138 Poverty and Hunger
researchers. I was one of those who succumbed to the temptations they presented.
I started my research on the fragile and vulnerable Perkerra Irrigation Settlement
in Kenya, and then concentrated on its stronger sibling, Mwea. The Mwea Irriga-
tion Settlement had several advantages. It was by most criteria more successful; it
had a stable water supply and in irrigated rice a reliable crop and a protected mar-
ket; it was better organized; it had a high profile and was frequently visited, being
a convenient distance from Nairobi for VIPs;^1 and, for the indolent PhD student
it had the advantage of being well documented, with time series tables which could
(I hasten to add, with due acknowledgement) be transposed easily to make a thesis
look good, at least to any examiner too pressed for time to look deeply. Many
researchers were attracted to Mwea, and the managers were so interested and artic-
ulate that we were able to combine to produce a book with 13 chapters and 529
pages (Chambers and Moris, 1973).^2
In parallel, numerous studies of settlement schemes were undertaken in other
countries, especially Sudan, Zimbabwe (Southern Rhodesia as it was), Zambia
(Northern Rhodesia as it was), Tanzania, Uganda, Nigeria and Ghana. These pro-
vided a wonderful collection for comparisons. Much of what was learnt is now of
mainly historical interest. However, analysis at the time pointed to three neglected
angles or themes which were important then and remain important and still rela-
tively neglected: commitment, continuity and irreversibility. The two extracts of
studies from this period, which follow, raise practical questions not only about
projects, but also about policies and programmes promoted and pursued by lend-
ers, donors and governments in the 2000s.
Conclusions from Settlement Schemes in Tropical Africa
(1969)
Risks and irreversibility of commitment^3
A neglected aspect of evaluation, which has far wider application than merely to
settlement schemes, concerns the relationship between risks and irreversibility of
commitment.
Settlement schemes, especially those that are more complex in system and
costly in capital, are high-risk undertakings. They share with non-settlement
approaches to agricultural development the uncertainties of innovations, weather,
pests and markets, and the disruptions of rapid turnovers in senior staff. In addi-
tion, however, they have to face other serious risks and difficulties which do not
have to be borne in non-settlement situations. The land on which settlement takes
place may be available for settlement for the simple reason that it is marginal or
unsuitable for cultivation. The locations of many settlement schemes, often with
poor communications and far from the services of urban centres, raise costs and
the difficulties of management. There is a danger that both organizational and
productive effectiveness will be restrained by the inbuilt incompatibilities of com-
plex schemes, by the cancelling out of managerial and settler efforts, in the games
of enforcing and beating the system. Moreover, adaptations of schemes to ensure