AGRICULTURAL INNOVATIONforce the late majority to adopt, but they gain little.
They have to adopt to stay in business, and some
late adopters may be forced out because they
cannot compete. This treadmill increases concen-
tration of agricultural production and benefits
large farmers, the suppliers of innovations, and
consumers. Indeed, it helps to create and to subsi-
dize cheap urban labor. When it comes to environ-
mental practices, however, large farms are not
early innovators (Pampel and van Es 1977; Buttel
et al. 1990).
The individual bias of adoption-diffusion re-
search is evident in its almost exclusive focus upon
individual farmers rather than upon industrial
farms or other agribusiness. There is also a tenden-
cy to blame the victim if anything goes wrong
(Rogers 1983). Change agents are too rarely criti-
cized for providing incomplete or inaccurate in-
formation, and governments and corporations are
too infrequently criticized for promoting inappro-
priate or harmful innovations. Empirical surveys
of individual farmers also lead to a number of
methodological problems. As noted above, if sur-
veys are retrospective, recall relies on fallible memo-
ry and renders unsuccessful innovations difficult
to study. These surveys are commonly combined
with correlational analysis that makes it difficult to
address issues of causality. After all, the farmer’s
attitudes and personality are measured at the time
of the interview, and the innovation probably
occurred some time before. As we have pointed
out, these issues can be addressed by prospective
designs that incorporate other methods, such as
qualitative case studies and available records data,
and focus on the social context of innovation.
Taking into account the social context of inno-
vation involves shifting levels of analysis from
individual farmers to the social, economic, and
political structures in which they are embedded.
Contextual analysis of social structures has evolved
in two directions, both of which have been in-
spired by the adoption-diffusion paradigm. The
first considers social structure as a set of social
relations among farmers, that is, a social network.
Typically, social network structure has been stud-
ied from the point of view of individual farmers.
For instance, farmers are more likely to innovate if
they are connected to others with whom they can
discuss new farming ideas (Rogers and Kincaid
1981; Rogers 1983; Warriner and Moul 1992). In
this type of analysis, ‘‘connectedness’’ becomes a
variable property of individual farmers that is
correlated with their innovativeness. It is much
less common to find studies that consider how
agricultural innovation is influenced by structural
properties of entire networks, such as the pres-
ence of subgroups or cliques, although other types
of innovation have been studied within complete
networks (see, e.g., Rogers and Kincaid 1981 on
the diffusion of family planning in Korean vil-
lages). Field studies of subcultural differences in
orientations to innovation report what amount to
network effects, though networks are rarely meas-
ured directly. For instance, studies of Amish farm-
ers have revealed that members of this sect restrict
certain kinds of social contacts with outsiders in
order to preserve their beliefs, which include envi-
ronmental orientations based on religious beliefs
(for a review, see Sommers and Napier 1993).
Given the growing importance of social network
analysis in contemporary sociology (Wasserman
and Faust 1995), and the demonstrated impor-
tance of networks of communication and influ-
ence in innovation research (Rogers and Kincaid
1981), future studies of agricultural innovation
could profitably incorporate network models and
data in their research designs.A second type of social structural analysis has
considered how agricultural innovation is influ-
enced by distributions of resources within farming
communities. Much of this research has focused
on the so-called ‘‘Green Revolution.’’ This term
refers to the increases in cereal-grain production
in the Third World, particularly India, Pakistan,
and the Philippines, in the late 1960s, through the
use of hybrid seeds and chemical fertilizer. In
Indian villages where knowledge of new farming
technology and agricultural capital were highly
concentrated, the rate at which individual farmers
translated their knowledge into trial was higher
(Gartrell and Gartrell 1979). Yet overall levels of
innovation tended to be lower in such high-ine-
quality villages. Had the primary goal of India’s
development programs been to maximize the rate
at which knowledge of new farming practices is
turned into innovation, then these results could
have been seen as a vindication of a development
strategy that concentrated on well-to-do cultiva-
tors and high-inequality villages. Yet, this and oth-
er assessments of the Green Revolution in the