Public Sector Governance and Accountability Series: Budgeting and ...

426 Salvatore Schiavo-Campo
By Objectives or by Results?
The classic approach to evaluation (assess the degree of achievement of the
objectives stated at inception of the task) and the pragmatic approach (assess
the results actually achieved, whether or not they match the initial objectives)
do not necessarily produce the same verdict. The classic approach has been
criticized for lending itself to excessive formalism and enabling a mutation
of simple and useful ideas into monsters of red tape. 18 But because M&E
capacity takes time to build, the pragmatic approach can degrade into an
alibi for perennial postponement of reckoning and accountability. On balance,
it is probably preferable to adopt the classic approach of evaluation by objectives
but complement it with some form of midcourse assessments. Thus,
evaluation shades into supervision.
In-house or External M&E Capacity?
The standard assumption is that M&E capacity should be created within the
government itself. Regardless of whether this assumption is correct, it is
surely fallacious to assume that because evaluation of government activities
is important it must be conducted by government. In-house evaluation has
the obvious advantage of inside expertise, savvy, and intimate operational
knowledge of the programs being evaluated (as in Australia). The other side
of the coin is a natural tendency to overstate results, and, where accountability
systems are weak or nonexistent, even to provide a coat of whitewash
to failed programs. 19 The advantages of external evaluation are, first, its
presumptively stronger independence and, second, the greater probability
that the evaluators are familiar with similar programs in other sectors or
other countries. 20
These advantages are not exclusive, however. In-house government evaluation
organs can also be assured of a degree of independence close to that
enjoyed by external entities. Conversely, if external evaluators contribute on
a regular basis, they will develop the intimate understanding of operations
that is needed for an informed assessment. The disadvantages, too, are not
exclusive: in particular, if the governance climate is not conducive to candid
evaluations, most probably even the best external evaluations will be suppressed
or distorted to produce the desired results. The choice is entirely
pragmatic. Thorough evaluations require a substantial input by economists,
researchers, and auditors—skills that are in limited supply in African developing
countries and are best used in designing and running sound programs,
not in evaluating them. Thus, evaluation in developing countries inevitably