Does culture matter in economic behavior? Ultimatum Game Bargaining among the Machiguenga of the Peruvian Amazon --- Henrich (2000)

Transcription

1 Does culture matter in economic behavior? Ultimatum Game Bargaining among the Machiguenga of the Peruvian Amazon --- Henrich (2000) Economists typically assume that humans everywhere deploy same cognitive machinery for making economic decisions Ø will respond similarly when faced with comparable economic circumstances Compare the Machiguenga with Los Angeles Ø economic reasoning and economic decisions heavily influenced by cultural differences Ø socially transmitted rules about how to behave in certain circumstances that may vary by group depending on cultural evolutionary trajectories

2 Simple bargaining game Ultimatum Bargaining Game (UG) Two players are allotted a sum of money (termed the stakes ) First player proposer offers a portion of the total sum to a second person responder Responder can either accept or reject the proposer s offer If responder accepts, she (or he) receives amount offered and proposer receives remainder (initial sum minus the offer) If the responder rejects the offer - neither player receives anything

3 Ultimatum Bargaining Game (UG) Players typically receive payments in real money Usually remain anonymous to other players (not to experimenters) experimental economists have extensively manipulated both of these variables Here players were anonymous to other players (but not the experimenter) and stakes were large relative to previous UG experiments

4 Previous UG experiments Game behavior deviates from predictions of positive game theory (under standard preferences) Positive game theory (subgame-perfect equilibrium and money maximization) Ø unambiguously predicts that proposers should offer the smallest nonzero amount possible Ø responders should always accept If $20 is allocated to a pair of players with the smallest unit being $1 -- then proposer should offer $1 to the responder and keep $19 Responders should always accept any nonzero offer

5 Experimental subjects from industrial societies Modal offer is typically 50 percent Mean offer averages between 40 and 50 percent Responders usually accept average offers Often reject offers lower than 20 percent Experimental economists have systematically studied influence of various factors on game s results: o stake size; degree of anonymity; context à each has little or no effect on players behavior

6 Experimental subjects from industrial societies People from many parts of the world (Europe, Asia, and North America) behave quite similarly in the UG o Ljubljana (Slovenia) o Pittsburgh o Tokyo o Yogyakarta (Indonesia) o Los Angeles Robust pattern of UG behavior led many economists to develop new models posit that humans possess: innate taste for costly punishment innate sense of fairness some combination of both

7 Economically independent at the family level, the people of Machiguenga possess little social hierarchy or political complexity, and most sharing and exchange occurs within extended kin circles.

8 The Machiguenga Traditionally lived in mobile single-family units in small extendedfamily hamlets scattered throughout tropical forests of SE Peruvian Amazon Subsisting on combination of hunting, fishing, gathering, and manioc horticulture (starchy root crop) Economically independent at the family level Possess little social hierarchy or political complexity Most sharing and exchange occurs within extended kin circles Cooperation above the family level is almost unknown (except cooperative fish poisoning)

9 The Machiguenga During last 30 years: missionaries, markets, government-administered schools à sedentized and centralized Machiguenga into a number of villages Machiguenga gradually intensified their reliance on horticultural products (manioc) Machiguenga farmers have begun to: o produce cash crops (coffee and cocoa) o raise domesticated animals o participate in limited wage labor (logging or oil companies)

10 The Machiguenga Although now live in small communities of 300 people Remain primarily a family-level society Families fully produce for their own needs (food, clothing, etc.) Do not rely on institutions or other families for their social or economic welfare Constant demand for market items such as machetes, salt, sugar, and steel axes Anonymous transactions are almost unknown

11

12 Methodology Modify the typical experimental procedures Gathered 12 men together (aged 18-30) under the auspices of playing a fun game for money Joe explained game to group in Spanish (second language) Bilingual school teacher (an educated Machiguenga) reexplained the game in Machiguenga language Display the money that we would be using to make payments

13 Methodology Each participant entered Joe s house (the guest hut) individually Explained game a third time, and asked a number of hypothetical, practice questions -- test participants comprehension of game Reexplained parts of the game as necessary Often numerous examples were necessary to make game fully understood After the individual confidently answered at least two hypothetical questions correctly Ø Joe would submit actual question with a pile of 20 soles in view

14 Methodology The following day began seeking randomly selected individuals to play the game Most people had already heard of the game and were eager to play Joe privately explained game to each individual (usually in his or her house) Ran through the same testing procedure as the previous day Several people were rejected -- after 301 minutes of explanation o could not understand the game o could not correctly answer hypothetical questions Initial 12 players were volunteers -- next 30 players selected at random

15 Methodology Players randomly assigned to their roles -- proposer or responder Told their anonymous partner was another member of their community (Camisea) Camisea: o contains 260 people from 36 households o about 70 adults o roughly divided into 12 extended families The player pool contains females and 28 males

16 LA Experiment Repeated nearly identical version of Machiguenga UG with UCLA graduate students Machiguenga s 20- soles stake equals 2.3 days pay UCLA stake at $160 = 2.3 days pay for a graduate student working Limited UCLA experiment to graduate students in the Department of Anthropology (also a community of about 70 adults) UCLA subjects o received game instructions (one-on-one situation with experimenter) o answer hypothetical test questions before actually playing o Joe primary experimenter

17 TABLE 1-SUMMARY OF CROSS-CULTURAL ULTIMATUM GAME DATA AND STATISTICAL TESTS Los Yogyakartab Data factors Angeles Machiguenga (high-stakes) Yogyakartab Tucsonc Pittsburgha Tokyoa Jerusalema Number of pairs Stake size $160 $160 $ $10-15 $10 $10 $10 $10 Mean Mode Standard deviation Rejection frequency Rej < 20 percent 0/0 1/10 = 0.1 0/0 9/15 = 0.6 0/1 2/4 = 0.5 5/7 = 0.71 EST p (LA)d EST p (Mach)d MW p (Mach)e 2.64E E E E EST p (Pitt)d

18 Results Machiguenga data differ substantially from usual patterns Machiguenga proposers offered only 26 percent Machiguenga responders more likely to accept: o offers less than 20 percent o nearly half of the total offers

19 Discussions Machiguenga made clear they would always accept any money o Rather than viewing themselves as being cheated by proposer o Feel it was just bad luck that they were responders not proposers o Responders did not expect a balanced offer Los Angeles players o claimed they would reject unfair offers (< 25%) o proposers said thinking of offering less but figured would be rejected o offered 50 percent to be fair Few Machiguenga proposers who offered 50 percent had greater exposure and dealings with Westerners -- North American evangelical missionaries

20 Conclusion First test of the UG s robustness outside of industrialized countries Machiguenga UG suggests that culturally transmitted behavioral variation may substantially affect decision making Machiguenga proposers seem to possess little or no sense of obligation to provide an equal share to responders Responders had little or no expectation of receiving an equal share nor any desire to punish unequal divisions Modal offer of 15 percent seemed quite fair to most Machiguenga

21 Further Questions Where do people get their rules, expectations, or notions of fairness from? Why do these rules, expectations, and notions seem to vary among groups of people? How much can these varying rules, expectations, and notions affect real economic behavior?

22 One Approach Treat humans as social animals who acquire many of their behavioral rules, rule calibrations, beliefs, and practices from other humans via social learning Specify the cognitive apparatuses, imitation rules, or interactional processes that maintain similarities within groups How important social learning is for economic behavior?

23 In Search of Homo Economicus: Behavioral Experiments in 15 Small-Scale Societies Henrich et. al. (2001) Economists' canonical assumption Ø individuals are entirely self-interested Many experimental subjects appear o to care about fairness and reciprocity o willing to change distribution of material outcomes at personal cost o willing to reward those who act in a cooperative manner while punishing those who do not even when these actions are costly to the individual

24 Questions Deviations from canonical model have important consequences for a wide range of economic phenomena o optimal design of institutions and contracts o allocation of property rights o conditions for successful collective action o analysis of incomplete contracts o persistence of noncompetitive wage premia Are deviations from canonical model evidence of universal patterns of behavior? Do the individual's economic and social environments shape behavior? Which economic and social conditions are involved?

25 Role of Culture Are there cultures that approximate the canonical account of selfregarding behavior? Much research cannot answer such questions because virtually all subjects have been university students Ø All WEIRD = Western, Educated, Industrialized, Rich, and Democratic societies

26 Large Cross-Cultural Study Large cross-cultural study of behavior in ultimatum, public good, and dictator games Twelve experienced field researchers, working in 12 countries on five continents, recruited subjects from 15 small-scale societies exhibiting a wide variety of economic and cultural conditions Sample consists of three foraging societies, six that practice slash-andburn horticulture, four nomadic herding groups, and three sedentary small-scale agriculturalist societies

27 Dolgan/Nganasan U.S., rural Missouri Au Sanquianga Samburu Sursurunga Accra Orma Maragoli Yasawa (Fiji) Shuar Isanga village Hadza Gusii Tsimane

28 RESEARCH SITES Sites span the spectrum of human variation in market integration and degree of incorporation in modern states Many of the small- scale societies are sufficiently remote that formal institutions of the states under which they reside have minimal impact In many of the remote populations studied -- state does not have the will or the institutional capacity to enforce its authority Ø Local norms and institutions still predominate

29 Hadza Example Hadza of Tanzania are purely foraging society Population subsists on their own hunting and gathering production for most of their calories Their camps average about 30 people and are quite nomadic Ø moving 4-10 times per year Entire ethnic group numbers about 1000 individuals None practice a world religion

30 INDUSTRIAL SOCIETIES Two industrialized populations Accra, Ghana -- experiments carried out in small firms Rural Missouri o drawn from a town of 1800 o all families are known to each other o population in the heart of the Bible belt

31 Table S2. Ethnographic Summary of Societies Group Nation/Region Language Family Environment Economic Base Residence Researcher Accra City Ghana Mixed Urban Wage Work Sedentary Barr Au Papua New Mountainous Tropical Foraging/ Torricelli/ Wapei Guinea/Sepik Forest Horticulture Sedentary Tracer Dolgan/ Hunting/Fishing/ Semisedentary Russia/Siberia Turkic/Samoyedic Tundra Taiga Nganasan Wages Ziker Gusii Kenya Ekegusii Fertile High Plains Mixed Farming/ Wage Work Sedentary Gwako Hadza Tanzania Khoisan/Isolate Savanna Woodlands Foraging Nomadic Marlowe Isanga Tanzania Bantu Mountainous Forest Agriculture/ Wage work Sedentary McElreath Maragoli Kenya Logoli Fertile Plains Mixed Farming/ Wage Work Sedentary Gwako Orma Kenya Cushitic Semi arid Savanna Pastoralism Semi nomadic Ensminger Samburu Kenya Nilotic Semi arid Savanna Pastoralism Semi nomadic Lesorogol Sanquianga Columbia/Pacific Coast Spanish Mangrove Forest Fisheries (Fish, clams, shrimp) Sedentary Cardenas Shuar Ecuador/Amazonia Jivaroan Tropical Forest Horticulture Sedentary Barrett Sursurunga Tsimane U.S., Rural Missouri Papua New Guinea/ New Ireland Bolivia/Amazonia Austronesian Coastal Tropical Island Horticulture Sedentary Bolyanatz Macro Panoan Isolate Tropical Forest Foraging/ Horticulture Semi nomadic Gurven U.S./Rural Missouri English Prairie Wage Work Sedentary Ensminger Yasawan Fiji/Yasawa Island Oceanic Coastal Tropical Island Horticulture/ Marine Foraging Sedentary Henrich

32 TABLE 1-THE ULTIMATUM GAME IN SMALL-SCALE SOCIETIES Lowoffer Mean Rejection rejection Group Country offera Modesb ratec rated Machiguenga Peru / (72) (1/21) (1/10) Hadza Tanzania (big camp) (28) (5/26) (4/5) Hadza Tanzania (small (38) (8/29) (5/16) camp) Tsiman6 Bolivia /0.3/ (65) (0/70) (0/5) Quichua Ecuador (47) (2/13) (1/2) Torguud Mongolia (30) (1/20) (0/1) Khazax Mongolia Mapuche Chile / (46) (2/30) (2/10) Au PNG (33) (8/30) (1/1) Gnau PNG (32) (10/25) (3/6) Sangu Tanzania farmers (35) (5/20) (1/1) Sangu Tanzania herders (40) (1/20) (1/1) Unresettled Zimbabwe villagers (56) (3/31) (2/5) Resettled Zimbabwe villagers (70) (12/86) (4/7) Achuar Ecuador (36) (0/16) (0/1) Orma Kenya (54) (2/56) (0/0) Ach6 Paraguay / (75) (0/5 1) (0/8) Lamelarae Indonesia (63) (3/8) (4/20)

33 Result Summary Canonical model is not supported in any society studied Considerably more behavioral variability across groups than had been found in previous cross-cultural research

34 Variables to explain differences across groups: Six variables used Sociopolitical Complexity (Johnson and Earle) amount of centralized decision making above household Settlement Size -- some less than 100 others over 1000 Payoffs to cooperation (potential) payoff to cooperative v. solitary or family activities (machiguenga and Tsimane lowest, Lamalera whale hunters highest) Market integration frequency of market exchange (Hadza low as little change in life if markets disappeared, Orma high because they trade livestock and work for wages) Anonymity how important are anonymous transactions? (Achuar never interact with strangers, Shona frequently do) Privacy how well can people keep their affairs secret from others? (Au, Gnau, Hadza small villages or bands and eat together - impossible to keep secrets or hide things of value, Hadza pants pockets are popular. Mapuche strict rules about approaching another s house)

35 are popular!!! pockets can hide things, Sociopolitical complexity Henrich and an outside blind expert (Johnson) 0.9 correlation. Correlation - MI, AN, SC, SS Table 2.4 Create an index aggregate market integration average of MI,SS,SC (AN same) OLS Table 2.5 Regress Mean Offers on PC and AMI R 2 =.5

36 Overview and Synthesis 31 Table 2.4. Correlation matrix among predictor variables PC AN MI PR SS SC PC Ð AN Ð MI Ð PR Ð SS Ð Table 2.5. Regression coefficients and statistics Unstandardized beta coefficients Standardized beta coefficients t- statistic Sig. Std. error (Constant) PC AMI SC ˆ Socio-Political Complexity; PC ˆ Payoffs to Cooperation; AN ˆ Anonymity; PR ˆ Privacy; SS ˆ Settlement Size; MI ˆ Market Integration the causal effects of this nexus of variables, we created a new index of `aggregate market integration' by averaging the ranks of Market Integration, Settlement Size, and Sociopolitical Complexity. (We did not include Anonymity because it is so similar to Market Integration; including it only changes the results slightly.) We estimated ordinary least squares regression equations for explaining group mean Ultimatum Game offer using the Payoffs to Cooperation and Aggregate Market Integration. Both their normalized regression coef cients are highly signi cant and indicate that a standard deviation difference in either variable results in roughly half a standard deviation difference in the group mean offers (Table 2.5, Figure 2.5). Together, these two variables account for 47 percent of the variation among societies in mean Ultimatum Game offers. All regressions using Payoffs to Cooperation and one of the other predictors (Anonymity, Market Integration,

37 Figures 2.5 In contrast individual variables sex, wealth, age do not explain substantial amount of the variation within or across groups some small counter-examples (i.e. sex marginally important in some group)

38 32 Joseph Henrich et al. 0.2 Lamalera Mean Ultimatum Game offer Resettled Shona Achuar Tsimane Orma Mapuche Unresettled Shona Khazax Quichua Torguud Machiguenga Au Gnau Hadza Ache Sangu Farmers Sangu Herders ` PC Mean Ultimatum Game offer Hadza Tsimane Achuar Mapuche Quichua Ache ` Au Sangu Farmers Gnau Machiguenga Sangu Herders Lamalera Orma Resettled Shona Unresettled Shona Khazax Torguud Aggregate Market Integration 4 6 Fig Partial regression plots of mean Ultimatum Game offer as a function of indexes of Payoffs to Cooperation and Market Integration Notes: The vertical and horizontal axes are in units of standard deviation of the sample. Because Aggregate Market Integration and Payoffs to Cooperation are not strongly correlated, these univariate plots give a good picture of the effect of the factors captured by these indexes on the Ultimatum Game behavior. Sociopolitical Complexity, and Settlement Size) yielded a signi cant positive coef cient for Payoffs to Cooperation and a positive, near signi cant coef cient for the other variable. If we use the Income- Maximizing Offer as a predictor of the Ultimatum Game offers

39 Public Good Games Data from public-goods games played in seven of these societies also show much greater variation than previously found Public-goods games ask subjects to contribute to a common pool that will be expanded by the experimenter and then redistributed to all subjects The canonical prediction is that everyone will free ride -- contributing nothing Typical distributions of public-goods game contributions with WEIRD o U-shape o mode at contributing nothing o secondary mode at full cooperation o mean contribution between 40-60%

40 Small Scale Societies Public Goods Game Machiguenga: o mode at contributing nothing o not a single subject cooperating fully o mean contribution of 22 percent Ache and Tsimane: o inverted distributions o few or no contributions at full free-riding or full cooperation

41 Everyday Life Behavior in experiments generally consistent with economic patterns of everyday life in these societies Plausible interpretation -- when faced with a novel situation (the experiment) o looked for analogues in their daily experience o "What familiar situation is this game like?" o acted in a way appropriate for the analogous situation

42 Examples Hyper-fair UG offers (greater than 50 percent) and frequent rejections of these offers among the Au and Gnau Ø reflect culture of gift-giving found in these societies Among these groups (like many in New Guinea) o Accepting gifts (even unsolicited ones) commits one to reciprocate at some future time to be determined by the giver o Receipt of large gifts also establishes one in a subordinate position o Gifts refused because of the anxiety about strings attached Low offers and high rejection rates of the Hadza Ø Reflect their reluctant process of sharing (termed "tolerated theft") Ø Share only because they fear the social consequences of not sharing

43 Examples Ache did not reject low offers Ø widespread meat-sharing and cooperation in community projects despite the absence of a fear of punishment Ache hunters quietly leave their kill at the edge of camp, often claiming that the hunt was fruitless; their catch is later discovered and collected by others and then meticulously shared among all in the camp Ache made high propose offers - stake in the game seemed analogous to their catch

44 Examples The Machiguenga show the lowest cooperation rates in public-good games Ø little cooperation, exchange, or sharing beyond the family unit Orma experimental subjects quickly dubbed the public-goods experiment a harambee game Ø widespread institution of village-level voluntary contributions for public-goods projects such as schools or roads Ø contributed generously (58 percent of the stake)

45 Evolution of Fairness and Punishment Scale of human communities dramatically expanded from kin-based foraging bands to: o complex o intensely cooperative societies o strangers frequently engage in mutually beneficial transactions Consistent with life in these large-scale societies -- behavioral experiments reveal: Ø fair, trusting, and cooperative behavior among strangers, even in one-shot encounters

46 Theoretical Approaches Two major theoretical approaches to explain o relatively rapid expansion of human societal scales o puzzlingly pro-social behavior observed in experiments Approach 1: Humans possess an innate social psychology calibrated to life in the small-scale societies of our Paleolithic ancestors Rooted in the evolutionary logic of kinship and reciprocity Ø these heuristics were mistakenly extended to nonkin and ephemeral interactants -- as societies expanded with emergence of agriculture Pro-social behavior observed in experiments reflects ancient heuristics

47 Approach 2 Crucial ingredient in rise of more-complex societies was the development of new social norms and informal institutions Ø capable of domesticating our innate psychology for life in everexpanding populations Larger and more-complex societies prospered and spread to the degree that their norms and institutions effectively sustained successful interaction in ever-widening socioeconomic spheres Ø well beyond individuals local networks of kin Modeling work shows that when these learning mechanisms are applied to different kinds of social interactions (large-scale cooperation or ephemeral exchange) Ø individually costly behaviors can be sustained by punishment, signaling, and reputational mechanisms

48 Role of Social Norms By sustaining such behaviors, norms can facilitate trust, fairness, and cooperation in a diverse array of interactions Allowing the most productive use of unevenly distributed skills, knowledge, and resources Increasing cooperation in exchange, public goods, and warfare Norms that enhance fairness among strangers are likely causally interconnected with diffusion of several kinds of institutions (1) Expansion of breadth and intensity of market exchange (2) Spread of world religions

49 Market Integration Efficiency of market exchange involving infrequent or anonymous interactions improves with trust, fairness, and cooperation Ø Lowers transaction costs, raises the frequency of successful transactions, and increases long-term rewards Propose market norms evolved as part of an overall process of societal evolution to sustain mutually beneficial exchanges in contexts where established social relationships (kin, reciprocity, and status) were insufficient If theory is correct: Ø measures of fairness should positively co-vary with market integration

50 World Religions Religious institutions, beliefs, and rituals may have coevolved with norms that support large-scale societies and broad exchange Inter-societal competition may have favored religious systems that galvanize pro-social behavior in broader communities Ø supernatural incentives (hell) Ø recurrent rituals that intensify group solidarity Ethnographic data show emergence of moralizing religions increases with greater societal size and complexity

51 World Religions In contrast to religions that dominated our evolutionary history Ø modern world religions (Christianity and Islam) may be unusual in ways that buttress norms and institutions that sustain larger-scale interaction If this theory is correct Ø greater fairness toward anonymous others should be associated with adherence to a world religion

52 Experiments If markets and world religions are linked to the norms that sustain exchange in large-scale societies à Expect experimental measures of fairness in anonymous interactions will positively covary with measures of involvement in these two institutions à Test this in 15 small-scale societies

53 Market Integration (MI) Measured at household level by calculating percentage of household s total calories purchased from the market Ø as opposed to homegrown, hunted, or fished Averaged to obtain a community-level measure Ø use average to remain consistent with our definition of norms (as local equilibria) and to remove day-to-day stochastic variation

54 World Religion (WR) Asked participants what religion they practiced Ø coded binary variable = 1 indicating Islam or Christianity = 0 marking the practice of a tribal religion or no religion.

55 opulation Location Environs Economic base Residence ccra City Ghana Urban Wage work Sedentary u* Papua New Guinea Mountainous forest Horticulture and foraging Sedentary olgan/ Ng. Siberia Tundra-taiga Hunting, fishing, and wages Semi-sedentary usii Kenya Fertile high plains Farming and wage work Sedentary adza* Tanzania Savanna-woodlands Foraging Nomadic 0 0 sanga village Tanzania Mountainous forest Farming and wage work Sedentary aragoli Kenya Fertile plains Farming and wage work Sedentary rma* Kenya Semi-arid savanna Pastoralism Semi-nomadic amburu Kenya Semi-arid savanna Pastoralism Semi-nomadic anquianga Colombia Mangrove forest Fisheries Sedentary huar Ecuador Tropical forest Horticulture Sedentary ursurunga Papua New Guinea Island Horticulture Sedentary simane* Bolivia Tropical forest Horticulture and foraging Semi-nomadic nited States* Missouri Prairie Wage work Sedentary asawa Fiji Coastal tropical pacific Horticulture and marine foraging Mean MI Mean WR Sedentary

56 Table S3: Mean Demographics by Society Society Market Integration Std. Dev (N) World Religion Std. Dev (N) Community Population Std. Dev (N) Female Proportion Std. Dev (N) Age Std. Dev (N) Education Std. Dev (N) Household Size Std. Dev (N) Income USD Std. Dev (N) Wealth USD Std. Dev (N) Accra Au Dolgan/Nganasan Gusii Hadza Isanga Maragoli Orma Samburu Sanquianga Shuar Sursurunga Tsimane U.S./Rural Missouri Yasawa Total

57 Experiments Used three experiments designed to measure individuals propensities for fairness and their willingness to punish unfairness o Dictator o Ultimatum, o Third-Party Punishment Dictator Game (DG) Ø two anonymous players are allotted a sum of money (the stake) in a one-shot interaction Ø Player 1 decides how to divide this sum between himself or herself and Player 2 Ø Player 2 receives the allocation (offer) and game ends Player 1 s offer to Player 2 provides a measure of Player 1 s behavioral fairness

58 Experiments Ultimatum Game (UG) Ø Player 1 offers a portion to Player 2 Ø Player 2, before hearing actual offer from Player 1, decides whether to accept or reject each of the possible offers (in10% increments) Ø Decisions are binding If Player 2 specified that he or she would accept the amount of the actual offer, then Player 2 receives the offered amount and Player 1 gets the remainder If Player 2 specified that he or she would reject the offered amount, then both players receive zero

59 Ultimatum Game (UG) If people are motivated purely by money maximization Ø Player 2 will always accept any positive offer Ø Realizing this, Player 1 should offer the smallest nonzero amount Because this is a one-shot anonymous interaction, Player 2 s willingness to reject provides a measure of punishment Player 1 s offer measures a combination of social motivations and an assessment of the likelihood of rejection Ø second behavioral measure of fairness

60 Third-Party Punishment Game (TPG) Now a third player also receives the equivalent of one half of the stake Player 1 must decide how much to allocate to Player 2 who has no choices Player 3, before hearing the amount that Player 1 allocated to Player 2 Ø has to decide whether to pay 20% of his or her allocation to punish Player 1 for each of the possible offers (in 10% increments) If punished, Player 1 loses triple the amount paid by Player 3

61 Suppose the stake is $100 Third-Party Punishment Game (TPG) Player 1 offers $10 to Player 2 (keeping $90) Player 3 wants to punish this offer amount Player 1 takes home $60 ($90 $30) Player 2 gets $10 Player 3 gets $40 ($50 $10) If Player 3 had instead decided not to punish offers of $10, Ø take-home amounts would be $90, $10, and $50, respectively.

62 Third-Party Punishment Game (TPG) Money maximizing -- Player 3 would never pay to punish Similarly motivated Player 1 should always give zero to Player 2 Ø Player 3 s willingness-to-pay provides another measure of punishment Player 1 s offer measures a mixture of his or her social motivations and an assessment of the punishment threat Ø provides a third behavioral measure of fairness

63 Table S4: Mean Summary Statistics on Offers and Rejections by Society Society DG Offer Std. Dev (N) UG Offer Std. Dev (N) TPG Offer Std. Dev (N) UG P2 Minimal Acceptable Offer Std. Dev (N) TPG P3 Minimal Offer Not Fined Accra Au Dolgan/Nganasan Gusii Hadza Isanga Maragoli Orma Samburu Sanquianga Shuar Sursurunga Tsimane U.S./Rural Missouri Yasawa Total Std. Dev (N)

64 able S5. Linear regressions for all offers (DG, UG, TPG) Variables (1) (2) (3) (4) (5) (6) Model Model Model Model Model Model MI 0.156*** 0.120*** *** *** *** *** (0.0244) (0.0231) (0.0192) (0.0187) (0.0182) (0.0182) (3.17e 10) (2.91e 07) (1.34e 05) (1.08e 05) ( ) ( ) World Religion *** 5.350*** 5.649*** 6.158*** 6.141*** (2.164) (2.038) (1.844) (1.807) (1.819) (1.817) (0.886) ( ) ( ) ( ) ( ) ( ) Income *** 0.198*** 0.218*** 0.227*** (1000 USD) (0.111) (0.0887) (0.0668) (0.0618) (0.0610) (0.0606) (0.384) (0.278) ( ) ( ) ( ) ( ) Sex (female = 1) (1.259) (1.280) (1.200) (1.187) (0.286) (0.234) (0.385) (0.290) Age (years) (0.0456) (0.0479) (0.0441) (0.0423) (0.143) (0.266) (0.115) (0.164) Household Size (0.211) (0.208) (0.187) (0.223) (0.241) (0.357) Education (std by pop) (0.576) (0.593) (0.573) (0.338) (0.374) (0.461) Community Size 0.135*** (100 people) (0.0471) (0.0522) ( ) (0.632) Wealth (1000 USD) ( ) ( ) (0.834) (0.910) UG (1.168) (1.167) (1.039) (1.029) (1.009) (0.426) (0.268) (0.219) (0.195) (0.181) TPG 4.049** *** 4.115*** 4.301*** 4.953*** (1.597) (1.586) (1.424) (1.400) (1.387) (1.254) (0.0115) (0.129) ( ) ( ) ( ) (8.62e 05)

65 Findings People living in small communities lacking market integration or world religions Ø display relatively little concern with fairness or punishing unfairness in transactions involving strangers or anonymous others This result challenges hypothesis that successful social interaction in large-scale societies arise directly from an evolved psychology that mistakenly applies kin and reciprocity-based heuristics to strangers in vast populations Findings suggest caution in interpreting behavioral experiments from WEIRD populations as providing direct insights into human nature

66 EVOLUTION OF MORALIZING (WORLD) RELIGIONS Recent work suggested that certain religious institutions, beliefs, and rituals may have coevolved with norms that support complex societies Competition among societies or religions may have favoured spread of beliefs, institutions, and rituals Ø because these emerging religious systems helped promote or sustain prosocial behaviour towards co- religionists (and the exploitation of non- co- religionists)

67 MODERN RELIGIONS ARE UNUSUAL RELIGIONS Comparative ethno- historical analyses show that emergence of potent moralizing religions and gods increases with greater societal size and complexity Cross- cultural analysis of 186 societies found that larger, more complex, societies were much more likely to subscribe to potent deities directly concerned with morality and willing to punish norm violators The gods of the smallest- scale societies are typically weak, whimsical, and morally ambiguous Religions of smallest scale societies Ø little or no overlap between domains of morality and the supernatural

68 MODERN RELIGIONS ARE UNUSUAL RELIGIONS Emergence of larger- scale chiefdoms o after origins of sedentary food production o tightly associated with changes in religious beliefs, rituals and institutions The ancestor gods of even simplest chiefdoms Ø punish errant individuals for violations like theft, murder, and adultery using instruments of illness, accidents, and bad luck Archaeological research indicates Ø expansion of regular formal rituals and construction of religiously significant monumental architecture Ø co- emerges with increasing societal size and political complexity

69 BELIEF IN MODERN RELIGIONS IMPACTS EXPERIMENTAL MEASURES OF ADHERENCE TO PROSOCIAL NORMS Behavioral experiments reveal that unconsciously activating religious concepts leads to increased generosity toward strangers except among atheists Ritual on its own seems sufficient to increase prosociality Studies found that strangers acting in synchrony marching, singing and dancing cooperated more in subsequent group exercises, even in situations requiring personal costs Synchronous action (rhythmically moving together) increases cooperation, even when no positive emotion was attached to the actions The ability of music,

70 FIELD EVIDENCE SUPPORTS NOTION THAT RELIGION EVOLVES VIA COMPETITION Frequency and intensity of costly rituals increases along with sociopolitical complexity Ø in face of increasing military and economic competition Islam spread into Africa, Islamic beliefs galvanized by costly displays involving abstaining from alcohol, blood and pork, limiting polygyny, Pilgrimages, and fasting Ø permitted greater trust, shared rules of exchange and the use of credit institutions among converted Muslims Ø facilitates more trade and greater economic success