384 Modern Agricultural Reforms
from the Yatenga region of Burkina Faso, where 120,000ha of earth bunds con-
structed at high cost with machine graders in the early 1960s have now all but
disappeared (Marchal, 1978, 1986). In the Majjia and Badéguicheri valleys of
Niger, most of the 6000ha of earth bunds constructed between 1964 and 1980 are
in an advanced state of degradation (Reij, 1988). In Sukumuland, Tanzania, where
contour banks, terraces and hedges were forced upon farmers in the 1950s, almost
no evidence remained of these conservation works by the early 1980s, and now
‘erosion is extremely severe’ (Stocking, 1985).
Graded and contour bunds developed for large-scale farming in the US are still
widely applied in programmes in India. Even under heavy subsidies, most small
farmers reject them (Kerr and Sanghi, 1992). These bunds leave corners in some
fields, and so there is a risk of losing the piece of land to a neighbour. The central
water course for drainage benefits only some farmers, while damaging the land of
others. Contour farming is inconvenient when farmers use multi-row implements,
and so is only suitable where the holding is large and tractors are available. Con-
tour bunding without facilities for dealing with surplus water commonly breach,
again concentrating water flow that quickly forms gullies. Therefore, it is not
uncommon for entire bunds to be levelled as soon as project staff shift to the next
village (Sanghi, 1987; Fernandez, 1993).
In Cape Verde, the state takes responsibility for erosion control by paying
farmers to work on their own land. The result is that traditional practices are
ignored as farmers take the money without influencing the project. Socalco ter-
races, for example, are built from top to bottom of steep slopes, with the result that
foundations are often left hanging in the air (Haagsma, 1990). As Haagsma put it
‘this does not stimulate ... good cooperation between farmers and MDRP [the
project]. It is difficult to eradicate the attitude “MDRP knows best”.’
A major project in Niger was described by the implementing agency in this
way: ‘People’s participation is the power behind the Keita project. From decision-
making – to planning – to action: local farmer-livestock owners have been con-
sulted and actively taken part in every step’ (FAO, 1992). Although some 2.76
million work-days were paid for with World Food Programme rations, which
served as ‘incentives to participate in land reclamation and training courses offered
by the project’, no farmers apply the technologies to their own lands, and replica-
bility is close to zero (IFAD, 1992).
In Ethiopia, where 1.5 million km of terracing were constructed during the
1980s with food for work, participation was ‘either compulsory via peasant asso-
ciation campaigns or paid through food for work’ (SIDA, 1984). A total of 34.3
million person-days of work was devoted to conservation, involving the ‘coopera-
tion of some 8000 Peasant Associations’ (FAO, 1986, in Östberg and Christians-
son, 1993). Apparently, ‘farmers’ participation was shown by their contributions
of labour for infrastructure development’, and the project expected these struc-
tures to be maintained because ‘training ... will help in sustaining activities when
the donor pulls out. The privilege of being trained will keep the individual respon-
sible in the activities he (sic) was trained for’ (reported in Oxfam, 1987).