DESCRIPTIVE RESULTS
(a) Awareness, tryout and adoption at village level
The total awareness, tryout and adoption rates of mineral
fertilizer in the research area are respectively 57%, 8%, and
4% (Table 2). Conditional on awareness, tryout increases to
13% and adoption to 7%. Conditional on tryout, the adoption
rate is 70%. The relatively large discrepancy between conditional
and unconditional tryout and adoption rates results
from incomplete awareness and tryout (Table 2).
We find much higher awareness, tryout, and adoption rates
in program and neighboring villages than in distant villages.
Despite a considerable rate of awareness (48%), there is virtually
no tryout and adoption in distant villages. In these villages,
there is limited access to information and inputs
because there are no program interventions and because technical
information beyond that provided in radio programs,
less easily spreads to these villages.
In neighboring villages, we observe slightly lower awareness
rates, but slightly higher tryout and adoption rates, than in
program villages (Table 2). The intensity of interventions is
much lower in neighboring villages. Yet, they benefit from
easy communication and exchange with program villages,
and have a considerably high share of households who are
member of program associations (10% compared to 17% in
program villages).
Figure 2 shows that the application of mineral fertilizer increased
rapidly after the start of the program in 2006, with
the increase being more immediate and rapid in program villages
and more smooth and delayed in neighboring villages.
Figure 2 and Table 2 point at some adverse selection in program
participation in selected villages. The program may
have selected villages where the likelihood of adoption is lower,
or farmers with a lower interest in technological improvement.
Also, there is possibly a positive self-selection by
farmers in neighboring villages caused by more motivated
farmers deciding to join program activities in a village nearby.
In Table 3, we compare households who are member of program
associations (program households) with households who
are member of other agricultural associations that are not directly
linked to the program, and with households who are not
a member of any agricultural association. The figures show
that awareness, tryout, and adoption rates are highest among