It took a long-term view, and included costs that could be externalized, such as
 land erosion and water quality and sustained quantity. FMR allowed for an
 approximation of experimental conditions by examining existing practices to see
 what was the most profitable; it then helped farmers to apply the results of thi
 quasi-experimental research through good record-keeping and problem analysis
 (McDermott, 1982, personal communication).

 In Thira World FSR, the researcher works with the farm family to decide
 upon appropriate innovations that can be introduced to increase farm family
 welfare, yet not contradict long-range societal goals. Generally, this involves
 increasing profits, but reducing risk is even more crucial for the small Third
 World farmer. Maintaining a fragile ecosystem, which may be antithetical to
 small farm profitability, must also be taken into account.

 In early FMR in the United Stares, it was demonstrated that the diversified
 family farm-was better able than the monocultural farm to withstand the
 inevitable dips and rises in the price structure. Yet, because of the
 constraints of introducing more efficient machinery to reduce labor costs and
 troubles, and because research was commodity oriented and not system oriented,
 agriculture steadily became.more oriented to monoculture-and more susceptible
 to market trends (Flora and Rodefeld, 1978). For the small, Third World family
 farm, the purchase of large, efficient equipment is impossible, even if the
 fuel and other inputs would continue to be available to run it. Further, the
 marginal Third World farmers do not have the ability to withstand several years
 of bad prices if a single crop is depended on. The family farm diversifies its
 sources of income, even if such a tactic may be inefficient in receiving the
 highest return on any single crop. (Research carried on by Instituto Colombiano
 Agropecuario (iCA) in Rio Negro, Antioquia, shows the superiority of a cropping
 system using beans, corn, and potatoes over maximizing the yield of any one of
 the crops.)

 Farm management, like FSR, starts with problem identificatcn. Identifying
 the problem- is the only way to go about solving it. However, methods of problem
 identification learned and shared, for example, by midwestern farmers, extension
 agents, and researchers, may be very different from both the logic and the
 perceived problems of farmers in developing countries. Farm management research,
 as farming systems research, stresses that data are necessray in order to make
 the proper decisions. Those data involve what people actually do and the
 implications of these actions related to their perceived goals.

 According to current farm management theory, the successful U.S. farmer
has specific and definable goals and objectives (Buller, 1976). Part of farm
management practice is to help people develop these goals and objectives, which
are primarily financial, generally within terms that are numerically operational,
often facilitated by computer.

 Less successful U.S. farmers do not have goals and are more passive
(Bulier, 1976). They wait to see what will happen to them. That is a more
Fatalistic approach, but one which is more congruent to people who may not
control directly all the resources necessary to farm. For example, in the
developing countries, if you are a tenant farmer it is difficult to set goals
because you cannot control what will happen (Griffin, 1976). If, as in many
cases, you are a woman farmer-which is increasingly likely as male migration
to seek off-farm employment increases-your ability to control labor as well
as other resources is greatly diminished vis a vis males (Bourque and Warren,