Joint Session American Economic Association
and Association for Comparative Economics
December 28, 1968 ,
iiicro-Development: The Transformation of Village-Level Economies
RISK, UNCERTAINTY AND THE SUBSISTENCE FARMER:
Technological Innovation and Resistance to Change
in the Context of Survival
Clifton R. Wharton, Jr.
The Agricultural Development Council, Inc.
630 Fifth Avenue
New York, New York 10020
"In general, farmers who are limited to traditional agricultural
factors are more secure in what they know about the factors they use
than farmers who are adopting and learning how to use new factors of
production. The new types of risk and uncertainty about the yield
inherent in factors embodying an advance in knowledge are of real
concern to farmers. They could be of critical importance to farmers
who are producing so little that there is barely enough production
for survival. But since traditional agriculture is not introducing
new factors, new elements of risk and uncertainty do not appear; they
arise only when the transformation gets under way." [Schultz, 1964, p.31]
A. The Dynamics of Micro-Development: A Prologue
B. The Subsistence Farmer
I. The Responsiveness of Subsistence Farmers
II. Sources of Resistance to Technological Change
III. Risk and Uncertainty in the Context of Subsistence Agriculture
A. The Causes of Risk and the Sources of Uncertainty
B. The Role of Subsistence Standards and Levels of Living
C. Types of Subsistence Farmers
D. The Dynamic Interaction of Risk, Uncertainty and Subsistence
upon Technological Innovation
A. Some Puzzles Explained
B. Some Policy and Program Implications
RISK, UNCERTAINTY, AND THE SUBSISTENCE FAPRER:
Technological Innovation and Resistance to Change
in the Context of Survival
Clifton R. Wharton, Jr.
The Agricultural Development Council, Inc.
A. The Dynamics of Micro-Development: A Prologue
The transformation of village-level economies is essentially the trans-
formation of peasant farmers. Few peasant farmers in the world are in a state
of pure subsistence where they buy and sell nothing outside of their farms.
However, the linkages which exist in a majority of cases are more heavily con-
centrated at a village economy level than with the wider economy outside of
the village. Contact with the wider economy does exist in most cases, but its
relative importance contrasted with the interface market contacts at the
village level is less. Thus, these village-level economies operate much as
self-contained, self-sufficient economic enclaves with communal goals, institu-
tions and processes designed far more for the preservation of human life than
for development. Since in its earliest phases, sedentary agriculture is a
productive process whose product may be eaten by the producer, there is
inevitably a strong attachment by the peasant farmer to the goals, institutions,
and processes associated with the economy, society, and polity of the village.
Despite the almost infinite variety of village-level institutions and
processes to be found around the world, they partake of three common character-
istics of importance for change: (a) they have historically proven to be
successful i.e. the members have survived; (b) they are relatively static, at
least the general pace of change is below that which is considered desirable
today; and (c) attempts at change are frequently resisted both because these
institutions and processes have proven dependable and because the various
elements constitute something akin to an ecologic unity in the human realm.
Yet the very essence of development at this micro level is an increase
in the linkages of the village and the peasant with the _ider-_orLd.- Link'?
-'- --- "** -"--' -- ~^~-- ------ -- ------ ~
is the path of modernization and dynamism whether through a sudden export boom
or through the increased availability of non-farm produced inputs like chemical
fertilizer. As these linkages increase, the range forces beyonithe control
of the peasant farmer and the village inrerase as _ell. Increased dependence
upon the outside world without the corresponding development of countervailing
power foci which can be manipulated by the peasantry is often at the root of
rural political unrest. Moreover, not all of the increased linkages with the
outside world necessarily contribute to sustained growth.
The current conventional wisdom about agricultural development to which
most economists would subscribe is:
1. Agricultural development is a function of a large number of complex
interrelated factors -- natural, physical, psychological, economic, social,
cultural and political [iillikan and Hapgood, 1967].
2. Certain factors which significantly affect development are largely
exogenous and not subject to short-run manipulation. For example, the size,
composition, distribution and rate of growth of population has a major effect
upon agricultural development at all levels.
3. The factors which can be manipulated so as to produce a more dynamic
agriculture fall into two groups:
(a) Those which rely upon exploiting existing unrealized opportunities
and the eliinination o-existing ..economicinefficiencgiesithout major changes
in such areas as available technology existing infrastructure levels of demand,
current agriculturalrices or motiva-tians--and-spirit-of-people.- Programs in
this area seek to improve the decision-making of farmers, their allocation of
resources, improvement in marketing information, and the elimination of monopsony
or monopoly gains. While such efforts can contribute to change in agriculture,
they rarely lead to a more dynamic or sustained rapid rate of growth.
(b) Those which involve significant changes in one or several of the
factors held constant under the first. This category of factors perhaps contains
elements where a dramatic change has on numerous occasions in the past led to
dynamic or discontinuous jumps. These variables are the most likely to provide
the greatest short-run gains or "shock" to a static system. As a rule most
of these change slowly, but the surprising feature of recent history is the
frequency with which major and rapid breakthroughs have occurred in a number
of these factors.
All "developers" love to play the game of "key factors." Let me also
play the game and list the seven which I currently consider to be the major
forces capable of contributing to dynamic change in agriculture. 1/
1. Changes in Technology. When technically feasible and economically
feasible, the diffusion and adoption of new technology can often lead to dramatic
changes and development in agriculture. These changes can be reflected in
There are numerous lists of "key" factors ranging from Mosher's 
five essentials and five accelerators to the Adelman and lIorris  empirical
finding on the overwhelming importance of improvement in effectiveness of
financial institutions, improvement in physical overhead capital, degree of
modernization in outlook, and leadership commitment to development. For a
fuller presentation of their analysis see Adelman and Morris [19671.
increases in yields and reductions in per unit cost which in:turn lead to
increases in quantities sold, and returns to the farmers. Dramatic and
significant changes in the technology available to farmers can also have
an important derivative effect upon their perceptions of opportunity and motiva-
tion to change in general.
2. Changes in Infrastructure. / Major changes in infrastructure,
particularly irrigation, roads, storage, and other marketing facilities, can
often be of signal importance in effectuating dynamic changes in the agricultural
sector served. For example; irrigation which gives greater assurance of water
and improves its availability and distribution or which permits the most
effective utilization of other farm inputs can often lead to major sustained
change in agriculture. The two-way facilitation provided by roads and improved
transport can occasionally alter the input cost/product price ratios leading to
significant increases in output and yield. Similarly, their role in heightening
marketing efficiency and access to markets in general need not be belabored.
3. Changes in Demand. Major shifts in the levels of effective demand
for agricultural output have usually led to dramatic responses, even by the
most traditional agriculture. At times, these dynamic changes have been
caused by strong shifts in domestic demand due to increases in per capital
income. (There is an important unexplored question here regarding the role
of population on which views differ sharply.) At other times, these shifts have
been the result of changes on the export side.
4. Changes in "Prices" or the Terms of Trade. Major changes in the terms
of trade between agriculture and non-agriculture leading to an increase in
the real value of agricultural product can sometimes serve.as a major stimulus
SThe term "infrastructure" is used here in the broader sense set out in my
earlier work [Wharton, 1967].
to a dramatic take-off in agriculture. This phenomenon is more likely to
occur when the increases in income lead to patterns of capital formation and
investment in agriculture which strengthen its productive capacity.
5. Changes in Motivation of People. The motivations of people, not
merely farmers but also those serving farmers and the elite, are often central
to development. Attitude toward change and the spirit with which individuals
approach development can be affected by a number of variables producing a
heightened sense of dedication, willingness to make changes, altered patterns
of behavior and changes in the objective functions (goals) of people. Causes
of such changes in the past have been wars, achievement of independence, and
major revolutions which altered the balance of political power.
6. Changes in Institutions. 1/ Changing the "rules of the economic game"
and "access to the game" can often have a major impact upon the pace of develop-
ment. Despite the mixed evidence,land reform is usually cited as an institutional
change which can effect dynamism in agriculture. Similarly cited examples
are an increase in the political franchise of rural people and the development
of local financial institutions adequate for external linkage to national
capital markets. Such changes significantly alter the institutional context
within which the economic processes are carried out.
7. Changes in Knowledge of the Development Process. A final variable
which we tend to ignore is the change in knowledge of the development process.
We are apt to forget how much our knowledge of the process of development has
grown in the past twenty years. While we do not know all the answers, there is
little doubt that there has been substantial progress in what is known about
SThe conceptual delineation of "institutions" and their role in development
is a needed, though difficult, task in which my colleague Dr. Weisblat
heightening the pace of development. This increase in knowledge is itself per-
haps one of the most significant inputs for accomplishing change in agriculture
as it is brought to bear on policies, plans and programs for agricultural change.
The present paper is hopefully a contribution in the seventh area while
focusing on problems associated with the first.
B. The Subsistence Farmer
A growing literature is emerging regarding the subsistence and peasant
farmer. While conceptions and definitions differ in detail, the population
embraced by most definitions of a "subsistence farmer" tend to result in
basically the same group. The core group comprised by these definitions 2/ is
a farmer who (a) consumes most of what he produces and/or sells very little of
his product in the case market; (b) buys very few items to use in farming and for
-/ For a more detailed rigorous exposition on the universe embraced by the terms
"subsistence" and "peasant" see Wharton [1969, Ch. 2]. The use of the term
"subsistence" is'sometimes'preferred in vide of the pejorative connotations of
"peasant" in some societies, but as used, the term includes more than the "pure"
Concern for peasantry is of long-standing among anthropologists. (For re-
cent summaries see Wolf [19661, Nash , and Firth (1969].) Economists have
only recently focused their research upon peasant farmers [Wharton, 1969, Ch. 14].
/- The critical factor is that these definitions do not seek to delineate a
"subsistence" sector in the traditional dual model sense, but to identify a sub-
set of farmers who share common characteristics which lead to common patterns of
economic behaviour. However, some economists [Miracle, 1968] feel that the term
"subsistence" however defined is too broadto cover the diversity and heterogeneity
embraced by the term and may be misleading by focusing undue attention on merely
one characteristic. While I readily applaud any stricture against univariate
classification which masks the fundamental heterogeneity of any universe, one
should be equally prepared to eschew the counsel of nc classification at all.
Subsistence agriculture and subsistence farmers are rarely classified on the
basis of a single characteristic [Beal, et al., 1967], despite sewing and stuffing
of straw men by objectors. The critical issues are whether or not the classifica-
tion system is appropriate for the analytical purpose at hand and whether or not
the classification is appropriately determined. "Women" is an appropriate category
for certain analyses; "blondes," "brunettes," and "redheads" for another. Sex
may or may not be relevant; similarly, hair color. The classification can not be
divorced from the purpose of analytical usage. Any caution to the contrary is a
disservice to the cause of scientific inquiry.
(c) uses verylittle, if any, non-family labor; (d) employs a level of technology
which is usually described as traditional or primitive and which changes
slowly if at all; (e) suffers a level of living which while not close to the
physiologic minimum is still judged as meager by most standards; and (f)
operates within a decision-makingcontext-where the issue of family survival
predominates and therefore where his primar- goal of production is for home
consumption .*- ..-. t"I' for the market.
The close inter-relationship between minimal subsistence levels of living
and the prevalence of high degrees of risk and uncertainty. which are so
frequently associated with subsistence and peasant farmers is a neglected but
highly important influence upon the economic behavior of such farmers. 1/ The
influence of risk and uncertainty upon economic behavior in agriculture has
received occasional attention, but the predominant focus has been upon the
commercial agriculture of modern, industrial nations rather than the subsis-
tence agriculture of the developing world. The notable exceptions are the
pioneering essays by Porter , Dutia , Gould , Myren 
The problem of risk was frequently raised at the recent International
Seminar on Change in Agriculture at the University of Reading, England
(September 3-14, 1968), Five of the eight specialists syndicates mentioned
risk as an important variable significantly affecting agricultural change.
The conference studied over 50 case studies of agricultural development
around the world. See also Uharton [1963a; 1963b].
2 Cochrane , Hildreth , Johnson, et al. , Reutlinger ,
Tedford , and lUerrill .
- This has been particularly true of empirical work: USA [Dillon and Heady,
1961; Johnson, 1962; Davidson and iiighell, 1963]; Australia [Officer, et.al.,
1967; Officer and Halter, 1968; Anderson and Dillon, 1968]; France [Boussard
and Petit, 1967]; and Sweden (Wolpert, 1964].
and the collection of articles in the 1964 Indian Journal of Agricultural
The present article will attempt to review the available literature and to
explore the conceptual and analytical issues involved in the relationship
between subsistence levels of production, plus the usually associated subsistence
levels of living, and the influence of risk and uncertainty factors upon the
adoption of new technology. In brief, the article is an attempt to provide a
more rigorous framework for analysing the frequent statement about farmers in
developing areas that "they are so close to minimal human survival that they
resist innovation and change." No attempt will be made to develop another new
model of peasant economic behavior / or to present still another "Game
Theoretic" framework of 'man against Nature.' 3 Rather the object of the
exercise is to explore more fully the inherent meaning, elements, and dimensions
of "subsistence" and survival as viewed by peasant farmers.
Even though the remainder of the article will concentrate upon economic
factors, it should be emphasized at the outset that economic factors are
neither predominant nor exclusive in the adoption or nonadoption of new tech-
nologies -- new seeds, new practices, new crop combinations, new inputs. Among
farmers in the developing world.economic factors are important in the process
of the diffusion of new innovations, but non-economic factors are often equally
SIn preparing this paper, I regret that I have been as yet unable to obtain
two theses which have come to my attention by Sipra Das Gupta and Judith Heyer
dealing with certain aspects on risk in the context of Indian and East African
SThe models to which I am most partial are Nakajima's  and Krishna's
 since both try to take account of subsistence levels of living and
/- The game theoretic framework has been frequently used for decision-making
under conditions of uncertainty. For an excellent summary see Krishna and
critical. None of the analysis which follows should be interpreted therefore
as a denial of the significance of such factors. Indeed, the paper may help
to redress what has been a past imbalance of overemphasis upon non-economic
explanations for resistance to change.
I. The Responsiveness of Subsistence Farmers
The earliest characterizations of subsistence or traditional farmers
described them as technologically backward, with deficient entrepreneurial
ability, and with limited aspirations. The influence of limited aspirations
is best summarized in the colonial stereotype of the "lazy natives" who refuse
to work for an income beyond what they require for their subsistence require-
ments [Black, 1953, p. 536]. Economists labeled such behaviour as the "back-
ward bending supply curve of labor." Although they pointed out the existence
of (backward bending labor supply curves in both developed and undeveloped
economies, the sophisticated theorizing about leisure income effects and the
utility of money was not terribly appealing when applied to primitive and
partially monetized societies. Other social scientists who were more culturally
sensitive and empathetic viewed any behaviour which superficially appeared
to constitute a denial of greater economic gain as merely instances where non-
economic variables dominated and swamped economic factors favorable to
Interestingly these early views of limited or negative peasant responses
to economic opportunity were held by many individuals who were witnesses to
or participated in instances of massive "response" by subsistence and peasant
farmers to improved economic opportunities, The colonial period in many of
today's developing nations offer considerable evidence of economic responsive-
- 10 -
ness. 1/ In most instances, the economic opportunity was the dynamic develop-
ment of new markets in the colonizing (or imperial) home country for the
beverages, food and industrial raw materials which could be produced in the
colony. Some but not all of the rapid dynamic response in these cases can
be explained by the coercion of colonialists, or by the development of infra-
structure facilities by the colonial power or by a crop's promotion by
"chartered companies" and cartel-type agency houses.
More recently, the idea of an economically inert peasantry has been
subjected to serious challenge. First, there are those economists led by
Professor Jones  and Professor T. W. Schultz , who find ample
evidence that subsistence farmers are economic men who do maximize in the
utilization of their available economic resources given the available technology.
Such farmers may be operating at low absolute levels of production but none-the-
less they are optimizing at the ceiling of the available technological
possibilities. This group argues that what is fundamentally lacking is improved
technology. The obvious solution under such circumstances is to give first
priority to the development of new technology to alter the production possibilities
available to the subsistence farmer. Sizeable investment in new technology, if
successful, would enable the subsistence farmer to apply his economizing skills
at a higher absolute level of production.
Second, a large number of economists have been conducting rigorous
empirical research to determine whether or not such farmers respond to economic
incentives. Despite the varieties of empirical and analytical measures used
i Andrus, 1948; Furnivall, 1957; Hill, 1963; Myint, 1964; Berry, 1967;
McHale, 1967; Pfanner, 1969.
- 11 -
and crops involved, the overwhelming evidence indicates that subsistence and
semi-subsistence farmers do in fact respond to economic incentives [Krishna,
1965, 1967; PSAC, 1967, Ch. 9; Behrman, 1969; Bateman, 1969]. They increase
the production of those crops whose relative economic returns have improved
and decrease those which have become disadvantageous. Some of the observable
response has come as a result of greater intensification in the use of available
resources without any significant alteration in the existing technology; others
have come through the adoption of new techniques and practices. Whether or
not the agricultural sector as a whole is responsive to economic incentives is
still open to debate and being investigated by a number of researchers.
Despite all the new .evidence on the economic responsiveness of subsistence
farmers, there is equally ample evidence, usually in semi-anecdotal or case
study form, where farmers have seemingly not responded to an "obvious" economic
opportunity [Borton, 1967]. Explanations of such cases vary. Some analyses
rely upon non-economic explanations -- the indigenous culture militated against
the new practice; there were serious religious prohibitions which would prevent
the adoption of a new technology; higher production would disrupt the fabric
of the traditional society. Others find that upon closer examination the economic
advantage turned out to be illusory -- the landlord secured all the gain; the
moneylender skimmed off the cream; the government guaranteed price was not in
fact paid; the cost structure made the new innovation unprofitable.
All of these explanations have varying degrees of merit in certain
instances. Yet the current pressures of burgeoning population on world food
supplies have thrown into relief the need for more rapid economic responsiveness
and the more rapid adoption of available new technology if the race is to be
- 12 -
won. Improved understanding of the-resitancen to adopt or to respond are
becoming critical One set of explanatory variables which deserves more
rigorous study is the-influence-e-f-ris-&-and-uncertainty juxtaposed against the
subsistence levels of living and subsistence production of such farmers. /
-- -------------- ---/--
II. Sources of Resistance to Technological Change
New technology is becoming the current vogue on which rest the hopes of
agricultural development to meet the challenge of the food/population race
[Stakman, et al., 1967; Noseman, 1967; Rockefeller, 1968]. Investment in
agricultural research on the physical and biological aspects of the plants and
animals which provide sustenance for the three fourths of the world engaged in
agriculture has become the new magic key. The massive concentration of research
talent and funds in crash programs to produce rapid payoffs with dramatic break-
throughs in agricultural technology have become the newest entrant in the parade
of solutions for underdeveloped agriculture.
For the economist and the social scientist, however, new technology is
neither an automatic panacea nor always an automatic source of dramatic growth
[Johnson, 1968]. For the economist, true technological change is that which
involves an increase in productive efficiency such that a larger output is ob-
tained with the same input bundle of land, labor, capital. Since differences
persist as to the appropriate definition of "innovation" among social scientists
[Ruttan, 1959; Rogers, 1962], I shall use throughout the term "technological
change." The term will be used in the same sense advocated by Ruttan [1959,
p. 606] "to desigte hangs in the coefficients of a function relating inputs
to outputs resulting from the practical application of innovations in technology
and in economic organization." Thus "technological change" will embrace those
changes in the previous relationship between outputs and inputs in a productive
- 13 -
process such that a greater output is now possible from the same bundle of inputs
from any of the following sources singly or in combination: (a) a change in
the quality of the input, i.e. the new technology is embedded in the input such
as a higher yielding variety of seed, or (b) a change in the economic organiza-
tion or use of the inputs, such as a recombination of the same inputs in a
fashion which results in higher output per unit of input, i.e. new knowledge on
the timingof fertilizer application without any change in amount or (c) a new
practice of farm operation, such as the time of planting or depth of planting,
which would result in higher yields without any other change in the quantity
and quality of inputs.
Naturally, this "pure" or neutral notion of technological change which
appeals to the empirical economist interested in analytical rigor rarely
approximates the situation in the real world. This is especially true where
dramatic new factor inputs are available within which the new technology is
embedded but which cannot be obtained without altering the other inputs and their
There is little question that dramatic technological breakthroughs have
been occurring for selected agricultural crops which in turn have been adopted
byfarmers and have had a dramatic impact upon agricultural productivity
[Griliches, 1958; Barletta, 1967; Evenson, 1967]. But there are also numerous
cases where a technical breakthrough has occurred on the research station
without adoption or significant impact on aggregate agricultural productivity
having taken place [Castillo, 1963]. A marvelous case study of where break-
throughs were made on two crops by the same organization with differing results
is the Rockefeller Foundation program in corn and in wheat in llexico [tyren
1969]. Similar evidence abounds in the literature and case studies of develop-
ment where a technological breakthrough was resisted by the farmers leading to
- 14 -
what some have called a "technological gap." This "technological gap" is seen
by the governmental change agents as a potential source.of higher productivity;
but despite efforts to close this "gap," the farmers have not adopted the new
In such typical situations, resistances to adoption tend to fall into
three broad categories:
(1) those which revolve around the farmers themselves;
(2) those which are associated with the technology or innovation itself;
(3) those which are external both to the farmer and the innovation but
which are situational, environmental, or institutional [Hsieh and Ruttan, 1967].
In examining each category, two careful distinctions must be kept in mind:
First, a careful distinction must be made among the different viewers of
the phenomena in question, especially between the farmer and the change agent.
Second, an equally important distinction is best seen as the difference
between "subjective perception" and "objective reality."
The "objective reality" may be that a new technology will increase yields
by 25 per cent with a 95 per cent probability, but the "subjective perception"
of the new technology may be a 15 per cent increase in yield with only a 75 per
cent probability. It should be emphasized that there is no reason why the
"objective reality" could not in fact be higher than the "subjective perception"
in various cases.
When these two distinctions are joined, one can readily see the divergences
which can take place when a new innovation is being promoted and
is being considered. A hypothetical example is given in Table 1 on rice yields
per hectare with a new technology. Let us assume that the outsider in this
case is the extension agent. His subjective perception of the new technology
is a yield per hectare of 5.4 metric tons and a probability of 95 per cent but
- 15 -
the objective reality is 5.1 with a probability of 90 per cent. The same
phenomenon when viewed by the farmer may have a different set of values subjec-
tively and objectively. The reason for the difference between the two objective
realities could be that the objective reality as seen by the change agent is for
all farmers within his area whereas the farmer's is for his specific farm. If
the change agent knew the farmer's farm as well as the farmer does, the objective
realities would be the same. Since the change agent does not, their respective
realities diverge. In this case as well there is no a prior reason why the
farmer's objective reality should be lower; it could be higher. Similarly,
subjective perceptions can diverge for a number of reasons. For example, a
farmer who is an early adopter and who has never previously tried the innovation
and who has only seen it on a nearby farm demonstration usually tends to add a
significant discount factor to any estimates of yield derived from his immediate
observable experience (see below).
Finally, it must be emphasized that there are important interactions
between one or more of the three categories. For example, lack of awareness
among the farmers regarding the true merits of the proposed innovation (category
1) may be due to improper methods of extension education (category 3). Similarly,
if the extension agents do not truly understand the new managerial practices
called for with a new variety (category 3), they may fail to provide the farmers
with the required managerial skills (category 1).
/ These interactions are often further complicated by disciplinary boxes. For
example, Edel  reports two interesting cases of Jamaican fishermen where an
income maximization and risk minimization model explains certain patterns of
economic behavior (viz. choice between shallows and deep-water fishing) but which
does not become generally valid in explaining other actions (viz. acceptance or
rejection of spear-fishing and cooperatives) until community solidarity and social
structure are explicitly introduced.
- 16 -
TABLE 1 Hypothetical Example of Subjective Perception and Objective
Reality Regarding Rice Yields from Standpoint of Different
(metric tons per
3.2 tons 5.4 tons
4.8 tons 5.1 tons
- 17 -
The relevant variables within each of the three categories, like so much
of the development problem constitute a wide analgam. Hany of these variables
are traditionally considered to fall within the purview of a particular discip-
line or branch of science and range from the social psychological variable of
farmer attitudes and motivations to the agronomic variable of soil texture and
porosity. No attempt will be made at integrating all the relevant variables,
even limiting the range to the social sciences, because there is no generally
accepted social science model of change and development. Nor will there be an
attempt at an exhaustive listing of all the relevant variables. However, it
may be possible and useful to list the six sets of variables which experience
reveals as the most frequent reasons for the failure of farmers to adopt a new
recommended technology. Again, these are not mutually exclusive in any given
situation, but are a useful check list for any change agent.
(A) The Farmer -----
S I. Not Known or Understood. T new technology may not be known
by the farmers. Despte efforts of the change agents, the bulk of
the farmers may not have heard of the new technology. 2 The media
SThose who have read my previous work will be surprised at the absence of
peasant motivations and attitudes within this rubric. The more that I have delved
into the subject the more I am becoming convinced that the problem in this partic-
ular area is not so much determining the attitudes and motivations which are dis-
incentive to change or dysfunctional for development as determining the precise
ordering or weights of relevant elements in the preference functions of peasants
in a particular situation so as to prepare the proper strategy -- especially
selection of targets their timing and sequences. Motivational problems are much
more critical in the case of political elites, planners and those serving farmers.
/ yren  believes that the critical factor to offset risk aversion among
peasant farmers is "adequate information about the new inputs which are proposed.
This includes potential adaptability to the farmer's own land and climate and a
vast number of details about the techniques to use with the new crop or practice."
- 18 -
and agents are just not reaching the farmers to make them "aware" of
the new technology -- a step which is unavoidably the first require-
ment in the normal progression toward ultimate adoption [Byrnes, 1966].
Or the farmers may be in touch with the agents, but the methods of ex-
tension employed may be unsuitable or ineffective.
2. Not Within Farmers' Hanagerial Competence. The farmers may
have heard about the new recommended technology, but the comprehension
of what it can do or the effective utilization of the new technology
may require new knowledge and skills on the part of the farmers which
they do not have. Unless the farmers are provided with these new
improved managerial skills, they will not adopt the recommendations.
3. Not Socially, Culturally or Psychologically Acceptable.
A great deal is made in the development literature of those cases where
a new practice or a new technique has not been adopted because it
would upset too severely the established patterns of social or economic
or political organization. A new planting or harvesting practice which
is labor saving might eliminate the traditional labor service of the
wives or relatives. The "gleaning" rights of wives in the harvest
fields is a frequent example. The importance of such forces in the
subsistence and semi-subsistence context is indeed considerable as
has been amply established by the work of anthropologists.
(B) The Innovation
4. Not Technically Viable or Adequately Adapted. Very often
the new recommended technology has not in fact been locally adapted or
- A great deal is made in the literature regarding the divisibility and com-
plexity of an innovation and the extent to which it requires major restructuring
of the productive process, For convenience, these attributes will be subsumed
under "technically viable."
- 19 -
tested under conditions which more closely approximate those faced
by the farmer. The government experiment station may have a location,
soils, and climate quite different from that of the farmers. Subsistence
farmers are sufficiently shrewd to discern whether the new variety or
new practice seems to have had enough adaptive research and local
testing to meet their unique local needs. If the new technology has
not had adequate adaptation or if a small trial run on a "demonstration
plot" reveals that it is not in fact technically viable, community
resistance to the new technology quickly develops.
5. Not Economically Feasible. Probably the biggest single cause
of resistance to change is the unprofitability of the new technology
as seen by the farmer. In many cases the new technology may be
physically better but not necessarily economically better. A new
technology usually requires an alteration in the input mix -- either
in its utilization or more important in its composition. Equally
often, the new technology requires the purchase of additional inputs
to achieve the higher productivity -- and these inputs have a cost.
And when the farmer compares the expected output plus its associated
income with the additional costs of the input, the balance sheet
employing the new technology is found wanting.
(C) The "Externalities"
6. Not Available. Often the new technology is embedded in a
physical item such as seed, pesticide, fertilizer, or equipment. But
unless the new item is readily available to the farmer in quantities
and at the time when he needs it, knowledge of its potential contri-
bution to his agricultural production will not result in its adoption.
- 20 -
Adequate provision of the new inputs must be assured and most often
these come in a combination or "package" so that not one but all must
be readily available before adoption will take place. A fertilizer
responsive new seed might be available, but it will be of limited
value to the farmer unless fertilizer is also available.
These six are neither mutually exclusive nor the sole determinants of resistance;
they also frequently interact with each other in a particular situation.
Some additional comments are in order regarding technical viability and
The history of past technical assistance is littered with instances
where a "new technology" which is labelled as both "new" and (automatically?)
"better" has in fact proven to be unsuited and not technically viable. Some
of these failures have been due to the improper recognition of the non-transfer-
ability of temperate zone technologies to tropical and sub-tropical zones. Less
crude but equally common have been those failures due to an inadequate recogni-
tion of the need for adaptive research which will take a new technology and
reshape it to fit specific growing conditions. Still other instances of failure
can be attributed to "slippage" which occurs due to such factors as poor seed
multiplication or improperly trained extension agents. While such considerations
are of the greatest importance, in the analysis which follows they will be
ignored because for our purposes they do not really constitute a "new technology"
at the farm gate. The following analysis, therefore, will assume that the
technology being discussed is in fact "new" in the sense that it is technically
viable, technically superior (at least in so far as higher average yields), and
adequately adapted and locally tested for use by individual farmers. Even under
these circumstances there remains a sizeable question regarding the economic
feasibility of the new technology.
- 21 -
One frequent objection to the introduction of economic feasibility in the
case of subsistence farming is that product is not monetized or only partially
monetized. But the relevance of profitability still is involved even where the
introduction of new technology is not the issue. For example, Robinson
Crusoe did not try to reach the absolute maximum of total physical production
on his garden plot because his labor had a value in other "economic" pursuits.
This simple fact is too frequently forgotten by those change agents who
criticize the farmer who does not try to achieve maximum total physical pro-
ductivity on each square inch of Us acreage. If the farmer is asked to plant
a new variety which is fertilizer responsive, the seed perhaps and certainly
the fertilizer will cost him money. If the new variety requires improved water
control, then he must invest in better distributaries or water controls or even
improved levelling of his fields -- which again involves a cost. If he is
asked to plant a shorter term variety, he must weigh the increase in output
(and hopefully income), against the added cost of its harvest during a period
when additional drying and/or storage costs may be incurred. Moreover, the
costs incurred by the low-income farmer are often greater than the nominal price
of the new input. If he is short of capital and must borrow, the interest
charges must be added and such charges may be considerable where money lenders
or merchants exercise monopoly or monopsony powers [Uharton, 1962; Long, 1968a].
/ This area is one where I suspect I may differ with my colleagues in economic
anthropology [Dalton, 1961]. While there may be differences as to the goals of
production and distribution in such situations, I would argue that the means
employed within the firm still follow economizing rules. For further treatment
of the issue in a cross-disciplinary context see Wharton [1969, Ch. 14].
- 22 -
Thus, the peasant farmer must compare the increased returns (greater
output times expected prices) against the increased costs (new required inputs
times known costs) before he is able to make a decision on the economic
feasibility of the proposed technology. If he feels that the new technology
is not economically viable for him, he will not adopt it. Any careful review
of the myriad case studies of resistance to change will unquestionably show
that the economic viatbiit- of h e proposed change has been a most important
determinant of adoption or of resistance.
Among the economic forces influencing technological adoption are the
variables risk and uncertainty which have been a somewhat neglected dimension
of resistance to change among subsistence farmers. The primary focus of the
remainder of the paper will be on the influence of risk and uncertainty and
their dynamic interaction with two sets of economic variables: (a) the absolute
levels of farm living as they relate to societal standards for the minimal
levels of subsistence and the average levels of farm productivity; (b) the
subjective (farmer's) expected variance in output associated with a proposed
new technological introduction compared with the historically determined
variance in output utilizing traditional practices.
- 23 -
III. Risk and Uncertainty in the Context of Subsistence Agriculture
A. The Causes of Risk and the Sources of Uncertainty
The basic conceptual distinction between "risk" and "uncertainty" owes its
origins to Frank H. Knight  where he set forth the notion that "measurable
uncertainty" to which one can assign numerical probabilities should be called
"risk" and that true "uncertainty" is where numerical probabilities can not be
applied. Knight's pioneering efforts have spawned a sustained debate on the
further refinements of the distinction, on the determinateness of the probabili-
ties, on questions of ambiguity, and on similar issues.
Rather than attempt to cope with the very extensive issues and arguments in
the literature on the validity of the distinction between the two concerts, I
would like to refine the precise content of these two concepts for the subsis-
tence farmer as decision-maker. Thus far, the arguments in the literature
have revolved around a distinction which focuses upon the ignorance of the
decision-maker as to the statistical frequencies of events related to his de-
cision. The assumption in most cases of "uncertainty" is that the decision-
maker can not assign a priori calculations to determine the probabilities or
where the relevant anticipated events are considered to be unique or of the
The two basic
distinctions which are relevant for the decision-making framework of the sub-
sistence farmer are (a) those future events to which he can assign probabilities
Despite the extensive debate on the point, I will assume throughout this
paper that peasant farmers behae rationally [Wise and Yotopoulos, 1968] and are
utility maximizers. The critical question then becomes the content of the
objective function and the relative weights assigned, especially to such
elements as profits (net returns) or security or social prestige.
- 24 -
based upon oast experience or personal knowledge, and (b) those future events
to which he can not assign probabilities or where the probabilities offered are
not those derived from his personal experience and which are based upon external
knowledge provided by others. Both can be said to have probabilities, but the
distinction being drawn is based upon the source of the estimates. In the former
case it is the farmer himself; in the latter, it is usually an outside agent.
who uses traditional techniques
SThe peasant farmer/has a set of customary probabilities regarding the out-
domes of his farming decisions which are based upon his own past experience and
that of his forebears. The farming practices, the technologies applied to his
production,havesignificant and predictable consequences not only for any given
year but on the average among several years. The proof of his skill in determin-
ing these probabilities is the simple fact of his family's survival generation
Even the most illiterate peasant farmer has a knowledge of the probabilities
which attach to his current, traditional practices. His current practices have
both a risk and an uncertainty dimension which relate to three major sources of
year to year variability in output:
First, the farmer faces yield variability. The acreage which he plants to
a given crop may be the concrete summation of his estimate of a variety of forces,
but the actual yield obtained is not solely dependent upon acreage but a whole
host of intervening factors between planting and harvest -- some subject to the
control of the farmer; others, in the laps of the Gods. Actual field or barn
yields obtained are a function of a wide range of variables -- sunshine, humidity,
rainfall and even their incidence and timing during the cropping season; pests
such as birds, rats, worms; blights, fungi and viruses; and even the unpredictable
- 25 -
acts of God and man such as wars, insurrections, and revolts. All of these forces
converge to determine the actual yield which will be obtained in any given crop
year. (Post-harvest losses while stored, processed or transported could also be
included if the loss is borne by the farmer.)
Second, there is cost variability. Despite the predominance of subsistence
and semi-subsistence type agriculture which heavily involves "subsistence
production," there are still inputs required for production which are/purchased.
These items range from minor farm tools and fertilizer to oxen rental and hired
labor. Whether or not actual "cash" is employed in the payment process is in-
consequential. The critical issue is the variability in the incidence of such
costs. The typical farm decision-maker faces a number of such expenses which
tend to fall into two categories: those which are endogenous in the sense that
they are subject to his decision-making control and those which are exogenous in
the sense that their costing and incidence are outside his control. In both
cases, however, predictable and unpredictable probabilities are involved, i.e.,
risk and uncertainty. Family labor is fundamentally an endogenous variable
whose utilization and remuneration is subject to the control of the farm decision-
maker. But, its utilization is affected by illness and even the availability of
off-farm employment. There are a number of exogenous factors which are subject
to year to year variability and for which the farm operator must predict specific
values in advance of or during the crop season. The costs of farm product pro-
cessing such as milling, and off farm storage and transport to market are equally
exogenous from the standpoint of the farmer. Knowing what these costs were in
Day's  rigorous study of yields using U.S. data showed that field crop
yields were non-normal and that the degree of skewness and kurtosis depended upon
the amount of available nutrients. Similar findings for yields in developing
areas havenot yet been established, but their apparent significance should not be
lost as one observes the necessary increases in fertilizer application required
by the newer varieties. The point is not that increased fertilizer application
should be avoided but that its effects upon yield distribution characteristics
should be recognized.
- 26 -
previous and the most immediate years helps him in formulating the probabilities
to assign to numerical values for the current year, but they are inherently ex-
ternal to his personal control. The same applies to production and consumption
credit, both before the crop is planted, during the crop season and after harvest.
Third, there is a product price variability. Choice of crop and crop com-
binations as well as intended levels of output are based upon price expectations
-- relative prices expected to obtain at harvest or time of product sale. The
actual price forthcoming is of course outside the control of the farmer (in the
usual textbook sense of an infinitely elastic demand curve or price being given).
The divergence between expected prices when crop choice and planting decisions
are made and actual prices may be considerable, both positively and negatively.
Free market forces, monopsony situations, government controls, price floors or
guarantees and similar factors may come into play. These calculations are further
complicated by seasonal, cyclical, trend or random forces.
The critical element is that these three variabilities combine in any given
crop period to affect the net return to the farm family. The extent to which the
farmer can reduce unintended fluctuations in each category is quite limited, but
every effort is made to reduce those subject to his control. Historical know-
ledge on the past variabilities in each does exist and he takes these into account
whether it is distrust of assured government prices or a locust cycle.
Under the usual conditions of traditional or subsistence agriculture, the
farmer's knowledge of the average yields and their probability distribution is
An interesting question is the extent to which the choice of crop and crop
combinations may not be wholly price determined for the semi-subsistence low-
income farmer who must first assure an adequate food supply for his family and
who is reluctant to be dependent upon market forces. A few economists have
argued that maximization in labor use and in farm family own farm inputs may be
the primary goal in production in cases of subsistence agriculture [Wharton,
1969, Ch. 141.
- 27 -
greatest. When a new technology or crop is introduced it affects the first two
variabilities most significantly.
The risk and uncertainties associated with these variabilities are the normal
decision-making and operational environment for agriculture generally. In the
case of subsistence and semi-subsistence agriculture, however, these variabilities
combine with low levels of income (output) or "subsistence" levels and standards
of living to produce a much stronger "survival" element in decision-making.
"Even though the average annual yield of the new factor is substan-
tially higher than that of the old factor which it replaces, it may vary 1/
much more from year to year because of weather, insects, and other pests.
Moreover, the true yield variability of the new factors from these sources
will not be known, while that of the old factor is well known from ex-
perience over many decades. Thus there would be inherent in the prospec-
tive yield of the new factors these new elements of risk and uncertainty.
They must be taken into account, too, in determining profitability,
especially so since farmers in a poor community are less able, in terms
of reserves and experience, to cope with such additional risk and un-
certainty than are farmers in high income countries." [Schultz, 1964,
Compared with commercial agriculture, subsistence agriculture has a stronger
X// "risk aversion" and "security preference." This fact results in economic choices
and institutions to reduce risk and to increase security ranging from the selection
of drought resistant varieties [Myren, 1964, p. 971 to reliance upon the extended
family (Wolf, 1966, p. 67] or upon the village [Georgescu-Roegen, 1969] to provide
In the present instance we would like to explore further the exact nature of
this interaction between risk and subsistence living so characteristic of peasant
dnate  has some interesting empirical evidence of increased variance
to be observed with higher yielding varieties of rice compared with the tradi-
- 20 -
B. The Role of Subsistence Standards and Levels of Living
A previous article [Wharton, 1963b] sought to clarify the ambiguous use of
the term "subsistence" by distinguishing between "subsistence production" and
"subsistence levels of living." Though both "subsistence production" and "sub-
sistence living" frequently coincide, they are not necessarily synonymous. A
model was presented to explain how minimum standards of subsistc e li"vi are
determined at any point in time and how they are changed through time. The
model was intended as an aid in describing and analysing the influence of mini-
mum standards of subsistence living at the micro-economic and individual farm
family level. The model may be briefly summarized.
S In any given socio-economic unit, there exists some social consensus as to
the standard minimum bundle of goods required per person per specified period of
time. The bundle is above the level needed for physical survival, i.e., no
marked deterioration causing death prior to the normal life span in that particu-
lar social economic unit. The absolute size and composition of such a bundle is
bounded on its lower side by the physiologic limiting requirements for human
survival, but the values above the lower limit are controlled by economic-socio-
cultural factors. There are subjective notions of minimum subsistence standards
of living and the content of a minimum standard varies between social aggregates
depending upon historical and current economic-cultural factors. The higher the
previous economic levels of well-being historically, the higher will be the cur-
rent subsistence living standards. Such a variable can be viewed as operating
in some lagged fashion, i.e., the current minimum standard is some function of
previous economic levels weighted in some fashion. Thus, a short-run drop in
- 29 -
current economic levels need not result in a lowering of the minimum standard;
such a change would come only as the result of prolonged and sustained drops in
The minimum subsistence standard of living per person within any particular
social aggregation of individuals such as a village, can be described as a function
of three types of variables: (1) physiologic and nutritional; (2) economic, es-
pecially current and previous levels of economic well-being; and (3) socio-cultural
in an historical and cross-cultural context.
For an individual in a specified homogeneous social aggregate and for any
specified time period the relevant variables may be symbolically defined as
Sas = achievement standard of living
L = actual level of living
Sms = minimum subsistence standard of living
Pm = minimum physiologic requirements (below which death)
The minimum subsistence standard of living, Sms, cannot be less than Pm, the
physiologic minimum below which death takes place, but note that Sms could still
be low enough to cause nutritional deficiencies and result in reduced capacity
for physical exertion.
1/ "Standard" and "level" are carefully distinguished. The latter refers to
the actual, existing components, whereas the former refers to the desired
2/ For a variety of reasons, I consider Nakajima's revision of my equation'an
improvement [Nakajima, 1969]. Sms = Pm + g(E, C) g O where E = economic well
being variable and C = cultural variable.
3/ Such minima have a long history in the literature. For some recent examples
see Mellor , Fei and Ranis , Miracle [19681, and Nakajima .
The role of such minima in determining an "agricultural surplus" has been the
source of debate among anthropologists [Dalton, 1960, 1963; Orans, 19661.
- 30 -
Any particular social aggregate, farm family or village, has an Sms, a
general consensus as to what is a minimum subsistence standard of living per
person. However, the Sms need not be the same for all social aggregates within
the same nation or geographic area and need not be the same for a given socio-
economic unit through time.
It is important to note that Sms is different from the traditional notions
of a standard of living as a goal which the individual always strives to reach.
The subsistence standard is a something which the individual strives to avoid
or not drop below. Thus, if we call the traditional standard the "achievement
standard of living," Sas, and the actual level of living per person, L, then
L is usually equal to or above Sms, the minimum subsistence standard, but always
below Sas, the achievement standard of living. Sms in turn is above the physio-
logic minimum, Pm. The relationships can be summarized as:
Sas > L > Ss 1 P
A central argument to much of what follows regarding risk is that the sig-
nificant operative variable with important behavioral implications is not Pm'
but the relationship of L to the Sms, the minimum subsistence living standard.
Many economists when talking about subsistence levels of living are usually
thinking of Pm, the minimum physiologic requirements, which are in fact fairly
constant from person to person, once account has been taken of age, sex, physical
activity, and climate.
Attention has also tended to focus upon the relationship between actual levels
and the achievement standard. In our view, however, the relationship between
actual levels and the minimum subsistence living standard may be even more important
I/ For an interesting recent critique of these standards from a nutritional
standpoint see Oshima .
- 31 -
in the case of subsistence and semi-subsistence production. This is particularly
true where the actual level of living is close to the minimum subsistence standard
jand where in turn both are very close to the physiologic minimum.
Behaviourally, the individual strives to attain Sas, the achievement standard
of living, and to stay above Sms, the minimum subsistence standard of living. The
closer that the actual level of living, L, of an individual comes to his minimum
subsistence standard, Sms, the more strongly it affects his behaviour to avoid
reaching it. Whenever an individual in the social group is at or below the Sms
for that group, he is aware of it and will strive to get and stay above it.
The same comments are valid for a "family group" as well as for an individual.
Since S is minimum subsistence standard of living per person, we can easily con-
vert this into a minimum subsistence standard of living per family by taking ac-
count of numbers in the family or operative social unit where the number of persons
in the family (the "common pot" definition) is counted using some standardized
adult unit basis. Thus family "numbers" would not be strictly the number of family
members, but a composite variable which takes account of such compositional factors
as age and sex. Thus, when dealing with a typical peasant or smallholder type of
agriculturist who predominates throughbuthbe underdeveloped world, he and his family
will have a family minimum subsistence standard of living as well as family achieve-
ment standard, and a family actual level of living.
Similarly, each of the standards, Sms and Sas, and level of living, L, can
be given a valuation per family per specified time period such as the total dollar
expenditures per year per family for an actual level of living.
What is particularly relevant for the present discussion are those farmers
(a) where the bulk of their actual level of income comes out of their farm
1/ Note similarity to Chayanov's concept of "self-exploitation" [Chayanov,
1966, Ch. 2].
- 32 -
production (i.e., those who are toward the pure subsistence production end of
the spectrum) and (b) where their actual level of living and minimum subsistence
standard are close to the physiologic minimum. For such farmers, the forces of
risk and uncertainty have a special influence.
C. Types of Subsistence Farms
Let us assume an area of subsistence farming (a) where the total product is
food, (b) where there are no cash production costs (such as taxes) and no pur-
chased inputs (land and labor are the major factors of production, and minor
equipment or farm capital structures constitute an embodiment of farm family
labor), and (c) where the annual level of living for the farm families is entirely
out of farm production. For simplicity, valuation of production and levels of
living (consumption) can be made in units of a food staple such as kilograms of
rice. Under these assumptions, L,which is a consumption variable and which is
traditionally a function of income ,becomes virtually a direct function of output.
Actually total output less production costs equals farm consumption plus farm
production sales plus savings. But we will ignore production costs and savings:
(a) We have assumed above that cash production costs are zero. (b) There is
some evidence of a constancy in the saving rate among peasants which "is consis-
tent with the fact that the marginal utility of consumption does not decrease very
fast in the relevant range of income per capital" [Lau, 1967, p.25]. The more
critical question concerns the relationship between consumption and sales. There
are some who argue that peasant farmers have a target farm familydemand out of
farm production and that sales are a residual. Others [Mathur and Ezekiel, 1961]
believe that such farmers have a target cash requirement (for non-farm produced
consumption goods, ceremonial or religious expenditures, taxes, and interest pay-
ments) which must be met thereby making consumption the residual.
- 33 -
Regardless of one's position on these contrasting views, there is little
doubt that the larger the fraction of total farm product consumed by the farm
family, the closer the direct functional relationship between output and con-
sumption or level of living.
This characteristic is of critical importance because it determines the
degree to which decisions regarding output, consumption, savings, and investments
are inter-related. T' ;u:': susstence peasant household output and consump-
tion are virtually equivalent so that "income" and consumption are jointly depen-
dent.Several models of peasant economic behavior relying upon this inter-related-
ness of production and consumption have been developed [Sen, 1966; Nakajima, 1969;
Krishna, 1969]. Each follow fairly similar assumptions regarding utility, pro-
duction, etc. and demonstrate that even in pure subsistence cases, farm output
is a function of the traditional production variables and level of technology
employed. Hence, the actual annual level of living experienced by a peasant
family through time is a function of the absolute level and quality of resources
employed in production and the technology employed. The closeness of this level
of living to the physiologic minimum and to the minimum subsistence standard is
consequently closely related to the farmers resource and technological endowment.
Under such circumstances, it would be possible to discern three types of
farmers or three different situations depicting the interrelationships through
time of levels of living, achievement and subsistence standards, and physiological
minima (Figure 1). (The farmers are all assumed to have the same numbers/sex/age
composition so that the minimum, Pm., is identical.) In situation A the typical
low-income subsistence case, the range among all four variables is quite narrow.
Sms is close to Pm and L is close to S'ms In this case, physical survival con-
siderations are paramount. In situation B, the intermediate case, Sms is
- 34 -
.. -- -- ... ..... .. ... ....
-A. A A A -A.A. AAA A.A-A 4-A********
- r- .-- -
" ------ --- ---. .
.: .T....... i m ...... ... ....
I ***** *AAA*' AAAAAAAA AAA A*-**
- 35 -
substantially higher than Pm. This may be due to a variety of reasons such as
larger or higher quality resources or better technology which provides an average
annual level of living considerably greater than the Pm. In this case, notions
of a minimum subsistence standard are dominant. In situation C, the high-income,
commercial case, Sms is higher than Pm, but now an increased difference between
Sms and L is evident. In this case, the achievement standard and the actual level
of living assume their normal roles as determinants of behavior. It must be
emphasized that in the real world, these three types are not discrete and that
there is in fact a continuum. However, the three cases highlight those situations
where the dominant behavioral considerations are different.
Our analysis is particularly concerned with farmers in type A and B situations
especially the former.
D. The Dynamic Interaction of Risk, Uncertainty and Subsistence Upon
Technological change occurs only when the innovation is actually adopted by
the farm operators. Since the decision-making or choice context of the farmer is
at the root of the adoption process, any new or suggested technological intro-
duction or innovation must be viewed from the standpoint of the farmers.
Any new technology or practice has associated with it certain expected
probabilities regarding yields per acre and their associated income figures.
These estimates are primarily derived from the results of experiment stations
and varietal trials. The variations in yield observed on the various trials
on the experiment station plots and in farm demonstrations need not be, and
1/ While field trials and farm demonstrations may help to reduce the subjective
variance which a peasant farmer may attach to the new technology, some variance
still remains. Thus the logic of the argument remains -- each farmer must still
live with his individual risk.
- 36 -
usually are not, those which will be experienced by the particular farmer who
is contemplating the new adoption or innovation. Associated with each experiment
station and field or farm trial are certain expected variations in
results, i.e., each practice or each new variety has a certain predictable varia-
tion in actual yield even under the conditions of the experiment station or field
trial. But this expected variance in yield is not the same as the variance (ex-
pected or derived) from the application of these techniques or new inputs on a
particular farm. Moreover, change agents who are promoting the new technology
usually sell it in terms of the maximum experiment station yield or the average
yields obtained by the station. "This new variety gave twice the yield which you
(the farmer) are now getting."
The typical subsistence or peasant farmer has his own subjective rate of
discount for the introduction of a new technology on his farm. Despite its
testing by the central, regional or local experiment station, or even on neigh-
boring farmers' fields, the. subsistence farmer has learned from bitter, historical
experience to be wary of "technological gifts," which as he sees them have been
insufficiently adapted and evaluated for his particular situation. The subsis-
tence farmer has learned that any new recommended technological introduction has
associated with it a different expected variance on his fields -- a variance
which may be wider than that on the fields of the research station. Under these
circumstances the determining fact is the comparison between the expected variance
of the new technology and the known variance of the traditional technology. When-
ever a change agent approaches a farmer with a new technology the farmer examines
the proposed changes within the framework of known vs. unknown expectations. He
compares the expected probabilities from continuing as he has done in the past
with its given degrees of risk and uncertainty, on the one hand, with the expected
- 37 -
probabilities from the new technology on the other. Critically important in the
latter is the fact that the probabilities are those provided externally to his
experience -- the extension agent tells the farmer that the average yield from
the new practices will be "x," but the farmer does not necessarily know that this
is so from his own directly observable experience. In the Knight sense, what may
appear to the extension agent as a risk in the calculable probability sense based
upon experiment station data, becomesfor the peasant farmer an "uncertainty."
Utilizing our earlier terminology, both the subjective perceptions and objective
of the farmer and the extension agent
realities/diverge. A critical element thus becomes the "degrees of belief" or the
firmness of reliance which the farmer can place upon the estimated probabilities
associated with any particular decision for a new course of action.
Let us assume that we are dealing with a type A subsistence farm family which
consumes 80 per cent or more of their annual average farm production. Let us
further oversimplify by assuming that this 80 per cent of production equals the
total bundle of consumption goods which constitutes the level of living for the
family. In good years and in bad the amount of production reserved out of the
staple food production on the family farm remains fairly constant. Thus, in good
years he sells a higher fraction of his total production and in poor years, less.
The critical factor is that at all times he seeks to preserve or guarantee the
minimal subsistence living of his family out of production; consequently, the
absolute minimal production reserved for family consumption remains fairly con-
stant. This level of consumption is related to my concept of a minimum subsistence
1/ We could use a food staple as rice as the sole product or numeraire.
2/ This statement is not quite accurate as is shown in my earlier formulation
where the current minimum subsistence standard is a function of a number of vari-
ables including previous levels of income. It also oversimplifies in.ignoring the
obvious effects of the income elasticity of demand for farm produced food which
though declining as development proceeds is still quite high for low income
producers [Stevens, 1965, p. 55].
- 38 -
standard of living described earlier. Given a particular Sms, a typical peasant
family will strive to reach the communal or societal achievement standard, Sas'
and most important it will struggle to avoid the Sms which is the minimum sub-
sistence standard. Even though the Pm is considerably lower than most individuals
realize, the crucial fact is that each farm family resists having a level of
living which is significantly below the Sms. In other words, it is not so much
whether the ingredients are nutritionally adequate for human physical survival
-- viz. roots and berries -- as it is the accepted community and societal standards
The next element in the farmer's decision-making package are the expected
variances in yield and hence in output and income associated with existing prac-
tices. The agricultural technology currently employed by most subsistence and
peasant farmers is traditional in the sense that it has been handed down from one
generation to the next. The farmer has a basis for determining the expected
variance in yield and in income derived from the experience of his father,
grandfather, and great grandfather. Hence, he knows what to expect in good,
average, and poor years. On the whole, these expected variances are not only
real, they are the product of centuries of experimentation by trial and error.
The final ingredient in the picture is the expected variances which attach
to the new technology and practices. The variances in yield are those which are
derived from the experiment station. From the farmer's viewpoint there is no
1/ An interesting related observation is the frequent finding that in areas
which suffer occasional droughts the local traditional varieties are often
drought resistant at the expense of not providing high yields in good years.
The farmers have chosen varieties which offer an insurance against severe loss
or ruin. My colleague, Arthur T. Mosher, has also pointed out another inter-
esting example from North India:where irrigation is not available the farmers
follow a practice of sowing wheat and gram or barley and gram mixed together
in the same field. If the season is "good" (adequate rainfall), the wheat or
barley will yield well. But if the season is unfavorable, the gram will still
mature and help compensate for poor wheat or barley yield.
- 39 -
guarantee that such variances will apply to his particular, unique combination
of ecological and human factors. Thus, even though the average expected yield
may be considerably higher than his average yields with current varieties and
practices, the variance in expected yields with the alternative technologies as
as viewed subjectively by the individual farmer are far more important in
determining the adoption of the new seed, practice or factor input.
The dominant failure in much of developmental assistance aimed at techno-
logical innovation has been the lack of understanding about the relationship
between the (a) expected variance in yields using current varieties and practices;
(b) the expected variance of yields using new varieties and practices; and (c)
and the minimum subsistence standards
relationship of these to the absolute levels of living/of the "clientelle."
Despite the considerably larger expected average yields from
new seeds and practices, the'critical issue for the peasant farmer is the ex-
pected variance in output under the new regime compared with the old. If the
expected variance in yield per acre shows a possibility of an output below
his Sms, then the farmer will resist the introduction and adoption of the new
change. He will not wish to gamble or to risk the life of his family. He will
prefer to stay with the present practice with which he is familiar and which is
known to him than shifting to that which is a new, untried and untested and
which therefore has a wider variance.
The situation may be summarized graphically for a typical farmer (Figure 2).
His customary average annual output, 00 has associated with it a standard
deviation +("where the one negative standard deviation is still above his
family's Sms. In time period t, he faces the possibility of adopting a new
1/ More correctly, the lower level of the variance associated with a one
or a two negative standard deviation.
- 40 -
Sm ... Sm s
S..s...... ....... ... ..... .. ... ....... .. ....... Sis
t 1 t t + i t -+ 2
variety. Associated with this new variety are a series of subjective expectations
-- average output, ON, which is higher than the traditional variety, but a vari-
ance, 7 which is considerably larger than when cultivating the traditional.
If the farmer's subjective expected negative standard deviation is below the Sms,
the farmer will resist the adoption. It should be emphasized that may
well be wider than the real variance in actual practice as determined on the
experiment station or in trials on farmers' fields. This will be especially true
during the early stages of adoption when there is a lesser degree of certainty
regarding expected yields as seen by the potentially innovative farmer. The
larger the number of farmers in his neighborhood who have tried the new variety
the closer the farmer's subjective variance will be to the real variance. Even
in those cases where the two variances have converged but the negative standard
deviation lies below the particular Sms of the farmer, he is likely to resist.
These are the cases usually referred to as a technological innovation which offers
limited gains -- insufficient to encourage adoption.
If on the other hand,the one:negative deviation, lies above the Sms with an
average output higher than the past output with traditional varieties, the farmer
is far more likely to adopt despite the wider variance which may attach to the
new practice (Figure 3). Under these circumstances, adoption is likely to pro-
ceed rapidly. As we shall see later, when the negative standard deviation is
above the traditional average output, then adoption will proceed very rapidly
indeed. (These are the cases which are usually referred to as situations of
dramatic output increase possibilities.)
1/ Besides the dearth of information of the basis of which to formulate his
expectations regarding yield and its variance, the farmer recognizes that there
is a greater likelihood of his performance being subject to wider error during
the early stages of adoption due to unfamiliarity with the variety, its re-
quirements, and the changed practices which may be required by the new technology
in order to realize its potential.
- 42 -
t 1 t t + 1 t 2
- 43 -
It should be emphasized that not all farms in a particular region will be
characterized by either Figure 2 or 3. Often in a given community, a potential
innovation will be juxtaposed against farms of both types. This fact explains
why certain innovations experience differential diffusion. Larger farms and
those farmers with larger assets are much more likely to be similar to Figure 3
or to have levels of output even utilizing traditional varieties which are
considerably above their Sms.
The recent experience with the new rice varieties from the International
Rice Research Institute is perhaps indicative. The-rapidity with which the new
varieties (especially IR8 and IR5) have been spreading in Asia refutes the stereo-
type of the non-economic peasant. The Philippines has traditionally been a rice
importer -- some 230,000 tons annually from 1961 to 1967. The very rapid adoption
of the new high yielding varieties first introduced in 1966 has already made the
Philippines self-sufficient in rice for the first time in recent history.
Some micro-economic data on the adoption is interesting from two standpoints
First, the variance in yields per hectare ( 67 ) associated with the newer
varieties is considerably larger than with the traditional varieties (6.86 to
2.76 metric tons per hectare in the dry season and 3.61 to 1.74 in the wet season).
Hence the new varieties result in a wider variation in possible results -- both
positive and negative. If the average yields for the new and old varieties had
been fairly similar, then the average farmer would probably/resist/adoption
because even though the average output with the new practice would offer a
bigger prize, the negative prospects would have him end up below the traditional
- 44 -
Second, what is especially significant in the present case is not merely
that the average yields with the new varieties are higher, but that the negative
standard deviation for the new varieties is higher than the average yields of the
old traditional variety: -- 3.24 metric tons per hectare for IR8 compared with
3.17 with the traditional varieties, and in the wet season 2.59 compared with
2.32. Thus, the farmer who adopts the new varieties faces a situation where
even allowing for a negative standard deviation he would experience a higher
yield than the average with the traditional varieties.
TABLE 2. Average Yields of IR6 and Local Rice Varieties in the Philippines,
Dry and Wet Seasons 1966-67. l/
(Metric Tons Per Hectare)
-C; x + G -
IR8 3.24 5.86 8.48
Binato 1.51 3.17 4.83
IR8 2.59 4.49 6.39
Local 1.00 2.32 3.64
1/ Data provided by Dr. Randolph Barker of the International Rice
- 45 -
This example helps to explain the startling phenomenon currently taking
place with the so-called "Green Revolution" -- the rapid adoption of the new
high yielding varieties of wheat and rice [Rockefeller, 1968; Brown 1968].
A second body of interesting evidence comes from the recent research of
Behrman [1966; 1969] on rice supply elasticities in Thailand. His study was
based on data from 50 Changwads (provinces) in Central and Northeastern
Thailand, 1940-1963. In estimating the short-run elasticity of area planted
to rice for the period 1940 to 1963, Behrman employed three independent measures
as surrogates for risk aversion: (a) the standard deviation in price over the
three preceding years; (b) the standard deviation in yields over three preceding
years, and (c) population residing in agricultural households. The choice of
the first two was based upon the strong presumption that they were suitable
proxies for the variances of the subjective probability distributions. For
both distributions, an asymptotically significant response was obtained in less
than half of the provinces. For both distributions, the mean value of the
implied elasticities with respect to the standard deviation is substantially
smaller than the mean value of the implied elasticities with respect to the
expected value of the same subjective probability distribution. The third
variable was seen as a reflection of possible peasant behavior in planting
enough area in rice to assure sufficient food for the household. For
approximately 70 per cent of the provinces the estimated response to this....
variable was symptotically significant. -/ The fact that the population/food
survival variable stood the test better than price or yield variability alone
1/ One final observation of Behrman's is that the pattern of rice supply
elasticities is highly correlated with the existence of profitable alternatives
[Behrman, 1966, pp. 366-691.
lends further support to our emphasis on the critical role of minimal subsis-
tence in the economic decision-making of subsistence peasant farmers.
Three points of caution as we move from the restrictive theoretical setting
tothe greater complexity of the real world. First, it would be a serious
mistake in most instances to look upon the absolute levels and variances in
yield for a single crop as the appropriate measure. To the extent that there
is some degree of diversification in production, the obviously relevant measures
should refer to the farm as a whole.
Second there is the obvious importance of introducing costs associated
with any new technology. For example, from studies conducted at IRRI , they
estimate that the total cost of production per hectare using traditional methods
and varieties is about US $200. On the average less than 10% of this is a
cash outlay. When IP.8 is grown greater use of fertilizer and chemicals is
required so that costs per hectare double and almost all the increase is in
cash. Thus, cash expenditures rise from US $20 to US $220. V- Fortunately,
in this case yield increases threefold, leading to a net return four times
greater than with traditional varieties and methods. Although most diffusion
studies do take such costs into consideration, the risk cost associated with
minimal subsistence, Sms, is not.
SPromoters of new innovations occasionally fail to recognize the double risks
associated with much of the new technology. Where the new technology requires the
purchase of a "package" of associated inputs to achieve maximum returns, the
peasant farmer not only subjects himself to risks in production but also in costs
which are of a much higher order of magnitude than he has previously experienced.
If this action also reduces his ability to depend upon traditional forms of
security (family or village) due to resentment at his innovative actions, the
peasant faces a third form of risk. Any valuation and summation of all three
often results in an extremely high risk factor.
Sadly, it is often hesitancy on the part of peasant farmers to incur the
full costs required by the optimal "package of practices" to secure the full
benefit of the new varieties which frequently leads to an actual loss. There are
several cases where farmers have failed to apply fertilizer at the higher levels
required with the result that traditional varieties outyield the new.
r 47 -
Third, the quantification of anins is admittedly difficult. A number of
possible proxies might be suggested to reflect this minimal standard: (a) for
any sample we might use the percentage of total farm product consumed in the
home to construct a community index and then select an arbitrary cut-off point
(ex. the lower decile or quartile) as the Sms for the community; (b) we might
estimate the fraction of total consumption (food plus non-food) out of farm
production to construct an index and then select an arbitrary cut-off as with
(a) above; or (c) we might merely use the estimated measures of farm family
consumption (farm plus non-farm produced) and then use the negative one
standard deviation for the sample as the approximate locus of the Sms. -
Each measure would be highly arbitrary but would at least render the Sms opera-
The above model as developed represents an extreme case of a subsistence
agricultural producer. Nevertheless, the model is instructive in helping to
explain differential response or adoption in the real world.
Given a close historical relationship between average annual food output
and a farm family's Sms, the degree of risk aversion or resistance to an
innovation will be reinforced by the following five factors:
1. the lesser the degree of food/non-food crop diversification on the
2. the lesser the availability of other food sources; -
iy preference is for the third simply because it is easier to estimate and
distribution of average consumption figures for rural samples tend to skew to
Such food sources may be provided through socio-cultural institutions, viz.
extended family, or ecologically, viz. proximate to easily secured wild/non-
cultivated food sources such as game, roots, berries.
- 48 -
3. the lesser the availability of alternative economic opportunities
for the employment of farm resources, especially family labor;
4. the closer to unity is the ratio of the value of annual family
Sms to the farmers' net worth (asset/debth structure); -
5. the tighter the capital rationing facing the farmer (including
relevant interest charges) [Dutia, 1961; Long, 1968a; 1968b].
Several recent models have been proposed and tested emDirically which
attempt to describe situations where the concentration in decision-making is
upon the risk taken rather than on the possible gains. 2/ Three efforts are
especially noteworthy. Boussard and Petit  have expanded and applied
Shackle's [1949; 1961] "focus of loss" approach. Using this approach the
assumption is made that "farmers want to maximize the 'normal,' or mean, value
of their incomes under the constraint that the focus of loss...is at least
equal to the permissible loss, that is, to the difference between the mean
income and the minimum income." [p. 873]. Perhaps the most interesting and
potentially useful of all is the recent formulation by Karl Borch 11968].
This point can be best seen in reverse. The existence of other economic
opportunities for the employment of available resources has often led to rapid
innovation and expansion in production. Study of development under such
situations in the past (Burma, West Africa) has led to the '"ent-for-surplus"
theories of development [yint, 1964; Eicher, 1967]. The influence of generally
limited economic opportunity as it affects innovation is another neglected
dimension which deserves further study.
/ In a crude sense the farmer's net worth constitutes his ultimate reserve,
either mortgagable or saleable, to cover minimal subsistence in the event of
massive production failure. Peasant farmers normally have very narrow capital
bases so that the margins for "risk experimentation" are also very narrow,
3/ The earlier pioneering efforts by Hildreth , Porter [19591, Johnson,
et al. , Gould , Reutlinger , Wolpert 11964] have a great deal
to commend them to future researchers in this field.
The six dominant characteristics of subsistence farmers with which this
essay began delineate a class of farmers who are influenced by special factors
or factors whose weights are different. These factors in turn lead to different
types of economic behavior and correspondingly different patterns of development,
One factor of importance in the context of subsistence agriculture is risk
because of its close inter-relationship with survival. The inter-relationship
is especially evident with technological innovation. Although new technology
constitutes a major dynamic force for the modernization of peasant agriculture,
its introduction among subsistence farmers often encounters resistance. Profit
maximization may not be as important in a subsistence or barter economy as the
maximization of security and survival.
Previous measures of risk which rely upon such variables as the cost of
the new input or its net returns and the year to year variability in these and
similar measures such as yield, production costs and product price are poor
surrogates when applied to subsistence agriculture. A better approach would be
to observe the interaction between two sets of variables: (a) the absolute
levels of farm family living as they relate to societal standards for the minimal
levels of subsistence, and the average levels of farm productivity and income;
and (b) the farmer's subjective expected variance in output associated with the
proposed new technological introduction compared with the historically determined
variance in output utilizing traditional practices. What must be taken into
consideration is the variability in the expected results with the new technology
not merely measured against costs as usually defined but against the minimum
subsistence standard and the variability of current technology. What is needed
is a risk aversion or security preference measure which takes this inter-
relationship into account.
- 50 -
A. Some Puzzles and Paradoxes Explained
The above analysis may help to explain a number of puzzles and paradoxes
which are commonly encountered with technological innovation in the developing
1. Differential Adoption within Same Community. In many agricultural
areas, one can find farmers who have adopted a new innovation coexisting with
neighboring farmers who have failed to adopt even though the latter see the new
technology every day and are aware of it. "Demonstration effects" and
"neighbor effects" seem to have no impact upon adoption by the remaining farmers.
2. Food Staple vs. Non-Food Staple Variations. A common experience is a
differential resistance to technological adoption between staple food crops and
non-staple food crops. Technological innovation tends to move more rapidly
among farmers specializing with non-food staples (especially commercial crops)
than is true with food staples, even where there is no significant relative price
3. The "Dual Farmer." One frequently encounters farmers who grow both a
food staple and a non-food staple on their farms. Yet they are willing to
innovate or to employ a new technology with a commercial crop but persist in
utilizing traditional practices with the food staple.
4. New Crops vs. Old Crops. Another common observation has been that the
introduction of new crops into the farming regime of a peasant requiring new
technology seems to be easier than changing the technology of a traditional,
In each of the four cases, a good deal of the variation in adoption can
be attributed to the relationship between subsistence standards of living, and
the expected variance in output of the food staple under the new technology.
In the first case, for example the non-adopters are most frequently those farmers
who are less commercial (both in product and input) and whose resources relative
to their Sms are extremely close. Where the proposed innovation and its associated
variance exceeds the Sms, such as was shown to be the case with the new rice
varieties in the Philippines, then adoption proves to be swift.
B, Some Program and Policy Implications
There are some obvious policy and program implications to the above
analysis. If risk and uncertainty are as important in the context of subsistence
as indicated, then certain steps are required to assure a greater rapidity and
extent of adoption of new technology.
First, information on the variance of any new technology is as important
as its average performance, Any determination of the economic feasibility of a
new practice or technique should pay equal, if not more, attention to the variance
in yields, especially the lower deviations as they relate to minimum subsistence
standards of living of potential innovating farmers.
Second, in developing new technologies, agricultural research organizations
should recognize the importance which subsistence farmers attach to the variances
associated with any possible innovation. Plant breeders, for example, should
pay greater attention to those specific characteristics which may help to reduce
negative deviation and offer greater dependability. /
Third, where only a narrow range covers the minimum subsistence standards,
levels of living, and physiologic minima (Type A situation in Figure 1),
programs designed to diffuse new technology need to pay much greater attention
to methods for "risk insurance" [tiarglin, 1965] or assuring the peasant who
innovates that failure (i.e., an output falling below his minimum subsistence
standard) will not result in a major penalty, viz. loss of life or loss of
Sy colleague/Albert H, IMoseman has pointed out that much of the plant
breeding work in the US during the past 25 years has paid close attention to
this factor in their work for US agriculture.
.- 52 -
property or indebtedness. Existing social structures and institutions (viz.
extended family) which already provide some degree of "risk insurance" should
be recognized as such and wherever possible treated as complementary to any new
Fourth, methods of technological introduction and trial in a peasant community
should recognize that in the early stages the typical farmer in the community
attaches a subjective variance to the expected yield of the new technology which
is considerably wider than the true variance. Extension and information
measures should concentrate just as much on reducing this subjective variance
in the minds of potential innovators as on spreading knowledge about the
average or maximum yields. Assurance as to the dependability of the practice
or technology may be more important to the peasant farmer than its dramatic
Peasantry throughout the world constitute the largest fraction of mankind
and their pattern of life is one of the oldest in human history. The peasant,
through his inherited institutions and his traditionally determined socio-
economic behavior, has developed a strategy to win the basic struggle for
survival. He will not relinquish this strategy easily.
One of the most dramatic and dynamic forces inducing change in traditional
societies is new technology. Subsistence economies and agrarian societies pro-
vide a security which has the force of historical certainty in the survival of
the peasantry and their community. Under these circumstances, attempts at
change, especially those which come into direct conflict with the fundamental
goals of security and survival, must take into account the degree of risk and
uncertainty associated with the change.
Our attempt has been to give greater precision and understanding to one
major variable which inhibits innovation among a very large number of farmers
in the developing world. Risk is not the only factor which retards development;
there are others, but its elimination or reduction should prove a major
stimulus for technological innovation and the modernization of subsistence
Adelman and Morris, 1967. Irma Adelman and Cynthia Taft Morris, Society,
Politics, and Economic Development (Baltimore, Maryland: Johns Hopkins
University Press, 1967).
1968. Irma Adelman and Cynthia Taft Morris, "Perform-
ance Criteria for Evaluating Economic Development Potential: An Opera-
tional Approach," Quarterly Journal of Economics, Vol. LXXXII, No.
Anderson and Dillon, 1968. J. R. Anderson and John L. Dillon, "Economic
Considerations in Response Research," American Journal of Agricultural
Economics, Vol. 50, No. 1 (February 1968).
Andrus, 1948. J. Russell Andrus, Burmese Economic Life,
(Stanford, California: Stanford University Press, 1948).
Barletta, 1967. N. Ardito-Barletta, "Costs and Social Returns
of Agricultural Research in Mexico," unpublished Ph. D. thesis, Univer-
sity of Chicago, 1967.
Bateman, 1969. Merrill J. Bateman, "Supply Relations for Peren-
nial Crops in the Less Developed Areas," in Clifton R. Wharton, Jr., ed.
Subsistence Agriculture and Economic Development (Chicago, Aldine Publish-
ing Company, 1969).
Beal et al,, 1967. George M. Beal, E. Walter Coward, Jr.,
Gerlad E. Klonglan, "Emerging Patterns of Commercial Farming in a Sub-
sistence Farm Economy: An Analysis of Indian Farmers in Guatemala,"
Rural Sociology Report 68, Department of Sociology and Anthropology,
Iowa State University, Ames, Iowa 1967.
Jere R. Berhman, "Supply Response in Under-
Agriculture: A Case Study of Four Major Annual Crops in
1937-1963," unpublished Ph.D. dissertation, Massachusetts
of Technology, September 1966.
1969. Jere R. Behrman, "Supply Response and the
Modernization of Peasant Agriculture: A Study of Four Major Annual
Crops in Thailand," in Clifton R. Wharton, Jr., ed., Subsistence Agri-
culture and Economic Development (Chicago: Aldine Publishing Co., 1969).
Berry, 1967. Sara Berry, "Cocoa in Western Nigeria, 1890-
1940: A Study of an Innovation in a Developing Country," unpublished
Ph.D. thesis, Michigan State University, 1967.
Wharton Page 2
Black, 1953. John D. Black, Introduction to
Agriculture (New York: Macmillan, 1953).
Borch, 1968. Karl Borch, "Decision Rules Depending on the
Probability of Ruin," Oxford Economic Papers, Vol. 20, No. 1 (March
Borton, 1967. Raymond E. Borton, ed., Case Studies to Accom-
pany Getting Agriculture Moving (New York: Agricultural Development
Council, Inc., 1967).
Boussard and Petit, 1967. Jean-Marc Boussard and Michel Petit, "Repre-
sentation of Farmers' Behavior under Uncertainty with a Focus-Loss
Constraint," Journal of Farm Economics, Vol. 49, No. 4 (November 1967).
Brown, 1968. Lester R. Brown, "New Directions in World Agri-
culture," paper presented at Second International Conference on War on
Hunger, Washington, D.C., February 1968.
Byrnes, 1966. Francis C. Byrnes, "Some Missing Variables in
Diffusion Research and Innovation Strategy," Philippine Sociological
Review, Vol. 14, No. 4 (October 1966).
1968. Gelia Tagumpay-Castillo, "Propensity to Invest
in Agriculture: Observations from a Developing Country, The Philippines,"
International Journal of Agrarian Affairs, Vol. V, No. 4 (July 1968).
1966. A. V. Chayanov, The Theory of Peasant Economy,
Edited by Daniel Thorner, Basile Kerblay and R. E. F. Smith (Homewood,
Illinois: Richard D. Irwin, 1966).
1955. Willard W. Cochrane, "Conceptualizing the Supply
Relation in Agriculture," Journal of Farm Economics, Vol 37, No. 5
Dalton, 1960. George Dalton, "A Note of Clarification on
Economic Surplus," American Anthropologist, Vol. 62, No. 3 (June 1960).
S1961. George Dalton, "Economic Theory and Primitive
Society," American Anthropologist, Vol. 63, No. 1 (February 1961).
1963. George Dalton, "Economic Surplus Once Again,"
American Anthropologist, Vol. 65, No. 2 (April 1963).
Davidson and Mighell, 1963.
ing Farmers' Reactions
Vol. 45, No. 3 (August
Jack R. Davidson and Ronald L. Mighell, "Trac-
to Uncertainty," Journal of Farm Economics,
Wharton Page 3
Day, 1965. Richard H. Day, "Probability Distribution of
Field Crop Yields," Journal of Farm Economics, Vol. 47, No. 3 (August
Dillon and Heady, 1961.
Vol. XLIII, No. 3
John L. Dillon and Earl 0. Heady, "Free Competi-
and Farmer Decisions," Journal of Farm Economics,
Dutia, 1961. B. P. Dutia, "Uncertainty, Capital Use and
Firm-Household Relationship," Indian Economic Journal, Vol. 8, No. 4
Edel, 1967. Matthew Edel, "Jamaican Fishermen: Two Ap-
proaches in Economic Anthropology," Social and Economic Studies, Vol.16,
No. 4 (December 1967).
Eicher, 1967. Carl K. Eicher, "The Dynamics of Long-Term
Agricultural Development in Nigeria," Journal of Farm Economics, Vol. 49,
No. 5 (December 1967).
Evenson, 1967. Robert Evenson, "The Contribution of Agricul-
tural Research to Production," Journal of Farm Economics, Vo. 49, No. 5
Fei and Ranis, 1964. John C. H. Fei and Gustav Ranis, Development of
the Labor Surplus Economy: Theory and Policy (Homewood, Illinois:
Richard D. Irwin, Inc., 1964).
Firth, 1969. Raymond Firth, "The Influence of Social Structure
Upon Peasant Economies," in Clifton R. Wharton, Jr., ed. Subsistence
Agriculture and Economic Development (Chicago: Aldine Publishing Co.,
Furnivall, 1957. J. S. Furnivall, An Introduction to the Political
Economy of Burma (Rangoon: Peoples; Literature Committee and House,
3rd ed. 1957).
Georgescu-Roegen, 1969. Nicholas Georgescu-Roegen, "The Institutional
Aspects of Peasant Communities: An Analytical View," in Clifton R.
Wharton, Jr., ed., Subsistence Agriculture and Economic Development
(Chicago: Aldine Publishing Company, 1969).
Gould, 1963. Peter R. Gould, "Man Against His Environment:
A Game Theoretic Framework," Annals of the Association of American Geog-
raphers, Vol. 53, No. 3 (September 1963).
Griliches, 1958. Zvi Griliches, "Research Costs and Social Re-
turns: Hybrid Corn and Related Innovations," Journal of Political Econ-
omy, Vol. 66, No. 5 (October 1958).
Wharton Page 4
Hildreth, 1957. C. Hildreth, "Problems of Uncertainty in Farm
Planning," Journal of Farm Economics, Vol. 39, No. 5 (December 1957).
Hill, 1963. Polly Hill, The Migrant Cocoa Farmers of
Southern Ghana: A Study in Rural Capitalism (Cambridge: Cambridge
University Press, 1963).
IRRI, 1967. International Rice Research Institute, "The
Economics of IR-8," The IRRI Reporter, Vol. 3, No. 4 (July 1967).
Ruttan, 1967. S. C. Hsieh and V. W. Ruttan, "Technological,
Institutional and Environmental Factors in the Growth of Rice Produc-
tion: Philippines, Thailand, and Taiwan," Food Research Institute
Studies, Vol. 7, No. 3 (1967).
Johnson, et al., 1961. Glenn L. Johnson, A. Halter, H. R. Jensen and
D. W. Thomas, A Study of Managerial Processes of Midwestern Farmers,
(Ames, Iowa: Iowa State University Press, 1961).
Johnson, 1968. Glenn L. Johnson, "Food'Supply, Agricultural
and Economic Development," Paper presented at Second Western Hemisphere
Nutrition Congress, San Juan, Puerto Rico, August 26-29, 1968.
Johnson, 1962. Paul R. Johnson, "Do Farmers Hold a Preference
for Risk?" Journal of Farm Economics, Vol. XLIV, No. 1 (February 1962).
Jones, 1960. William 0. Jones, "Economic Man in Africa,"
Food Research Institute Studies, Vol. 1, No, 2 (May 1960).
Knight, 1965. Frank H. Knight, Risk, Uncertainty and Profit
(New York: Harper and Row, 1965). First published in 1921.
for a Subsistence Crop:
Annual Number, Vol. 17,
Raj Krishna, "The Marketable Surplus Function
An Analysis with Indian Data," Economic Weekly,
Nos. 5, 6, and 7 (February 1965).
1967. Raj Krishna, "Agricultural Price Policy and
Economic Development," in Herman M. Southworth and Bruce F. Johnston,
eds., Agricultural Development and Economic Growth (Ithaca, New York:
Cornell University Press, 1967).
1969. Raj Krishna, "Models of the Family Farm,"
in Clifton R. Wharton, Jr., ed. Subsistence Agriculture and Economic
Development (Chicago: Aldine Publishing Co., 1969).
Krishna and Desai, 1964. jai Krishna and D. K. Desai, "Economic Models
of Farm Planning Under Uncertainty," Indian Journal of Agricultural
Economics, Vol. 19, No. 1 (January-March 1964).
Wharton Page 5
Lau, 1967. Lawrence J. Lau, "Peasant Consumption, Saving
and Investment in Mainland China," Paper presented at conference on
The Agrarian Question, University of Washington, August 23-26, 1967.
Long, 1968a. Millard Long, "Interest Rates and the Struc-
ture of Agricultural Credit Markets," Oxford Economic Papers, Vol. 20,
No. 2 (July 1968).
Long, 1968b. Millard F. Long, "Why Peasant Farmers Borrow,"
American Journal of Agricultural Economics, Vol. 50, No. 4 (November
McHale, 1967. T. R. McHale, Rubber and the Malaysian Econ-
omy (Singapore: M. P. H. Publications Sendrian Berhad, 1967).
Marglin, 1965. Stephen A. Marglin, "Insurance for Innovators,"
Appendix B in David Hapgood and Max F. Millikan, eds., Policies for
Promoting Agricultural Development (Cambridge, Massachusetts: Center
for International Studies, Massachusetts Institute of Technology,
Mellor, 1963.- John W. Mellor, "The Use and Productivity of
Farm Family Labor in Early Stages of Agricultural Development," Journal
of Farm Economics, Vol. 45, No. 3 (August 1963).
Merrill, 1965. William C. Merrill, "Alternative Programming
Models Involving Uncertainty," Journal of Farm Economics, Vol. 47,
No. 3 (August 1965).
Millikan and Hapgood, 1967. Max F. Millikan and David Hapgood, No Easy
Harvest (Boston, Massachusetts: Little, Brown and Company, 1967).
Miracle, 1968. Marvin P. Miracle," 'Subsistence Agriculture':
Analytical Problems and Alternative Concepts," American Journal of
Agricultural Economics, Vol. 50, No. 2 (May 1968).
Moseman, 1967. A. H. Moseman, "The Role of Science and
Technology in Meeting World Food Needs," Paper presented at the Annual
Meeting of the American Phytopathological Society of Nematologists,
Washington, D.C., August 21, 1967.
Mosher, 1966. Arthur T. Mosher, Getting Agriculture Moving
(New York: Praeger, 1966).
Myint, 1964. H. Myint, The Economics of the Developing
Countries (London: Hutchinson, 1964).
Wharton Page 6
Myren, 1964. Delbert T. Myren, "The Role of Information
in Farm Decisions under Conditions of High Risk and Uncertainty,"
Proceedings, Symposium on Role of Communications in Agricultural
Development (Mexico City, Mexico, October 5-13, 1964), pp. 94-100.
1969. Delbert T. Myren, "The Rockefeller Foundation
Program in Corn and Wheat in Mexico," in Clifton R. Wharton, Jr., ed.,
Subsistence Agriculture and Economic Devdlopment (Chicago: Aldine
Publishing Company, 1969).
Nakajima, 1969. Chihiro Najajima, "Subsistence and Commercial
Family Farms: Some Theoretical Models of Subjective Equilibrium,"
in Clifton R. Wharton, Jr., ed.', Subsistence Agriculture and Economic
Development (Chicago: Aldine Publishing Company, 1969).
Nash, 1966. Manning Nash, Primitive and Peasant Economic
Systems (San Francisco, California: Chandler Publishing Company, 1966).
Officer et al., 1967. R. R. Officer, A. N. Halter and John L. Dillon,
"Risk, Utility and the Palatability of Extension Advice to Farmers,"
The Australian Journal of Agricultural Economics, Vol. 11, No. 2
Officer and Halter, 1968. R. R. Officer and A. N. Halter, "Utility
Analysis in a Practical Setting," American Journal of Agricultural
Economics, Vol. 50, No. 2 (May 1968).
1967 (?). Burton T. Onate, "Risk and Uncertainty:
Distributions of Rice Yields," International Rice Research Institute,
Los Banos, Philippines, Mimeographed (Date uncertain, 1967).
Orans, 1966. Martin Orans, "Surplus," Human Organization,
Vol. 25, No. 1 (Spring 1966).
67. Harry T. Oshima, "Food Consumption, Nutrition,
and Economic Development in Asian Countries," Economic Development
and Cultural Change, Vol. 15, No. 4 (July 1967).
PSAC, 1967. Panel on World Food Supply, The World Food
Problems, A report of the President's Science Advisory Committee, The
White House, Washington, D.C., May 1967.
Pfanner, 1969. David E. Pfanner, "Case Study: A Semi-Sub-
sistence Village Economy in Lcwer Burm&," in Clifton R. Wharton, Jr.,
ed., Subsistence Agriculture and Economic Development (Chicago: Aldine
Publishing Company, 1969).
Wharton Page 7
Porter, 1959. R. C. Porter, "Risk, Incentive, and the
Technique of the Low-Income Farmer," Indian Economic Journal, Vol. 7,
No. 1 (July 1959).
Reutlinger, 1963. Shlomo Reutlinger, "Alternative Uncertainty
Models for Prediction Supply Response," Journal of Farm Economics,
Vol. 45, No. 5 (December 1963).
Rockefeller, 1968. The Rockefeller Foundation Symposium, Strat-
egy for the Conquest of Hunger, Proceedings of Symposium at Rockefeller
University, April 1 and 2, 1968 (New York: The Rockefeller Foundation,
Rogers, 1962. Everett M. Rogers, Diffusion of Innovation
(New York: Free Press, 1962).
Ruttan, 1959. Vernon W. Ruttan, "Usher and Schumpeter on
Invention, Innovation, and Technological Change," Quarterly Journal
of Economics, Vol. LXXIII, No. 293 (November 1959).
Schultz, 1964. T. W. Schultz, Transforming Traditional Agri-
culture (New Haven, Connecticut: Yale University Press, 1964).
Sen, 1966. Amartya K. Sen, "Peasants and Dualism with
and without Surplus Labor," Journal of Political Economy, Vol. LXXIV,
No. 5 (October 1966).
Shackle, 1949. G. L. S. Shackle, Expectations in Economics
(Cambridge: Cambridge University Press, 1949).
1961. G. L. S. Shackle, Decision, Order, and Time
in Human Affairs (Cambridge: Cambridge University Press, 1961).
Stakman, et al., 1967. E. C. Stakman, Richard Bradfield and Paul C.
Mangelsdorf, Campaigns Against Hunger (Cambridge, Massachusetts:
Belknap Press of Harvard University Press, 1967).
Stevens, 1965. Robert D. Stevens, Elasticity of Food Con-
sumption Associated with Changes in Income in Developing Countries,
Foreign Agricultural Economic Report No. 23 (March 1965). (Washington,
D.C.: Development and Trade Analysis Division, Economic Research
Service, U.S. Department of Agriculture).
Tedford, 1964. John R. Tedford, "Analytics of Decision Mak-
ing," Journal of Farm Economics, Vol. 46, No. 5 (December 1964).
Wharton Page 8
Wharton, 1962. Clifton R. Wharton, Jr., "Marketing, Mer-
chandising and Moneylending: A Note on Middleman Monopsony in Malaya,"
Malayan Economic Review, Vol. 7, No. 2 (October 1962).
1963a Clifton R. Wharton, Jr., "Economic and Non-
Economic Factors in the Agricultural Development of Southeast Asia:
Some Research Priorities," A/D/C Paper (New York: Agricultural
Development Council, June 1963).
1963b. Clifton R. Wharton, Jr., "The Economic Mean-
ing of Subsistence," Malayan Economic Review, Vol. 8, No. 2 (October
1967. Clifton R. Wharton, Jr., "The Infrastructure
for Agricultural Growth," in Herman M. Southworth and Bruce F. Johnston,.
Agricultural Development and Economic Growth, (Ithaca, New York:
Cornell University Press, 1967.
1969. Clifton R. Wharton, Jr., ed., Subsistence
Agriculture and Economic Development (Chicago; Aldine Publishing
Wise and Yotopoulos, 1968. John Wise and Pan Yotopoulos, "A Test of the
Hypothesis of Economic Rationality in a Less Developed Economy: An
Abstract," American Journal of Agricultural Economics, Vol. 50,
No. 2 (May 1968).
Wolf, 1966. Eric Wolf, Peasants (Englewood Cliffs, New
Jersey: Prentice.Hall, 1966).
Wolpert, 1964. Julian Wolpert, "The Decision Process in a
Spatial Context," Annals of the Association of American Geographers,
Vol. 54, No. 4 (December 1964).