152 Poverty and Hunger
With aid agencies shifting to sector support and policy influence, development
projects were being abandoned during the early 2000s. In Uganda, between one
financial year and the next, DFID’s (the UK Department for International Devel-
opment’s) ratio of project-to-programme funds shifted with astonishing abrupt-
ness from 5:2 to 2:5, without a significant change in total.^18 Ironically, this was at
a time when, in my view, lenders, donors and governments were getting better at
learning from projects. But as and when these were terminated, governments and
aid agencies lost the precious opportunities they had had for innovation and co-
learning. With one project in Tanzania, local-level staff had devoted years to project
preparation and negotiation only to find that they had been led up the garden path
and there would be no project (Groves, 2004). With another in Brazil, after rela-
tionships with NGOs (non-governmental organizations) and communities had
been built up over three years, people’s time and energy had been invested, and
enthusiasm and expectations had been raised, a decision was taken in DFID to
withdraw, leading to much anger, anguish and disillusionment. The hidden costs
of such abandonment can be incalculable. Local people who have engaged in par-
ticipatory planning and have been led to expect support are left in the lurch and
reconfirmed in their resentment and cynicism about government. Field staff are
seen to have misled their clients; they are made to look foolish, if not duplicitous,
and are furious, embittered and demotivated. Decent and perceptive aid agency
staff, too, are demoralized, embarrassed and ashamed, but do not have to face the
people on the ground. Centrally isolated office-bound policy makers in northern
capital cities may be blithely or wilfully^19 unaware of the distant damage they have
done. All too often, abandoning projects was unethical and anti-poor.
With experience from western India comes David Mosse’s (2003, 2005) fasci-
nating, perceptive and subtle analysis of the IBRFP (Indo-British Rain-fed Farm-
ing Project) with which he was closely involved until 1998, and which he revisited
and reviewed three years later in 2001. Despite his early criticisms, and despite
managerially exacting contradictions in the project, he shows that much had been
achieved, not least through participatory seed breeding and selection. This was a
core project innovation that challenged the prevailing regulatory frameworks and
bureaucratic practice of Indian agricultural research (Witcombe et al, 1996), with
huge implications for policy and practice. The project had also ‘brought a version
of “development” more meaningful than any previous to Bhil tribal communities
excluded from even the most basic state services’ (ibid, p24). However, by 2001 it
had fallen from favour in DFID and was threatened with closure, not because it
was failing but because projects had become unfashionable. As Mosse (2003) puts
it:
Project practice seemed to me unchanged – meaningful engagements between staff and
villagers still produced important local benefits even under conditions of severe drought.
But a fundamental change had occurred, not in the project but in donor policy ...
DFID’s India programme had become reorganized around the funding of state-wide
government programmes, sectoral reform and donor–government partnerships. Unable