore remains general.
mechanisms that conceal personal wealth information might induce lower economic inequality
The experiment characterizes a mechanism under controlled conditions and supplements it with agent-based simulations. It does not test an intervention in a workplace, platform, community, or policy setting.
laboratory experiments do not translate directly into the real world
Transfer is constrained by one payoff scheme, random-network structure, rewiring rule, group-size range, and ten-round duration. The paper explicitly identifies these structural choices as unresolved boundaries.
whether the payoff structure, group size, network topology, or rewiring rate matter
The study extends an established sequence of dynamic-network cooperation experiments and offers a visibility mechanism relevant to mixed inequality findings. This creates a clear route for subsequent field and parameter-robustness research.
our results help shed light on these findings
Lower confidence on Trajectory — domain match limited.
AI-generated, human-governed. Something look off? Contact us to request a review.