Evaluating development policies and programmes in the third world
| Date | 01 April 1985 |
| Author | Thomas B. Smith |
| Published date | 01 April 1985 |
| DOI | http://doi.org/10.1002/pad.4230050204 |
PUBLIC ADMINISTRATION AND DEVELOPMENT,
Vol.
5,
NO.
2, 129-144 (1985)
Evaluating development policies and programmes in the
Third World
THOMAS
B.
SMITH
The Australian National University
SUMMARY
The issues and methodologies of programme evaluation have largely been developed within
the North American context of an open and competitive political process. The policy cycle in
Third World nations, especially authoritarian regimes, is
a
closed system with little
consultation with affected target groups, suppression of criticism of government officials and
policies, and
a
severe weakness in policy implementation. Third World governments prefer
evaluation methods such as cost-benefit analysis which will not undermine fragile regimes by
indicating problems in the policy formulation and implementation process. However, the
most useful form of evaluation is implementation analysis, which can pinpoint the reasons
why policies succeed or fail.
INTRODUCTION
It is important
to
consider that the initiative and development
of
policy and
programme evaluation is essentially a North American experience. Campbell
(1979,
p.
69)
has expressed the hope for an evaluation methodology to assess the impact
of
planned social change ‘that would be useful to any political system’. This may be
possible if one deals only with evaluation
techniques.
Although some
of
the
strategies and methodologies have been exported
to
overseas countries with
reasonably similar societies and political processes (such as Western Europe and
Australasia), the basic fact remains that the methodologies, assumptions and the
current debates are based upon the North American experience and as such
represent a very culture-bound set of assumptions, problems and methodologies.
Programme evaluation specialists who have addressed themselves
to
Third World
problems (Hoole,
1978;
Freeman
ef
al.,
1979;
Weiss and Fuller,
1983;
Apthorpe and
Gasper,
1982;
Shumavon,
1978)
have not,
I
feel, adequately considered the impacts
on evaluation strategies and utility of the very different environments which exist in
the Third World. What may be an appropriate evaluation strategy for the American
experience may have very little relevance or utility in very different societies
or
political systems such as in authoritarian Third World regimes. It is illustrative to
note the key features of the North American experience in which policy evaluation
strategies and methods have been developed and flourished:
1.
A stability in political and administrative institutions and processes with
regular elections for political leadership and supremacy of the political sector
over the bureaucratic.
Thomas Smith
is
Senior Lecturer in Political Science at the Australian National University, PO
Box
4,
Canberra ACT 2600, Australia.
027
I
-2075/85/020129- 16%01.60
0
1985
by John Wiley
&
Sons, Ltd.
130
Thomas
B.
Smith
2.
Democratic traditions and practices of individual and group freedom, lack of
suppression, reasonably ‘open’ government, an active investigative media,
freedom to criticize political leaders, administrators and government policies
and programmes.
3.
A diversity of centres
of
power with divisions and rivalries between executive
and legislative branches of government and between federal, state and local
levels.
4.
A slow, deliberative, incremental policy formulation process which, in some
policy arenas, leads to competition between organized interests.
5.
Public policies which are incremental in scope and impact, which do not
usually demand substantial change on the part
of
target groups.
6.
A reasonably ‘satisfied’, well paid, professional, non-corrupt public service
which expects evaluation and review as part of the job.
7.
An active and powerful political party out of power seeking to gain electoral
advantage.
8.
A lack
of
fundamental ideological cleavages with major political groups
more pragmatic than ideological in matters
of
policy, and with those groups
strongly supportive of existing social and political processes.
9.
An array of governmental and non-governmental bodies which support, fund
or
conduct evaluations and a large number
of
professional social scientists
from various disciplines who teach about evaluation and conduct
evaluations.
10.
A large database of social and economic information readily accessible
to
policy researchers.
Most
of
these conditions also prevail in Western Europe and Australasia and the
various debates and issues over the ethics
of
evaluation, research methodologies,
and use of evaluation reflect this relatively open policy environment. Evaluation
specialists when prescribing evaluation strategies seem content
to
recommend what
is regarded as sound practice to be applied universally. Hoole
(1978)
advocates a
standard quantitative impact assessment approach to evaluation using cost-benefit
analysis and social impact assessment. His prescriptions
for
a Third World
development programme evaluation strategy may apply readily to the North
American
or
Western European type of environment, but whether it fits the brief
description he provides
of
the context
of
Third World development policies and
programm; is debatable:
Development activities frequently take place in an unstable and highly
political setting. This will result in changing programs and priorities and in
problems in implementing and evaluating development activities. Factors
such as high turnover of office-holders and bureaucrats, domestic conflict
and strife, an inflationary economy, uncertain funding for the budget,
changes in the international economic order, famines, and unemployment
may mean that development planning is not meaningful, that development
activities are not implemented as planned, that changes are made as the
activity is implemented,
or
that the activity never occurs. Furthermore,
many development activities involve substantial changes in the traditional
way
of
life and result in resistance by the population to both the program
and its implementation (Hoole,
1978,
p.
126).
To continue reading
Request your trial