Words and Ideas: Commitment, Continuity and Irreversibility 145which requires ongoing government involvement, and on the other programmes
like the introduction of a new seed variety which can be one-shot efforts. In a
major project the risks and liabilities are shouldered by government: if the project
succeeds, government is obliged to continue to service and manage it; if it fails, it
may prove politically and administratively impossible to abandon it. In the one-
shot programme, however, the risks and liabilities are accepted by the individual
farmers: if the innovation succeeds, it is propagated without further government
intervention; if it fails, it is quickly and easily abandoned by the farmers without
additional cost or administration for government.
There are, of course, a great many other considerations which bear on policy
choices; but capital and administrative capacity are typically scarce resources to be
used sparingly; and in choosing between alternative approaches to agricultural
development there is a case, other things being equal, for preferring those which
are cheap, simple, administration sparing and easy to withdraw from to those
which are expensive, complex, administration intensive and committing.
It is not enough merely to be aware of these considerations; there must also
be a climate and machinery in government to make sure that they are taken into
account. In British colonial government in East Africa in the 1950s there was a
relative absence of economic criteria in official thinking and a readiness to sup-
port the initiative of officers at the local level when they promoted projects.
There was sympathetic backing in the Kenya central government for the vision
and enthusiasm of the civil servants at provincial and district level who energeti-
cally launched the Perkerra Irrigation Scheme. Entrepreneurial capacity of the
sort which they demonstrated is certainly an asset to a government, but as the
Perkerra story shows it can be dangerous unless it is controlled. What is needed
is a powerful and perceptive presence in governments which, while not stifling
local initiatives, ensures that schemes as unpromising as Perkerra are never begun;
for it is far easier to prevent a bad project than, once it has been started, to close
it down.
Part 2: Developments, Concepts and Discourse (2004)
The legacy of Perkerra and its need for subsidies has continued into the 21st century.
After the 1960s, agricultural settlement projects became less favoured in most parts of
the world. Big dams and other projects were increasingly recognized to have high human
costs in the many who were displaced, dispossessed and inadequately compensated.
Champions within the World Bank and internationally networking activists made
lenders, donors and, to a lesser degree, governments, more aware of human costs and
more reluctant to fund dams and other projects which displaced people.
Commitment to projects and continuity of aid agency and developing country
government staff, created conditions for innovation, learning and changes in practices
and policies. In the 1990s, as aid agencies and governments shifted their priorities to