Assessing private solutions to collective action problems in a 34-nation study

Assessing private solutions to collective action problems in a 34-nation study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Social Sciences - Article Assessing private solutions to collective action problems in a 34-nation study Eugene Malthouse, Charlie Pilgrim, Daniel Sgroi, Michela Accerenzi, and 68 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4924647/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Collective action problems emerge when individual incentives and group interests are misaligned, as in the case of climate change 1 – 5 . Individuals involved in collective action problems are often considered to have two options: contribute towards a public solution or free-ride. But they might also choose a third option of investing in a private solution such as local climate change adaptation 6 – 8 . Here we introduce a collective action game featuring wealth inequality caused by luck or merit and both public and private solutions with participants from 34 countries. We show that the joint existence of wealth inequality and private solutions has a consistent effect across countries: participants endowed with higher income choose the private solution almost twice as often as those endowed with lower income; and this finding cannot be explained by different sources of wealth (luck vs. merit) or by cultural or economic factors. We also show that preferences for private solutions undermine support for public solutions, resulting in wealth inequality increasing in every country. In contrast, we identify two universal pathways to successful public solution provision: early contributions to public solutions and conditional cooperation. Our findings highlight the ubiquity of the ‘private solution problem’ and its potential consequences for global collective action problems. Earth and environmental sciences/Environmental social sciences/Psychology and behaviour Scientific community and society/Social sciences/Climate change/Climate-change mitigation Scientific community and society/Social sciences/Economics Scientific community and society/Social sciences/Psychology/Human behaviour Scientific community and society/Social sciences/Decision making Introduction Human beings possess a unique ability to solve shared problems together 9 – 11 . Across the world we have developed health systems to protect us from disease, school systems to educate our young, and transport systems to help us move around. These are all examples of public goods that generally benefit everyone in the community, regardless of who has contributed to their provision. On account of this property of ‘non-excludability’, public goods are vulnerable to free-riding 1 . Providing public goods therefore means dealing with the free-rider problem – a problem that received much attention in recent years 12 – 16 . But it also means dealing with a second problem that has received far less attention during the same period. We call this the private solution problem . This problem can occur whenever private solutions exist as substitutes for public solutions to shared problems 6 – 8 . Real-world examples abound: just as many people choose between private healthcare, education, and transport and their public alternatives, many governments choose between climate change adaptation and abatement 17 . The adoption of private solutions, in turn, tends to undermine support for public solutions 6 – 8 . The private solution problem is therefore different from the free-rider problem, reflecting a preference for self-reliance and self-protection rather than selfishness, but threatens public good provision in a similar manner. Despite the increasing relevance of the private solution problem, particularly in the context of climate change, very little research has aimed to identify factors that may influence preferences for private rather than public solutions to shared problems. This pre-registered study ( https://osf.io/3ubt9 ) empirically investigates three factors proposed as relevant: 1) people’s ability to pay for private solutions 7 ; 2) the source of their wealth (luck vs. merit) 18 – 23 ; and 3) their cultural and economic background 24 – 27 . To achieve this we adapt a public good game known as the collective-risk social dilemma , which was originally designed to represent the collective action problem of limiting global warming to prevent catastrophic climate change 28 . In the collective-risk social dilemma , participants are assigned to groups and given an endowment that they can contribute (or not) towards a group target sum over 10 successive rounds. If as a group they contribute enough to achieve this target within the 10 rounds, all group members get to keep their remaining funds (i.e., those not contributed towards the target). If they fail to achieve the group target, simulated catastrophic climate change occurs and each participant loses their remaining funds with a certain probability. The dilemma for participants is therefore essentially the same as the one faced by individuals and governments alike: the more they contribute to mitigation efforts, the more likely their group is to avoid catastrophic climate change, but the fewer resources they will have left. The main change we made to this standard version of the game was to introduce a private target as an alternative to the group target, as has previously been done in similar experimental games 6 , 7 (see Fig. 1 ). The private target represented a private good, just as the group target represented a public good. We refer to these goods as private and public solutions, respectively. Unlike the public solution in our game, the private solution was excludable: if a player achieved it, they alone got to keep their remaining funds. We assigned participants to groups of four, gave them an endowment of Monetary Units (MU), and asked them how much they would like to contribute towards the private solution (costing 60 MU) and public solution (costing 160 MU, or 40 MU per group member on average) over the course of 10 successive rounds. All contributions were spent, or non-refundable. To avoid losing their remaining funds at the end of the game, players needed to achieve either the public or private solution within the 10 rounds. If they did so, their remaining funds were converted into cash and sent as a bonus payment after the study. In this way, the game offered the same opportunities for cooperation, free-riding, and self-reliance that exist in many collective action problems today. In our adapted game, groups were made up of two rich players (in red) who started the game with 120 MU and two poor players (in blue) who started the game with 80 MU. They either had to collectively contribute 160 MU to the public solution or individually contribute 60 MU to their private solution to avoid losing their remaining funds at the end of the game. Players could contribute up to a total of 20 MU towards the public solution and private solutions in each of 10 successive rounds, and all contributions were non-refundable. If the group achieved the (non-excludable) public solution, all members kept their remaining MU. Individuals who achieved their private solution also kept their remaining MU. These MU were converted into cash and paid as a bonus after the study. We made two further changes to the game in order to investigate whether people’s ability to pay for a private solution and the source of their wealth influenced their preferences for private vs. public solutions. The first was to introduce wealth inequalities within groups so that the private solution was more affordable for some players than others. Specifically, each group of four was made up of two rich players who started the game with 120 MU and two poor players who started with 80 MU. To compare rich and poor players’ preferences for private solutions, we measured whether they adopted the private solution by contributing sufficiently towards their private target. To compare their preferences for public solutions, we measured the proportion of their wealth that they contributed towards the public solution. We compare proportional rather than absolute contributions because this provides a clearer understanding of who is contributing their fair share. The second change was to introduce different sources of wealth inequality (luck vs. merit) by randomly assigning groups to one of three treatments: one in which participants’ endowments were determined by a lottery (the luck treatment); one in which they were determined by performance in an effort task (the merit treatment); and one in which they were determined either by a lottery or by task performance (the uncertain treatment). Before the game started, we also asked players what they thought would be fair for rich and poor players to contribute towards the public solution. Lastly, we recruited 7,504 participants from universities in 34 countries to investigate whether people’s cultural and economic background influenced their preferences for private vs. public solutions. We measured the influence of cultural background by testing for relationships between mean private and public solution success rates in each country and scores on the most widely used cultural dimensions in social science research. These included factors recorded by Hofstede’s 6D model (individualism, power distance, masculinity, uncertainty avoidance, long-term orientation, and indulgence) 24 ; the Inglehart-Welzel Cultural Map (survival vs. self-expression values, traditional vs. secular-rational values) 29 ; the tightness vs. looseness index (measuring the strength of social norms) 25 ; the World Values Survey (trust in others and acceptance of inequality) 30 ; and the OECD (trust in government) 31 . We also ran equivalent tests on participants’ beliefs about the relative importance of luck and merit in life and about individual responsibility, elicited via questions drawn from the World Values Survey and presented at the end of the experiment (see Methods). In addition, we compared the variation in private solution adoption rates both within and between cultures by classifying countries according to the Inglehart-Welzel World Cultural Map 32 , 33 . We followed a similar approach to measure the influence of economic background, testing for correlations between mean private and public solution success rates in each country and economic measures previously shown to influence behaviour in collective action problems 26 , 27 including individual subjective socioeconomic status 34 and World Bank measures of national GDP per capita, income inequality, and control of corruption (see Supplementary Information Table S1 for all measures and sources). Results Private Solution Adoption Rich players in our collective action game, who were better able to afford the private solution than poor players, adopted this solution 62.1% of the time. In contrast, poor players did so only 32.0% of the time ( χ² (1) = 681.8, p < .001, BF 10 = 1.1.291558e + 149, n = 7,504, two-sided χ² -tests). This represented decisive evidence that rich players were more likely to adopt the private solution than poor players. Figure 2a shows that this was the case in all 34 countries, while Fig. 2b shows that rich players contributed more towards the private solution than poor players in all 10 rounds of the game. This wealth effect remained strong after taking variability at the group and country levels into account ( ß = -1.61, 95% CI = [-1.80, -1.42], p < .001, n = 7,504, generalised logistic mixed-effects model; see Supplementary Information Table S7). Calculating the odds ratios here revealed that when players were rich the odds of adopting the private solution were 85% higher compared to when they were poor; and when players were poor, the odds of private solution success were around 20% of those when players were rich. This preference for private solutions among rich players emerged independently of whether they earned their higher wealth through luck or merit (or an uncertain one of the two). In fact, overall private solution adoption rates among rich players were almost the same in the luck (62.0%), merit (62.7%), and uncertain (61.6%) treatments ( χ² (2) = 0.3, p = .865, BF 01 = 422.5; standard and Bayesian two-sided chi-squared tests) and we did not detect any differences between treatments within countries (see Supplementary Information Table S8). At the same time, private solution adoption rates among poor players were also similar both overall in the luck (32.1%), merit (30.9%), and uncertain (32.9%) treatments ( χ² (2) = 1.2, p = .544, BF 01 = 287.6) and between treatments within countries. For additional robustness we fitted a mixed-effects model on private solution adoption with fixed wealth and treatment effects as well as random group and country effects ( n = 7,504). This model detected a poor wealth effect ( ß = -1.66, 95% CI = [-1.86, -1.47], p < .001) but no luck treatment effect ( ß = -0.03, 95% CI = [-0.24, 0.18], p = .777) or uncertain treatment effect ( ß = -0.03, 95% CI = [-0.24, 0.18], p = .809) or treatment x wealth interactions (see Supplementary Information Table S9). These results provided further evidence that the source of wealth did not influence preferences for private solutions among either rich or poor players. Preferences for private solutions also emerged independently of cultural or economic background. Kendall’s rank correlation tests did not reveal any relationship between private solution success rates in each country and scores on five of Hofstede’s six dimensions of national culture (individualism, τ = -0.13, p = .348, n = 27; power distance, τ = 0.23, p = .085, n = 27; masculinity, τ = -0.04, p = .786, n = 27; long-term orientation, τ = -0.12, p = .332, n = 31; and indulgence, τ = -0.04, p = .746, n = 31). The exception here was uncertainty avoidance ( τ = -0.30, p = .030, n = 27) although this relationship was not statistically significant after controlling for multiple testing ( p = .180, Bonferroni-Holm method). Equally, we did not detect a relationship between private solution success rates and survival vs. self-expression values ( τ = -0.06, p = .721, n = 22); secular-rational vs. traditional values ( τ = -0.11, p = .417, n = 29); tightness vs. looseness ( τ = 0.28, p = .111, n = 18); national-level beliefs about the acceptability of inequality ( τ = 0.22, p = .105, n = 28), the trustworthiness of others ( τ = -0.04, p = .752, n = 28), or trust in government (( τ = 0.05, p = .837, n = 12); or participant-level beliefs about the importance of luck and merit ( τ = 0.11, p = .382, n = 32) and individual responsibility ( τ = 0.18, p = .142, n = 32). Our evaluation of private solution adoption rates across cultural regions (as defined by the Inglehart-Welzel World Cultural Map) also revealed differences both between cultures ( F (7,26) = 9.46, p < .001) and within cultures ( F (26,1842) = 3.97, p < .001) with greater variability in the former than the latter (see Extended Data Fig. 1 ). Lastly, we did not detect a relationship between private solution success rates and either our individual economic measure of subjective socioeconomic status ( τ = 0.07, p = .624, n = 34) or our national economic measures (GDP per capita, τ = -0.21, p = .091, n = 34; income inequality, τ = 0.23, p = .067, n = 34; control of corruption, τ = -0.12, p = .319, n = 34). In summary, our finding that rich players adopted the private solution almost twice as often as the poor was evident regardless of the source of their wealth or their cultural and economic background. Public Solution Contributions Private solution adoption also undermined public solution contributions and provision. This was evident in the negative relationship between private and public solution success rates in each country ( ρ = -0.81, p < .001) as shown in Fig. 3 a. This pattern emerged because individuals adopting private solutions generally stopped contributing towards public solutions, even though they may have had the funds available to do so, which made these solutions more difficult for their group to achieve (see Extended Fig. 2). This was particularly evident among rich players, who in total contributed a lower proportion of their wealth towards the public solution than poor players in every country except Ghana (see Fig. 3 b). Overall, rich players contributed 31.3% of their wealth whereas poor players contributed 39.8% of theirs ( W = 5753438, p < .001, BF 10 = 2.600205e + 58; two-sided Mann-Whitney-Wilcoxon [MWW] tests). In absolute terms, this equated to mean contributions of 37.6 MU from rich players and 31.8 MU from poor players. This meant that rich players, unlike poor players, did not contribute towards the public solution in line with what participants judged to be fair (see Extended Data Fig. 3 ). This finding of lower proportional but higher absolute contributions from the rich is in line with several previous studies 35 – 37 . It was further supported by the results of a mixed-effects model of players’ contributions as a proportion of their wealth in each round that controlled for the effect of rounds, included random individual, group and country effects, allowing the effect of wealth to vary within countries ( ß = 2.71, 95% CI = [2.20, 3.20], p < .001, n = 60,270; see Supplementary Information Table S10). In addition, Fig. 3 c shows that these differences in contributions from rich and poor players emerged from the very first round (Rich: 7.1% vs. Poor: 11.3%, W = 5413782, p < .001, BF 10 = 2.695783e + 109, two-sided MWW tests) with poor players contributing more than rich players even in absolute terms (Rich: 8.5 MU vs. Poor: 9.0 MU, W = 6746814, p = .001, BF 10 = 1.35, two-sided MWW tests). This pattern of rich players contributing a lower proportion of their wealth than poor players also emerged independently of different sources of wealth. On average, rich players contributed a similar proportion of their wealth in each round in the luck (5.1%), merit (4.7%), and uncertain (4.6%) treatments. Likewise, poor players contributed a similar proportion of theirs in the luck (6.9%), merit (6.8%), and uncertain (6.8%) treatments (see Extended Data Fig. 4 for a comparison over the course of the 10 rounds). We tested for a treatment effect with a linear mixed-effects model of players’ contributions as a proportion of their wealth in each round ( n = 60,270) that did not detect any treatment effects or treatment x wealth interactions (see Supplementary Information Table S11). These similarities in contribution behaviours between treatments help to explain why public solution provision was almost exactly the same among groups in the luck (63.6%), merit (60.8%) and uncertain (59.5%) treatments ( n = 1,876, χ² (2) = 2.41, p = .300, BF 01 = 72.7, two-sided standard and Bayesian χ² -tests). As a result of rich players contributing a lower proportion of their wealth towards the public solution in all three treatments and in almost all countries, wealth inequality within groups tended to increase over time. At the start of the game, wealth inequality (as measured by the Gini coefficient) was the same in every group (0.1). At the end of the game, after taking losses resulting from failures to achieve either the private or public solution into account, wealth inequality increased to an average of 0.52 ( V = 1504068, p < .001, one-sample Wilcoxon signed rank test). This level of inequality is equivalent to the current level of inequality in Brazil, which at present has the second highest level of inequality out of the G20 countries 38 . Figure 3 d shows that this trend towards increasing inequality emerged in every country. It also remained similar (0.49) when losses from private and public solution failures were excluded ( V = 1624098, p < .001). Lastly, we investigated whether the variation in public solution provision across countries could be explained by our selection of cultural and economic factors by testing for correlations between these variables. We did not detect statistically significant relationships with any of Hofstede’s six dimensions of national culture (individualism: τ = 0.24, p = .083, n = 27; power distance: τ = -0.20, p = .144, n = 27; masculinity: τ = 0.04, p = .754, n = 27; uncertainty avoidance: τ = 0.24, p = .080, n = 27; long-term orientation: τ = -0.06, p = .646, n = 31; indulgence: τ = 0.03, p = .798, n = 31). Nor did we detect a relationship between public solution provision in a country and its survival vs. self-expression values ( τ = 0.23, p = .163, n = 22); secular-rational vs. traditional values ( τ = -0.02, p = .905, n = 29); tightness vs. looseness ( τ = -0.139, p = .425, n = 18); beliefs about the acceptability of inequality ( τ = -0.11, p = .429, n = 28) and the trustworthiness of others ( τ = -0.04, p = .782, n = 28); or trust in government ( τ = 0.17, p = .450, n = 12). Furthermore, public solution provision across countries did not correlate with participants’ beliefs about the importance of luck and merit as determinants of life outcomes ( τ = -0.18, p = .142, n = 32) or with any of our economic factors (individual subjective socioeconomic status: τ = 0.11, p = .441, n = 34; national GDP per capita: τ = 0.17, p = .173, n = 34; national income inequality: τ = -0.16, p = .189, n = 34; national control of corruption: τ = 0.13, p = .291, n = 34). Finally, public solution provision differed both between cultures ( F (7) = 14.64, p < .001) and within cultures ( F (26) = 6.66, p < .001) with greater variability in the former than the latter (see Extended Data Fig. 5). In summary, none of the above cultural or economic factors could help to explain differences in public solution provision rates between countries, in line with some previous research 39 . There was, however, one set of cultural values that did help to explain differences in public solution provision between countries. This was the strength of the belief within our samples that ‘people should take greater responsibility to provide for themselves’. We found that the average level of agreement with this statement among participants in each country was negatively associated with the rate of public solution provision in each country ( τ = -0.29, p = .014, n = 32; see Extended Data Fig. 6a). Furthermore, the same pattern emerged at a group level: the stronger the beliefs in individual responsibility within a group, the less likely it was to achieve the public solution ( ß = -0.12, 95% CI = [-0.19, -0.04], p = .001, n = 1,850, generalised logistic mixed-effects model; see Supplementary Information Table S12). In this way, beliefs about individual responsibility served as a universal pathway to public solution non-provision across countries. One possible explanation for this effect is that beliefs in individual responsibility predicted higher first round contributions to private solutions ( ß = -0.08, 95% CI = [0.02, 0.14], p = .015, n = 7,445, linear mixed-effects model; see Supplementary Information Table S13). In turn, these higher private contributions were universally associated with lower rates of public solution provision ( ß = -0.06, 95% CI = [-0.08, -0.05], p < .001, n = 1,876, see Extended Data Fig. 6b and Supplementary Information Table S14 for generalised logistic mixed-effects model details). In summary, people’s beliefs in individual responsibility were associated with higher first round contributions to private solutions and undermined public solution provision at both a national and a group level. Universal Pathways for Public Solutions Given the number of present-day collective action problems that span international borders, identifying factors that might undermine and underpin public solution provision across countries is an important empirical question. In the previous section we identified beliefs in individual responsibility and first round contributions to private solutions as universal pathways to public solution non-provision. In this section we outline two universal pathways to public solution provision. The first was first round contributions towards the public solution. As shown in Fig. 4 a, groups with higher public contributions towards the public solution in the first round were more likely to achieve the public solution across all countries ( ß = 0.11, 95% CI = [0.09, 0.12], p < .001, n = 1,876, generalised logistic mixed-effects model; see Supplementary Information Table S15). Furthermore, comparing the distributions of these first-round contributions between groups that provided vs. did not provide the public solution revealed that there was a 69% chance that a randomly selected successful (vs. unsuccessful) group would exhibit higher contributions in the first round ( W = 704291, p < .001, two-sided MWW test; rank-biserial correlation r b = 0.58; Cliff’s delta d = 0.69). In short, early contributions towards public solutions presaged public solution provision across countries. The second universal pathway to public solution provision was conditional cooperation within groups. We measured conditional cooperation by evaluating the relationship between a player’s contribution to the public solution in the present round and the average public solution contribution from their group members in the previous round 40 , 41 . Figure 4 b shows that levels of conditional cooperation were significantly higher in groups that successfully achieved the public solution compared with groups that did not. This finding was supported by a linear mixed-effects model that revealed a significant average group contribution on previous round x public solution provision interaction ( ß = 0.23, 95% CI = [0.20, 0.25], p < .001, n = 52,892; see Supplementary Information Table S16). In other words, players’ responses to contributions from their fellow group members universally underpinned public solution provision. The Private Solution Problem We reported that rich players in our collective action game adopted the private solution almost twice as often as poor players. They did so regardless of both the source of their wealth (luck vs. merit vs. luck or merit) and their cultural and economic background. Furthermore, rich players’ preferences for private solutions saw them contribute proportionally less than the poor towards public solutions, also independently of both the source of their wealth and their cultural and economic background. Wealth inequality within groups increased as a consequence. These results point towards the possibility of a universal private solution problem whereby the existence and widespread adoption of private solutions by those who can afford them creates an externality that exacerbates wealth inequality within groups and undermines public solution provision 7 , 8 , 42 . While we identified a universal preference for self-reliance among rich players, we also identified two universal pathways to public solution provision that may be relevant for global collective action problems such as climate change mitigation. The first was via early contributions to the public solution: countries in which contributions in the first round were higher (e.g., France, Denmark, and Colombia) were more likely to achieve the public solution than places in which they were lower (e.g., UK, Ghana, and the Dominican Republic) in line with findings from previous studies 32 , 43 . The second was conditional cooperation: groups that succeeded in achieving the public solution were generally made up of players who responded more reciprocally to the contributions of others. Generalising these findings would suggest that early commitments, as well as reciprocal responses to these commitments, may disproportionately influence group outcomes. Overall, we observed few cultural or economic explanations of preferences for private and public solutions across our samples – even though our participants came from a wide range of Western, Educated, Industrialised, Rich and Democratic (WEIRD) and non-WEIRD countries 44 . While we detected greater variability in preferences for private and public solutions between cultures than within cultures 32 , we were mostly unable to explain these differences from either a cultural or economic perspective. This may be on account of our recruitment of university student participants, who were unlikely to be nationally representative and were perhaps more likely to behave similarly to their country’s elites rather than its ordinary citizens 45 , 46 . Further research might therefore explore how universities may shape people’s approaches to collective decision-making in ways that do not reflect their local culture. Lastly, we propose that private solutions are more costly than public solutions in ways that extend far beyond economic inefficiency. What we mean is that global collective action problems never exist in spatial or temporal isolation – they are invariably just one of several contemporaneous problems that all extend into the future. And not all of them can be solved privately. Methods Participants A total of 7,504 participants completed this study, an average of 55 groups of four in each country (see Supplementary Information for further participant details). We recruited participants from student pools accessible to local universities in each country between July 2022 and March 2024. In universities with more diverse student pools, we aimed to predominantly recruit local students by restricting participation to national citizens or by translating the experiment into the local language. We recruited 20 groups in each treatment in each country, or 15 groups where this was not possible. All participants provided informed consent prior to taking part in the experiment and were free to withdraw at any time. The experiment was approved by ethics committees in every country where this was required by local universities. To ensure consistent implementation across countries, local researchers followed the same set of instructions for the experimental procedure (see Supplementary Information section 1). Participants were randomly assigned to groups of four, which in turn were randomly assigned to the luck, merit, or uncertain treatments. At the end of the experiment, participants’ final earnings (in MU) were converted to cash and added to their participation fee. Conversion rates and participation fees were adjusted for purchasing power parity (PPP) across countries (see Supplementary Information section 1). On average, participants earned around GBP £5.20 (£7.75 PPP equivalent). The experiment generally lasted 30–45 minutes, depending on the location and implementation method (in-person vs. online). The vast majority of participants were undergraduates: 83.6% were aged 18–24; 14.7% were aged 25–34; and 1.6% were aged 35 and over. They were recruited from a range of academic disciplines, with the greatest representation from Economics (16%). 81.7% attended public universities, while 16% attended private universities. Around half (44.4%) of participants completed the experiment in a lab setting; the remainder (55.5%) completed the experiment in an online setting. Overall, 51.9% of our sample identified as female and 46.4% identified as male; 1.7% identified as non-binary or another gender (see Supplementary Information Table S4 for breakdown by country). The entire study was programmed in oTree 47 . Instructions for participants were translated into local languages where appropriate. Translations can be found in the experimental code, which can be accessed via https://osf.io/3ubt9 . Experimental Design After reading a participant information sheet and completing a consent form, participants provided their basic demographic data (age and gender). On the next page, they were informed that the study was made up of two sections: 1) an arithmetic task to be completed individually; and 2) a problem-solving game to be completed in groups of four and played with Monetary Units (MU) that would be converted into cash at the end of the study. Participants were then told that two players in each group would start the game with 120 MU while the other two players in 80 MU. Participants in the merit treatment were informed that the two players in the group who scored highest in the mental arithmetic task would start with 120 MU and the two players who scored lowest would start with 80 MU. In the luck treatment, participants were informed that a lottery would determine which two players would start the game with 120 MU and which two players would start with 80 MU. In other words, their performance in the mental arithmetic task had no bearing on their endowment. To incentivise effort here for consistency across treatments, we informed these participants that the highest-scoring player in each group would receive a bonus payment of £1 (or approximate PPP equivalent) at the very end of the study. Participants in the uncertain treatment were told that either their performance in the arithmetic task or a lottery would determine their endowments. In practice, luck was chosen for two players in each group, one of whom became rich and one of whom became poor. For the other two players in the group, the one who scored highest in the effort task became rich and the one who scored lowest became poor. We included this treatment because it represents the real world, in which the cause of wealth is always an uncertain combination of luck and merit. It also enabled us to test for self-serving bias, which would be evident if rich players in the uncertain treatment behaved similarly to rich players in the merit treatment and if poor players in the uncertain treatment behaved similarly to poor players in the luck treatment. The mental arithmetic task was one that has been used by several researchers in the past 37 , 48 . Participants were asked to add up sequences of five random numbers from 1-100 (e.g., 81 + 25 + 2 + 52 + 13 = ?) for five minutes. They were awarded one point per correct answer. After they had finished the task they were informed whether they had become rich or poor and reminded that this was either determined by their performance in the task (in the merit treatment) or by a lottery (in the luck treatment). Participants were then provided with more detailed instructions about the game (see Supplementary Information for materials) and then had to correctly answer four comprehension questions in order to proceed to the game. The quiz also included two questions about what participants judged to be fair contributions from rich and poor players. Once all four participants within a group had completed the pre-game quiz, the first round of the game commenced. At the start of each round participants were told the round number, how many MU they had left, how much they had contributed to their private solution, how much more they needed to contribute to achieve the private solution, how much the group as a whole had contributed to the public solution, and how much more the group needed to contribute to achieve the public solution. They were asked how much they would like to contribute (up to 20 MU in total) towards the private and public solutions. The round concluded once all four players had made their contribution decisions. The cost of the public solution was 160 MU, or on average 40 MU per player, while the cost of the private solution was 60 MU. We chose this cost ratio of 1:1.5 because previous authors 6 , 7 found that this creates uncertainty among participants regarding the optimal decision. In other words, if the private solution is any less expensive then people will generally choose that over the public solution; if it is any more expensive, then people will generally choose the public solution. All funds that participants contributed to either the private or public solution were spent (i.e., non-refundable). If they failed to solve the collective action problem either privately or publicly they lost all their remaining MU. Once participants had completed the game they responded to two statements drawn from the World Values Survey. We asked them to place their views on a scale from 1–10, where 1 indicated complete agreement with a statement on the left and 10 indicated complete agreement with a statement on the right. In the first statement about merit and luck, the statement on the left (1) was ‘In the long run, hard work usually brings a better life’; the statement on the right (10) was ‘Hard work doesn’t generally bring success – it’s more a matter of luck and connections’. In the second statement about individual responsibility, the statement on the left (1) was ‘The government should take more responsibility to ensure that everyone is provided for’; the statement on the right (10) was ‘People should take more responsibility to provide for themselves’. After responding to these statements, participants in most countries reported their household income, field of study, and perceived position on the MacArthur Scale of Subjective Social Status 34 . Finally, they were debriefed and thanked for their participation. Data Analysis Data analysis was conducted in R. Our hypotheses relating to private and public solution preferences were pre-registered at https://osf.io/3ubt9 (we added hypotheses relating to cultural and economic factors after the number of international labs and participants involved increased). To test these hypotheses we mainly assessed how private solution success rates, contributions to the public solution, and public solution success rates compared between wealth levels, treatment levels, and countries. To compare private and public solution success rates we used standard and Bayesian chi-squared tests. To compare contributions to the public solution we used a combination of MWW tests and mixed-effects models. The reason for using MWW tests was that the distribution of contribution behaviour deviated from normality. The reason for using mixed-effects models was that our data had a nested structure as rounds were nested in individuals, individuals were nested in groups, and groups were nested in countries 49 . Full details of all statistical tests and models can be found in our Supplementary Information and coding script (see below). Declarations Data and Code Availability All data will be uploaded to https://osf.io/zg8nj/ upon acceptance for publication. All code used for data analysis is available at https://osf.io/zg8nj/ . References Olson, M. The Logic of Collective Action: Public Goods and the Theory of Groups. (Harvard University Press, 1965). Hardin, G. The tragedy of the commons. Science 162, 1243–1248 (1968). Ostrom, E. Collective action and the evolution of social norms. J. Econ. Perspect. 14, 137–158 (2000). Barrett, S. Why Cooperate?: The Incentive to Supply Global Public Goods. (Oxford University Press, 2007). Carattini, S., Levin, S. & Tavoni, A. Cooperation in the climate commons. Rev. Environ. Econ. Policy (2019). Gross, J. & De Dreu, C. K. Individual solutions to shared problems create a modern tragedy of the commons. Sci. Adv. 5, eaau7296 (2019). Gross, J., Veistola, S., De Dreu, C. K. & Van Dijk, E. Self-reliance crowds out group cooperation and increases wealth inequality. Nat. Commun. 11, 5161 (2020). Gross, J. & Böhm, R. Voluntary restrictions on self-reliance increase cooperation and mitigate wealth inequality. Proc. Natl. Acad. Sci. 117, 29202–29211 (2020). Henrich, N. & Henrich, J. P. Why Humans Cooperate: A Cultural and Evolutionary Explanation. (Oxford University Press, 2007). Bowles, S. & Gintis, H. A cooperative species: Human reciprocity and its evolution. in A Cooperative Species (Princeton University Press, 2011). Nowak, M. & Highfield, R. Supercooperators: Altruism, Evolution, and Why We Need Each Other to Succeed. (Simon & Schuster, 2011). Brubaker, E. R. Free ride, free revelation, or golden rule? J. Law Econ. 18, 147–161 (1975). Nordhaus, W. Climate clubs: Overcoming free-riding in international climate policy. Am. Econ. Rev. 105, 1339–1370 (2015). Barclay, P. & Willer, R. Partner choice creates competitive altruism in humans. Proc. R. Soc. B Biol. Sci. 274, 749–753 (2007). Bó, P. D. Cooperation under the shadow of the future: experimental evidence from infinitely repeated games. Am. Econ. Rev. 95, 1591–1604 (2005). Fehr, E. & Gächter, S. Altruistic punishment in humans. Nature 415, 137–140 (2002). Barrett, S. Dikes versus windmills: Climate treaties and adaptation. Clim. Change Econ. 11, 2040005 (2020). Alesina, A. F., Glaeser, E. L. & Sacerdote, B. Why Doesn’t the US Have a European-Style Welfare System? (2001). Cappelen, A. W., Sørensen, E. Ø. & Tungodden, B. Responsibility for what? Fairness and individual responsibility. Eur. Econ. Rev. 54, 429–441 (2010). Mijs, J. J. The paradox of inequality: Income inequality and belief in meritocracy go hand in hand. Socio-Econ. Rev. 19, 7–35 (2021). Mounk, Y. The age of responsibility: Luck, choice, and the welfare state. Harv. Univ. Press (2017). Frank, R. H. Success and Luck. (Princeton University Press, 2016). Sandel, M. J. The Tyranny of Merit: What’s Become of the Common Good? (Penguin UK, 2020). Hofstede, G. Dimensionalizing cultures: The Hofstede model in context. Online Read. Psychol. Cult. 2, 8 (2011). Gelfand, M. J. et al. Differences between tight and loose cultures: A 33-nation study. Science 332, 1100–1104 (2011). Alesina, A. & La Ferrara, E. Participation in heterogeneous communities. Q. J. Econ. 115, 847–904 (2000). Bardhan, P., Ghatak, M. & Karaivanov, A. Wealth inequality and collective action. J. Public Econ. 91, 1843–1874 (2007). Milinski, M., Sommerfeld, R. D., Krambeck, H. J., Reed, F. A. & Marotzke, J. The collective-risk social dilemma and the prevention of simulated dangerous climate change. Proc. Natl. Acad. Sci. 105, 2291–2294 (2008). Inglehart, R. & Welzel, C. Value Change and the Persistence of Cultural Traditions. in Modernization, Cultural Change, and Democracy 48–76 (Cambridge University Press, 2012). Haerpfer, C. et al. World Values Survey: Round Seven – Country-Pooled Datafile. (JD Systems Institute & WVSA Secretariat, 2020). OECD. Trust in Government. https://www.oecd.org/en/data/indicators/trust-in-government.html (2024). Gächter, S., Herrmann, B. & Thöni, C. Culture and cooperation. Philos. Trans. R. Soc. B Biol. Sci. 365, 2651–2661 (2010). Inglehart, R. & Welzel, C. The Inglehart–Welzel World Cultural Map—World Values Survey 7. (2023). Adler, N. E., Epel, E. S., Castellazzo, G. & Ickovics, J. R. Relationship of subjective and objective social status with psychological and physiological functioning: Preliminary data in healthy, White women. Health Psychol. 19, 586 (2000). Tavoni, A., Dannenberg, A., Kallis, G. & Löschel, A. Inequality, communication, and the avoidance of disastrous climate change in a public goods game. Proc. Natl. Acad. Sci. 108, 11825–11829 (2011). Vicens, J. et al. Resource heterogeneity leads to unjust effort distribution in climate change mitigation. PloS One 13, Article e0204369 (2018). Malthouse, E., Pilgrim, C., Sgroi, D. & Hills, T. T. When fairness is not enough: The disproportionate contributions of the poor in a collective action problem. J. Exp. Psychol. Gen. (2023). World Bank. GINI Index (World Bank Estimate). (2024). Romano, A., Balliet, D., Yamagishi, T. & Liu, J. H. Parochial trust and cooperation across 17 societies. Proc. Natl. Acad. Sci. 114, 12702–12707 (2017). Gächter, S. Conditional Cooperation: Behavioral Regularities from the Lab and the Field and Their Policy Implications. in Psychology and Economics: A Promising New Cross-Disciplinary Field (eds. Frey, B. S. & Stutzer, A.) (The MIT Press, Cambridge, 2007). Fischbacher, U. & Gächter, S. Social preferences, beliefs, and the dynamics of free riding in public goods experiments. Am. Econ. Rev. 100, 541–556 (2010). Mori, R., Hanaki, N. & Kameda, T. An outside individual option increases optimism and facilitates collaboration when groups form flexibly. Nat. Commun. 15, 5520 (2024). Gürdal, M. Y., Torul, O. & Yahşi, M. The seeds of success: the pivotal role of first round cooperation in public goods games. J. Econ. Sci. Assoc. 1–23 (2024). Henrich, J., Heine, S. J. & Norenzayan, A. The weirdest people in the world? Behav. Brain Sci. 33, 61–83 (2010). Hugh-Jones, D. Honesty, beliefs about honesty, and economic growth in 15 countries. J. Econ. Behav. Organ. 127, 99–114 (2016). Fisman, R., Jakiela, P., Kariv, S. & Markovits, D. The distributional preferences of an elite. Science 349, aab0096 (2015). Chen, D. L., Schonger, M. & Wickens, C. oTree—An open-source platform for laboratory, online, and field experiments. J. Behav. Exp. Finance 9, 88–97 (2016). Oswald, A. J., Proto, E. & Sgroi, D. Happiness and productivity. J. Labor Econ. 33, 789–822 (2015). Aarts, E., Verhage, M., Veenvliet, J. V., Dolan, C. V. & Van Der Sluis, S. A solution to dependency: using multilevel analysis to accommodate nested data. Nat. Neurosci. 17, 491–496 (2014). Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformation.pdf Supplementary Information Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4924647","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Social Sciences - Article","associatedPublications":[],"authors":[{"id":343362895,"identity":"48b4fb29-9666-4d3c-b1f5-95aaad8f7e9d","order_by":0,"name":"Eugene Malthouse","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0001-5665-7501","institution":"University of Warwick","correspondingAuthor":true,"prefix":"","firstName":"Eugene","middleName":"","lastName":"Malthouse","suffix":""},{"id":343362896,"identity":"9796f50f-3310-4cc1-8d37-3a0d0fc107f4","order_by":1,"name":"Charlie Pilgrim","email":"","orcid":"","institution":"University of Warwick","correspondingAuthor":false,"prefix":"","firstName":"Charlie","middleName":"","lastName":"Pilgrim","suffix":""},{"id":343362897,"identity":"b321ba3d-00e5-4c2b-9bb8-c87c0d67f213","order_by":2,"name":"Daniel Sgroi","email":"","orcid":"https://orcid.org/0000-0001-9575-2640","institution":"University of Warwick","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Sgroi","suffix":""},{"id":343362898,"identity":"ce69b894-a688-4bdc-874d-6428f0508fd9","order_by":3,"name":"Michela Accerenzi","email":"","orcid":"","institution":"Fundación ETEA para el Desarrollo y la Cooperación","correspondingAuthor":false,"prefix":"","firstName":"Michela","middleName":"","lastName":"Accerenzi","suffix":""},{"id":343362899,"identity":"b7962372-1b9c-4540-92c9-87d2b5a81746","order_by":4,"name":"Antonio Alfonso","email":"","orcid":"https://orcid.org/0000-0002-5396-3711","institution":"Universidad Loyola Andalucía","correspondingAuthor":false,"prefix":"","firstName":"Antonio","middleName":"","lastName":"Alfonso","suffix":""},{"id":343362900,"identity":"84133502-76df-437f-b02f-12acf66dc368","order_by":5,"name":"Rana Umair Ashraf","email":"","orcid":"https://orcid.org/0000-0002-0152-8419","institution":"Faculty of Economic Sciences, University of Warsaw","correspondingAuthor":false,"prefix":"","firstName":"Rana","middleName":"Umair","lastName":"Ashraf","suffix":""},{"id":343362901,"identity":"c624cff6-2013-4cd9-bfb6-12d90fb7f95d","order_by":6,"name":"Max Baard","email":"","orcid":"","institution":"University of Cape Town","correspondingAuthor":false,"prefix":"","firstName":"Max","middleName":"","lastName":"Baard","suffix":""},{"id":343362902,"identity":"06dee0d5-6265-42aa-a48f-127f97a81cba","order_by":7,"name":"Sanchayan Banerjee","email":"","orcid":"","institution":"Vrije Universiteit Amsterdam","correspondingAuthor":false,"prefix":"","firstName":"Sanchayan","middleName":"","lastName":"Banerjee","suffix":""},{"id":343362903,"identity":"65255528-30ce-4842-b8eb-1b4914ebff29","order_by":8,"name":"Alexis Belianin","email":"","orcid":"","institution":"National Research University Higher School of Economics, Moscow","correspondingAuthor":false,"prefix":"","firstName":"Alexis","middleName":"","lastName":"Belianin","suffix":""},{"id":343362904,"identity":"67d42ade-3ca0-411a-9501-4564db861cd6","order_by":9,"name":"Swagata Bhattacharjee","email":"","orcid":"","institution":"Ashoka University","correspondingAuthor":false,"prefix":"","firstName":"Swagata","middleName":"","lastName":"Bhattacharjee","suffix":""},{"id":343362905,"identity":"1e1a9482-5f1f-4c16-8df6-ea8c9c4b87ad","order_by":10,"name":"Mihir Bhattacharya","email":"","orcid":"","institution":"Ashoka University","correspondingAuthor":false,"prefix":"","firstName":"Mihir","middleName":"","lastName":"Bhattacharya","suffix":""},{"id":343362906,"identity":"0dd02e1b-f225-43de-b7f2-967eb63b5fd1","order_by":11,"name":"Pablo Branas-Garza","email":"","orcid":"https://orcid.org/0000-0001-8456-6009","institution":"Universidad Loyola Andalucia","correspondingAuthor":false,"prefix":"","firstName":"Pablo","middleName":"","lastName":"Branas-Garza","suffix":""},{"id":343362907,"identity":"4d86ff0e-9347-4678-bf02-120f35b775dd","order_by":12,"name":"Juan-Camilo Cardenas","email":"","orcid":"https://orcid.org/0000-0003-0005-7595","institution":"Universidad de los Andes","correspondingAuthor":false,"prefix":"","firstName":"Juan-Camilo","middleName":"","lastName":"Cardenas","suffix":""},{"id":343362908,"identity":"05d0fca7-f1b2-4247-958a-6ec19175417f","order_by":13,"name":"Miguel Carriquiry","email":"","orcid":"","institution":"Universidad de la República de Uruguay","correspondingAuthor":false,"prefix":"","firstName":"Miguel","middleName":"","lastName":"Carriquiry","suffix":""},{"id":343362909,"identity":"20f8149c-5624-4f86-8d38-f3790435c55b","order_by":14,"name":"Syngjoo Choi","email":"","orcid":"","institution":"Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Syngjoo","middleName":"","lastName":"Choi","suffix":""},{"id":343362910,"identity":"cbe3c1ee-14a1-4c01-b9d7-87f654e2e4c4","order_by":15,"name":"Gwen-Jiro Clochard","email":"","orcid":"","institution":"Institute of Social and Economics Research, Osaka University; Joint Initiative for Latin American Experimental Economics","correspondingAuthor":false,"prefix":"","firstName":"Gwen-Jiro","middleName":"","lastName":"Clochard","suffix":""},{"id":343362911,"identity":"f96702c2-caba-434d-a4cd-1329d98970b0","order_by":16,"name":"Eduardo Ezekiel Denzon","email":"","orcid":"","institution":"Ateneo de Manila University","correspondingAuthor":false,"prefix":"","firstName":"Eduardo","middleName":"Ezekiel","lastName":"Denzon","suffix":""},{"id":343362912,"identity":"e21ad1b2-db80-4a50-afe0-e8eac2d89826","order_by":17,"name":"Bartlomiej Dessoulavy-Sliwinski","email":"","orcid":"https://orcid.org/0000-0002-4395-9815","institution":"Faculty of Economic Sciences, University of Warsaw","correspondingAuthor":false,"prefix":"","firstName":"Bartlomiej","middleName":"","lastName":"Dessoulavy-Sliwinski","suffix":""},{"id":343362913,"identity":"9d12e933-91e4-403c-9593-69851de020c7","order_by":18,"name":"Giorgio Dini","email":"","orcid":"https://orcid.org/0009-0000-2514-3129","institution":"University of Bologna","correspondingAuthor":false,"prefix":"","firstName":"Giorgio","middleName":"","lastName":"Dini","suffix":""},{"id":343362914,"identity":"566c35d1-8f53-4346-b2ff-7cdb36c4d318","order_by":19,"name":"Lu Dong","email":"","orcid":"","institution":"Southern University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Lu","middleName":"","lastName":"Dong","suffix":""},{"id":343362915,"identity":"af249d11-7004-4254-838a-123593639518","order_by":20,"name":"Antal Ertl","email":"","orcid":"","institution":"Corvinus University of Budapest","correspondingAuthor":false,"prefix":"","firstName":"Antal","middleName":"","lastName":"Ertl","suffix":""},{"id":343362916,"identity":"e21a0a4c-3327-4ba4-98cf-39d7f812df63","order_by":21,"name":"Filippos Exadaktylos","email":"","orcid":"https://orcid.org/0000-0001-7339-8847","institution":"Institut de Ciència i Tecnologia Ambientals","correspondingAuthor":false,"prefix":"","firstName":"Filippos","middleName":"","lastName":"Exadaktylos","suffix":""},{"id":343362917,"identity":"8820dae0-8211-4cef-b51c-f97132298ddd","order_by":22,"name":"Emel Filiz-Ozbay","email":"","orcid":"","institution":"University of Maryland","correspondingAuthor":false,"prefix":"","firstName":"Emel","middleName":"","lastName":"Filiz-Ozbay","suffix":""},{"id":343362918,"identity":"ba71001f-f96c-432a-b6c5-bddaa03729b8","order_by":23,"name":"Sarah Flecke","email":"","orcid":"","institution":"University of Innsbruck","correspondingAuthor":false,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Flecke","suffix":""},{"id":343362919,"identity":"20bb55a0-6818-4df7-b4dc-d77ddfba6e5c","order_by":24,"name":"Fabio Galeotti","email":"","orcid":"","institution":"National Center for Scientific Research (CNRS)","correspondingAuthor":false,"prefix":"","firstName":"Fabio","middleName":"","lastName":"Galeotti","suffix":""},{"id":343362920,"identity":"51bd59e5-5d05-48b4-9113-9f651f441fa0","order_by":25,"name":"Teresa Garcia","email":"","orcid":"https://orcid.org/0000-0003-1778-254X","institution":"GLOBE: Univ. Granada","correspondingAuthor":false,"prefix":"","firstName":"Teresa","middleName":"","lastName":"Garcia","suffix":""},{"id":343362921,"identity":"a2070e2e-6549-4c13-abe4-da5792069697","order_by":26,"name":"Nobuyuki Hanaki","email":"","orcid":"https://orcid.org/0000-0003-3127-690X","institution":"Osaka University","correspondingAuthor":false,"prefix":"","firstName":"Nobuyuki","middleName":"","lastName":"Hanaki","suffix":""},{"id":343362922,"identity":"17083090-a982-4470-80ad-32387ed4e242","order_by":27,"name":"University of Warwick Hollard","email":"","orcid":"","institution":"CREST, École polytechnique, Institut Polytechnique de Paris - CNRS","correspondingAuthor":false,"prefix":"","firstName":"University","middleName":"of Warwick","lastName":"Hollard","suffix":""},{"id":343362923,"identity":"2566053a-409b-438a-a117-b971ef6deef8","order_by":28,"name":"Daniel Horn","email":"","orcid":"","institution":"HUN-REN KRTK \u0026 Corvinus University of Budapest","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Horn","suffix":""},{"id":343362924,"identity":"ff6318cd-a904-43b4-8077-c54ba5e88ea0","order_by":29,"name":"Lingbo Huang","email":"","orcid":"https://orcid.org/0000-0003-2713-6867","institution":"Shandong University","correspondingAuthor":false,"prefix":"","firstName":"Lingbo","middleName":"","lastName":"Huang","suffix":""},{"id":343362925,"identity":"2d5d6726-d3de-4160-b21d-0e0a22b4053f","order_by":30,"name":"Doruk İriş","email":"","orcid":"https://orcid.org/0000-0001-6829-5691","institution":"Sogang University","correspondingAuthor":false,"prefix":"","firstName":"Doruk","middleName":"","lastName":"İriş","suffix":""},{"id":343362926,"identity":"c2baef7d-a9e1-4a86-85d4-034a4bb25f84","order_by":31,"name":"Hubert Janos Kiss","email":"","orcid":"","institution":"HUN-REN KRTK \u0026 Corvinus University of Budapest","correspondingAuthor":false,"prefix":"","firstName":"Hubert","middleName":"Janos","lastName":"Kiss","suffix":""},{"id":343362927,"identity":"cbe0577b-e562-42b0-abce-8e9860746196","order_by":32,"name":"Juliane Koch","email":"","orcid":"","institution":"University of Hamburg","correspondingAuthor":false,"prefix":"","firstName":"Juliane","middleName":"","lastName":"Koch","suffix":""},{"id":343362928,"identity":"5f559f84-0072-434e-8b4f-8ac37877726f","order_by":33,"name":"Jaromír Kovářík","email":"","orcid":"","institution":"University of the Basque Country (UPV/EHU) \u0026 University of West Bohemia","correspondingAuthor":false,"prefix":"","firstName":"Jaromír","middleName":"","lastName":"Kovářík","suffix":""},{"id":343362929,"identity":"59eb541a-09b6-47c6-b77c-fe48061bd631","order_by":34,"name":"Osbert Kwarteng","email":"","orcid":"","institution":"University of Ghana","correspondingAuthor":false,"prefix":"","firstName":"Osbert","middleName":"","lastName":"Kwarteng","suffix":""},{"id":343362930,"identity":"9d513113-ec77-41c2-b71e-6006b3ad16ad","order_by":35,"name":"Andreas Lange","email":"","orcid":"https://orcid.org/0000-0001-9518-5308","institution":"University of Hamburg","correspondingAuthor":false,"prefix":"","firstName":"Andreas","middleName":"","lastName":"Lange","suffix":""},{"id":343362931,"identity":"255891fa-2bf2-44ac-921d-9d0e5ae7a484","order_by":36,"name":"Martin Leites","email":"","orcid":"","institution":"Universidad de la República de Uruguay","correspondingAuthor":false,"prefix":"","firstName":"Martin","middleName":"","lastName":"Leites","suffix":""},{"id":343362932,"identity":"62580cbb-48c0-4f43-aabc-ff764d1547e8","order_by":37,"name":"Ho-Fung Leung","email":"","orcid":"","institution":"University of Arizona","correspondingAuthor":false,"prefix":"","firstName":"Ho-Fung","middleName":"","lastName":"Leung","suffix":""},{"id":343362933,"identity":"c76fdbc6-8d70-4487-aae9-9db275b5ae3e","order_by":38,"name":"Wooyoung Lim","email":"","orcid":"","institution":"Hong Kong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Wooyoung","middleName":"","lastName":"Lim","suffix":""},{"id":343362934,"identity":"1b61442c-432e-4a0d-ba39-1242a1bedaca","order_by":39,"name":"Meike Morren","email":"","orcid":"","institution":"Vrije Universiteit Amsterdam","correspondingAuthor":false,"prefix":"","firstName":"Meike","middleName":"","lastName":"Morren","suffix":""},{"id":343362935,"identity":"18e1e1ef-00e6-4453-9732-bf2e2d9e5ca6","order_by":40,"name":"Laila Nockur","email":"","orcid":"","institution":"Aarhus University","correspondingAuthor":false,"prefix":"","firstName":"Laila","middleName":"","lastName":"Nockur","suffix":""},{"id":343362936,"identity":"2dee5430-01d5-4b3a-b991-40468cb85b31","order_by":41,"name":"Charles Okyere","email":"","orcid":"","institution":"University of Ghana","correspondingAuthor":false,"prefix":"","firstName":"Charles","middleName":"","lastName":"Okyere","suffix":""},{"id":343362937,"identity":"0900e6c8-3956-4305-9dbc-3e40f5d7d460","order_by":42,"name":"Mayada Oudah","email":"","orcid":"https://orcid.org/0000-0003-3141-6159","institution":"New York University Abu Dhabi","correspondingAuthor":false,"prefix":"","firstName":"Mayada","middleName":"","lastName":"Oudah","suffix":""},{"id":343362938,"identity":"53861e86-5f2d-486d-9021-5712603365f9","order_by":43,"name":"Ali Ozkes","email":"","orcid":"https://orcid.org/0000-0002-8720-2494","institution":"SKEMA Business School, GREDEG, Université Côte d'Azur","correspondingAuthor":false,"prefix":"","firstName":"Ali","middleName":"","lastName":"Ozkes","suffix":""},{"id":343362939,"identity":"de7d7cda-5e25-413c-9600-97a1ecb49a6a","order_by":44,"name":"Lionel Page","email":"","orcid":"","institution":"University of Queensland","correspondingAuthor":false,"prefix":"","firstName":"Lionel","middleName":"","lastName":"Page","suffix":""},{"id":343362940,"identity":"d18c0c19-b004-44b6-af02-c933f9c1b20f","order_by":45,"name":"Junghyun Park","email":"","orcid":"","institution":"Seoul National University","correspondingAuthor":false,"prefix":"","firstName":"Junghyun","middleName":"","lastName":"Park","suffix":""},{"id":343362941,"identity":"987d56b4-049d-4874-a762-a0a3c6b14449","order_by":46,"name":"Stefan Pfattheicher","email":"","orcid":"https://orcid.org/0000-0002-0161-1570","institution":"Aarhus University","correspondingAuthor":false,"prefix":"","firstName":"Stefan","middleName":"","lastName":"Pfattheicher","suffix":""},{"id":343362942,"identity":"dc05ff92-6d6c-4370-afa9-d1ee91b86ce6","order_by":47,"name":"Antonios Proestakis","email":"","orcid":"","institution":"Helenic Mediterranean University","correspondingAuthor":false,"prefix":"","firstName":"Antonios","middleName":"","lastName":"Proestakis","suffix":""},{"id":343362943,"identity":"00e45519-5296-4f9e-ace4-ee046858f778","order_by":48,"name":"Carlos Ramos","email":"","orcid":"https://orcid.org/0000-0001-6744-4376","institution":"Instituto Tecnológico de Santo Domingo","correspondingAuthor":false,"prefix":"","firstName":"Carlos","middleName":"","lastName":"Ramos","suffix":""},{"id":343362944,"identity":"95e9a1cc-3a97-42c5-9bba-3b5c92b7c8e0","order_by":49,"name":"Mapi Ramos-Sosa","email":"","orcid":"","institution":"Universidad Loyola Andalucía","correspondingAuthor":false,"prefix":"","firstName":"Mapi","middleName":"","lastName":"Ramos-Sosa","suffix":""},{"id":343362945,"identity":"90f9e7d1-03fd-423a-9e30-efe78e291de4","order_by":50,"name":"Muhammad Saeed Ashraf","email":"","orcid":"","institution":"Jiangsu University","correspondingAuthor":false,"prefix":"","firstName":"Muhammad","middleName":"Saeed","lastName":"Ashraf","suffix":""},{"id":343362946,"identity":"f598989a-9569-4d32-93bb-f147fa3bb1c5","order_by":51,"name":"Muhammad Ryan Sanjaya","email":"","orcid":"https://orcid.org/0000-0001-9891-0835","institution":"Universitas Gadjah Mada","correspondingAuthor":false,"prefix":"","firstName":"Muhammad","middleName":"Ryan","lastName":"Sanjaya","suffix":""},{"id":343362947,"identity":"ed830c69-21e1-4fd7-9a54-40730ca7a6e6","order_by":52,"name":"Rene Schwaiger","email":"","orcid":"","institution":"University of Innsbruck","correspondingAuthor":false,"prefix":"","firstName":"Rene","middleName":"","lastName":"Schwaiger","suffix":""},{"id":343362948,"identity":"c2e0e906-39d7-427f-bcc3-92aeed0f5103","order_by":53,"name":"Omar Sene","email":"","orcid":"","institution":"Alioune Diop University of Bambey","correspondingAuthor":false,"prefix":"","firstName":"Omar","middleName":"","lastName":"Sene","suffix":""},{"id":343362949,"identity":"e22d5139-c687-491f-9998-9791b69c5c75","order_by":54,"name":"Fei Song","email":"","orcid":"","institution":"Toronto Metropolitan University","correspondingAuthor":false,"prefix":"","firstName":"Fei","middleName":"","lastName":"Song","suffix":""},{"id":343362950,"identity":"3f41d25a-e768-426b-b4d2-d4bb6b41362f","order_by":55,"name":"Sarah Spycher","email":"","orcid":"","institution":"University of Bologna","correspondingAuthor":false,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Spycher","suffix":""},{"id":343362951,"identity":"87954e0d-c5ba-461e-9800-a62c7a6209fd","order_by":56,"name":"Rostislav Staněk","email":"","orcid":"https://orcid.org/0000-0002-6310-4777","institution":"Masaryk University","correspondingAuthor":false,"prefix":"","firstName":"Rostislav","middleName":"","lastName":"Staněk","suffix":""},{"id":343362952,"identity":"ff32ba50-f659-4d6e-97b7-1174785b5af0","order_by":57,"name":"Norman Tanchingco","email":"","orcid":"","institution":"Ateneo de Manila University","correspondingAuthor":false,"prefix":"","firstName":"Norman","middleName":"","lastName":"Tanchingco","suffix":""},{"id":343362953,"identity":"08bf522f-8b14-4925-a5da-43a2c7940670","order_by":58,"name":"Alessandro Tavoni","email":"","orcid":"https://orcid.org/0000-0002-2057-5720","institution":"University of Bologna","correspondingAuthor":false,"prefix":"","firstName":"Alessandro","middleName":"","lastName":"Tavoni","suffix":""},{"id":343362954,"identity":"b16ea0fb-3f5f-47a5-8fa2-31dfa635f9cf","order_by":59,"name":"Vera te Velde","email":"","orcid":"","institution":"University of Queensland","correspondingAuthor":false,"prefix":"","firstName":"Vera","middleName":"te","lastName":"Velde","suffix":""},{"id":343362955,"identity":"dcd57596-a5b2-4033-9a36-d75ebab68a38","order_by":60,"name":"María José Vázquez-De Francisco","email":"","orcid":"","institution":"Fundación ETEA para el Desarrollo y la Cooperación","correspondingAuthor":false,"prefix":"","firstName":"María","middleName":"José Vázquez-De","lastName":"Francisco","suffix":""},{"id":343362956,"identity":"5daadaae-1b6e-42f6-a362-9f8832fc23d4","order_by":61,"name":"Martine Visser","email":"","orcid":"","institution":"University of Cape Town","correspondingAuthor":false,"prefix":"","firstName":"Martine","middleName":"","lastName":"Visser","suffix":""},{"id":343362957,"identity":"304d40b8-d005-409b-b020-473a2c089638","order_by":62,"name":"Joseph Tao-yi Wang","email":"","orcid":"https://orcid.org/0000-0001-6786-8099","institution":"National Taiwan University","correspondingAuthor":false,"prefix":"","firstName":"Joseph","middleName":"Tao-yi","lastName":"Wang","suffix":""},{"id":343362958,"identity":"02e05e9c-c8be-4656-a3fe-b0986f736375","order_by":63,"name":"Willy Wang","email":"","orcid":"","institution":"National Taiwan University","correspondingAuthor":false,"prefix":"","firstName":"Willy","middleName":"","lastName":"Wang","suffix":""},{"id":343362959,"identity":"19377b4e-abc5-44a9-a753-4aaf264b0e0f","order_by":64,"name":"Wei-Chien Weng","email":"","orcid":"https://orcid.org/0009-0006-4780-4737","institution":"National Taiwan University","correspondingAuthor":false,"prefix":"","firstName":"Wei-Chien","middleName":"","lastName":"Weng","suffix":""},{"id":343362960,"identity":"81ef8c61-eead-4544-b3ec-d7d9b7b7b3df","order_by":65,"name":"Katharina Werner","email":"","orcid":"https://orcid.org/0000-0002-7096-5890","institution":"University of Cologne","correspondingAuthor":false,"prefix":"","firstName":"Katharina","middleName":"","lastName":"Werner","suffix":""},{"id":343362961,"identity":"30124c55-3618-4ea6-bc72-acdf6ea786a2","order_by":66,"name":"Amanda Wijayanti","email":"","orcid":"","institution":"Universitas Gadjah Mada","correspondingAuthor":false,"prefix":"","firstName":"Amanda","middleName":"","lastName":"Wijayanti","suffix":""},{"id":343362962,"identity":"d54c397c-3b7f-4d1d-b274-43a9f54c3a27","order_by":67,"name":"Ralph Winkler","email":"","orcid":"","institution":"University of Bern","correspondingAuthor":false,"prefix":"","firstName":"Ralph","middleName":"","lastName":"Winkler","suffix":""},{"id":343362963,"identity":"cdea985f-bb70-4955-94b1-446455e1884b","order_by":68,"name":"John Wooders","email":"","orcid":"","institution":"Division of Social Science, New York University Abu Dhabi","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Wooders","suffix":""},{"id":343362964,"identity":"b7c06c37-01b0-47d8-b4cf-b64629dd020a","order_by":69,"name":"Li Ying","email":"","orcid":"","institution":"Key Laboratory of Behavioral Science, Institute of Psychology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Ying","suffix":""},{"id":343362965,"identity":"e0bc8335-3146-4d6a-a46c-93d733b4ce41","order_by":70,"name":"Wei Zhen","email":"","orcid":"","institution":"Institute of Psychology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Zhen","suffix":""},{"id":343362966,"identity":"fd67a6cd-3d78-456b-8588-86b3cb972f0e","order_by":71,"name":"Thomas Hills","email":"","orcid":"https://orcid.org/0000-0003-3842-2076","institution":"The University of Warwick","correspondingAuthor":false,"prefix":"","firstName":"Thomas","middleName":"","lastName":"Hills","suffix":""}],"badges":[],"createdAt":"2024-08-16 11:05:33","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4924647/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4924647/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":69069765,"identity":"aa187e12-594b-4167-aeb3-68c644610df4","added_by":"auto","created_at":"2024-11-15 09:52:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":761622,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4924647/v1/f87ed3bc-01db-4862-96db-9d5a6d964eea.pdf"},{"id":69069505,"identity":"9bc4657c-d7ae-47d6-b00b-4365ead79f49","added_by":"auto","created_at":"2024-11-15 09:44:45","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1571277,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"SupplementaryInformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4924647/v1/e1d22fbd3c196d5af0c54ac5.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Assessing private solutions to collective action problems in a 34-nation study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eHuman beings possess a unique ability to solve shared problems together\u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Across the world we have developed health systems to protect us from disease, school systems to educate our young, and transport systems to help us move around. These are all examples of public goods that generally benefit everyone in the community, regardless of who has contributed to their provision.\u003c/p\u003e \u003cp\u003eOn account of this property of \u0026lsquo;non-excludability\u0026rsquo;, public goods are vulnerable to free-riding\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Providing public goods therefore means dealing with the \u003cem\u003efree-rider problem\u003c/em\u003e \u0026ndash; a problem that received much attention in recent years\u003csup\u003e\u003cspan additionalcitationids=\"CR13 CR14 CR15\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. But it also means dealing with a second problem that has received far less attention during the same period. We call this the \u003cem\u003eprivate solution problem\u003c/em\u003e. This problem can occur whenever private solutions exist as substitutes for public solutions to shared problems\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Real-world examples abound: just as many people choose between private healthcare, education, and transport and their public alternatives, many governments choose between climate change adaptation and abatement\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. The adoption of private solutions, in turn, tends to undermine support for public solutions\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. The private solution problem is therefore different from the free-rider problem, reflecting a preference for self-reliance and self-protection rather than selfishness, but threatens public good provision in a similar manner.\u003c/p\u003e \u003cp\u003eDespite the increasing relevance of the private solution problem, particularly in the context of climate change, very little research has aimed to identify factors that may influence preferences for private rather than public solutions to shared problems. This pre-registered study (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://osf.io/3ubt9\u003c/span\u003e\u003cspan address=\"https://osf.io/3ubt9\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e empirically investigates three factors proposed as relevant: 1) people\u0026rsquo;s ability to pay for private solutions\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e; 2) the source of their wealth (luck vs. merit)\u003csup\u003e\u003cspan additionalcitationids=\"CR19 CR20 CR21 CR22\" citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e; and 3) their cultural and economic background\u003csup\u003e\u003cspan additionalcitationids=\"CR25 CR26\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. To achieve this we adapt a public good game known as the \u003cem\u003ecollective-risk social dilemma\u003c/em\u003e, which was originally designed to represent the collective action problem of limiting global warming to prevent catastrophic climate change\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. In the \u003cem\u003ecollective-risk social dilemma\u003c/em\u003e, participants are assigned to groups and given an endowment that they can contribute (or not) towards a group target sum over 10 successive rounds. If as a group they contribute enough to achieve this target within the 10 rounds, all group members get to keep their remaining funds (i.e., those not contributed towards the target). If they fail to achieve the group target, simulated catastrophic climate change occurs and each participant loses their remaining funds with a certain probability. The dilemma for participants is therefore essentially the same as the one faced by individuals and governments alike: the more they contribute to mitigation efforts, the more likely their group is to avoid catastrophic climate change, but the fewer resources they will have left.\u003c/p\u003e \u003cp\u003eThe main change we made to this standard version of the game was to introduce a private target as an alternative to the group target, as has previously been done in similar experimental games\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The private target represented a private good, just as the group target represented a public good. We refer to these goods as private and public solutions, respectively. Unlike the public solution in our game, the private solution was excludable: if a player achieved it, they alone got to keep their remaining funds. We assigned participants to groups of four, gave them an endowment of Monetary Units (MU), and asked them how much they would like to contribute towards the private solution (costing 60 MU) and public solution (costing 160 MU, or 40 MU per group member on average) over the course of 10 successive rounds. All contributions were spent, or non-refundable. To avoid losing their remaining funds at the end of the game, players needed to achieve either the public or private solution within the 10 rounds. If they did so, their remaining funds were converted into cash and sent as a bonus payment after the study. In this way, the game offered the same opportunities for cooperation, free-riding, and self-reliance that exist in many collective action problems today.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn our adapted game, groups were made up of two rich players (in red) who started the game with 120 MU and two poor players (in blue) who started the game with 80 MU. They either had to collectively contribute 160 MU to the public solution or individually contribute 60 MU to their private solution to avoid losing their remaining funds at the end of the game. Players could contribute up to a total of 20 MU towards the public solution and private solutions in each of 10 successive rounds, and all contributions were non-refundable. If the group achieved the (non-excludable) public solution, all members kept their remaining MU. Individuals who achieved their private solution also kept their remaining MU. These MU were converted into cash and paid as a bonus after the study.\u003c/p\u003e \u003cp\u003eWe made two further changes to the game in order to investigate whether people\u0026rsquo;s ability to pay for a private solution and the source of their wealth influenced their preferences for private vs. public solutions. The first was to introduce wealth inequalities within groups so that the private solution was more affordable for some players than others. Specifically, each group of four was made up of two rich players who started the game with 120 MU and two poor players who started with 80 MU. To compare rich and poor players\u0026rsquo; preferences for private solutions, we measured whether they adopted the private solution by contributing sufficiently towards their private target. To compare their preferences for public solutions, we measured the proportion of their wealth that they contributed towards the public solution. We compare proportional rather than absolute contributions because this provides a clearer understanding of who is contributing their fair share. The second change was to introduce different sources of wealth inequality (luck vs. merit) by randomly assigning groups to one of three treatments: one in which participants\u0026rsquo; endowments were determined by a lottery (the luck treatment); one in which they were determined by performance in an effort task (the merit treatment); and one in which they were determined either by a lottery or by task performance (the uncertain treatment). Before the game started, we also asked players what they thought would be fair for rich and poor players to contribute towards the public solution.\u003c/p\u003e \u003cp\u003e Lastly, we recruited 7,504 participants from universities in 34 countries to investigate whether people\u0026rsquo;s cultural and economic background influenced their preferences for private vs. public solutions. We measured the influence of cultural background by testing for relationships between mean private and public solution success rates in each country and scores on the most widely used cultural dimensions in social science research. These included factors recorded by Hofstede\u0026rsquo;s 6D model (individualism, power distance, masculinity, uncertainty avoidance, long-term orientation, and indulgence)\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e; the Inglehart-Welzel Cultural Map (survival vs. self-expression values, traditional vs. secular-rational values)\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e; the tightness vs. looseness index (measuring the strength of social norms)\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e; the World Values Survey (trust in others and acceptance of inequality)\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e; and the OECD (trust in government)\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. We also ran equivalent tests on participants\u0026rsquo; beliefs about the relative importance of luck and merit in life and about individual responsibility, elicited via questions drawn from the World Values Survey and presented at the end of the experiment (see Methods). In addition, we compared the variation in private solution adoption rates both within and between cultures by classifying countries according to the Inglehart-Welzel World Cultural Map\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. We followed a similar approach to measure the influence of economic background, testing for correlations between mean private and public solution success rates in each country and economic measures previously shown to influence behaviour in collective action problems\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e including individual subjective socioeconomic status\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e and World Bank measures of national GDP per capita, income inequality, and control of corruption (see Supplementary Information Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e for all measures and sources).\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n\u003ch2\u003ePrivate Solution Adoption\u003c/h2\u003e\n\u003cp\u003eRich players in our collective action game, who were better able to afford the private solution than poor players, adopted this solution 62.1% of the time. In contrast, poor players did so only 32.0% of the time (\u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e (1)\u0026thinsp;=\u0026thinsp;681.8, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003eBF\u003c/em\u003e\u003csub\u003e\u003cem\u003e10\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.1.291558e\u0026thinsp;+\u0026thinsp;149, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7,504, two-sided \u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e-tests). This represented decisive evidence that rich players were more likely to adopt the private solution than poor players. Figure\u0026nbsp;2a shows that this was the case in all 34 countries, while Fig.\u0026nbsp;2b shows that rich players contributed more towards the private solution than poor players in all 10 rounds of the game. This wealth effect remained strong after taking variability at the group and country levels into account (\u003cem\u003e\u0026szlig;\u003c/em\u003e = -1.61, \u003cem\u003e95% CI\u003c/em\u003e = [-1.80, -1.42], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7,504, generalised logistic mixed-effects model; see Supplementary Information Table S7). Calculating the odds ratios here revealed that when players were rich the odds of adopting the private solution were 85% higher compared to when they were poor; and when players were poor, the odds of private solution success were around 20% of those when players were rich.\u003c/p\u003e\n\u003cp\u003eThis preference for private solutions among rich players emerged independently of whether they earned their higher wealth through luck or merit (or an uncertain one of the two). In fact, overall private solution adoption rates among rich players were almost the same in the luck (62.0%), merit (62.7%), and uncertain (61.6%) treatments (\u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e (2)\u0026thinsp;=\u0026thinsp;0.3, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.865, \u003cem\u003eBF\u003c/em\u003e\u003csub\u003e\u003cem\u003e01\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;422.5; standard and Bayesian two-sided chi-squared tests) and we did not detect any differences between treatments within countries (see Supplementary Information Table S8). At the same time, private solution adoption rates among poor players were also similar both overall in the luck (32.1%), merit (30.9%), and uncertain (32.9%) treatments (\u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e (2)\u0026thinsp;=\u0026thinsp;1.2, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.544, \u003cem\u003eBF\u003c/em\u003e\u003csub\u003e\u003cem\u003e01\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;287.6) and between treatments within countries. For additional robustness we fitted a mixed-effects model on private solution adoption with fixed wealth and treatment effects as well as random group and country effects (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7,504). This model detected a poor wealth effect (\u003cem\u003e\u0026szlig;\u003c/em\u003e = -1.66, \u003cem\u003e95% CI\u003c/em\u003e = [-1.86, -1.47], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) but no luck treatment effect (\u003cem\u003e\u0026szlig;\u003c/em\u003e = -0.03, \u003cem\u003e95% CI\u003c/em\u003e = [-0.24, 0.18], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.777) or uncertain treatment effect (\u003cem\u003e\u0026szlig;\u003c/em\u003e = -0.03, \u003cem\u003e95% CI\u003c/em\u003e = [-0.24, 0.18], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.809) or treatment x wealth interactions (see Supplementary Information Table S9). These results provided further evidence that the source of wealth did not influence preferences for private solutions among either rich or poor players.\u003c/p\u003e\n\u003cp\u003ePreferences for private solutions also emerged independently of cultural or economic background. Kendall\u0026rsquo;s rank correlation tests did not reveal any relationship between private solution success rates in each country and scores on five of Hofstede\u0026rsquo;s six dimensions of national culture (individualism, \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.13, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.348, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27; power distance, \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.23, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.085, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27; masculinity, \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.04, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.786, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27; long-term orientation, \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.12, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.332, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;31; and indulgence, \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.04, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.746, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;31). The exception here was uncertainty avoidance (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.30, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.030, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27) although this relationship was not statistically significant after controlling for multiple testing (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.180, Bonferroni-Holm method). Equally, we did not detect a relationship between private solution success rates and survival vs. self-expression values (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.06, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.721, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;22); secular-rational vs. traditional values (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.11, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.417, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;29); tightness vs. looseness (\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.28, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.111, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;18); national-level beliefs about the acceptability of inequality (\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.22, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.105, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;28), the trustworthiness of others (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.04, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.752, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;28), or trust in government ((\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.05, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.837, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;12); or participant-level beliefs about the importance of luck and merit (\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.11, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.382, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;32) and individual responsibility (\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.18, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.142, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;32). Our evaluation of private solution adoption rates across cultural regions (as defined by the Inglehart-Welzel World Cultural Map) also revealed differences both between cultures (\u003cem\u003eF\u003c/em\u003e(7,26)\u0026thinsp;=\u0026thinsp;9.46, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) and within cultures (\u003cem\u003eF\u003c/em\u003e(26,1842)\u0026thinsp;=\u0026thinsp;3.97, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) with greater variability in the former than the latter (see Extended Data Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). Lastly, we did not detect a relationship between private solution success rates and either our individual economic measure of subjective socioeconomic status (\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.07, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.624, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34) or our national economic measures (GDP per capita, \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.21, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.091, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34; income inequality, \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.23, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.067, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34; control of corruption, \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.12, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.319, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34). In summary, our finding that rich players adopted the private solution almost twice as often as the poor was evident regardless of the source of their wealth or their cultural and economic background.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n\u003ch2\u003ePublic Solution Contributions\u003c/h2\u003e\n\u003cp\u003ePrivate solution adoption also undermined public solution contributions and provision. This was evident in the negative relationship between private and public solution success rates in each country (\u003cem\u003e\u0026rho;\u003c/em\u003e = -0.81, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) as shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea. This pattern emerged because individuals adopting private solutions generally stopped contributing towards public solutions, even though they may have had the funds available to do so, which made these solutions more difficult for their group to achieve (see Extended Fig.\u0026nbsp;2). This was particularly evident among rich players, who in total contributed a lower proportion of their wealth towards the public solution than poor players in every country except Ghana (see Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb). Overall, rich players contributed 31.3% of their wealth whereas poor players contributed 39.8% of theirs (\u003cem\u003eW\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5753438, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003eBF\u003c/em\u003e\u003csub\u003e\u003cem\u003e10\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;2.600205e\u0026thinsp;+\u0026thinsp;58; two-sided Mann-Whitney-Wilcoxon [MWW] tests). In absolute terms, this equated to mean contributions of 37.6 MU from rich players and 31.8 MU from poor players. This meant that rich players, unlike poor players, did not contribute towards the public solution in line with what participants judged to be fair (see Extended Data Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). This finding of lower proportional but higher absolute contributions from the rich is in line with several previous studies\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e35\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. It was further supported by the results of a mixed-effects model of players\u0026rsquo; contributions as a proportion of their wealth in each round that controlled for the effect of rounds, included random individual, group and country effects, allowing the effect of wealth to vary within countries (\u003cem\u003e\u0026szlig;\u003c/em\u003e = 2.71, \u003cem\u003e95% CI\u003c/em\u003e = [2.20, 3.20], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;60,270; see Supplementary Information Table S10). In addition, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ec shows that these differences in contributions from rich and poor players emerged from the very first round (Rich: 7.1% vs. Poor: 11.3%, \u003cem\u003eW\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5413782, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003eBF\u003c/em\u003e\u003csub\u003e\u003cem\u003e10\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;2.695783e\u0026thinsp;+\u0026thinsp;109, two-sided MWW tests) with poor players contributing more than rich players even in absolute terms (Rich: 8.5 MU vs. Poor: 9.0 MU, \u003cem\u003eW\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6746814, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.001, \u003cem\u003eBF\u003c/em\u003e\u003csub\u003e\u003cem\u003e10\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.35, two-sided MWW tests).\u003c/p\u003e\n\u003cp\u003eThis pattern of rich players contributing a lower proportion of their wealth than poor players also emerged independently of different sources of wealth. On average, rich players contributed a similar proportion of their wealth in each round in the luck (5.1%), merit (4.7%), and uncertain (4.6%) treatments. Likewise, poor players contributed a similar proportion of theirs in the luck (6.9%), merit (6.8%), and uncertain (6.8%) treatments (see Extended Data Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e for a comparison over the course of the 10 rounds). We tested for a treatment effect with a linear mixed-effects model of players\u0026rsquo; contributions as a proportion of their wealth in each round (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;60,270) that did not detect any treatment effects or treatment x wealth interactions (see Supplementary Information Table S11). These similarities in contribution behaviours between treatments help to explain why public solution provision was almost exactly the same among groups in the luck (63.6%), merit (60.8%) and uncertain (59.5%) treatments (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1,876, \u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e (2)\u0026thinsp;=\u0026thinsp;2.41, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.300, \u003cem\u003eBF\u003c/em\u003e\u003csub\u003e\u003cem\u003e01\u003c/em\u003e\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;72.7, two-sided standard and Bayesian \u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e-tests).\u003c/p\u003e\n\u003cp\u003eAs a result of rich players contributing a lower proportion of their wealth towards the public solution in all three treatments and in almost all countries, wealth inequality within groups tended to increase over time. At the start of the game, wealth inequality (as measured by the Gini coefficient) was the same in every group (0.1). At the end of the game, after taking losses resulting from failures to achieve either the private or public solution into account, wealth inequality increased to an average of 0.52 (\u003cem\u003eV\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1504068, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, one-sample Wilcoxon signed rank test). This level of inequality is equivalent to the current level of inequality in Brazil, which at present has the second highest level of inequality out of the G20 countries\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ed shows that this trend towards increasing inequality emerged in every country. It also remained similar (0.49) when losses from private and public solution failures were excluded (\u003cem\u003eV\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1624098, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001).\u003c/p\u003e\n\u003cp\u003eLastly, we investigated whether the variation in public solution provision across countries could be explained by our selection of cultural and economic factors by testing for correlations between these variables. We did not detect statistically significant relationships with any of Hofstede\u0026rsquo;s six dimensions of national culture (individualism: \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.24, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.083, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27; power distance: \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.20, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.144, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27; masculinity: \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.04, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.754, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27; uncertainty avoidance: \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.24, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.080, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;27; long-term orientation: \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.06, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.646, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;31; indulgence: \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.03, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.798, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;31). Nor did we detect a relationship between public solution provision in a country and its survival vs. self-expression values (\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.23, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.163, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;22); secular-rational vs. traditional values (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.02, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.905, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;29); tightness vs. looseness (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.139, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.425, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;18); beliefs about the acceptability of inequality (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.11, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.429, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;28) and the trustworthiness of others (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.04, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.782, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;28); or trust in government (\u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.17, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.450, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;12). Furthermore, public solution provision across countries did not correlate with participants\u0026rsquo; beliefs about the importance of luck and merit as determinants of life outcomes (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.18, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.142, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;32) or with any of our economic factors (individual subjective socioeconomic status: \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.11, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.441, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34; national GDP per capita: \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.17, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.173, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34; national income inequality: \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.16, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.189, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34; national control of corruption: \u003cem\u003e\u0026tau;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.13, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.291, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;34). Finally, public solution provision differed both between cultures (\u003cem\u003eF\u003c/em\u003e(7)\u0026thinsp;=\u0026thinsp;14.64, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) and within cultures (\u003cem\u003eF\u003c/em\u003e(26)\u0026thinsp;=\u0026thinsp;6.66, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) with greater variability in the former than the latter (see Extended Data Fig.\u0026nbsp;5). In summary, none of the above cultural or economic factors could help to explain differences in public solution provision rates between countries, in line with some previous research\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eThere was, however, one set of cultural values that did help to explain differences in public solution provision between countries. This was the strength of the belief within our samples that \u0026lsquo;people should take greater responsibility to provide for themselves\u0026rsquo;. We found that the average level of agreement with this statement among participants in each country was negatively associated with the rate of public solution provision in each country (\u003cem\u003e\u0026tau;\u003c/em\u003e = -0.29, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.014, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;32; see Extended Data Fig.\u0026nbsp;6a). Furthermore, the same pattern emerged at a group level: the stronger the beliefs in individual responsibility within a group, the less likely it was to achieve the public solution (\u003cem\u003e\u0026szlig;\u003c/em\u003e = -0.12, \u003cem\u003e95% CI\u003c/em\u003e = [-0.19, -0.04], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.001, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1,850, generalised logistic mixed-effects model; see Supplementary Information Table S12). In this way, beliefs about individual responsibility served as a universal pathway to public solution non-provision across countries. One possible explanation for this effect is that beliefs in individual responsibility predicted higher first round contributions to private solutions (\u003cem\u003e\u0026szlig;\u003c/em\u003e = -0.08, \u003cem\u003e95% CI\u003c/em\u003e = [0.02, 0.14], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.015, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7,445, linear mixed-effects model; see Supplementary Information Table S13). In turn, these higher private contributions were universally associated with lower rates of public solution provision (\u003cem\u003e\u0026szlig;\u003c/em\u003e = -0.06, \u003cem\u003e95% CI\u003c/em\u003e = [-0.08, -0.05], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1,876, see Extended Data Fig.\u0026nbsp;6b and Supplementary Information Table S14 for generalised logistic mixed-effects model details). In summary, people\u0026rsquo;s beliefs in individual responsibility were associated with higher first round contributions to private solutions and undermined public solution provision at both a national and a group level.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n\u003ch2\u003eUniversal Pathways for Public Solutions\u003c/h2\u003e\n\u003cp\u003eGiven the number of present-day collective action problems that span international borders, identifying factors that might undermine and underpin public solution provision across countries is an important empirical question. In the previous section we identified beliefs in individual responsibility and first round contributions to private solutions as universal pathways to public solution non-provision. In this section we outline two universal pathways to public solution provision.\u003c/p\u003e\n\u003cp\u003eThe first was first round contributions towards the public solution. As shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea, groups with higher public contributions towards the public solution in the first round were more likely to achieve the public solution across all countries (\u003cem\u003e\u0026szlig;\u003c/em\u003e = 0.11, \u003cem\u003e95% CI\u003c/em\u003e = [0.09, 0.12], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1,876, generalised logistic mixed-effects model; see Supplementary Information Table S15). Furthermore, comparing the distributions of these first-round contributions between groups that provided vs. did not provide the public solution revealed that there was a 69% chance that a randomly selected successful (vs. unsuccessful) group would exhibit higher contributions in the first round (\u003cem\u003eW\u003c/em\u003e\u0026thinsp;=\u0026thinsp;704291, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, two-sided MWW test; rank-biserial correlation \u003cem\u003er\u003c/em\u003e\u003csub\u003e\u003cem\u003eb\u003c/em\u003e\u003c/sub\u003e = 0.58; Cliff\u0026rsquo;s delta \u003cem\u003ed\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.69). In short, early contributions towards public solutions presaged public solution provision across countries.\u003c/p\u003e\n\u003cp\u003eThe second universal pathway to public solution provision was conditional cooperation within groups. We measured conditional cooperation by evaluating the relationship between a player\u0026rsquo;s contribution to the public solution in the present round and the average public solution contribution from their group members in the previous round\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb shows that levels of conditional cooperation were significantly higher in groups that successfully achieved the public solution compared with groups that did not. This finding was supported by a linear mixed-effects model that revealed a significant average group contribution on previous round x public solution provision interaction (\u003cem\u003e\u0026szlig;\u003c/em\u003e = 0.23, \u003cem\u003e95% CI\u003c/em\u003e = [0.20, 0.25], \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;52,892; see Supplementary Information Table S16). In other words, players\u0026rsquo; responses to contributions from their fellow group members universally underpinned public solution provision.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n\u003ch2\u003eThe Private Solution Problem\u003c/h2\u003e\n\u003cp\u003eWe reported that rich players in our collective action game adopted the private solution almost twice as often as poor players. They did so regardless of both the source of their wealth (luck vs. merit vs. luck or merit) and their cultural and economic background. Furthermore, rich players\u0026rsquo; preferences for private solutions saw them contribute proportionally less than the poor towards public solutions, also independently of both the source of their wealth and their cultural and economic background. Wealth inequality within groups increased as a consequence. These results point towards the possibility of a universal private solution problem whereby the existence and widespread adoption of private solutions by those who can afford them creates an externality that exacerbates wealth inequality within groups and undermines public solution provision\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWhile we identified a universal preference for self-reliance among rich players, we also identified two universal pathways to public solution provision that may be relevant for global collective action problems such as climate change mitigation. The first was via early contributions to the public solution: countries in which contributions in the first round were higher (e.g., France, Denmark, and Colombia) were more likely to achieve the public solution than places in which they were lower (e.g., UK, Ghana, and the Dominican Republic) in line with findings from previous studies\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. The second was conditional cooperation: groups that succeeded in achieving the public solution were generally made up of players who responded more reciprocally to the contributions of others. Generalising these findings would suggest that early commitments, as well as reciprocal responses to these commitments, may disproportionately influence group outcomes.\u003c/p\u003e\n\u003cp\u003eOverall, we observed few cultural or economic explanations of preferences for private and public solutions across our samples \u0026ndash; even though our participants came from a wide range of Western, Educated, Industrialised, Rich and Democratic (WEIRD) and non-WEIRD countries\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. While we detected greater variability in preferences for private and public solutions between cultures than within cultures\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, we were mostly unable to explain these differences from either a cultural or economic perspective. This may be on account of our recruitment of university student participants, who were unlikely to be nationally representative and were perhaps more likely to behave similarly to their country\u0026rsquo;s elites rather than its ordinary citizens\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. Further research might therefore explore how universities may shape people\u0026rsquo;s approaches to collective decision-making in ways that do not reflect their local culture.\u003c/p\u003e\n\u003cp\u003eLastly, we propose that private solutions are more costly than public solutions in ways that extend far beyond economic inefficiency. What we mean is that global collective action problems never exist in spatial or temporal isolation \u0026ndash; they are invariably just one of several contemporaneous problems that all extend into the future. And not all of them can be solved privately.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003eA total of 7,504 participants completed this study, an average of 55 groups of four in each country (see Supplementary Information for further participant details). We recruited participants from student pools accessible to local universities in each country between July 2022 and March 2024. In universities with more diverse student pools, we aimed to predominantly recruit local students by restricting participation to national citizens or by translating the experiment into the local language. We recruited 20 groups in each treatment in each country, or 15 groups where this was not possible. All participants provided informed consent prior to taking part in the experiment and were free to withdraw at any time. The experiment was approved by ethics committees in every country where this was required by local universities. To ensure consistent implementation across countries, local researchers followed the same set of instructions for the experimental procedure (see Supplementary Information section 1).\u003c/p\u003e \u003cp\u003eParticipants were randomly assigned to groups of four, which in turn were randomly assigned to the luck, merit, or uncertain treatments. At the end of the experiment, participants\u0026rsquo; final earnings (in MU) were converted to cash and added to their participation fee. Conversion rates and participation fees were adjusted for purchasing power parity (PPP) across countries (see Supplementary Information section 1). On average, participants earned around GBP \u0026pound;5.20 (\u0026pound;7.75 PPP equivalent). The experiment generally lasted 30\u0026ndash;45 minutes, depending on the location and implementation method (in-person vs. online).\u003c/p\u003e \u003cp\u003eThe vast majority of participants were undergraduates: 83.6% were aged 18\u0026ndash;24; 14.7% were aged 25\u0026ndash;34; and 1.6% were aged 35 and over. They were recruited from a range of academic disciplines, with the greatest representation from Economics (16%). 81.7% attended public universities, while 16% attended private universities. Around half (44.4%) of participants completed the experiment in a lab setting; the remainder (55.5%) completed the experiment in an online setting. Overall, 51.9% of our sample identified as female and 46.4% identified as male; 1.7% identified as non-binary or another gender (see Supplementary Information Table S4 for breakdown by country).\u003c/p\u003e \u003cp\u003eThe entire study was programmed in oTree\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Instructions for participants were translated into local languages where appropriate. Translations can be found in the experimental code, which can be accessed via \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://osf.io/3ubt9\u003c/span\u003e\u003cspan address=\"https://osf.io/3ubt9\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eExperimental Design\u003c/h2\u003e \u003cp\u003e After reading a participant information sheet and completing a consent form, participants provided their basic demographic data (age and gender). On the next page, they were informed that the study was made up of two sections: 1) an arithmetic task to be completed individually; and 2) a problem-solving game to be completed in groups of four and played with Monetary Units (MU) that would be converted into cash at the end of the study.\u003c/p\u003e \u003cp\u003eParticipants were then told that two players in each group would start the game with 120 MU while the other two players in 80 MU. Participants in the merit treatment were informed that the two players in the group who scored highest in the mental arithmetic task would start with 120 MU and the two players who scored lowest would start with 80 MU. In the luck treatment, participants were informed that a lottery would determine which two players would start the game with 120 MU and which two players would start with 80 MU. In other words, their performance in the mental arithmetic task had no bearing on their endowment. To incentivise effort here for consistency across treatments, we informed these participants that the highest-scoring player in each group would receive a bonus payment of \u0026pound;1 (or approximate PPP equivalent) at the very end of the study. Participants in the uncertain treatment were told that either their performance in the arithmetic task or a lottery would determine their endowments. In practice, luck was chosen for two players in each group, one of whom became rich and one of whom became poor. For the other two players in the group, the one who scored highest in the effort task became rich and the one who scored lowest became poor. We included this treatment because it represents the real world, in which the cause of wealth is always an uncertain combination of luck and merit. It also enabled us to test for self-serving bias, which would be evident if rich players in the uncertain treatment behaved similarly to rich players in the merit treatment and if poor players in the uncertain treatment behaved similarly to poor players in the luck treatment.\u003c/p\u003e \u003cp\u003eThe mental arithmetic task was one that has been used by several researchers in the past\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. Participants were asked to add up sequences of five random numbers from 1-100 (e.g., 81\u0026thinsp;+\u0026thinsp;25\u0026thinsp;+\u0026thinsp;2\u0026thinsp;+\u0026thinsp;52\u0026thinsp;+\u0026thinsp;13 = ?) for five minutes. They were awarded one point per correct answer. After they had finished the task they were informed whether they had become rich or poor and reminded that this was either determined by their performance in the task (in the merit treatment) or by a lottery (in the luck treatment).\u003c/p\u003e \u003cp\u003eParticipants were then provided with more detailed instructions about the game (see Supplementary Information for materials) and then had to correctly answer four comprehension questions in order to proceed to the game. The quiz also included two questions about what participants judged to be fair contributions from rich and poor players. Once all four participants within a group had completed the pre-game quiz, the first round of the game commenced. At the start of each round participants were told the round number, how many MU they had left, how much they had contributed to their private solution, how much more they needed to contribute to achieve the private solution, how much the group as a whole had contributed to the public solution, and how much more the group needed to contribute to achieve the public solution. They were asked how much they would like to contribute (up to 20 MU in total) towards the private and public solutions. The round concluded once all four players had made their contribution decisions.\u003c/p\u003e \u003cp\u003eThe cost of the public solution was 160 MU, or on average 40 MU per player, while the cost of the private solution was 60 MU. We chose this cost ratio of 1:1.5 because previous authors\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e found that this creates uncertainty among participants regarding the optimal decision. In other words, if the private solution is any less expensive then people will generally choose that over the public solution; if it is any more expensive, then people will generally choose the public solution. All funds that participants contributed to either the private or public solution were spent (i.e., non-refundable). If they failed to solve the collective action problem either privately or publicly they lost all their remaining MU.\u003c/p\u003e \u003cp\u003eOnce participants had completed the game they responded to two statements drawn from the World Values Survey. We asked them to place their views on a scale from 1\u0026ndash;10, where 1 indicated complete agreement with a statement on the left and 10 indicated complete agreement with a statement on the right. In the first statement about merit and luck, the statement on the left (1) was \u0026lsquo;In the long run, hard work usually brings a better life\u0026rsquo;; the statement on the right (10) was \u0026lsquo;Hard work doesn\u0026rsquo;t generally bring success \u0026ndash; it\u0026rsquo;s more a matter of luck and connections\u0026rsquo;. In the second statement about individual responsibility, the statement on the left (1) was \u0026lsquo;The government should take more responsibility to ensure that everyone is provided for\u0026rsquo;; the statement on the right (10) was \u0026lsquo;People should take more responsibility to provide for themselves\u0026rsquo;. After responding to these statements, participants in most countries reported their household income, field of study, and perceived position on the MacArthur Scale of Subjective Social Status\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Finally, they were debriefed and thanked for their participation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eData Analysis\u003c/h2\u003e \u003cp\u003eData analysis was conducted in R. Our hypotheses relating to private and public solution preferences were pre-registered at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://osf.io/3ubt9\u003c/span\u003e\u003cspan address=\"https://osf.io/3ubt9\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (we added hypotheses relating to cultural and economic factors after the number of international labs and participants involved increased). To test these hypotheses we mainly assessed how private solution success rates, contributions to the public solution, and public solution success rates compared between wealth levels, treatment levels, and countries. To compare private and public solution success rates we used standard and Bayesian chi-squared tests. To compare contributions to the public solution we used a combination of MWW tests and mixed-effects models. The reason for using MWW tests was that the distribution of contribution behaviour deviated from normality. The reason for using mixed-effects models was that our data had a nested structure as rounds were nested in individuals, individuals were nested in groups, and groups were nested in countries\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. Full details of all statistical tests and models can be found in our Supplementary Information and coding script (see below).\u003c/p\u003e \u003c/div\u003e "},{"header":"Declarations","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eData and Code Availability\u003c/h2\u003e \u003cp\u003eAll data will be uploaded to \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://osf.io/zg8nj/\u003c/span\u003e\u003cspan address=\"https://osf.io/zg8nj/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e upon acceptance for publication. All code used for data analysis is available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://osf.io/zg8nj/\u003c/span\u003e\u003cspan address=\"https://osf.io/zg8nj/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eOlson, M. The Logic of Collective Action: Public Goods and the Theory of Groups. (Harvard University Press, 1965).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHardin, G. The tragedy of the commons. Science 162, 1243\u0026ndash;1248 (1968).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOstrom, E. Collective action and the evolution of social norms. J. Econ. Perspect. 14, 137\u0026ndash;158 (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarrett, S. Why Cooperate?: The Incentive to Supply Global Public Goods. (Oxford University Press, 2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarattini, S., Levin, S. \u0026amp; Tavoni, A. Cooperation in the climate commons. Rev. Environ. Econ. Policy (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGross, J. \u0026amp; De Dreu, C. K. Individual solutions to shared problems create a modern tragedy of the commons. Sci. Adv. 5, eaau7296 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGross, J., Veistola, S., De Dreu, C. K. \u0026amp; Van Dijk, E. Self-reliance crowds out group cooperation and increases wealth inequality. Nat. Commun. 11, 5161 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGross, J. \u0026amp; B\u0026ouml;hm, R. Voluntary restrictions on self-reliance increase cooperation and mitigate wealth inequality. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 117, 29202\u0026ndash;29211 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenrich, N. \u0026amp; Henrich, J. P. Why Humans Cooperate: A Cultural and Evolutionary Explanation. (Oxford University Press, 2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBowles, S. \u0026amp; Gintis, H. A cooperative species: Human reciprocity and its evolution. in A Cooperative Species (Princeton University Press, 2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNowak, M. \u0026amp; Highfield, R. Supercooperators: Altruism, Evolution, and Why We Need Each Other to Succeed. (Simon \u0026amp; Schuster, 2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrubaker, E. R. Free ride, free revelation, or golden rule? J. Law Econ. 18, 147\u0026ndash;161 (1975).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNordhaus, W. Climate clubs: Overcoming free-riding in international climate policy. Am. Econ. Rev. 105, 1339\u0026ndash;1370 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarclay, P. \u0026amp; Willer, R. Partner choice creates competitive altruism in humans. \u003cem\u003eProc. R. Soc. B Biol. Sci.\u003c/em\u003e 274, 749\u0026ndash;753 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eB\u0026oacute;, P. D. Cooperation under the shadow of the future: experimental evidence from infinitely repeated games. Am. Econ. Rev. 95, 1591\u0026ndash;1604 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFehr, E. \u0026amp; G\u0026auml;chter, S. Altruistic punishment in humans. Nature 415, 137\u0026ndash;140 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarrett, S. Dikes versus windmills: Climate treaties and adaptation. Clim. Change Econ. 11, 2040005 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlesina, A. F., Glaeser, E. L. \u0026amp; Sacerdote, B. Why Doesn\u0026rsquo;t the US Have a European-Style Welfare System? (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCappelen, A. W., S\u0026oslash;rensen, E. \u0026Oslash;. \u0026amp; Tungodden, B. Responsibility for what? Fairness and individual responsibility. Eur. Econ. Rev. 54, 429\u0026ndash;441 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMijs, J. J. The paradox of inequality: Income inequality and belief in meritocracy go hand in hand. Socio-Econ. Rev. 19, 7\u0026ndash;35 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMounk, Y. The age of responsibility: Luck, choice, and the welfare state. Harv. Univ. Press (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrank, R. H. Success and Luck. (Princeton University Press, 2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSandel, M. J. The Tyranny of Merit: What\u0026rsquo;s Become of the Common Good? (Penguin UK, 2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHofstede, G. Dimensionalizing cultures: The Hofstede model in context. Online Read. Psychol. Cult. 2, 8 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGelfand, M. J. \u003cem\u003eet al.\u003c/em\u003e Differences between tight and loose cultures: A 33-nation study. Science 332, 1100\u0026ndash;1104 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlesina, A. \u0026amp; La Ferrara, E. Participation in heterogeneous communities. Q. J. Econ. 115, 847\u0026ndash;904 (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBardhan, P., Ghatak, M. \u0026amp; Karaivanov, A. Wealth inequality and collective action. J. Public Econ. 91, 1843\u0026ndash;1874 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMilinski, M., Sommerfeld, R. D., Krambeck, H. J., Reed, F. A. \u0026amp; Marotzke, J. The collective-risk social dilemma and the prevention of simulated dangerous climate change. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 105, 2291\u0026ndash;2294 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInglehart, R. \u0026amp; Welzel, C. Value Change and the Persistence of Cultural Traditions. in Modernization, Cultural Change, and Democracy 48\u0026ndash;76 (Cambridge University Press, 2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaerpfer, C. et al. World Values Survey: Round Seven \u0026ndash; Country-Pooled Datafile. (JD Systems Institute \u0026amp; WVSA Secretariat, 2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOECD. Trust in Government. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.oecd.org/en/data/indicators/trust-in-government.html\u003c/span\u003e\u003cspan address=\"https://www.oecd.org/en/data/indicators/trust-in-government.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026auml;chter, S., Herrmann, B. \u0026amp; Th\u0026ouml;ni, C. Culture and cooperation. Philos. Trans. R. Soc. B Biol. Sci. 365, 2651\u0026ndash;2661 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInglehart, R. \u0026amp; Welzel, C. The Inglehart\u0026ndash;Welzel World Cultural Map\u0026mdash;World Values Survey 7. (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAdler, N. E., Epel, E. S., Castellazzo, G. \u0026amp; Ickovics, J. R. Relationship of subjective and objective social status with psychological and physiological functioning: Preliminary data in healthy, White women. Health Psychol. 19, 586 (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTavoni, A., Dannenberg, A., Kallis, G. \u0026amp; L\u0026ouml;schel, A. Inequality, communication, and the avoidance of disastrous climate change in a public goods game. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 108, 11825\u0026ndash;11829 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVicens, J. \u003cem\u003eet al.\u003c/em\u003e Resource heterogeneity leads to unjust effort distribution in climate change mitigation. \u003cem\u003ePloS One\u003c/em\u003e 13, Article e0204369 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalthouse, E., Pilgrim, C., Sgroi, D. \u0026amp; Hills, T. T. When fairness is not enough: The disproportionate contributions of the poor in a collective action problem. J. Exp. Psychol. Gen. (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWorld Bank. GINI Index (World Bank Estimate). (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRomano, A., Balliet, D., Yamagishi, T. \u0026amp; Liu, J. H. Parochial trust and cooperation across 17 societies. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e 114, 12702\u0026ndash;12707 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026auml;chter, S. Conditional Cooperation: Behavioral Regularities from the Lab and the Field and Their Policy Implications. in Psychology and Economics: A Promising New Cross-Disciplinary Field (eds. Frey, B. S. \u0026amp; Stutzer, A.) (The MIT Press, Cambridge, 2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFischbacher, U. \u0026amp; G\u0026auml;chter, S. Social preferences, beliefs, and the dynamics of free riding in public goods experiments. Am. Econ. Rev. 100, 541\u0026ndash;556 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMori, R., Hanaki, N. \u0026amp; Kameda, T. An outside individual option increases optimism and facilitates collaboration when groups form flexibly. Nat. Commun. 15, 5520 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026uuml;rdal, M. Y., Torul, O. \u0026amp; Yahşi, M. The seeds of success: the pivotal role of first round cooperation in public goods games. J. Econ. Sci. Assoc. 1\u0026ndash;23 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenrich, J., Heine, S. J. \u0026amp; Norenzayan, A. The weirdest people in the world? Behav. Brain Sci. 33, 61\u0026ndash;83 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHugh-Jones, D. Honesty, beliefs about honesty, and economic growth in 15 countries. J. Econ. Behav. Organ. 127, 99\u0026ndash;114 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFisman, R., Jakiela, P., Kariv, S. \u0026amp; Markovits, D. The distributional preferences of an elite. Science 349, aab0096 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, D. L., Schonger, M. \u0026amp; Wickens, C. oTree\u0026mdash;An open-source platform for laboratory, online, and field experiments. J. Behav. Exp. Finance 9, 88\u0026ndash;97 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOswald, A. J., Proto, E. \u0026amp; Sgroi, D. Happiness and productivity. J. Labor Econ. 33, 789\u0026ndash;822 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAarts, E., Verhage, M., Veenvliet, J. V., Dolan, C. V. \u0026amp; Van Der Sluis, S. A solution to dependency: using multilevel analysis to accommodate nested data. Nat. Neurosci. 17, 491\u0026ndash;496 (2014).\u003c/span\u003e\u003c/li\u003e \u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4924647/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4924647/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCollective action problems emerge when individual incentives and group interests are misaligned, as in the case of climate change\u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3 CR4\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Individuals involved in collective action problems are often considered to have two options: contribute towards a public solution or free-ride. But they might also choose a third option of investing in a private solution such as local climate change adaptation\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Here we introduce a collective action game featuring wealth inequality caused by luck or merit and both public and private solutions with participants from 34 countries. We show that the joint existence of wealth inequality and private solutions has a consistent effect across countries: participants endowed with higher income choose the private solution almost twice as often as those endowed with lower income; and this finding cannot be explained by different sources of wealth (luck vs. merit) or by cultural or economic factors. We also show that preferences for private solutions undermine support for public solutions, resulting in wealth inequality increasing in every country. In contrast, we identify two universal pathways to successful public solution provision: early contributions to public solutions and conditional cooperation. Our findings highlight the ubiquity of the \u0026lsquo;private solution problem\u0026rsquo; and its potential consequences for global collective action problems.\u003c/p\u003e","manuscriptTitle":"Assessing private solutions to collective action problems in a 34-nation study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-15 09:44:40","doi":"10.21203/rs.3.rs-4924647/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-human-behaviour","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nathumbehav","sideBox":"Learn more about [Nature Human Behaviour](http://www.nature.com/nathumbehav/)","snPcode":"","submissionUrl":"","title":"Nature Human Behaviour","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"f8b1b07f-0f89-4efb-906b-c9c72bdb7f83","owner":[],"postedDate":"November 15th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":36383271,"name":"Earth and environmental sciences/Environmental social sciences/Psychology and behaviour"},{"id":36383272,"name":"Scientific community and society/Social sciences/Climate change/Climate-change mitigation"},{"id":36383273,"name":"Scientific community and society/Social sciences/Economics"},{"id":36383274,"name":"Scientific community and society/Social sciences/Psychology/Human behaviour"},{"id":36383275,"name":"Scientific community and society/Social sciences/Decision making"}],"tags":[],"updatedAt":"2024-11-15T09:44:41+00:00","versionOfRecord":[],"versionCreatedAt":"2024-11-15 09:44:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4924647","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4924647","identity":"rs-4924647","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}