Intergenerational Wealth Transmission and the Dynamics of Inequality in Small-Scale Societies

Intergenerational Wealth Transmission and the Dynamics of Inequality in Small-Scale Societies
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Intergenerational Wealth Transmissionand the Dynamics of Inequality inSmall-Scale Societies Monique Borgerhoff Mulder, 1 * †  Samuel Bowles, 2 *  Tom Hertz, 3 *  Adrian Bell, 4 Jan Beise, 5 Greg Clark, 6 Ila Fazzio, 7 Michael Gurven, 8 Kim Hill, 9 Paul L. Hooper, 10 William Irons, 11 Hillard Kaplan, 12 Donna Leonetti, 13 Bobbi Low, 14 Frank Marlowe, 15 Richard McElreath, 16 Suresh Naidu, 17 David Nolin, 18 Patrizio Piraino, 19 Rob Quinlan, 20 Eric Schniter, 21 Rebecca Sear, 22 Mary Shenk, 23 Eric Alden Smith, 24 Christopher von Rueden, 25 Polly Wiessner 26 Small-scale human societies range from foraging bands with a strong egalitarian ethos to moreeconomically stratified agrarian and pastoral societies. We explain this variation in inequality usinga dynamic model in which a population ’ s long-run steady-state level of inequality depends onthe extent to which its most important forms of wealth are transmitted within families acrossgenerations. We estimate the degree of intergenerational transmission of three different typesof wealth (material, embodied, and relational), as well as the extent of wealth inequality in21 historical and contemporary populations. We show that intergenerational transmission of wealthand wealth inequality are substantial among pastoral and small-scale agricultural societies (on apar with or even exceeding the most unequal modern industrial economies) but are limited amonghorticultural and foraging peoples (equivalent to the most egalitarian of modern industrialpopulations). Differences in the technology by which a people derive their livelihood and in theinstitutions and norms making up the economic system jointly contribute to this pattern. I nvestigations of the dynamics of economicinequality across distinct economic systemshavebeenlimitedbythepaucityofdataonall but contemporary market-based industrial socie-ties. They are also hampered by the lack of anempirically based model applicable to the dif-fering institutions and technologies characteristicof the broad range of economic systems, rangingfromhunter-gatherersthroughpastoralandagrariansocieties to modern economies. Here we present empiricalestimatesofthe extentofinheritanceof wealth across generations and of the degree of wealth inequality, along with a descriptive modelof the relation between the two. We support our model with data on three distinct wealth classes  —  material, embodied, and relational, to be defined below  —  in 21 contemporary and recent hunter-gatherer, horticultural, pastoral, and agricultural populations.The key thesis to be explored is that for somekindsof wealth and some economic systems (but not others) the parents ’  wealth strongly predictsthe wealth of the offspring. In particular, the cat-tle, land and other types of material wealth of  pastoral and agricultural economies are directlytransmitted by simple transfers, often buttressed by social conventions of inheritance. By contrast the somatic wealth and skills and the social net-work ties central to foraging and horticulturallivelihoods are more subject to the vagaries of learning, genetic recombination, and childhooddevelopment. Moreover, in foraging and horti-culturaleconomies,suchmaterialwealthasexiststends to circulate through broad social networksrather than being vertically transmitted to off-spring. A corollary of the thesis is that, if our model is correct, economies in which materialwealth is important will show substantial levelsof wealth inequality.Both the thesis and the corollary find strongsupport in our data. We focus on small-scalesocieties because they offer the greatest variationin both the technologies by which a livelihood isgained and the basic institutions that provide theincentives and constraints regulating economiclife, including the dynamics of inequality and theinheritance process. (We use the term  “ small-scale ”  to refer to populations in which the influ-ence of modern national states is limited). Thesesocieties thus provide the most powerful lens for exploring hypotheses concerning the importanceof technologies (kinds of wealth) and institutions(kinds of society) in explaining the dynamics of inequality and, thus, may also illuminate long-termtrends in contemporary and future economies.Theconnectionbetween wealthinheritanceandwealth inequality (explained more precisely in themodelbelow)isthefollowing:Ifwealthisstronglytransmitted across generations, chance shocks tothe economic fortunes of a household due to dis-ease or accident, luck in a hunt or harvest, andother environmental disturbances or windfallswill be reproduced in the next generation. Theseeffects will thus accumulate over time and there- by counteract the widely observed inequality-dampening tendency of regression to the mean( 1  –  3 ). We seek to understand the effects of this process by examining how the offsetting effectsof random shocks and imperfect transmissionacross generations jointly determine a steady-state distribution of wealth for differing kinds of wealth and across the four different economicsystems ( 4 ). The institutions and norms that characterize distinct economic systems and thenature of the wealth class alike will affect the de-gree of intergenerational transmission. The extent of shocks will also differ across wealth classesand economic systems.For a number of modern economies, thereare quantitative estimates and comparisons of theintergenerational transmission of education, oc-cupational prestige, nonhuman physical capital,andotherformsofembodiedandmaterialwealth( 3 ,  5 ,  6  ). For small-scale populations, associa-tions between reproductive success and materialforms of wealth have been studied ( 7  ), and thereexist piecemeal estimates of intergenerational trans-mission of, for example, fertility ( 8 ) and height ( 9 ). But there are no estimates allowing a com- parison across populations of the inheritance of the distinctive kinds of wealth that are central tothe livelihoods of small-scale communities of for-agers, horticulturalists, herders, and farmers. Herewe present a new set of data and conduct a quan-titative comparative analysis of the transmissionof distinct types of wealth among the 21 popula-tions shown in Fig.1 and Table 1. Further infor-mation is provided in ( 4 ). RESEARCH ARTICLES 1 Department of Anthropology and Center for PopulationBiology, University of California, Davis, CA 95616, USA. 2 Santa Fe Institute and University of Siena, Santa Fe, NM87501, USA.  3 International University College of Turin,10121 Turin, Italy.  4 Graduate Group in Ecology, Universityof California, Davis, CA 95616, USA.  5 Department of Eco-nomic and Social Affairs, United Nations, New York, NY10017, USA.  6 Department of Economics, University of Cal-ifornia, Davis, CA 95616, USA.  7 Center for Economic Per-formance, London School of Economics, London WC2A 2AE,UK.  8 Integrative Anthropological Sciences Program, Univer-sity of California, Santa Barbara, CA 93106, USA.  9 School ofHuman Evolution and Social Change, Arizona State Univer-sity, Tempe, AZ 85281, USA.  10 Department of Anthropology,University of New Mexico, Albuquerque, NM 87131, USA. 11 Department of Anthropology, Northwestern University,Evanston, IL 60208, USA.  12 Department of Anthropology,University of New Mexico, Albuquerque, NM 87131, USA. 13 Department of Anthropology, University of Washington,Seattle WA 98195, USA.  14 School of Natural Resources andEnvironment, University of Michigan, Ann Arbor, MI 48109,USA.  15 Department of Anthropology, Florida State Univer-sity, Tallahassee, FL 32306, USA.  16 Department of Anthro-pology and Center for Population Biology, University ofCalifornia, Davis, CA 95616, USA.  17 Harvard Academy forInternational Studies, Harvard University, Cambridge, MA02138, USA.  18 Carolina Population Center, University ofNorth Carolina, Chapel Hill, NC 27516, USA.  19 100 Tunney'sPasture Driveway, Ottawa, ON K1A 0T6, Canada.  20 Depart-ment of Anthropology, Washington State University, Pullman,WA 99164, USA.  21 Integrative Anthropological SciencesProgram, University of California, Santa Barbara, CA 93106,USA.  22 Department of Social Policy, London School of Eco-nomics, London WC2A 2AE, UK.  23 Department of Anthro-pology, University of Missouri, Columbia, MO 65211, USA. 24 Department of Anthropology, University of Washington,Seattle, WA 98195, USA.  25 Integrative Anthropological Sci-ences Program, University of California, Santa Barbara, CA93106, USA.  26 Department of Anthropology, University ofUtah, Salt Lake City, UT 84112, USA.*The first three authors contributed equally to this article. † To whom correspondence should be addressed. 30 OCTOBER 2009 VOL 326  SCIENCE 682    o  n   M  a  r  c   h   1   5 ,   2   0   1   2  w  w  w .  s  c   i  e  n  c  e  m  a  g .  o  r  g   D  o  w  n   l  o  a   d  e   d   f  r  o  m   The dynamics of wealth inequality.  To clar-ify our model, we initially consider just a singletype of wealth and show that the degree of in-equality in its steady-state distribution depends onthe extent to which wealth is transmitted acrossgenerations.Supposethatahousehold ’ swealthisacquired in two ways. The first is transmissiondirectly from the parents, in the form of material bequests, clients, skills, private information,genotype, conditions affecting development,network connections, and so on. The secondway of acquiring wealth is from the resourcesavailabletoall members of the population, in theform, say, of equal access to common resourcesor public information.We summarize these two influences on ahousehold ’ s wealth by expressing the expectedwealth of household  i  as  b w ip  + (1  –   b ) w , wherewealth is measured in natural logarithms,  w ip  isthewealthlevelofhousehold i ’ sparents,and w isthe population-average wealth level (normalizedto be the same across generations). The inter-generationaltransmissioncoefficient, b  (0 ≤ b  <1),measures the extent to which the wealth of ahousehold in one generation depends on thewealth in the previous generation, and (1  –   b )represents regression to the mean as introduced by Galton in his study of human stature ( 10 ).In each generation, the realized wealth of ahousehold is its expected wealth (above) plus adisturbance term,  l , reflecting exogenous shocksthat over time are assumed to be independent of the wealth of previous generations, with meanzero and variance  s 2 l : w i  ¼ b w ip þð 1 −  b Þ w  þ  l i  ð 1 Þ We are interested in the variance of the loga-rithm of a population ’ s wealth (a standard unit-free measure of inequality) in the long run. Todetermine this, we use Eq. (1) to write the var-iance of wealth in generation  t   as:var  ð w it  Þ ≡  m t   ¼ b 2 m t   −  1 þ s l 2 ð 2 Þ where  m t  -1  is the variance of the logarithm of wealth in the parental generation. We then solvefor   m , the steady-state (stationary, or long run)variance of the logarithm of wealth, by setting m t  -1  =  m t   =  m , giving:var  ð w i Þ ≡  m ¼  s 2 l = ð 1 −  b 2 Þ ð 3 Þ Thissteady-statelevelof wealthinequalitymay be interpreted as the effect of stochastic shocks(the numerator), blown up by the intergenerationaltransmission multiplier, (1  –   b 2 ) − 1 , which is in-creasing in  b , the extent of intergenerational trans-mission of wealth. As  b  approaches one, the effectsof windfalls of wealth, accidents of health, theft,and the like, dissipate more slowly over time sothat the shocks of even the distant past contributeto inequality in a given generation, resulting inhigh levels of steady-state inequality. For   b  > 1,there is no steady state, and inequality will increaseovertime.Thedeterminationofthissteadystateisillustrated in Fig. 2.A population exposed to greater wealth shocksis represented by a larger intercept on the verticalaxis ( s 2 l ), whereas greater intergenerational trans-mission of wealth is represented by a steeper sol-id line [the slope of which is  b 2 , see Eq. (2)]. Touse this model, we need not assume that thesteady-state wealth distribution is typically at-tained. What is important for our approach isthat, for a given society, the fluctuations aroundthe steady-state value are small relative to thedifferences in steady-state inequality across soci-etiescharacterizedbydifferenteconomicsystemsand different kinds of wealth.By a household ’ s wealth, we mean any of itsattributes thatcontributetoitswell-being asmea-sured by consumption levels, social status, or other ends that are valued in the particular society. Totake account of many kinds of wealth simulta-neously, we define the importance of each classof wealth as follows. Let   E  ,  M  , and  R  be positivenumbers representing the amount of a household ’ sembodied, material and relational wealth. Thewell-being of the household,  W  , is a weighted product of these classes of wealth, the weights beingtherelativeimportanceofeachwealthclassin the economic system in which the householdlives: W   ¼  g  E  e  M  m  R r  ð 4 Þ where  g  is a positive constant and the exponents e , m ,and r  (theweights)arethederivativesofthelogarithm of well-being with respect to the loga-rithms of the three respective wealth classes or,equivalently, the percent difference in well-beingassociated with a 1% difference in the amount of each class of wealth. The weighted product is preferred (to the weighted sum, for example) be-cause it implies, plausibly, that the wealth classesarecomplements;thatis,the contribution ofeachclassof wealthtoindividualwell-beingisenhanced by the extent of the other classes of wealth.We assume constant returns to scale (dou- bling the amount of all three classes of wealth of a household will double its well-being) implyingthat   e  +  m  +  r   = 1. This motivates our inter- pretationoftheseexponentsasweightsindicatingthe relative importance of each class of wealth.We refer to these weights as  a  ≡  { e ,  m ,  r  }. Tocombine this information on the importance of wealth classes with our measures of the extent of transmission of each wealth class across genera-tions,weestimatean a -weightedaverage b  foreacheconomic system. We also calculate an  a -weightedaverage measure of wealth inequality (the Ginicoefficient) for each economic system (see below).Ideally,onewouldhavecomparablemeasuresofthe relative importance ( a ) anddegree oftrans-mission ( b ) of each class of wealth, the degreeof inequality in the distribution of each kind of wealth (Gini), and the extent of wealth shocks( s 2 l ). Measuring the last-mentioned is difficult in any economy and impossible in the economiesunder study, as the estimate requires long time- Fig.1. Populationsstud-ied. Note:Circle indicateshunter-gatherers; star,horticulturalists; square,pastoralists; andtriangle,agriculturalists.  SCIENCE  VOL 326 30 OCTOBER 2009  683 RESEARCH ARTICLES    o  n   M  a  r  c   h   1   5 ,   2   0   1   2  w  w  w .  s  c   i  e  n  c  e  m  a  g .  o  r  g   D  o  w  n   l  o  a   d  e   d   f  r  o  m   series data for individual wealth, which, with fewexceptions, are simply nonexistent. We are able,however, to measure the other three quantities,and this permits us to gauge the extent to whichintergenerational wealth transmission allows theeffect of shocks to accumulate over time, andto explore differences in both intergenerationalwealthtransmissionandwealthinequalityacrosseconomic systems and wealth classes. The nature of wealth and the varieties of economic systems.  Since the development of human capital theory a half-century ago, it has been conventional to treat wealth as a multi-dimensional attribute, as evidenced by the adjec-tivesnowroutinelyappliedto the word “ capital, ” namely, social, somatic, material, cultural, and net-work ( 11  –  13 ). We identified three broad classesof wealth in our populations, namely, embodied Table 1.  Population characteristics and estimates of the intergenerational transmission of 43 measures of embodied, relational, and material wealth across 5hunter-gatherer, 4 horticultural, 4 pastoral and 8 agricultural populations. For wealth type, the number of parent-child pairs is in parentheses.  b  T  SE. Economic systemand population Wealth type Wealthclass  b  General description (ref.) Hunter-gathererAche Hunting returns (49) E 0.081  T  0.273 Mobile foragers (Paraguay 1982 – 2008) (  30 )Ache Body weight (137) E 0.509  T  0.128Hadza Body weight (227) E 0.305  T  0.076 Mobile foragers (Tanzania 1982 – 2008) (  31 )Hadza Grip strength (196) E  – 0.044  T  0.050Hadza Foraging returns (33) E 0.047  T  0.193Ju/  ’ hoansi Exchange partners (26) R 0.208  T  0.114 Mobile foragers (Botswana 1973 – 75) (  32 )Lamalera Quality of housing (121) M 0.218  T  0.099 Sedentary fishers, trade, and some farming (Indonesia 2006) (  33 )Lamalera Boat shares (121) M 0.122  T  0.093Lamalera Food share partners (119) R 0.251  T  0.052Lamalera Reproductive success (121) E 0.161  T  0.174Meriam Reproductive success (91) E 0.088  T  0.247 Sedentary fishers, some farming (Australia 1998) (  34 )HorticulturalDominicans Land (62) M 0.137  T  0.140 Subsistence farming, fishing, bay oil production, limited wages(Dominica 2000 – 08) (  35 )Gambians Body weight (1274) E 0.391  T  0.041 Subsistence rice and cash farmers (Gambia 1950 – 80) (  36 )Gambians Reproductive success (967) E 0.088  T  0.086Pimbwe House/farm utensils (283) M 0.107  T  0.318 Subsistence farming, some cash farming, some foraging(Tanzania 1995 – 2006) (  37  )Pimbwe Farming skill (217) E  – 0.015  T  0.097Pimbwe Body weight (148) E 0.377  T  0.096Pimbwe Reproductive success (599) E  – 0.057  T  0.107Tsimane Household utensils (110) M 0.024  T  0.071 Subsistence farming, some foraging (Bolivia 2002 – 08) (  38 )Tsimane Labor cooperation (67) R 0.181  T  0.106Tsimane Allies in conflict (45) R 0.338  T  0.103Tsimane Knowledge/skill (181) E 0.111  T  0.094Tsimane Grip strength (490) E 0.070  T  0.042Tsimane Body weight (383) E 0.253  T  0.069Tsimane Hunting returns (26) E 0.384  T  0.130Tsimane Reproductive success (849) E 0.128  T  0.073PastoralDatoga Livestock (135) M 0.622  T  0.127 Transhumant pastoralism, some farming (Tanzania 1987 – 89) (  39 )Datoga Reproductive success (133) E 0.066  T  0.060Juhaina Arabs Camels (21) M 0.535  T  0.226 Transhumant pastoralism (Chad 2003) ( 40 )Sangu (Ukwaheri) Cattle (108) M 0.957  T  0.424 Pastoralism, some farming (Tanzania 1997 – 2000) ( 41 )Yomut (Charwa) Patrimony (livestock) (22) M 0.564  T  0.167 Transhumant pastoralism, some farming (Turkmenistan/Iran1965 – 74) ( 42 )AgriculturalBengali Reproductive success (382) E  – 0.074  T  0.057 Farmers with wage labor (India 2000 – 01) ( 43 )Bengaluru In-law networks (249) R 0.114  T  0.073 Farmers, merchants, wage labor, urban (India 1910 – 1973) ( 44 )East Anglians Estate value (land) (210) M 0.642  T  0.073 Farmers, wage labor, merchants; rural and urban (England1540 – 1845) ( 45 )East Anglians Reproductive success (200) E 0.171  T  0.150Khasi Reproductive success (650) E 0.165  T  0.045 Farmers with wage labor (India 2000 – 01) ( 43 )Kipsigis Land (270) M 0.357  T  0.041 Farmers with livestock (Kenya 1981 – 1990) ( 46 )Kipsigis Livestock (270) M 0.635  T  0.098Kipsigis Cattle partners (102) R 0.041  T  0.139Kipsigis Reproductive success (270) E 0.213  T  0.106Krummhörn Land (1602) M 0.610  T  0.043 Farmers with diverse off-farm occupations (German 17th to19th century) ( 47  )Skellefteå Reproductive success (2515) E 0.010  T  0.028 Farmers with diverse off-farm occupations (Sweden 1800 – 88) ( 48 )Yomut (Chomur) Patrimony (land) (58) M 0.528  T  0.147 Farmers with livestock (Turkmenistan/Iran 1965 – 74) ( 42 ) 30 OCTOBER 2009 VOL 326  SCIENCE 684 RESEARCH ARTICLES    o  n   M  a  r  c   h   1   5 ,   2   0   1   2  w  w  w .  s  c   i  e  n  c  e  m  a  g .  o  r  g   D  o  w  n   l  o  a   d  e   d   f  r  o  m   (body weight, grip strength, practical skills, and,in predemographic transition populations, repro-ductive success); material (land, livestock, andhousehold goods); and relational (social ties infood-sharing networks and other forms of as-sistance).Wehavenomeasuresofotherheritabledeterminants of well-being such as ritual knowl-edge, an important source of institutionalized in-equality in some populations. By linking the levelof wealth of parents and adult offspring, mea-sured as appropriate for individuals (e.g., bodyweight) or households (e.g., land), we are able toestimate the degree of intergenerational persist-ence for particular types of wealth and then tocreate averages for each broad class of wealth.We classify economic systems according tothe conventions of anthropology ( 14 ). Hunter-gatherer economic systems are those that makeminimaluseof domesticated species(either plant oranimal),whereaspastoralistsrelyheavily,thoughrarely exclusively, on livestock kept for subsist-ence and sometimes commercial purposes. Al-though both horticulturalists and agriculturalistsuse domesticated plants and animals, horticultur-alists do not typically use ploughs, their cultiva-tion is labor- not land-limited, and land marketsare absent or limited. As with all classificatorysystems, there are some ambiguities of assign-ment of our populations to these classes, but theleast improbable reclassifications do not affect our results [see ( 4 ), section 4].Transmission of wealth across generationsneed not take the form of bequests, or the literal passingonofphysicalobjects(suchaswhenlandis transmitted from father to son). What mattersfor the long-run dynamics of inequality is any-thing that results in a statistical association be-tween the wealth of parents and children. Thisstatistical association may be enhanced by posi-tiveassortmentinmatingorineconomicpursuitsas occurs when skilled hunters pursue prey to-gether, or when successful herders cooperate inlivestock management. The same is true of in-creasing returns or other forms of positive feed- backs, for example when those who invest asubstantial amount earn higher than average re-turns, or when childhood developmental effectsassociated with modest genotypic differences re-sult in substantial phenotypic differences. Nega-tivefeedbacks,suchassharingnormsthatextract substantial transfers from the wealthy, or wealthshocks that are inversely correlated with one ’ swealth (such as occur when cattle thieves target large herds), by contrast, heighten regression tothe mean by reducing  b , thereby attenuating the persistence of inequality over time and hencereducing steady-state inequality.Our three wealthclasses differintheextenttowhich these transmission mechanisms  —  transfers,assortment,andpositivefeedbacksindevelopment or accumulation  —  are at work. Material wealth isreadily transferred to the next generation by be-quests sanctioned by cultural rules. Moreover, be-cause it is typically observable, material wealthcan facilitate deliberate marital or economic assort-ment. For some types of material wealth (storagefacilities, herds of livestock, and irrigated land,for example), the correlation of material wealthlevels across generations is further enhanced bythe presence of increasing returns to scale or other positive feedbacks. Network ties can easily be passed from parent to child, but the offspringof less well-connected parents can usually gainaccess to allies and helpers more readily than alandless son in a farming community can acquireland, for example, through savings or systemsof patronage. As a result we expect the inter-generational transmission of relational wealth to be limited, at least by comparison with materialwealth.Embodied wealth is transmitted by a combi-nation of genetic inheritance, socialization, and parent-offspring similarity in the conditionsaffecting childhood development. The knowledgecomponent of embodied wealth is readily trans-mitted to offspring, but, unless restricted by reli-gious or other constraints, it is typically availableto other members of a population as well (thecommon knowledge of the behavior of prey spe-cies, for example, or farming practices). Geneticand psychometric evidence from industrial soci-eties suggests that parent-offspring transmissionof economically relevant personality and behav-ioral characteristics, such as risk-taking, trust-worthiness, conscientiousness, and extroversionis limited ( 4 ). We do not have similar evidenceacross generations in the small-scale populationsunder study, but industrial-society estimates sup- port our expectation that the degree of inter-generational transmission will differ markedlyamong our three wealth classes, with substantialtransmission of material wealth and more limitedtransmission of relational and embodied wealth.Ethnographic evidence suggests that the four economicsystemsalsodifferintheimportanceof  µ t µ t = β 2 µ t-1 +   σ λ  µ t-1 µ t = µ t-1 σ λ  22 Fig. 2.  Steady-state wealthdistribution. The dashedline is the steady-state condition requiring wealthinequality to be unchanging from one period to thenext. The solid line (Eq. 2) is the combined effect ofthis period ’ s variance of shocks (the constant) aug-mented by the inequalities in wealth transmittedfrom the previous period (the slope). Table 2.  Summarystatistics:Intergenerationaltransmissionofwealth( b ),byeconomicsystemandwealthclass. Cell-means were estimated in a regression against a full set of dummy variables for each cell, withconventional standard errors. See ( 4 ), section 1, for a discussion of alternative approaches to estimatingthese cell-means and their standard errors, and tables S11 and S12 for the alternative results. Reported P valuescorrespondtotwo-tailedtestsofthehypothesisthatthetrue b orGinicoefficientiszeroforthatcell.Averages across wealth classes (final two columns) are calculated after weighting the cell-mean  b  valuesand Ginis by the values of  a  shown. NA, data not available. Economic systemsWealth classes  a -weightedaverageof   b  values a -weightedaverageof GinisEmbodied Relational Material Hunter-gatherer  a  0.46 0.39 0.15 b  0.16  T  0.06 0.23  T  0.11 0.17  T  0.011 0.19  T  0.05 0.25  T  0.04 P  0.01 0.04 0.12 0.00 0.00Horticultural  a  0.53 0.26 0.21 b  0.17  T  0.05 0.26  T  0.11 0.09  T  0.09 0.18  T  0.04 0.27  T  0.03 P  0.00 0.02 0.31 0.00 0.00Pastoral  a  0.26 0.14 0.61 b  0.07  T  0.15 NA †  0.67  T  0.07 0.43  T  0.06 †  0.42  T  0.05 † P  0.66 0.00 0.00 0.00Agricultural  a  0.27 0.14 0.59 b  0.10  T  0.07 0.08  T  0.11 0.55  T  0.07 0.36  T  0.05 0.48  T  0.04 P  0.16 0.47 0.00 0.00 0.00Average acrossall economicsystems a  0.38 0.23 0.39 b  0.12  T  0.05 0.19  T  0.06 0.37  T  0.04 0.29  T  0.03 0.35  T  0.02 P  0.01 0.00 0.00 0.00 0.00 † The  b  and Gini for Kipsigis cattle partners (see Table 1 and table S4) are used in the pastoral/relational cell for the calculationof the  a -weighted average across wealth classes.  SCIENCE  VOL 326 30 OCTOBER 2009  685 RESEARCH ARTICLES    o  n   M  a  r  c   h   1   5 ,   2   0   1   2  w  w  w .  s  c   i  e  n  c  e  m  a  g .  o  r  g   D  o  w  n   l  o  a   d  e   d   f  r  o  m   the three classes of wealth. A successful hunter-gathererorhorticulturalistdependsheavilyonhisor her strength, practical knowledge, and socialnetworks, while making little use of material re-sources that are not in the public domain. Bycontrast, the well-being of a herder or farmer isclosely tied to the amount of stock or land under his or her command, which makes materialwealth a more important influence on livelihoodsin these economic systems. Estimating the intergenerational transmis-sion of wealth.  To estimate our model of wealthtransmission, we need two pieces of information:the degree of intergenerational transmission ( b )for each wealth type and the importance of eachwealth class in a given economic system ( a  ≡ { e ,  m ,  r  }). Note that we do not require iden-tification of the causal paths by which transmis-sion takes place, as might be represented in amultiequation structural model ( 15 ). Our modelinstead requires a single estimate of the magni-tude of the statistical association between paren-talandoffspringwealth( b )foreachdataset.Thisrequirement, along with the absence of robust evidence of nonlinearities, motivated our consist-ent use of linear models. Functional forms, esti-mation procedures, robustness checks, weighting procedures, and other aspects of our statistical tech-niques and results are described in ( 4 ), section 1. Note that the populations studied were not se-lected at random; instead, we included all popu-lationswewereawareof forwhichintergenerationalwealth transmission estimates are feasible and theresearchers agreed to share data. Table 1 presentsour individual estimates of   b ; Table 2 presentsthe summary statistics for both the intergenera-tional transmission ( b ) and the importance ( a )of the three wealth classes in the four economicsystems.Across the four economic systems, the esti-mated  b  for 14 measures of material wealth, in-cluding agricultural and horticultural land, livestock,shares in sea mammal  –  hunting boats, quality of housing, and household utensils averages 0.37(Table 2). For farm land (5 data points), the de-gree of transmission is substantial, averaging 0.45(calculatedfromthedatainTable1),thusequalingor exceeding the intergenerational transmissionof most forms of wealth in modern industrialeconomies ( 16  ). Livestock are even more highlytransmitted across generations (Table 1,  b  valuesaveraging 0.66).Our 23 estimates of the transmission of embodied wealth across generations average0.12. The highest estimates are for body weight (for which  b  averages 0.37). We also find a verymodest level of intergenerational transmissionof reproductive success (number of offspringsurviving to age 5); it is entirely absent in threesocieties, has a maximum value of 0.21, andaverages 0.09, similar to low correlations be-tween parental and offspring fertility in many pre-demographic transition populations ( 17  ). Gripstrength is weakly transmitted across generations.The transmission of hunting success is highlyvariable (0.08 for the Ache, 0.38 for the hor-ticultural Tsimane, and 0.05 for hunting andforaging yields in the Hadza), averaging 0.17.Knowledge and skill, such as the production andmanagement of horticultural crops in the Pimbweor proficiency in subsistence tasks and culturalknowledge in the Tsimane, are only weakly trans-mitted from parents to offspring.The six estimates of relational wealth trans-mission indicate that the extent to which network links are transmitted across generations is mod-est, averaging  b  =  0.19.To measure the importance of each wealthclass in the four economic systems ( a ) we usedethnographers ’  judgments (for each wealth classin the population they studied) of the percentagedifference inhousehold well-beingassociatedwitha 1% difference in the amount of a given wealthclass, holdingother wealth classes constantat theaverage for that population, and requiring these percentage effects to sum to one. The averagevalues of   a  bywealth classand economic systemalso appear in Table 2. Consistent with descriptiveethnographies of these and other populations, em- bodied and relational wealth are relatively impor-tantforhunter-gatherers,whereasmaterialwealthis key in pastoral and agricultural populations.Statistical estimates of the importance of eachclass of wealth across the economic systems ( a )wouldhavebeenpreferable,butareprecludedbythe absence for most populations of a single rela-tively homogeneous measure of well-being. How-ever, we were able to econometrically estimate m  —  the importance of material wealth  —  from anequationsimilarto (4)usingdata (mostofitfromhalf a century ago) from populations not repre-sented in our study, including one horticultural,two pastoral, and seven small-scale agriculturaleconomies. These estimates [see ( 4 ) section 1] areclose to our ethnographers ’  estimates and suggest that, if anything, we have understated the differ-ence in the importance of material wealth between pastoral and agricultural economies, on the onehand, and horticultural economies on the other.Correctingthisunderstatementwouldonlystrength-en our main conclusions. Results.  Ourfirstfindingisthatthe a -weightedaverages of the  b  values (the importance-weightedaverage transmission coefficients) for the four eco-nomic systems differ markedly (Table 2). Inter-generational transmission of wealth is modest inhunter-gatherer and horticultural systems and sub-stantial in agricultural and pastoral systems. How-ever, even the smaller   b  values of the former implythat being born into the top 10% of the wealthdistribution confers important advantages. In thesesocieties, a child of parents in the highest wealthdecile is on average more than three times aslikely to end up in the top decile as is the child of the bottom decile [( 4 ), section 3 and table S7].Although hardly a level playing field, intergen-erational transmission in these economic systemsis modest when compared with the agriculturalsystems, where the child of the top decile is onaverage about 11times more likely than thechildof the poorest decile to end up in the richest decile, or to the pastoral systems, where the ratioexceeds 20.Our second finding is that economic systemsin which wealth is more heritable are indeed moreunequal, as predicted by our model. For each pop-ulation and type of wealth, we estimated the Ginicoefficient, which is a measure of inequality rang-ing from 0 (equal wealth) to approximately 1 [allwealth held by a single household, see table S4and discussion in ( 4 ), section 1]. To calculate anoverall measure of wealth inequality for a giveneconomic system we again weight the results for eachwealthclassinthatsystembyitsimportance( a ). These estimates of overall wealth inequalityappear in the last column of Table 2, and in moredetail in table S5. They exhibit the same patternas the transmission coefficients ( b  values): hunter-gatherer and horticultural populations are bothrelatively egalitarian; pastoral and agriculturalsocieties are characterized by substantial wealthinequality (see also fig. S2).A third finding is that neither the overallintergenerational transmission of wealth nor thelevel of inequality is greater in horticultural thanin hunter-gatherer populations. This result chal-lenges a long-standing view ( 18 ) that foragersare uniquely egalitarian among human societies.Thus, it may be ownership rights in land and live-stock, rather than the use of domesticated plantsandanimalsperse,thatarekeytosustaininghighlevels of inequality. Our finding that pastoraliststransmit wealth across generations to an extent equal to if not greater than farmers, and likewisedisplay similar Gini coefficients, will also chal-lengewidely held views thatherdersarerelativelyegalitarian ( 19 ).Are the relative intergenerational mobility of the hunter-gatherer and horticultural systems andthe high levels of intergenerational wealth trans-mission of the pastoral and agricultural systemsdue primarily to technology (the differing impor-tance of the distinct classes of wealth across eco-nomic systems) or to institutions (differences inintergenerational transmission, independent of dif-ferences in the importance of the wealth classes)?To answer this question, we take advantage of thefact that both the importance of the wealth classesand degree of intergenerational transmission of wealth are similar in the hunter-gatherer and hor-ticultural populations, on the one hand, and the pastoralandagriculturalpopulationsontheother.This allows us to reduce the four systems to two.Forty-five percent of the large (namely 0.21) andstatistically significant difference (  P   < 0.001) be-tween the average  a -weighted  b  values of thesetwogroupsofeconomicsystemsisaccountedfor  by differences in technology, reflected primarilyin the greater importance of material wealth in producing the herders ’  and farmers ’  livelihoods[forthedecompositionformula,see( 4 )section1;for the paired economic systems results,see tableS3]. The remaining 55% is due to differences ininstitutions, reflected primarily in the lesser de-gree of transmission of material wealth in the 30 OCTOBER 2009 VOL 326  SCIENCE 686 RESEARCH ARTICLES    o  n   M  a  r  c   h   1   5 ,   2   0   1   2  w  w  w .  s  c   i  e  n  c  e  m  a  g .  o  r  g   D  o  w  n   l  o  a   d  e   d   f  r  o  m 
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks