ST/ESA/78 |
Department of Economic and Social Affairs |
SYSTEMATIC
MONITORING |
AND
EVALUATION OF INTEGRATED |
DEVELOPMENT
PROGRAMMES: |
A
SOURCE‑BOOK |
UNITED
NATIONS New York,
1978 |
Monitoring
and evaluation have become widely used terms in the literature on programme
planning) implementation and review. They are frequently used as
interchangeable synonyms, but actually refer to different activities and
processes. Although closely interrelated and interdependent, monitoring and
evaluation need to be distinguished clearly if they are to be systematic and
reliable tools for decision‑making in programme management, programme
planning and policy formulation.
Before
entering into a discussion about the methodological issues and practical
problems involved in designing and implementing monitoring and evaluation
procedures, it is necessary to define precisely the key terms to be used and to
explain the implications of the definitions set forth. This should help to
avoid terminological ambiguities and misunderstandings, particularly when the
discussion concerns other works on programme monitoring and evaluation.
Monitoring
and evaluation systems are discussed with special reference to integrated socio‑economic
development programmes. Thus, the definition of monitoring and evaluation has
to be related to a description and specification of the basic characteristics
of such integrated programmes. Section 1 deals with the essential attributes of
development programmes in general and the distinguishing features of integrated
socio‑economic development programmes in particular. The meaning of the
terms monitoring and evaluation is the topic of section 2
The
term "'programme" is meant to refer to a planned complex of
activities, including utilization of resources, undertaken to reach specified
objectives within a certain time‑limit and cost‑frame.
For
operational purposes, programmes must be planned as a sequence of subunits
comprising defined sets of implementation activities for defined implementation
purposes. These subunits are called "projects", 1/[1]
Development programmes and the strategies from which
they are derived, can be said to be integrated if they take full account of the
close interrelationship between economic and social factors and the
interdependence between the various sectoral subprocesses of development in:
(a) analysing the existing impediments to socio‑economic change and
progress; (b) formulating development goals and objectives; and (c) planning
practical measures needed for promoting developmental change. Owing to their
multifaceted approach to development analysis and planning, integrated
development programmes typically comprise several sectoral project components
designed to be mutually supportive and/or functionally interrelated.[2]
Example An example
of an integrated development effort is the activities undertaken by the Chilalo
Agricultural Development Unit (CADU) in Ethiopia.[3] The CADU programme was
intended to serve small‑scale landowners and tenant farmers. It included the following project
components:[4]
(a) Improvement of milk collection
and grain distribution in order to establish markets with fair prices that
would make it attractive for farmers to increase their production;
(b) Use of extension workers
and model farmers for stimulating the target group to buy and use high‑yield
varieties of wheat and fertilizers;
(c) Provision of credit at a
reasonable rate in order to enable farmers to buy the improved seeds and
fertilizers;
(d) Engagement in
agricultural research and experimentation for the purpose of obtaining a
continuous flow of well‑tested methods and means as a basis for extension
work;
(e) Motivation of the local
people to assume increased responsibility for the development work by helping
them to learn to help themselves, in order to make development a self‑sustaining
process;
(f) Training of staff for similar, subsequent
programmes.
Another example worthy of mention in this context is
the Matourkou programme in Upper Volta. It comprises three main sets of
activity: "Firstly, it is a pilot project for rural development. In a
model village and 10 communities or villages in the surrounding areas, it has
encouraged farmers to adopt a settled form of agriculture instead of the
prevailing cropping systems, and has introduced some modern farming techniques
including animal traction, improved seeds and tools, and the use of fertilizers
and pesticides. In addition, it has provided the communities with health and
educational assistance. Secondly, the project trains extensionists at various
levels, and 'technical agents' in agriculture, forestry, irrigation and
livestock. Thirdly, it has established and equipped a 1,000 hectare farm for
experimentation and applied research in horticulture.) seed multiplication,
livestock breeding and training for animal traction.[5]
As can be seen from the
examples, the projects included in a programme may be linked to each other in
either a causal or supportive way:
Causal (means‑end
linkage
In
this case, a project is intended to produce the inputs to, or the pre‑conditions
for the functioning of, subsequent ones. It thus relates to the programme
objective only indirectly. That is, the project affects the overall programme
objective through its impact on other projects included in the programme. This
type of linkage can be seen in the second example presented above, the
Matourkou programme. The results of the applied research projects will provide
inputs to the training of extensionists and technical agents. Their improved
knowledge and skills are, in turn, inputs to the training of farmers which may
help to improve their production techniques, increase their productivity, and
ultimately lead to an improvement of their socio‑economic living
conditions. This chain of events, i.e., the causal relationships that exist
between applied research, training of trainers, extension work with farmers
etc. and the programme objective, can graphically be depicted as follows:
Supportive linkage
When the projects comprising
a programme are mutually reinforcing so that one project supports the effects
and impact of the others and vice versa, the linkage can be termed supportive.
An example is the relationship that exists between the above‑listed CADU
projects (a), (b) and (c). on the one hand, and (e), on the other. The linkage can be illustrated
as follows:
Programme objective I in this case is to increase farm
production the strengthening of popular participation in the development effort
is stated as programme objective II.[6]
Programme objectives, as already indicated, state the
reason or purpose for the programme or, in other words, the changes the
programme is expected to initiate and accomplish in its environment. If the
ultimate objective is to satisfy human needs and improve peoples living
conditions, that is, if the programme seeks economic growth and change for
social purposes, it is said to be a socio‑economic programme".
Under present circumstances, programmes aiming at such
objectives as satisfaction of needs, improvement of living standards and levels
of well‑being must be primarily oriented towards alleviating the poverty
of the masses of low‑income people residing in rural areas and making
their development self‑sustaining.[7] It is in the rural areas of
the developing nations that poverty has assumed the most desperate dimensions,
and that, according to World Bank estimates,[8] 80 per cent of the 750 million people live who are
considered to exist either in absolute or relative poverty.[9] The most affected groups,
the poorest among the rural poor, are small‑scale farmers, sharecroppers,
landless workers and their families. A target‑group‑oriented
approach to rural development has been advocated with increasing vigour in
recent times as a direct and immediate way of reducing, and ultimately
eradicating, poverty, hunger, malnutrition, illiteracy, underemployment and
unemployment in these groups.[10] The notion of target groups
is intended to assist in defining the size, shape and developmental needs of
the rural poor. It is also intended to introduce into the programme planning,
monitoring and evaluation processes specific considerations about the way in
which the poor are benefiting from developmental efforts.
Integrated socio‑economic
development programmes, when defined as suggested here, pose two specific
monitoring and evaluation problems:
(a) Their multidisciplinary
and multisectoral approach to development requires that at certain critical
points in the programme implementation process, careful assessments should be
made as to whether and to what extent the linkages between the various project
components have been established in the intended way;
(b) Since their main purpose
is to improve the welfare of described target groups, any meaningful monitoring
and evaluation will have to provide an answer to such basic questions as what
impact the programmes have on the living conditions of their intended
beneficiaries and whether that impact is self-sustaining, i.e., of a relatively
lasting nature.
In view of these questions
that monitoring and evaluation must answer, it is useful to define
"programme result as the anticipated and. intended, as well as
unanticipated and unintended, observable changes issuing from the programme in the
respective target group or area. These programme results can
be classified as follows:
Outputs These are (a) physical outcomes
produced by the programme, e.g., number of schools built, kilometres of road
constructed, hectares of land irrigated, number of co‑operatives
established; and (b) measurement of services provided, e.g., in terms of
persons trained, home visit made, credits granted;
Effects These are direct and
immediate consequences of the programme outputs, e.g., higher crop yields,
increased school attendance rates, changes in levels of information,
preferences or skills;
Impact These are changes in the
standard of living of the target group or within the target area stemming from
the programme. They may take place at two levels: (a) the individual or family level, where they may include changes
in income, housing, nutrition or health status, and (b) the community level,
where they may comprise changes in the
structure of social services Ð including health care ‑ in the economic
system ‑ including production and infrastructure ‑ or in social
interrelationships and patterns of communication.
The results produced
by one project component may be inputs to other projects of the same programme
if, as was pointed out earlier, there exists a means‑end linkage between
them. In an integrated development programme, for example, a principal input
could be the work of field extension workers. It is often necessary to provide
these workers with a preliminary orientation and technical training that is
geared to the specific needs of the programme. The results of such a training
project, namely, the knowledge or skills imparted to extension workers, will
constitute at the same time inputs to subsequent projects undertaken within the
framework of the programme. In such a situation, it will be extremely important
to evaluate very carefully the results of the training project in order to be
sure what the input to the subsequent projects is and whether it meets the
requirements laid down in the programme plan.
Within a given
programme, two types of factors are important: programme inputs and programme
activities. Programme inputs consist of resources and manpower that are
provided for the purpose of attaining programme objectives. Programme
activities are actions undertaken by the programme staff in order to execute
the programme plan. These actions involve the utilization of programme resources,
such as machinery, technical instruments, medicine, teaching or training
materials.
Programme activities
could also be called 'conversion factors" because programme inputs are
transformed by them into results. Fertilizer, school books or credits by
themselves do not produce any developmental change. They need to be put to use
in a specific way, i.e., combined with certain activities. For example,
"twelve months of extension work" as an input into an agricultural
development programme can produce a variety of results, depending on the method
of action adopted by the extension worker. He/she may rely on local leaders as
change agents or may organize group discussions and demonstrations for farmers.
Different techniques tend to cause different effects. This example illustrates
that inputs and activities determine programme results; one is ineffective
without the other.
No programme operates in a socio‑economic vacuum.
Rather, it operates under certain
economic, social and political conditions which it seeks to infIcan1 which
influence it. These conditions can be termed "programme environment".
Within its environment, the programme will often be just one of several factors
causing change. Programme monitoring and evaluation are more complicated when
outside factors can be expected to produce the same results as programme
activities.
Example Studies
of community development have often found that exposure to ordinary mass
communications media, such as radio and newspapers, is frequently related to
participation in voluntary communal activities. The question for the programme
evaluator then becomes whether any increase in community participation was a
result of action by community development personnel or merely of routine
exposure to mass media. In order to determine which is the real cause of
change, information on both contact with community development workers and
level of exposure to mass media must be related to a particular activity. For
such an analysis, data must be collected on both the programme input (here,
community development work) and the other change factor (here, exposure to
radio).
It is generally useful to
distinguish between those factors of the programme environment which support
developmental change and those which impede programme progress. The former,
referred to above as "other change factors", can he called
"external support factors, the latter 'external constraint factors. To
mention just one example, the ethnic homogeneity of the programme target group
has often proved to be an external support factor, since it facilitates social
interaction and intercommunication, while extreme ethnic heterogeneity is
usually a major external constraint factor.
The relationships between the
various programme elements and the programme environment, as they have been
defined above, are depicted graphically in figure I.
The term evaluation is understood to
mean a process by which programme inputs, activities and results are analysed
and judged against explicitly stated norms. The norms may be the stated
programme objectives, work schedule, budget etc.
Three types
of evaluation can be distinguished according to the programme phase or element
focused upon:[11]
(a) Ex ante (or pre‑programme
evaluation is undertaken before programme implementation. It is
usually carried out in order to assess the developmental needs and potentials
of the target group or region, test programme hypotheses or determine the
feasibility of a planned programme or project;
(b) Ongoing
(or concurrent evaluation is undertaken during programme
implementation. It analyses the relationship between outputs on the one hand
and effects or, to the extent possible, impact, on the other. Its main purpose
is twofold: (i) flexible adaptation of the programme to changes either in its
environment or in the over‑all development goals, and (ii) early
detection of programme deficiencies and immediate design of intervention
strategies;
(c) Ex post evaluation is
carried out after programme implementation. It assesses programme
effects and impact, and aims at obtaining information on (i) the effectiveness
of the programme in achieving its stated objectives; (ii) it contribution to
the achievement of sectoral or national planning targets an development goals;
and (iii) the self‑sustaining character of the changes resulting rom the
programme.
|
The term monitoring usually refers
to the process of routine periodic measurement of programme inputs, activities
and outputs undertaken during programme implementation. Monitoring is normally
concerned with the procurement, delivery and utilization of programme
resources, adherence to the work schedule or progress made in the production of
outputs. Although primarily concerned with measuring these aspects of programme
performance, monitoring also aims at determining the following causal
relationships: (a) the relation between programme inputs, activities and
outputs, and (b) the influence of external constraint/support factors on
programme performance and outputs. The main purpose of monitoring is to
indicate as early as possible any short‑comings with regard to delivery
of inputs, execution of activities or production of outputs, in order that
corrective measures can be undertaken in time. As such, monitoring is primarily
a device for improving programme management.
The
relationship between the different programme phases and various steps in the
monitoring and evaluation process is shown in figure II.
There are
usually three categories of information that monitoring and evaluation are
expected to provide:
(a) Information
on programme performance which would indicate whether programme
implementation is proceeding in accordance with the programme plan and budget;
(b) Information on
programme effectiveness which would provide answers to such questions as
whether and to what extent the programme has achieved its objectives, and on
what external conditions it depended;
(c) Information on
programme efficiency which would permit determination as to whether
programme results were produced in the most economical way, i.e., by maximizing
outputs or minimizing inputs.
The main
focus of this source book will be the evaluation of effectiveness. This
involves such questions as: What impact does the programme have on the living
conditions of its intended beneficiaries? How were its results achieved? Are
the results of a self‑sustaining nature? How do they compare with the
stated objectives of the programme?
|
Systematic
monitoring and evaluation as presented here combine three basic assumptions:
1. Monitoring
and evaluation should be a tool for decision making;
2. Monitoring
and evaluation should be a continuous process of problem definition,
measurement, analysis and judgement;
3. Monitoring
and evaluation should be an integral part of the programme process.
Not all of
these assumptions are universally accepted. They will be explained in greater
detail, since they underlie the conceptual basis of the following discussion on
designing and implementing monitoring and evaluation systems.
The basic premise here is that monitoring and evaluation should be
designed in such a way that the information obtained facilitates and supports
rational decision making on the programme.[12] In the definitions given above, explicit
reference has therefore been made to the way in which the data secured by the
various types of monitoring and evaluation could be fed back into the
decision-making process.
Decisions on development programmes are taken at different political and administrative levels and in different functional contexts. Some are concerned with policy formulation, others with programme planning and still others with programme management. Decision makers who participate in the programme process in different capacities will each have distinct information requirements. Nevertheless, all information secured must fulfil three general criteria for monitoring and evaluation to serve as a tool for rationalizing decision-making. It must be:
(a) Relevant i.e., geared to the
specific needs of decision makers;
(b) Timely, i.e., available and accessible at the time
decisions are taken;
(c) Accurate i.e., reliable and
empirically verifiable.[13]
Policy makers may, for
example, he interested in obtaining reports on the contributions made by
various types of programmes to the achievement of national development goals.
Such reports would provide them with guidelines on the selection of future
development strategies, setting of development priorities and allocation of
scarce resources. On the other hand, programme planners, who are primarily
concerned with the technical soundness of programmes, might ask for information
that would indicate an eventual need for replanning of ongoing programme
operations and allow formulation of new programme designs in the light of past
experience. Programme managers would most likely have rather comprehensive
requirements for information relating to all critical aspects of the programme
process.
In addition to being geared
to the information requirements of decision makers, the data secured by
monitoring and evaluation have to be available at the right time if they are to
serve as a basis for decision making. Moreover, the basic purpose of monitoring
and evaluation would largely be undermined if the information provided was not
accurate. Reliability and validity are two important aspects of data accuracy.
Procedures used for data collection and analysis are said to be reliable to the
extent that repeated measurements of a single variable lead to similar results.
The validity of data depends on whether they really measure what they are
expected to measure.[14]
In order to produce relevant,
timely and accurate data, monitoring and evaluation activities should he
designed as a repetitive cycle comprising four essential elements:
(a) Problem definition i.e.,
specification of the topics to be studied and the information to be obtained;
(1) Measurement i.e., collection of data;
(c) Analysis i.e., review, categorization and
tabulation of data;
(a) Judgement i.e.,
comparison of findings against programme norms (e.g., objectives) and
assessment of their degree of concordance.
On the basis of judgements that are made, decision
makers select future actions from alternatives. The implementation of these
actions will pose new problems to be studied in subsequent rounds of monitoring
and evaluation. Although the topics and content of the various study cycles may
differ, their basic structure will in most cases consist of the elements shown
in figure III.[15]
Monitoring and
evaluation should he designed as a repetitive process because programme results
‑ the assessment of which is of primary concern here ‑ are produced
throughout the entire life cycle of a programme. They are influenced by factors
and forces that operate in all programme phases. Thus, in order to determine
precisely which elements contribute to or impede the progress of a programme, a
steady flow of information on its performance and achievements must be
obtained. This means that monitoring and evaluation have to be carried out on a
continuous basis covering the entire life cycle of the programme.
Of course, continuous
monitoring and evaluation of a development programme is an ideal. In actual
practice, this ideal will inevitably be modified, especially in the light of
the available financial and manpower resources. The data acquisition procedures
actually established will usually be close approximations of continuous
monitoring and evaluation. For various reasons, however, it may often be
preferable to collect the information on a periodic, rather than a continuous,
basis.
If the emphasis is on
assessing programme effects and impact, the data acquisition system that is
designed should minimally plan for (a) execution of a base‑line study;
(b) periodic replications of the base‑line study. A simple type of
"before‑after" measurement, consisting of a base‑line
study followed by a single replication, would not be sufficient. The results
obtained from such a measurement would not permit an assessment of which
programme inputs and activities were ineffective or which linkages between
project components were missing, and why this was so. The available information
could not be used for decision-making in connexion with programme management.
Consequently, many advantages of monitoring and evaluation would be lost and
the reason for having a data acquisition system at all would be subject to
question.[16]
Base‑line
studies, the first step in implementing monitoring and evaluation, are
undertaken either before or immediately after the commencement of programme
implementation. They are meant to serve two main purposes: diagnosis of the
existing situation for programme planning purposes; establishment of a bench‑mark
against which to measure future changes in the living conditions of the target
group or in the target area.
Data on changes
deriving from the programme will be collected through the replication of the
base‑line study at later points in time. The replication needs to be done
very carefully so that differences between the original and new data can be
interpreted to indicate programme effects or impact.
The number of
replications or restudies undertaken will depend mainly upon three factors: (a)
the nature of the programme process; (b) the information requirements of the
decision makers; (c) the resources available. These factors also determine the
number of and intervals between the routine measurements for monitoring
purposes.
To be a successful and
effective tool for decision making by providing a continuous flow of relevant
data, a monitoring and evaluation system must first be designed and established
as an integral part of the programme process.[17] There are three main reasons for this.
(a) The salient
feature of evaluation that distinguishes it from other research processes is
that facts (e.g., about programme performance or results) are judged against
explicitly stated norms. Thus) successful evaluation will only be possible if
the programme plan is formulated in a precise and detailed way, facilitating a
clear determination of the programme objectives and procedures needed to
achieve them. In designing the programme plan, consideration must therefore be
given to the way in which the programme should he evaluated;
(b) In order to
ensure continuity and comprehensiveness of data acquisition) adequate resource
provisions have to he made in the programme budget;
(a) In order to facilitate a prompt
feedback of results to programme managers, responsibility for monitoring and
ongoing evaluation should be assigned to the programme staff, since they are
the principal users of the information during the programme implementation.
This last point is a rather controversial one in the
literature on programme monitoring and evaluation. It is sometimes maintained
that evaluators should be emotionally detached from the programme and
consequently that studies on programme performance and achievement are best
performed by outside experts. The various reasons for one or the other strategy
are discussed in greater detail in chapter II, section C. It is emphasized that
the two approaches are not mutually exclusive but, rather) complementary to one
another since each serves a different purpose.
[1] For definitions of these two terms and those introduced below, see United States of America Agency for International Development, Project Evaluation Guidelines, 3rd ed. (Washington, D,C., l97b); and World Health Organization, Evaluation of Environmental Health Programmes Report of a WHO Scientific Group, Technical Report Series No. 526 (Geneva, l973)
[2] An extensive discussion of the concept of an integrated or, as it is also called, unified approach to development analysis and planning is to be found in "'Report on a unified approach to development analysis and planning; note by the Secretary‑General" (E/CN.5/519).
[3] Bengt Nekby, ed., CADU An Ethiopian Experiment in Developing Peasant Farming (Stockholm, Prisma Publishers, 1971).
[4] Ibid., pp. 53-94.
[5] Food and Agriculture Organization of the United Nations, Review of Field Programmes 1971‑1975 (Rome, August 1975), pp. ll6-117.
[6] Bengt Nekby, ed., op._cit., p. 17.
[7] For a more detailed discussion of this issue, see Uma Lele, The
Design of Rural Development Lessons from Africa (Baltimore and London, Johns Hopkins
University Press, 1975), pp. 12l.
[8] World Bank, Rural Development Sector Policy Paper (Washington, D.C., February 1975), p. 4.
[9] A widely accepted definition of absolute and relative poverty is that suggested by the World Bank. As a criterion for identifying absolute poverty, the Bank proposes an annual per capita income equivalent to $50 or less. The standard set for determining relative poverty is an income above $50 but less than one third of the national average. See World Bank, op cit. pp. 4 and 17.
[10] See, for example, International Labour Office, Employment, Growth and Basic‑Needs; A One‑world Problem (Geneva, 1976); Jan Tinbergen, co‑ordinator, Reshaping the International Order A Report to the Club of Rome (New York, E. P¥ Dutton 1916);, "Basic services for children in developing countries (E/ICEF/L.1342).
[11] Similar definitions are proposed in World Bank, Technical Workshop on Monitoring and Evaluation of Rural Development Projects and Programs Copenhagen, Denmark, 6‑10 December l976 SummaryReport (Washington, D.C., 1977), p.2.
[12] A decision‑making approach to monitoring and evaluation is presented in the following studies: Michael C. Latham, Planning and Evaluation of Applied Nutrition Programmes (Rome, Food and Agriculture Organization of the United Nations, 1972); B. J. Mardel, "Building a better MIS", Financial Executive, August 1970, pp. 65‑82; Herbert 0. Turner, "Principles and methods of program evaluationÓ, Focus Technical Co‑operation No. 3, 1976, pp. 26‑30; Carol H. Weiss, Evaluation Research Methods for Assessing Program Effectiveness (Englewood Cliffs, New Jersey, Prentice‑Hall, 1972); World Health Organization, op cit.
[13] These three requirements of a successful monitoring and evaluation system are strongly emphasized by B, J. Mandel, loc. cit, pp. 66‑68. A fourth requirement mentioned by Mandel, namely, economy, is discussed in a later context; see chap. II, sect. A.
[14] See Claire Selltiz, Lawrence S. Wrightsman and Stuart W. Cook, Research Methods in Social Relations (New York, Holt, Rinehart and Winston, 1976). chap. VI.
[15] A similar model depicting the evaluation mechanism in relation to the system being evaluated is set forth in World Health Organization, op cit. p. 13.
[16] This point is especially stressed by Michael C. Latham, op cit, p. 71.
[17] See, in this connexion Herbert D. Turner, loc, cit., pp. 26‑27; and World Bank, Technical Workshop on Monitoring and Evaluation... pp. 4-5.