Attempts at bureaucratic reform have sometimes succeeded in establishing new administrative structures but have often failed to improve efficiency or change bureaucratic behaviour. This suggests that governments need to choose their candidates for reform carefully and then concentrate their efforts on them. Far from being mere tinkering, such an approach should be seen as the best way of achieving a longterm strategy for institutional development. It requires a political commitment to those strategic goals-a commitment that is difficult to sustain, particularly since it must contend with bureaucratic inertia and resistance. But in many countries this combination of selective radicalism and incremental change will achieve more than plunging into wholesale reform of the entire administration, and can gradually build support for change by showing results.
Persistence is fundamental to bureaucratic reform. This requires a permanent capacity, though not necessarily a single agency, to provide analysis and operational support for reforms. If that capacity exists, governments will be better placed to seize the occasional opportunity to make fundamental reforms because the preparatory and technical work will have been done. The experience of developing countries indicates that public service commissions, central personnel agencies, and the like are inappropriate overseers of administrative reform, being too limited in scope and preoccupied with detail. Institutes of public administration tend to be too remote from power, though they can help to diagnose the kind of reform that is needed. In several smaller developing countries, technical offices concerned with organization and methods or management services have been useful instruments of reform, but are rarely able to deal with the larger structural and performance issues.
Experience has also shown how persistence can pay off. Countries have avoided dealing with all administrative problems and none of them would claim complete success. Instead, they have concentrate on a limited range of objectives and persist with them. It is probably wise to link reforms to the budgetary process so that leaders can impose their priorities on the administration and redirect its activities.
In many developing countries it makes sense to base reforms on two broad principles: first, reducing the management intensity of development, rather than adding new managerial burdens to an already overextended bureaucracy; second, instituting reforms that make the bureaucracy more responsive, both to political authority and to the public at large.