FIELD METHODS
Kuznar
/ ESTIMATING VALUE IN NON-WESTERN SOCIETIES
Application of General Utility
Theory for Estimating Value
in Non-Western Societies
LAWRENCE A. KUZNAR
Indiana University—Purdue University at Fort Wayne
Understanding value and being able to monitor it is crucial to many anthropological
theories of economy. The author presents a method, developed in economics, of evaluating peoples’ subjective values. While this method is derived from neoclassical
economics, its assumptions are minimal and its use is not necessarily tied to any particular economic paradigm. Therefore, the method may be useful to anthropologists
working within neoclassical, ecological, Marxist, or interpretive paradigms.
Value is an elusive and difficult concept to operationalize in social science
research. Yet, the concept of value is indispensable in many anthropological
theories, whether they concern the energy maximizer of optimal foraging
theory, the laborer of neo-Marxist theory, or the symbolically enriched individual of interpretive anthropology. For example, the Soviet economist
Chayanov asserted that a key to understanding peasant agriculture is analysis
of their “subjective evaluation of the values obtained by [their] marginal
labor” (quoted in Durrenberger 1984:9). Here, I propose a method, derived
from utility theory in economics, of evaluating peoples’ subjective values
over goods. This method circumvents the particular assumptions of many
paradigms and may lead to greater testing and evaluation of their explanatory
worth. I also provide an example of the method’s application among Aymara
herders in the Peruvian Andes.
Economists often assume that people make decisions based on subjective
values of goods (Hirshleifer 1980:66; Cowell 1986:210). Utility theory was
designed to provide empirical measures of such values. Anthropologists can
provide crucial tests of this economic assumption by using utility theory to
measure value and then to see if such value is truly arbitrary and subjective or
if it correlates with underlying material factors such as labor inputs, material
wealth, energy, or reproductive fitness. In this way, utility theory transcends
different economic paradigms and, despite its genesis in neoclassical economics, can be used to evaluate the comparative usefulness of labor theories
Field Methods, Vol. 12, No. 4, November 2000 334–345
© 2000 Sage Publications, Inc.
334
Kuznar / ESTIMATING VALUE IN NON-WESTERN SOCIETIES
335
of value, sociobiological theories of value, materialist theories of value, or
culture-based theories of value. Economic anthropologists (Durrenberger
1984:12) and human behavioral ecologists (Cashdan 1990:15) have called
for such operationalizations of basic concepts in economic theory. My application among Andean herders indicates that material factors that influence
men’s and women’s prestige and ability to care for offspring seem to influence the values individuals assign to animals.
GENERAL UTILITY THEORY
General utility theory is a perspective on how people assign value based
on their ordering of preferences (von Neumann and Morgenstern 1944; Luce
and Raiffa 1957; Raiffa 1968; Myerson 1979). Utility is a measure of the satisfaction a person derives from choosing a good, a course of action, or a lottery that offers chances of attaining a good.1 Utility theory methods take
advantage of individuals’ subjective risk perception to derive values for
objects or decisions. The application of utility theory methods does not
require that a decision maker have any explicit idea of probability or make
mathematical calculations (Rapoport 1966:30; Raiffa 1968:274). More
importantly, application of this approach does not require that a decision
maker be concerned with maximizing any “objectively” measured currency
such as energy, money, or labor; this independence from assumptions of
what constitutes value gives utility theory methodology its potential for evaluating theories of value. Utility theory allows measurement of peoples’ subjective values of goods, or even of subjective states such as rank, self-fulfillment, and so on. The decision maker need only make decisions based on his
or her subjective perception of probabilities that certain events may occur.
Several assumptions underlie utility theory methodology, namely that a
decision maker’s preferences (over goods, alternative jobs, alternative
courses of action) are complete, transitive, and continuous (von Neumann
and Morgenstern 1944, Luce and Raiffa 1957; Myerson 1979).
1.
2.
3.
Completeness means that a decision maker can compare any alternatives
under consideration.
Transitivity means that if a decision maker prefers an alternative—A to B, and
B to C—then the decision maker will also prefer A to C.
Continuity means that a decision maker’s utility increases in a continuous
manner, such that if A is preferred to C, any option, B, which is ranked
between A and C, can be represented by a randomized combination between
A and C. A lottery is a randomized combination yielding probabilities of win-
336
FIELD METHODS
ning the goods A and C, and so the use of lotteries to elicit values is consistent
with the continuity assumption.
Provided a decision maker’s preferences meet these requirements,
researchers can use utility theory methods to monitor preferences and to
model decision making. Alternative methods of monitoring risk-averse
behavior and decision processes have been proposed in anthropology,
including methods based on risky decisions and linguistic analysis of decision sequences (Gladwin 1975, 1989; Quinn 1978; Ortiz 1980, 1983; Werner
and Schoepfle 1987). All of these methods, and the one I am proposing,
deserve more attention to delimit their relative strengths and limitations.
One methodology that has gained much notoriety in environmental economics is contingent valuation. Contingent valuation involves asking
respondents how much they would pay for some nonmarket good, such as
preserving a habitat or renovating a public landmark (Carson 2000). This
method became controversial because environmentalists and government
agencies began using the technique to levy fines against industries that pollute as in the Exxon Valdez oil spill (Portney 1994). Conservative business
interests enlisted their own economists to argue against the method.
The National Oceanic and Atmospheric Administration (NOAA)
appointed a special task force led by two Nobel laureates in economics (Kenneth Arrow and Robert Solow) to evaluate the scientific validity of contingent valuation. Their landmark report pointed out that contingent valuation
could provide realistic estimates of value for nonmarket goods, but only if the
methodology adhered to a number of strict guidelines. These guidelines
include face-to-face interviewing, using referendum or closed-ended questions, and using follow-up questions to ensure that respondents understand
the interview task, to name a few (Arrow et al. 1993:4611–13). The NOAA
panel findings did little to appease conflicting economists, and the debate
continues (for pro opinions, see Portney 1994; Carson 1997, 2000; Whittington 1998; for con, see McFadden 1994; Hanley and Milne 1996; Gatto and
De Leon 2000).
While contingent valuation may yield valid utilities under appropriate circumstances, and while it has also been applied in non-Western contexts in
Africa, Asia, and the Caribbean (Whittington et al. 1990; Whittington,
Lauria, and Mu 1991; Choe, Whittington, and Lauria 1996; Echessah et al.
1997; Whittington 1998), the method is further limited in ways that basic utility theory methodology is not. A typical application of contingent valuation
involves asking about the worth of nonmarket goods in terms of monetary
currencies such as dollars. In Western, industrialized settings where people
truly use currency as a medium of exchange, it is probably reasonable for
Kuznar / ESTIMATING VALUE IN NON-WESTERN SOCIETIES
337
people to estimate value in terms of currency. However, where currency is
not much used or not used at all, monetary currency is not appropriate. For
instance, when I made purchases among Aymara herders in Peru, they
insisted that payment be made in food (flour, sugar, rice) or utilitarian goods
(sleeping bags). Monetary currency was little valued because Peruvian currency devalues at an alarming rate, and markets where it could be used were
distant. One could conceivably use an indigenous currency (e.g., llama-units
in the Andes, or polished stone axe-units in New Guinea), but given the indivisible nature of such valuables, estimating the precise value of a good would
be difficult if not impossible (a respondent is not likely to accept a half a
llama or stone axe). Utility theory methodology provides interval-scale measurements of value without asking respondents to accept nonsensical payments or currencies they do not value. For these reasons, the method I propose in this article probably has broader applicability.
EXAMPLE 1—
CHECKING THE ASSUMPTIONS
The first step in applying utility theory methods among Aymara pastoralists was to establish whether a utility theory approach was warranted. First, I
asked herders, “Do you like some types of animals better than others?” Then I
had the herders order their preferences by asking, “Which animal do you like
best? Which animal do you like the least?” In addition, I had herders explain
the relative merits and problems of each herd species to gain a more thorough
understanding of why they had their preferences. The Aymara herders I interviewed in southern Peru had no trouble rank ordering their preferences
among different herd animals, satisfying the completeness axiom.
The second step was to test whether a herder’s preferences were transitive.
Testing transitivity was easy in my application because only three species of
animals were herded in each of the communities where I worked. For example, if a person stated that alpacas were most valued and sheep least valued in
step 1, then llamas had to be of intermediate value. I tested this by then asking, “Is a llama better than a sheep?” “Is an alpaca better than a llama?” “Is an
alpaca better than a sheep?” If they replied “yes” to these questions, they
demonstrated transitivity. All herders replied consistently with their initially
stated preferences, so their preferences were transitive. This method resembles paired comparisons and triad tests, in which researchers count incidences where items are paired or ask respondents to rank the similarity of
items (Weller 1998:390). These counts not only test for transitivity, but also
provide similarity values among items based on ranking scales established by
338
FIELD METHODS
the researcher (Weller 1998:392). However, utility theory methodology does
not depend on a researcher’s a priori assignation of a ranking scale since a
respondent’s own perceptions of probabilities are used to establish value.
To ascertain whether the continuity axiom seemed reasonable, I asked the
herders whether they understood the difference between having prizes of animals for certain versus lotteries that offered chances of winning animals. I did
this by asking them directly whether they understood the questions. I also
inquired whether they had played lotteries, which most Peruvian cities run.
All the herders affirmed their understanding of lotteries, and many had actually played them. The important point here is that respondents have an idea of
subjective probability, of the comparative likelihood of different events.
Since utility theory methodology is based on peoples’ subjective probabilities, no precise understanding of probability is required.
EXAMPLE 2—
MEASURING VALUE
Once I ascertained that utility theory assumptions were reasonable, I
could apply utility theory methods for measuring the value of animals. This
involves presenting people with different options, or lotteries, that have different long-run expected utilities. Expected utility is important because a
probabilistic return (e.g., a 50% chance of winning $100) either awards an
individual, or not, depending on whether an event (winning) occurs. However, if one were to repeat the trial many times, the expected value would
equal the prize, multiplied by the probability of winning (expected utility =
0.5 × $100 = $50). It is a person’s subjective evaluation of the probability of
winning a preferred item, and therefore that person’s subjective evaluation of
expected value, that enables a researcher to measure that person’s subjective
value of the item. The general formula that measures utility values for items
is the following: Value of intermediately preferred good = (Value of most
preferred good) × (Probability of winning most preferred good in a lottery) +
(Value of the least preferred good) × (Probability of winning least-preferred
good in a lottery). The probabilities must add up to one.
To obtain initial measures of the relative values of herd animals, I first
assigned the most highly regarded animal a value of one, and the
least-regarded animal a value of zero. Since I set the most- and least-preferred goods values of one and zero respectively, the formula can be simplified to the following: Value of intermediately preferred good = 1 × (Probability of winning most-preferred good in a lottery).
Kuznar / ESTIMATING VALUE IN NON-WESTERN SOCIETIES
339
I measured the relative values of the other species between one and zero
by presenting a herder with a choice between a species of intermediate preference for certain, or a lottery offering probabilistic chances of winning the
most- and least-preferred species. For example, if a herder most preferred a
llama and least preferred a sheep, I asked the herder, “Which would you
choose? A llama for certain, or a lottery ticket with 50% probability of winning an alpaca?”
If the herder preferred the certain prize, the llama, then I would ask the
question again, increasing the probability of winning the most-preferred animal until the herder had a difficult time deciding which was the better
choice—the certain prize, or the lottery offering a chance of winning the
most-preferred animal. When a respondent changed his or her mind about a
lottery, or replied, “I am not sure,” indifference was implied.
Continuing the example, if a herder was indifferent between having a
llama for sure or a lottery with a 70% chance of winning an alpaca (the most
preferred) and a 30% chance of winning a sheep (the least preferred), then
this indicated that the relative value for a llama is 0.70 = 0.70 (1, the value of
an alpaca) + 0.30 (0, the value of a sheep). The person’s preference ordering
can thus be transformed to an interval scale where the relative values of
alpaca, llama, and sheep are 1, 0.70, and 0 respectively. These are arbitrary
values, especially the assignation of zero to the least preferred animal that, in
fact, has some value to a respondent. This arbitrariness is not a problem for
modeling because the values can be changed with any linear transformation
(e.g., multiplying and adding constants) and still preserve their ordering
(Rapoport 1966:28; Myerson 1979).
DIFFICULTIES
Several difficulties may arise with this method. One concerns the probability perception of consultants. In some attempts to incorporate uncertainty
in economic anthropology, the probabilities people use to make decisions are
assumed to be accurately and objectively perceived quantities of the environment (Stephens 1990). Calavan (1984:58, 66) more realistically treats decision-maker probabilities as given quantities (accurate or not) based on a person’s evaluation of various sources of information. Similarly, utility theory
does not require decision makers to make any specific probability estimates
or mathematical calculations, respondents simply must offer their subjective
reactions to differing levels of probability. In my experience, most Aymara
herders are very familiar with evaluating subjective probabilities and readily
340
FIELD METHODS
make decisions and compare their desires with the probabilities of attaining
those desires (Kuznar 1991a, 1991b, 1991c, 1994). Since everyone must survive in an uncertain world, I suspect people in many non-Western economies
will reasonably be able to respond to questions concerning probabilities
(Calavan 1984:55). Once again, a respondent need not (and in fact should
not) have a mathematically precise understanding of probability to respond
in a manner consistent with his or her subjective feelings about choices.
Another problem is subtler, yet very important. A tacit assumption underlying utility theory methodology is that the data researchers receive reflect
peoples’ goals and aspirations. The necessity of separating goals (that may be
entirely culturally defined) from constraints (that may be physical, social, or
ideological) is central in models of decision making. Separating goals and
constraints may be difficult to achieve in the field since, when a person is presented with alternative animals, the person may state his or her preferences
not according to his or her aspirations but with consideration of the resources
that the person has available to support the animals in question. This problem
has been recognized and discussed by Cancian (1980:165, 1989:151) and
Ortiz (1983:276).
EXAMPLE 3—
SEPARATING GOALS AND ASPIRATIONS
A few herders demonstrated the problem of separating goals and aspirations with respect to their preferences for cows. Cows demand a large amount
of high-quality forage, and their stocking rates (amount of land required for
one animal) are at least nine times higher than those for goats in the region
where I worked (Kuznar 1991c:98, 1991d:110–11). Some families do not
have access to enough high-quality pasture and alfalfa to support a cow; the
option of having a cow is not feasible for them, although most would very
much like to own one. Therefore, some herders assigned a low value to cows,
reflecting their preevaluation that they could not raise a cow, not their desire
to own a cow.
I had to be careful that people were giving me their actual preferences and
not the result of their own optimizing calculations. Untrained researchers
unfamiliar with a respondent’s language and culture cannot apply utility theory methods in a formulaic fashion. It is crucial that a dialogue takes place to
ensure that all parties (anthropologists, translators, and people being consulted) understand the meaning of the questions asked. This dialogue goes
beyond a momentary rapport. There is no substitute for an ethnographer’s
Kuznar / ESTIMATING VALUE IN NON-WESTERN SOCIETIES
341
familiarity with his or her consultants and, perhaps more important, with the
culture in which he or she works. Only through such familiarity can ethnographers understand what questions have appropriate relevance in a particular cultural setting (see Werner and Schoepfle 1987 on questionnaires).
EXAMPLE 4—
AN APPLICATION
Using these methods, I examined risk aversion in two Andean herding
communities located in contrasting environmental settings; one located in
the high sierra (2,500 m–3,800 m above sea level), and one in the Andean
puna (above 3,800 m). The high sierra has deep valleys, abrupt terrain, seasonal rains, and highly unpredictable precipitation (Kuznar 1991c). Aymara
pastoralists follow forage up and down the sierra valleys seasonally with
their herds of goats, sheep, and cattle. Aymara herders in the puna keep indigenous herd animals, the llama (Lama glama) and the alpaca (Lama pacos).
The puna has extremely cold temperatures, rolling terrain, and pastures
restricted to marshes (Kuznar in press). Puna settlement is restricted to these
marshes.
Two trained Aymara assistants and I interviewed twelve herders among
the high sierra families (50% of adults in the community) and eleven herders
in the puna community (17.5% of adults). We conducted interviews in the
language with which a respondent was most comfortable to evaluate the relative values of the potential herd animals in each major environmental zone
(goats, sheep, and cows in the high sierra; llamas, alpacas, and sheep in the
puna). All male and most female respondents were fluent in Spanish, and one
woman was monolingual in Aymara.
High sierra pastoralists demonstrated a wide range of preferences for specific animals. In general, they preferred cows to goats, and sheep had the lowest preference. Some individuals preferred goats (average utility value .51,
maximum value 1) above other animals, others preferred sheep (average
utility value .27), and yet others preferred cows (average utility value .69, see
Table 1).
There were interesting gender differences in animal preference. Women
had a tendency to value animals with a direct domestic household use, preferring sheep because they provide useful wool and cows because they provide
milk. When men preferred cows, it was because of the cow’s high potential market value, or high prestige value. This is only a tendency, and women also noted
the high market values of these animals. Other researchers (e.g., McCorkle
342
FIELD METHODS
TABLE 1
Mean Utility Values of Herd Animals among Aymara Pastoralists
a
Sheep
0.51
0.44
12
0.27
0.44
12
Goat
Mean value
SD
n
a
Cow
a
0.69
0.63
12
b
Alpaca
0.59
0.314
11
0.98
0.034
11
Llama
b
a. High sierra utility values.
b. Puna utility values. The value of sheep for puna herders was uniformly zero.
1990:16) have shown gender to be a crucial variable in determining the goals
of Andean agro-pastoralists. Clearly, there is a dichotomy of values in which
high sierra women value animals that produce goods useful for immediate
household consumption, whereas men tend to prefer animals that provide
benefits in the public prestige and market sector. Men and women are not
unaware of the values of these animals for the opposite sex, but they have
their own gendered reasons for preferring certain animals.
Puna pastoralists almost universally preferred alpacas to llamas, with the
average utility value for alpacas being .98 (maximum value 1), compared to
an average utility value for llamas of .59. Puna pastoralists universally devalued sheep based on these animals’ inability to survive snow and predation
hazards. No significant gender differences in preference appeared to exist
among puna herders.
CONCLUSIONS
My applications of utility theory methods in the Andes allowed evaluation
of peoples’ subjective values of animals and testing of anthropological theories of risk aversion and utility maximization (Kuznar 1991a, in press, n.d.),
and these methods can be applied to any situation where completeness, transitivity, and continuity may apply. Since value presumably drives peoples’
economic decisions, and value is often asserted to be subjective, this method
could prove useful for anthropologists in two ways. First, the method, provided its assumptions are met, can provide measurements of subjective values. Second, by collecting measurements of subjective value and comparing
them to etic phenomena (kcal yields, number of offspring, monetary values,
etc.), anthropologists could potentially test competing theories of value.
Kuznar / ESTIMATING VALUE IN NON-WESTERN SOCIETIES
343
NOTE
1. If one were working among the Yanomamo, who count 1, 2, and more than 2 (Chagnon
1992), asking about lottery games would be pointless. Researchers use a variety of means to deal
with consultants’ lack of formal mathematical training. In one Brazilian case, previous interviews were used to see if the peasants could describe the likelihood of yield probabilities as possibilities out of ten (Dillon and Scandizzo 1978:426). In another study conducted in India, the
researchers based their games on coin tosses (Binswanger 1980:396). I could see the first
approach being most generally useful—discussing stochastic events with which people were
familiar, and then asking people to express the likelihood of their occurrence. This is no guarantee that people will understand probabilities, nor is it a guarantee that people who do understand
probabilities will be able to express what they know. However, these methods should identify
those who understand likelihood and have a vocabulary with which to describe their knowledge
and can therefore be used to evaluate their subjective values.
REFERENCES
Arrow, K., R. Solow, E. Leamer, P. Portney, R. Radner, and H. Schuman. 1993. Report to the
NOAA panel on contingent valuation. U.S. Federal Register 58 (10): 4601–14.
Binswanger, H. P. 1980. Attitudes toward risk: Experimental measurement in rural India. American Journal of Agricultural Economics 62:395–407.
Calavan, M. 1984. Prospects for a probabilistic reinterpretation of Chayanovian theory: An
exploratory analysis. In Chayanov, peasants, and economic anthropology, edited by E. P.
Durrenberger, 51–71. Orlando, FL: Academic Press.
Cancian, F. 1980. Risk and uncertainty in agricultural decision making. In Agricultural decision
making: anthropological contributions to rural development, edited by P. Bartlett, 161–76.
New York: Academic Press.
. 1989. Economic behavior in peasant communities. In Economic anthropology, edited
by S. Plattner, 127–70. Stanford, CA: Stanford University Press.
Carson, R., 1997. Contingent valuation: Theoretical advances and empirical tests since the
NOAA panel. American Journal of Agricultural Economics 79:1501–8.
. 2000. Contingent valuation: A user’s guide. Environmental Science & Technology
34:1413–19.
Cashdan, E. 1990. Introduction. In Risk and uncertainty in tribal and peasant economies, edited
by E. Cashdan, 1–16. Boulder, CO: Westview.
Chagnon, N. 1992. Yanomamo. 4th ed. Fort Worth, TX: Harcourt Brace.
Choe, K., D. Whittington, and D. Lauria. 1996, The economic benefits of surface water quality
improvement in developing countries: A case study of Davao, Phillipines. Land Economics
72:519–37.
Cowell, F. 1986. Microeconomic principles. New York: Oxford University Press.
Dillon, J. L., and P. L. Scandizzo. 1978. Risk attitudes of subsistence farmers in northeast Brazil:
A sampling approach. American Journal of Economics 60:425–35.
Durrenberger, E. P. 1984. Introduction. In Chayanov, peasants, and economic anthropology,
edited by E. P. Durrenberger, 1–25. Orlando, FL: Academic Press.
344
FIELD METHODS
Echessah, P., B. Swallow, D. Kamara, and J. Curry. 1997. Willingness to contribute labor and
money to tsetse control: Application of contingent valuation in Busia District, Kenya. World
Development 25:239–53.
Gatto, M., and G. De Leon. 2000. Pricing biodiversity and ecosystem services: The never-ending story. Bioscience 50:347–56.
Gladwin, C. 1975. A model of smoked fish from Cape Coast to Kumasi. In Formal methods in
economic anthropology, edited by S. Plattner, 77–128. Washington, DC: American Anthropological Association.
. 1989. On the division of labor between economics and economic anthropology. In Economic anthropology, edited by S. Plattner, 397–425. Stanford, CA: Stanford University
Press.
Hanley, N., and J. Milne. 1996. Ethical beliefs and behaviour in contingent valuation surveys.
Journal of Environmental Planning & Management 39:255–73.
Hirshleifer, J. 1980. Price theory and applications. 2d ed. Englewood Cliffs, NJ: Prentice Hall.
Kuznar, L. 1991a. Herd composition in an Aymara community of the Peruvian altiplano as a linear programming problem. Human Ecology 19:369–87.
. 1991b. Mathematical models of pastoral production and herd composition in traditional
Andean herds. Journal of Quantitative Anthropology 3 (1): 1–17.
. 1991c. Transhumant pastoralism in the high sierra of the south central Andes: Human
responses to environmental and social uncertainty. Nomadic Peoples 28:93–104.
. 1991d. El medio ambiente y la capacidad de carga de la sierra alta de los Andes sur central en el departamento de Moquegua, Perú. Diálogo Andino 10:100–12.
. 1994. Pastoreo en las sierras altas de la zona centro surandina: El caso de Moquegua,
Perú. Diálogo Andino 13:57–64.
. In press. Altiplano human ecology/altiplano past. In Quelcatani: The evolution of a pastoral lifeway, edited by M. Aldenderfer. Washington, DC: Smithsonian Institution Press.
. n.d. Risk aversion among Andean pastoralists: Models, measures, and empirical tests.
Unpublished manuscript.
Luce, R. D., and H. Raiffa. 1957. Games and decisions: Introduction and critical survey. New
York: John Wiley.
McCorkle, C. M., ed. 1990. Improving Andean sheep and alpaca production. University of Missouri, Columbia: Small Ruminant Collaborative Research Support Program.
McFadden, D. 1994. Contingent valuation and social choice. American Journal of Agricultural
Economics 76:689–709.
Myerson, R. B. 1979. An axiomatic derivation of subjective probability, utility, and evaluation
functions. Theory and Decision 11:339–52.
Ortiz, S. 1980. Forecasts, decisions, and the farmer’s response to uncertain environments. In
Agricultural decision making: Anthropological contributions to rural development, edited
by P. Bartlett, 177–202. New York: Academic Press.
. 1983. What is decision analysis about? The problems of formal representations. In Economic anthropology: Topics and theories, edited by S. Ortiz, 249–97. Lanham, MD: University Press of America.
Portney, P. 1994. The contingent valuation debate: Why economists should care. Journal of
Economic Perspectives 8:3–18.
Quinn, N. 1978. Do Mfantse fish sellers estimate probabilities in their heads? American Ethnologist 5:206–26.
Raiffa, H. 1968. Decision analysis: Introductory lectures on choices under uncertainty. New
York: Random House.
Kuznar / ESTIMATING VALUE IN NON-WESTERN SOCIETIES
345
Rapoport, A. 1966. Two-person game theory: The essential ideas. Ann Arbor: University of
Michigan Press.
Stephens, D. W. 1990. Risk and incomplete information in behavioral ecology. In Risk and
uncertainty in tribal and peasant economies, edited by E. Cashdan, 19–46. Boulder, CO:
Westview.
von Neumann, J., and O. Morgenstern. 1944. Theory of games and economic behavior. New
York: John Wiley.
Weller, S. C. 1998. Structured interviewing and questionnaire construction. In Handbook of
methods in cultural anthropology, edited by H. R. Bernard, 365–409. Walnut Creek, CA:
AltaMira.
Werner, O., and G. M. Schoepfle. 1987. Systematic fieldwork vol. 1: Foundations of ethnography and interviewing. Newbury Park, CA: Sage.
Whittington, D., 1998. Administering contingent valuation surveys in developing countries.
World Development 26:21–30.
Whittington, D., J. Briscoe, X. Mu, and W. Barron. 1990. Estimating the willingness to pay for
water services in developing countries: A case study of the use of contingent valuation surveys in southern Haiti. Economic Development and Culture Change 38:293–311.
Whittington, D., D. Lauria, and X. Mu. 1991. A study of water vending and willingness to pay
for water in Onitsha, Nigeria. World Development 19:179–98.
LAWRENCE A. KUZNAR is an associate professor of anthropology at Indiana University—Purdue University at Fort Wayne. His research interests include the ecology and
economics of traditional pastoralists, ethnoarchaeology, and quantitative methods. He
has worked in the Andean highlands and among the Navajo. His recent books include
“The Ethnoarchaeology of Andean South America” (edited, Monographs in Prehistory,
2000), Reclaiming a Scientific Anthropology (AltaMira, 1997), and Awatimarka: The
Ethnoarchaeology of an Andean Herding Community (Harcourt Brace, 1995).