You are on page 1of 123

Beyond Logframe; Using Systems Concepts in Evaluation

Foundation for Advanced Studies on International Development

The Foundation for Advanced Studies on International Development (FASID) was established in April 1990. FASID and its affiliate, International Development Research Institute (IDRI), conduct research, facilitate interaction among researchers and practitioners, and offer training programs to development specialists. These activities are aimed for improvement in the quality of development programs and policies.

Copyright 2010 by FASID Published in 2010 in Japan by the Foundation for Advanced Studies on International Development, 1-6-17 Kudan-minami, Chiyoda-ku, Tokyo 102-0074, Japan e-mail: pub@fasid.or.jp URL: http://www.fasid.or.jp

Editors Note

The 2010 Issues and Prospects of Evaluations for International Development employs systems concepts as clues to re-assess the conventional ways of conducting evaluations and to explore how development evaluation can potentially be made more useful. In Japan, development evaluation predominantly relies on the Logical Framework (logframe) when conducting evaluations. Evaluations based on a logframe often face difficulties. One such difficulty arises from the futile attempt to develop an evaluation framework based on a logframe, which, in many cases, was prepared as part of the early-stage planning of the project and which then does not necessarily reflect a projects real situation at the time of evaluation. Although a logframe can be utilised initially as a tentative project plan, logframes are rarely revised even when the situation has changed. By the end of the project, the original logframe may not be an accurate embodiment of what the project is about and therefore logframes do not particularly help in terminal or ex-post evaluations. Still, having been institutionalized by clients, logframe-based evaluations are common practice and in extreme cases, evaluators face the danger of evaluating the logframe instead of the actual project. Although widely used for its simplicity, logframes can end up becoming a cumbersome tool, or even a hindrance to evaluation. Various attempts have been made to overcome the limitations of the logframe and some aid organizations such as USAID 1, UNDP 2, CIDA 3 and the World Bank have shifted from the logframe to Results-Based Management (RBM). Now GTZ 4 is in the process of shifting to a new project management approach designed on RBM and systems ideas. In the first article, Beyond logframe: Critique, Variations and Alternatives, Richard Hummelbrunner, an evaluator/consultant from Austria, sums up the critique of logframe and the Logical Framework Approach (LFA), and explores some variations employed to overcome specific shortcomings of LFA. He then outlines a systemic alternative to logframe
1 2 3 4 United States Agency for International Development United Nations Development Programme Canadian International Development Agency Deutsche Gesellshaft fr Technische Zusammenarbeit (The German Technical Cooperation)

and introduces the new GTZ management model for sustainable development called Capacity WORKS. Richard has dealt with LFA and possible alternatives to LFA at various points along his career, and he is currently involved in GTZs rollout of Capacity WORKS as it becomes the standard management model for all BMZ 5 projects and programmes. What does he mean by systemic alternative? In the second article, Systems Thinking and Capacity Development in the International Arena, Bob Williams, a consultant and an expert in systems concepts, explains what thinking systemically is about and how it might help evaluation. He boils down systems ideas into three core concepts (inter-relationships, perspectives, and boundaries), and relates these concepts to various systems methods. In December 2009, FASID offered a training course and a seminar on this topic in Tokyo. Through the exchange of numerous e-mails with the instructors prior to the seminar, it occurred to me that the concepts might be more easily understood presented as a conversation. That is what we tried to do in the third article, Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams . These two instructors of the FASID training course and workshop explain in simple conversational style where and how we can start applying systems concepts in development evaluation. This issue also carries a report of two collaborative evaluations of Japanese Official Development Assistance (ODA) projects. The first case presents an innovative joint evaluation conducted collaboratively with Vietnamese stakeholders. The evaluation took place in 2009 2010 as the last year of a three-year evaluation capacity development project coordinated by the Japan International Cooperation Agency. The second case covers a joint evaluation study of another Japanese ODA project in Lao PDR with a local Lao administration for which neither logframe nor OECD DAC five criteria was used. Instead, an evaluation framework was developed from scratch, based entirely on the beneficiaries interests and perspectives. In both cases, a partner countrys participation in the evaluation necessitated considerable changes in perspectives of evaluation practice. I hope they provide examples of how boundaries and perspectives, as discussed theoretically in the first three articles, relate to development evaluation in practice.

The German Federal Ministry for Economic Cooperation and Development

I would like to express my sincere gratitude to the contributors. Each article or report is based on the opinions of each author and does not represent the opinions of the organizations to which the author belongs. It would be my pleasure if this publication contributes to the ongoing discussions in the development evaluation practice. March 2010 Nobuko Fujita FASID

Issues and Prospects of Evaluations for International Development Series IV

Beyond Logframe; Using Systems Concepts in Evaluation


Contents
Editors Note Articles I. Beyond logframe: Critique, Variations and Alternatives Richard Hummelbrunner 1 II. Systems Thinking and Capacity Development in the International Arena Bob Williams 35 III. Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams Patricia Rogers, Bob Williams, FASID 55 Report Enriching Evaluation with Perspectives of Local People Two Cases of Collaborative Evaluation Nobuko Fujita 75 Annex: The Socio-Economic Impacts of the Second Mekong International Bridge Nobuko Fujita, Eriko Takahashi 87 About the Contributors 119

1
Beyond logframe:
Critique, Variations and Alternatives
Richard Hummelbrunner 1

Over the last decades, the Logical Framework Approach (LFA) has become universally known and has assumed a key role for planning and managing development interventions 2. LFA, however, is not uncontroversial and the approach has been subject to criticism, concerning both its theoretical foundations and practical use. Despite these criticisms LFAs position has not been fundamentally weakened and while many donors acknowledge its limits and weaknesses, they maintain (some would say impose) its use as a planning and monitoring tool. This chapter reviews some of the experience gained with LFA and outlines major attempts to develop variations or move beyond it altogether. The first section briefly describes the LFA concept and summarizes the main points of critique. Section 2 then explores some variations which have been developed in response to this critique and to improve LFA as a management tool. Section 3 proposes a systemic alternative to logframe and Section 4 outlines alternatives to LFA which have recently been introduced in German development aid.

1. The Logical Framework Approach and its limitations


Over the past few decades LFA has come to play a central role in planning, managing and monitoring development interventions. Its origins lie in a US military planning approach, which USAID adopted for development projects sometime in the late 1960s. In the 1980s it was used by several international

1 2

Senior Associate, AR Regionalberatung, Austria Intervention is used here as a summary term for projects and programmes alike.

CHAPTER 1

development organizations (e.g. UNDP, ILO) and also taken on board by several European development organizations. In the late 1980s it was introduced as a planning and management tool for European development aid. And by the end of the 1990s it had become the standard approach required by many donors and directorates of the European Commission. 1.1. The Logical Framework - concept and rationale Originally the logframe was a tool for intra-organisational management in military and business contexts of the 1960s, which were marked by strong central authority and control around a relatively clear set of goals. It reflects a management style which demands precisely structured and quantifiable objectives (management by objectives), assuming that the actors dispose of all relevant information and operate in rather stable environments. The focus is on the delivery of activities and outputs, and on the achievement of intended effects through intended routes. It is useful to distinguish the Logical Framework (logframe) the matrix which summarises the main elements of an intervention and connects them to each other from the Logical Framework Approach (LFA) the process by which these elements are formulated: Although there are variations in terminology and structure, the Logical Framework as a matrix has a reasonably standard form which integrates two types of logic: i) a vertical logic as a hierarchy of objectives activities deliver outputs, which contribute to outcomes, which help bring about the overall goal; ii) a horizontal logic showing how progress against each objective can be assessed (indicators and means of verification) and the external factors (assumptions and risks) which might affect the achievement of objectives at the next level. The Logical Framework Approach is concerned with the procedures of problem analysis, the development of objectives and indicators, and the identification of risks and assumptions, which feed into the matrix. In general, this process should be a participatory one, involving key stakeholders in order to reach consensus on an intervention, which is then summarised in a logical framework. Thus the LFA is more than the logframe, but there are many variations using (or combining) these two aspects. Bakewell and Garbutt (2005) have concluded that the terms connotations fall into three broad categories: the LFA as a formal system, the LFA as a way of thinking and the LFA as a
2

Beyond logframe: Critique, Variations and Alternatives

brand to comply with donor requests. In practice, LFA can be used to produce a logframe (i.e. the ideal type described in the literature), but there are also cases where logframes are established without LFA (i.e. a participatory process) or LFA takes place without a logframe (i.e. fitting the agreements reached into a matrix format). LFA definitely has a series of advantages: It helps to think about and conceptualize interventions in a structured manner. It imposes a uniform way of thinking and demands a formalized way of representing an intervention that everyone who has been initiated into the method can read and understand. By providing a common terminology, it can facilitate discussion and exchange between the various stakeholders. The logframe provides a convenient overview of the main features of an intervention, as well as the information needed for monitoring and evaluation. This overview is particularly useful for (busy) senior officials, funders or supervisors, who require and prefer clear and simple descriptions of an intervention. However, as with any model, logframes are inevitably simplifications, which become dangerous when not seen as such; they can help logical thinking, not substitute for it. This danger particularly holds true when the logframe (matrix) is mistaken for the design of an intervention and not as a visual aid that summarises its most important aspects. 1.2. Critique of the Logical Framework Approach In public interventions (e.g. development aid) it is difficult and dangerously misleading to assume or attempt imposing the same sort of clarity and order for which LFA was originally designed: differences in opinion among stakeholders are likely to exist, which makes the assumption of consensual objectives difficult and often highly problematic. A single centre of authority rarely exists and in practice clearly agreed to objectives are rare to find. When pursuing these objectives in situations in which there are multiple and diverse stakeholders or changing or uncertain context conditions, the LFA tends to over-specify objectives and to over-emphasize control as opposed to flexibility. In his critique of LFA, Gasper concluded that it has been a practice with relatively little accompanying theory...while awareness of alternative possible formats and styles within LFA has increased in the 1990s, understanding of the nature of what LFA attempts, what it achieves and where it fails has been limited, compared to the remarkable spread in its use (Gasper, D. 2000, p.18). Based on an analysis of LFA use, he has pointed to three recurrent fail3

CHAPTER 1

ings which inhibit the potential of the approach from being fully exploited (Gasper, D. 2000, p. 21f): logic-less frame: When the use of logframes is imposed by donors, they are often invented after a project has been prepared. Thus only an illusion of logic is provided because the logframe format is used to accommodate a pre-existing design, rather than to help create a logical design in an appropriate format. lack-frame: The logframe is frequently too simple and omits vital aspects of a project, as not everything of importance can be captured in one table. Many users have underestimated that a frame includes some things and leaves others out, and that a frame-work is to frame (=help) the required work, not substitute for it. lock-frame: after an LFA has been prepared, it tends to be fixed and not up-dated, thus blocking learning and adaptation. Therefore oversimplified plans (matrices) become treated as blueprints that dictate outcome, and as a control tool to ensure that the required outcome is achieved. As a consequence, LFA is inherently easy to misuse. Its balance of advantages vs. disadvantages will depend on the characteristics of an intervention, its context conditions, the nature of the problem(s) faced and the characteristics of involved actors. Therefore an essential skill in dealing with logframes will be to know when not to use them or when to supplement them with other methods. The use of LFA in development aid has revealed some further problems, which have added to and aggravated the shortcomings inherent to the approach (Gasper 1997, Chambers 1997): Often LFA is imposed externally (e.g. by donors), and then tends to be applied in an over-standardized, rigid and top-down manner. Donors in particular tend to consider the logframe (matrix) as something definite against which aid recipients can be held accountable. LFA turns into a rigid directive and unbendable administrative rule, turning lack-frames into lock-frames against all logic. Power imbalances, low trust and existential distances between partners in aid programmes have contributed towards the lock-frame syndrome, fearing a loss of accountability and control if receivers are allowed to modify what has earlier been agreed. Use of LFA in a cross-cultural context has often led to the domination of an external concept and the development ideology on which it is based. Local management traditions or skills have been neglected and participa4

Beyond logframe: Critique, Variations and Alternatives

tion in the use of the approach has been severely hindered due to cultural alienation. And the use of LFA often ends with the funding, undermining attempts to induce sustainable learning and capacity building. LFAs has a mechanistic rationale, assuming a linear progression of effects which takes place quasi-automatic, i.e. irrespective of the actors involved or contextual conditions. But for instance, carrying out activities as planned is by no means a guarantee that expected results let alone impacts will be achieved, as there are many other important (external or internal) factors. It is very tempting to claim observable effects, whether or not an intervention has actually contributed to their achievement. This is often the case with higher-level objectives, in which contributions of single factors are easy to claim but difficult to (dis)prove. Or in the case of long impact chains, where causes and effects are rather distant from each other, either in time or in their functional relations. As a result, LFA often fails to reflect the messy realities facing development actors, thus producing confusion rather than clarity. In addition, context dependent and far from universally applicable terminology and language aggravates the situation. It is often difficult for different actors to agree on terms like output or outcome because they are not naturally occurring divisions in time. It is particularly difficult to communicate this way of thinking to actors with different logic or cultural backgrounds. LFA was introduced as the core management tool of the EUs PHARE Programme (Poland and Hungary: Assistance for Restructuring their Economies) in the early 1990s and was later expanded to provide financial and technical assistance to the countries of Central and Eastern Europe and prepare them for EU membership. In fact, during the 1990s it became the worlds most important and voluminous aid programme. Having worked as consultant and manager of PHARE funded projects, I gained much hands-on experience with LFA at the time. It had been introduced in one of its earlier formats, and had not taken into account LFAs conceptual limits as well as the experience gained with its use up to that date. Therefore it was little surprising that a lot of the deficiencies and problems outlined above (re)appeared when LFA was used with PHARE. LFAs deficiencies were aggravated by the fact that it was applied in a period for which this tool is not well suited: intense economic and social transformation in Central and Eastern Europe, i.e. contexts of high dynamic and complexity. In addition, LFAs usefulness was further dampened because
5

CHAPTER 1

implementation of the PHARE programme was notoriously slow, leading to considerable delays between the planning and implementation of projects. As the cumulative result of these factors, some of the most significant effects were (Hummelbrunner 2001): The tendency of the responsible authorities to enforce strict implementation of predetermined plans (as defined in the Terms of Reference ToR) led to a tunnel vision of projects. Projects were dissociated from their context and prior assumptions, regarding their implementation, were often neglected. Important developments in the implementation context were either overlooked or downplayed as undesired effects. Even though periodic reviews of plans were foreseen, the requirements and approval procedures had the effect that modifications were only considered at periodic, pre-defined review steps: if at all. This was often inappropriate, either too late or as in the case of inception reports too early. Given the rather short implementation periods for projects, approval times for project changes were particularly excessive (up to a year). Strict separation between planning and implementation of projects contributed to a mechanistic style of implementation, seen as merely the realisation of previously defined plans. This separation was aggravated by the fact that due to tendering procedures and considerations of fair competition planning and implementation were usually carried out by different actors. Due to this mechanistic style, the assumption side of logframes was often superficial, if not neglected altogether. In particular when project documents needed to be drawn up quickly, these external factors were considered almost irrelevant, further reinforcing the tendency for tunnel vision. LFA was implemented in a context of cultural alienation and poor understanding. External experts who were usually not sufficiently familiar with Iocal circumstance and culture dominated its use. Moreover, logframes were often grafted upon projects in retrospect by (foreign) experts and participation in this exercise by local actors was rather poor. Despite the rhetoric of partnership, implementation of PHARE often took place in a climate of dominance by EU authorities and mistrust towards national partners. Furthermore this led to excessive rules and guidelines, which created a climate that diminished the chances for a more flexible use of LFA and became an obstacle for its application.
6

Beyond logframe: Critique, Variations and Alternatives

As a consequence of this implementation framework, conformity to preestablished plans instead of performance in relation to changing needs and circumstance became the central criteria against which EU officials, national administrators and experts were judged and also rewarded each other. It was more advantageous for all of these actors to implement interventions according to original plans (as exemplified by the ToR), than to seek modifications when new insights or changes in context conditions warranted them. Collective behaviour patterns based on these values led to satisfactory reports, but hardly to better and more sustainable projects. Since the LFA is intended for assessing progress towards objectives, it is often used as a tool for monitoring and evaluation: However, due to conceptual flaws and limitations, it is not well suited for these purposes (Hummelbrunner 1997): Since the focus of LFA is on achieving intended effects via intended routes, its utility for monitoring & evaluation is rather limited. By restricting attention to these elements, the tunnel vision aspects are reinforced and effects other than those specified are systematically neglected. And even if unforeseen routes or unintended effects are taken into account, they are usually treated in an isolated manner and not linked to the original intervention logic, thus they have little explanatory value. This accountability bias also reflects LFAs fundamental assumptions of perfect advance knowledge and full control of implementation - despite the existence and conceptual importance of the Assumptions & Risks column in a logframe. But this part is usually taken the least seriously, especially if compared to outcomes and indicators. When it is not envisaged that events can develop in unforeseen ways, no need is seen for a learning attitude and a corresponding evaluation style, which could help in fine-tuning implementation and would need to pay attention to a broader range of effects. LFA can become a serious hindrance in situations where unintended effects or routes are important for understanding and assessing interventions. Or when the relations between an intervention and its context are entangled and hard to foresee, intended routes are not well understood or based on prior experience. In short, when situations become increasingly complex or require innovative routes. Gasper (1997) noted that the LFAs audit focus is an obstacle for learning and reinforces a climate of control. Under such conditions care is taken that
7

CHAPTER 1

interventions are made auditable: structured in line with the need to be monitored. This leads to a preference for measurable variables and short-term effects and to the detriment of information which is qualitative and can only be captured in the longer term. It can also turn into a veritable obsession with indicators (indicator-itis 3) and their quantification, forgetting that indicators are only observation tools and should not be confused with what they are intended to measure. But thats exactly what happens when indicators are (mis)used as substitutes for stated objectives. In recent years there has been less critical writing on LFA, or at least no essentially new arguments were brought forth. In their review of International Development NGOs Experiences with LFA, Bakewell and Garbutt (2005) have concluded that the individuals attitude to the LFA is related to how useful it is for their work: managers dealing with multiple projects like it as a useful summary to help simplify the complexity they face. However, those closer to the messy realities of development are less convinced. LFAs promise of improved management often remains unfulfilled and respondents in this survey lamented LFAs dominant role as a management tool. People operate with a much higher level of complexity than can possibly be included in a logframe, so the neat logic does not work in reality. Moreover, the LFA requires reaching a consensus on one theory of change, one logical path from activities to goals, which ignores that there may be alternative paths to get to the same goal. And it is usually the path that is most acceptable to donors which dominates. Where NGOs have invested in participatory LFA processes, they ironically found that the more participatory the approach, the more difficult it is to revise this consensus later on during implementation (lockframe). Whatever attempts were made to make the LFA participatory and flexible, it seemed unable to change as quickly as the environment within which it has been placed. And again: the more people participate in developing the logic and completing the matrix, the more difficult it is to adapt the programme to a rapidly changing environment. But despite such difficulties, the blame is often placed on those who use the LFA, rather than the approach itself. There is a common chorus...that the problem with the LFA is not the framework itself, but the way it is used (Bakewell and Garbutt, 2005 p. 12).

A word coined by Gasper the disease of being obsessed with indicators.

Beyond logframe: Critique, Variations and Alternatives

2. Variations of the Logical Framework Approach


Awareness of the need for different formats and styles within LFA has increased throughout the 1990s and attempts were made to adapt it in various ways. At the most basic level, the terminology was adapted to fit better with the specific conditions. Some have made changes to the matrix, for instance by adding additional rows to reduce the size of steps between levels in the objective hierarchy. Others have used nested logframes to break down large programme frameworks into smaller parts. But while such adaptations make it easier to map the reality of development programmes, they do not tackle any of the more profound objections to LFA. Another, more radical adaptation of the LFA is to abandon the matrix altogether, while retaining the basic elements. This has the possible advantages of avoiding the box format or allowing for better visualization, as is the case with Logic Models, for instance, which moved from a tabular (matrix) structure to diagrams. Using Logic Models, which already began in the 1970s, has become quite popular with the rise of theory-based evaluations. Although there are some methodological differences, programme logic models use a similar hierarchy of objectives and linear sequence of events. Three variations are outlined in this section: One, PCM which was developed early on as a reaction to initial critique of LFA, and two others related to Outcome Mapping and Social Network Analysis which link LFA with methods that have been developed more recently to deal appropriately with complicated or complex situations. 2.1. Project Cycle Management (PCM) In 1992 the Directorate General for External Aid (DG VIII) of the European Commission adopted Project Cycle Management (PCM) as its primary set of project design and management tools. PCM, which was developed at the end of an extensive evaluation exercise, was adopted with a rather sobering outcome. In his account on the making of PCM, the Director of DG VIII at the time, H. Eggers (1998), pinpoints the key reasons: There was a general tendency to confuse the project with the people that were meant to profit from its implementation and functioning; In many cases, vitally important aspects were overlooked in project preparation, implementation and follow-up; and All too often, decisions were taken without being subject to the required
9

CHAPTER 1

decision-making discipline appropriate to each phase in the cycle. These three root causes of project failure were then transformed into future guidance in the form of three principles, which remain the cornerstones of PCM today: 1. Express the project purpose in terms of sustainable benefits for the target group: This is the most basic principle of PCM, which should allow making a clear distinction between the project and the people affected by it. The Project Purpose is also called Specific Objective and its place in a log-frame is between Results and Overall Objective(s). 2. Devise a basic format setting out the vital aspects of a project: These are the criteria that must be observed if a project is to achieve sustainable benefits. The Basic Format is PCMs first technical tool and structures the criteria under the following headings: Background: presenting the policy guiding the project, the beneficiaries concerned and the problems to be solved; Intervention: spelling out the objectives of the project, notably the kind of sustainable benefits to be created, and the activities designed to achieve the objectives; Assumptions: alluding to the risks to be faced and identifying the circumstances or events considered essential for the projects success but that it cannot influence; Implementation: containing the concrete project description means to be employed, their cost, procedures to be applied, the implementation timetable; Sustainability factors: policy, technological, environmental, socio-cultural, management and economic aspects that are designed to ensure project survival after completion; Monitoring/evaluation: allowing the project to be kept on track and be reoriented, if necessary. 3. Devise a mechanism to guide sound decision-making throughout the project cycle: This principle requires the application of the same criteria (as expressed in the Basic Format) throughout the entire project cycle. The cycle of operations for managing the EUs external assistance projects has five phases, as shown in Figure 1 below 4. Whereas the duration and importance of each phase will vary for different
4 In the first version, the project cycle included six phases, but later Financing was removed as a single stage in the cycle, given that the financing decision is taken at different times depending on the EC Regulation under which projects are financed.

10

Beyond logframe: Critique, Variations and Alternatives

Figure 1
Programming

Evalution & Audit

Identification

Implementation Formulation

projects (depending on their scale, scope and operating modalities), three main principles are to be followed: Decision making criteria and procedures are defined at each phase (including key information requirements and quality assessment criteria); The phases in the cycle are progressive each phase should be completed for the next to be tackled with success; and New programming and project identification draws on the results of monitoring and evaluation as part of a structured process of feedback and institutional learning. The Format of Phases and Decisions is PCMs second technical tool. It provides guidance for the production of good-quality key document(s) in each phase (with commonly understood concepts and definitions) and the related decision-making mechanisms which they support. The logframe is PCMs third technical tool. But it is only part of the method and was integrated as an analytical tool, also taking into account its weaknesses. Therefore PCM tries to limit the use of the logframe for project design, and emphasizes the importance of the LFA (as a participatory process) as much as the matrix product. Logframe analysis is to be carried out during the Identification phase and involves three main elements: Problem Analysis, Analysis of Objectives and Analysis of Strategies (Stakeholder Analysis was added later on as a forth element). And the logframe is to be drawn up as a planning tool during the Formulation phase. The Basic Format and the logframe do overlap in parts and should be used complimentary: Whereas the Basic Format is employed in structuring the Terms of Reference for studies or reports along the project cycle, the
11

CHAPTER 1

logframe is a means for summarising the project design as well as a checking mechanism for the internal logic of the project and each document. PCM has been conceived as an instrument for improving development projects and it differs at least in theory from LFA due to a mutual learning philosophy, a more participatory approach and a positive debating culture. It should facilitate the management of larger, complex interventions and allow implementing them in a more flexible, learning-oriented manner. But in PCM practice the logframe is often given a (too) dominating role, which leads to overly rigid applications and overrides some of the benefits expected from PCM. The father of PCM, Eggers (1998), was very clear on this threat, when he formulated the following warning after several years of PCM practice: Dont be tempted to fall back into the routine of the logframe tradition. LFA and PCM are not the same thing. Applying only the logframe would, in my view, be a recipe for failure. Unfortunately, later versions of PCM (e.g. European Commission 2004) did not correct this flaw: They did take new aid delivery modalities into account, introduced institutional and organisational capacity assessments and provided a set of quality attributes, criteria and standards (the Quality Frame). But they did not fundamentally alter the role and function of LFA and the logframe in particular within PCM. As consequence, when applying PCM (especially with larger, long term interventions and/or under more dynamic context conditions) some of the weaknesses associated with LFA become even more visible and acute: Rigid planning: Although reviews are foreseen in principle, the procedural requirements often prohibit modifications (especially at higher levels), thus original plans turn into rigid tracks which are hard to leave or bypass, despite changes or better insights. Problem orientation: The focus is on problems and their solutions, as exemplified by the Problem (Tree) Analysis in the Identification phase. This is not only a retrospective leaning, but also an emphasis on obstacles or threats and neglects prospective creativity. Linear causality: The intervention logic (of the first logframe column) assumes a linear means-end causality between the various levels and a sequence of stages over time. Although there can be several elements at each level, there is no meaningful way of showing how they interact within or between levels. The assumed process of causation is also one way and therefore circular causality (feedback) and connections between ele12

Beyond logframe: Critique, Variations and Alternatives

ments or delays are neglected. But these interrelationships are essential for understanding and taking account of the dynamic features of interventions. 2.2. From Logical to Social Frameworks Social Network Analysis (SNA) is a set of techniques for analyzing social systems. It can be used to understand networks and their participants, i.e. grasp and describe the organization of the network as a whole as well as the position of individual actors. It offers a variety of techniques (e.g. matrices, network graphs) for visualizing, measuring and simulating relationships and allows for analyzing these relationships in visual as well as mathematical terms. Since SNA lends itself well for capturing complex relationships, it can also be used as an alternative to prevailing linear stage models for representing theories of change. Programmes involving various actors and operating at different levels of scale can be conceptualized as networks, in terms that are easy to understand and possible to verify. The UK based evaluation and monitoring consultant Rick Davies suggests using an SNA perspective to overcome some of the problems with logframes. It is useful to think of events described in the Logical Framework in an actor-centred way, not simply as events involving disembodied and abstract processes of change. When this is done most theories of change documented within a Logical Framework becomes clearer, along with their plausibility or lack there of (Davies 2005, p.3). He points out that the connections between each of the adjacent levels of the Logical Framework (and the associated types of actors) are far from linear. They can often be quite complex and the various linkages may overlap leading to networks of relationships. The same holds true for the indicators in the Means of Verification column. It is useful to conceive them as interconnected networks, rather than as a series of parallel one-to-one linkages because there are usually several indicators found at each level,. He has also developed several potential applications for SNA-based models of change (see his website for corresponding material): 5 The first and foremost is to use SNA for moving from Logical to Social Frameworks. A network perspective replaces the stages in time of Logical

His website http://mande.co.uk/special-issues/network-models/ has two special sections for SNA and LFA, each of them containing a range of documents and other material.

13

CHAPTER 1

Frameworks (e.g. from Activities to Goals) by a sequence of actors, connected by their relationships (e.g. intended beneficiaries, partner organization, and project staff). This can be seen as a potential impact pathway along which influence, money and material objects can pass in both directions. The difference between the two views is captured in the figure below, which was developed while working on a Monitoring and Learning Plan for an NGO (ALNAP). It shows how the traditional rows in a logframe can be refocused to describe the expected change in each actor within a larger chain of actors (a Social Framework). The other columns of the matrix, describing Objectively Verifiable Indicators, Means of Verifications and Assumptions, can still be used to describe what should happen with each actor in a Social Framework.
Figure 2. Relationships between rows in Social vs. Logical Framework view of ALNAP
New view: Actors who are part of a pathway Intended beneficiaries of humanitarian actions Old view: Stages in time

Super-Goal

Other organisations

Goal

ALNAP members

Purpose

ALNAP Secretariat

Outputs

Steering Committee

Activities

Nodes = Actors, or categories of actors Lines = Relationships (to other actors and amongst members of a category of actors Thick line = Main relationship documented in the Social Framework Thin line = Other relationships that are expected to be important, described under Assumptions
Source: Davies, R. (2009), p. 9

Such a social framework view has several advantages: Responsibility for achieving results is distributed along the whole chain of actors and the respective changes expected from each actor can be described. This form allows the building of several response levels into each row and the delega14

Beyond logframe: Critique, Variations and Alternatives

tion and the clarification of details of the framework to the actors concerned at each level. Since this pathway takes place within a wider network of actors, the map can show the range of possible relations between actors. This in turn can serve to identify alternative pathways for achieving desired changes that can then be used for defining viable options and guiding subsequent monitoring and evaluation efforts. Secondly, SNA can also be used to move beyond overly simple logic models, e.g. by connecting events which take place at different levels. For instance, complex causal relationships can be represented by linking output (indicators) and results (respectively objectives) and by assembling the expected contributions of each output. This is best done in a participatory manner, by using matrices projected on screens in workshops, and progressively filled in by participants, line by line, and revised, as discussion of each relationship proceeds. If the matrix becomes too difficult to handle, it could be simplified by focusing only on the most important linkages (those above a specific threshold value) and transforming the information into a network diagram. In this way, SNA tools can be used alongside logframes or logic models and applied to aspects that cannot be easily captured by the logframes or logic models. Whereas the two uses outlined above are best applied within a single organisation, SNA can also be used in situations where multiple organizations work together without a joint plan or theory of change. Here network structures can be considered emergent outcomes, resulting from the decision making of individual actors. For example, individual organizations often make their own decisions about which issue coalitions to support, but these individual decisions can result in an aggregate structure that may or may not be optimal from the network point of view. The process of mapping those networks, and feeding back the results to the network members, has the potential to facilitate decentralized planning and evaluation processes, without impinging on each actors autonomy. These uses exemplify ways in which SNA tools can be used for developing new methods of participatory and decentralized planning or evaluation, in situations which involve a diversity of objectives and actors. Strategies can be built bottom-up, by aggregating the views of individual actors, identifying their consequences, and resolving gaps through reiteration. Such a process can also provide useful information about overall objectives to all as well as guidance to individual actors about who they need to work with about what.

15

CHAPTER 1

2.3. LFA and Outcome Mapping In response to the weaknesses and criticism of other monitoring frameworks in complex change processes, Outcome Mapping (OM) was developed in 2001 by the International Development Research Centre (IDRC) in Canada with research partners in Asia, Africa and Latin America. By too narrowly focusing their planning, monitoring and evaluation activities, results-based management frameworks (e.g. logframe) have been found to be counterproductive: hindering interventions. This is an obstacle for innovative approaches, learning and flexibility and a strong marker for failure in complex situations. OM is an approach to planning, monitoring, and evaluating social change initiatives. It is a set of tools and guidelines that steer project or programme teams through an iterative process to identify their desired change and to work collaboratively to bring it about. Its originality lies in the shift away from assessing the development impact of a programme and toward changes in the behaviours, relationships, actions or activities of the people, groups, and organizations with whom a development programme is working directly and seeking to influence and of the programme being influenced by these interactions. OM differs from other logic models in several ways: Foremost, it recognizes the importance of perspectives, i.e. that actors operate within different logic and responsibility systems. It is not based on a linear cause-effect framework but assumes that multiple (often nonlinear) causes lead to change. And it departs from the notion of attributing that change to specific intervention(s), but assumes that only contributions are made and tracks these by looking at the logical links between interventions and behavioural change. The focus of OM is on one specific type of outcome, changes in the behaviour of boundary partners, with whom a programme works directly and anticipates opportunities for mutual influence. These outcomes should at least be logically linked to programme activities (but not necessarily directly caused by them). While recognizing that challenges beyond the scope of the programme exist, OM limits performance assessment to a programmes direct sphere of influence. It monitors and evaluates whether a programme has contributed to intended behaviour changes and does so in ways which can be sustained in the future. In that sense it is also one of the most boundary-aware monitoring and evaluation approaches. In operational terms OM defines three distinct but interrelated sets of
16

Beyond logframe: Critique, Variations and Alternatives

activities and changes, and provides tools to monitor them: changes in partners, programme strategies and organizational practices. By linking the programmes organizational assessment with monitoring the changes of their partners, OM assumes that a programme needs to grow and develop in order to make a significant contribution. Programmes are encouraged to think of themselves as dynamic organizations, improving their ability to work with their partners by reviewing and adjusting their own goals and methods. OM assumes that change is controlled by the partners of an initiative, and that development (aid) programmes, as external agents, can only facilitate the process for a certain period of time by providing access to new resources, ideas, or opportunities. A focus on partners behaviour emphasizes the need to effectively devolve power and responsibility to endogenous actors, as a condition for success. Although data is gathered on the programmes actions and on changes in its partners, OM does not attempt to imply a causal relationship between the two. The programme can make a logical argument regarding its contributions to change, but cannot claim sole credit. By combining information on external outcomes with data on internal performance, a programme will be able to tell learning stories illustrating how the programme has improved its work in order to encourage the transformation of its partners. And it can document the extent to which the partners have moved in the desired directions. Lately some attempts were made towards a fusion of LFA and OM, in order to address the gaps that are perceived in each approach. Ambrose and Roduner (2009) postulate that such a fusion can integrate LFAs results-oriented focus with OMs process-oriented learning pathways. In their paper they define four cornerstones of the fusion model: Focus on different scales of results (behavioural changes, capacity building, and impact): Orientation on capacity building and on results are to be seen as complementary approaches but not as mutually exclusive options (unfortunately this is often the case). Looking for the common denominator: Instead of seeing LFA and OM as competing approaches, the intention is to combine them and thus use their respective strengths, i.e. the most convincing and effective elements of each approach. Capacity building: Since the OM approach was originally conceived for projects in which the focus is on changes in behaviours and capacities of partners, a fusion model will be most applicable for projects where capacity building plays a major role.
17

CHAPTER 1

Fusion = LFA+ or OM+: The goal of the fusion is to combine the advantages and strengths of both approaches, so that it is applicable in different institutional contexts in the most multifaceted way. The details of a fusion model should be defined in response to context and stakeholders, either by enhancing the LFA or OM. The focus consists of orientation towards an overall goal and explicit consideration of changes in behaviour of project partners. The fusion model should make it possible to determine and display the distribution of roles and responsibilities of development actors directly in the logic model. So far as it is reasonable and possible, this should be presented in synoptic form (e.g. as a table or matrix), outlining the summary of core elements of the intervention. A document containing detailed descriptions of all elements is needed for a more profound understanding. Key requirements and demands of a fusion model are (see also fig. below): Programme Goal: In a fusion model these goals are defined as results of the behaviour changes of the partners and following the logic of OM the responsibility for verifying whether these goals are achieved rests primarily with the programme partners. Outcome Challenges: They are formulated for each partner and describe the tasks, responsibilities and activities that they must carry out in order to contribute to programme goals within their system, including what they must do beyond programme support. Progress markers: Qualitative and quantitative indicators are defined for each partner for monitoring changes in their practice or behaviour. Progress markers may also be defined for several partners at once and need to be monitored at specific stages or times. Strategy maps and outputs: Clear and concrete description of the strategies (activities and outputs), roles and responsibilities. Outputs must have a plausible relation to outcome challenges and progress markers and should be verifiable through indicators. Mission of the external change agent: This is useful for defining the intended support to the partners and clarifying the change agents role (which is limited in time and scope). A fusion model can lead to con-fusion and more work, therefore it should only be applied under conditions where it can provide added-value. For instance in situations where there is a need to harmonise between various levels (i.e. from ministries to communities), when information for various
18

Beyond logframe: Critique, Variations and Alternatives

Figure 3: The Fusion Model at a Glance


Formulation of the desired situation as well as the practices and behaviour of project partners Description of concrete changes to be attained by the project (situational changes) Description of tasks and responsibilities of program partners Definition of tasks and roles of the project team & outputs that are provided to partners Mission: Defining the intended overall support by the external change agent Overall Goal (Beneficiaries)

Impact indicators

Program Goal

Goal indicators

Program Partner A Outcome Challenge

Program Partner A Outcome Challenge

Progress Markers (Outcome Measurement)

Output

Output

Output indicators
Strategy map Strategy map

Mission of the Change Agent & Project Management (Budget, HR, Organisational Practices)

Source: Ambrose and Roduner (2009)

partners or for different needs and accountabilities is needed, or if a programme should be improved in an LFA environment.

3. Systemic Project Management


The approaches described in the previous section are variations of LFA, which may be needed to overcome some LFAs shortcomings and render it a more effective management tool. A systemic management perspective, on the other hand, is an alternative approach that is based on profoundly different concepts and uses instruments other than the logframe. Since systems concepts and systems thinking are explored in other chapters of this volume, focus will be placed here on the implications for managing interventions. The systemic approach to project management outlined here is largely based on ideas developed at the University of St. Gallen, Switzerland, and the writings of Prof. Schwaninger and his colleagues at the Institute for Management Science in particular (2006, 2001 a, b). They have taken systems ideas, notably Stafford Beers Viable System Model (VSM), and elaborated an evolutionary management model, which has influenced management thinking in German speaking countries and beyond. A central aspect of the Viable System Model is the principle of recursion.
19

CHAPTER 1

This principle states that, in a constellation of systems which are themselves composed of sub-systems, each system, regardless of the level on which it is situated, will display the same structure. This means that systems that are structured according to this principle will not display the pyramid-type structure familiar from organizational theory. Rather, they will be like mutually interlocking and nested systems that resemble Russian Matryoshka dolls. Systemic Project Management consists of three main elements and corresponding tools: At the core are Primary (value creating) Processes, i.e. those activities that are directly responsible for producing desired outputs and outcomes. In development interventions these primary processes usually consist of activities, which are implemented by (public or private) actors/owners for whom an external agent (donor) provides resources, mostly in the form of funding, but also advisory services, provision of information, co-ordination etc. These processes are conceived as networks or loops, which connect the various elements (or variables) that are considered decisive for producing the desired effects. Causal Loop Diagrams (CLDs) are an appropriate and widely used representational tool. They are based on the concept of feedback (originally developed in cybernetics), and their building blocks are feedback loops, i.e. closed sequences of causes and effects: variable X is affecting Y and Y in turn affecting X, due to the relations between them. Feedback loops can appear in two types, reinforcing (positive) or balancing (negative) feedback. CLDs visualize variables and their relationships over time, thus permitting the analysis of relational patterns of a primary process. But they also allow looking beyond current states and reaching a dynamic and systemic level of understanding, by mapping the structure that is responsible for producing recurring patterns of events over time. On the other hand, the primary processes can also be conceived as the product of interacting social systems, which are involved in producing the desired outputs or outcomes (e.g. donors, project staff, beneficiaries, external experts). These social systems are semi-autonomous agents who act according to their own logic and values. In cooperation systems (without a central command entity) their interactions are essentially self-organized. Thus the behaviour patterns do not only depend on the properties of the individual actors, but also on their relationships. However, self-organized behaviour is neither predictable nor determined by the starting conditions alone,
20

Beyond logframe: Critique, Variations and Alternatives

but will evolve over time. Therefore the management task should be conceived as the steering of interlinked value-creating processes, which needs to take into account the logics of the actors involved. Communication and managing relationships are important ingredients of such steering processes, which should also aim at reconciling the various logics. If the logic of one actor (e.g. the donor) becomes too dominant, sub-optimal outputs or outcomes are likely and might even jeopardize the achievement of objectives altogether. Although this steering task is essentially fulfilled by those who are in direct contact with beneficiaries (e.g. project staff), they act within a framework established by others. Such a framework can be regarded as a set of nested systems, and steering should be done in line with the principles of Recursive Management, which is the second element of the method. This means that the interfaces between a system and its subsystem are managed in similar ways. Therefore the flow of information follows the same self-similar (fractal) pattern at all organisational levels. The following figure illustrates the management levels of a cooperation system (e.g. Donor, Programme and Project)
Figure 4: Recursive management model

Interface Donor

Programme (e.g. Country, Sector)

Interface Project
Source: Adapted from Schwaninger, M. (2001a)

The interfaces between these levels can be designed according to the principles of management by objectives: Contracts or agreements determine what other systems (e.g. donors, partners) expects from a project, for instance the objectives to be met or the quality criteria to be achieved; fur21

CHAPTER 1

thermore it defines the most important framework conditions for implementation (e.g. rules, milestones, issues to be taken into account). What should not be specified, however, is how this contract is to be fulfilled, namely the activities and operations envisaged. This is entrusted to the sole responsibility of the project and its respective cooperation system, i.e. external agents and their partners. Thus the agreement is not concluded any more at the level of outputs and outcomes (as in a logframe), but instead at the meta-level of objectives, core processes and rules. Such an approach not only drastically reduces the workload for planning and monitoring, it also makes a project more adaptive, flexible and realistic. If interventions are designed and their implementation is organised in such a recursive manner, monitoring and control requirements will be substantially reduced and their structure considerably altered. Instead of a hierarchic top-down manner, control will be organized systemically, whereby each level monitors its own performance, but within the framework defined by the level(s) above. Systemic Control is the third element of the method and should follow similar principles as management, essentially linking self-control processes carried out at various levels in a recursive manner. Thus responsibility for the entire system is assumed jointly and is translated into rules for self-control of the various sub-systems, to be supervised by the corresponding higher level. But standardizing these processes is difficult, which essentially require sensitivity for change. To this end, control tasks must be carried out in a decentralized manner and by the same actors who are responsible for planning and implementation. They have the best knowledge on planning assumptions or implementing conditions and are therefore well placed to detect relevant changes, including weak or earlywarning signals. These can foreshadow important changes, but often remain unnoticed in centralized control systems. In more general terms, the tasks of Systemic Control can be associated with different management levels. These are not to be confused with the subsystems of the recursive management model, but the tasks can be distributed across the subsystems. Systemic control is carried out at three levels of management (derived from the VSM concept), which act in support of the primary processes: The task of the normative level is to assure that things are done at the service of the larger whole: It establishes policy in light of competing demands between the present and the future, and between internal and external perspectives. It ensures long term viability and development,
22

Beyond logframe: Critique, Variations and Alternatives

Figure 5: Model of Systemic Control and corresponding management levels


Normative Management

Strategic Management Environment/ Context

Operative Management

Primary (Valuecreating) Processes

Source: Adapted from Schwaninger M. (2001a)

identity and vision. The task of the strategic level is to do the right things: It decides on resources for the primary processes and controls their effective use. To this end it acts as interface between the normative and operative level, but also monitors the environment and assures organizational stability through adaptation. It is carried out through supervisory entities (e.g. steering committees, boards) and makes use of intelligence tools like foresight or scans. The task of the operative level is to do things right: It provides information and communication which are useful to coordinate the various primary processes, assuring stability and conflict resolution. This is usually carried out through coordination teams, information systems or internal service providers. In spite of rising complexity from the operative to the normative level, the steering tasks at each level follow similar patterns and are essentially relying on the self-organising capacity of these sub-systems. In this model, hierarchy is a logical construct and does not imply a hierarchy of command and control. Therefore authority is not rooted in chains of command, but in the relevance of information. And continuous, real time feedback should ensure that the information is kept relevant during the entire process duration. Schwaninger and Krner have tested elements of this model with development projects (2001b). And together with colleagues from AR Regionalberatung I have applied this model to the management challenges
23

CHAPTER 1

posed by the EUs Structural Fund Programmes (Baumfeld, Hummelbrunner, Lukesch 2002). These programmes are characterized by multiple objectives and involve several implementing agents located at different levels. They normally consist of a set of support measures with specific objectives and budget that are implemented through a large number of projects, with a given time frame and pre-defined funding conditions. More recently, we have also applied these ideas to the management tasks of territories (regional governance) and development programmes implemented in multi-level, multi-actor contexts.

4. Alternatives to LFA in German Development Aid


The German Association for Technical Co-operation (GTZ) has been a European pilot user of LFA. After initial experiments with the logical framework in the 1970s GTZ have developed their own variant called ZOPP (objectives-oriented project planning). Some of the criticism on earlier LFA versions has been taken into account and lead to the incorporation of new analytical elements (e.g. stakeholder and problem analysis). ZOPP was also conceived as a participatory process, which consisted in a series of stakeholder workshops that were facilitated by trained moderators and involved the use of metaplan (cards) technique. ZOPP was formerly adopted in 1983 as a compulsory method and rapidly became GTZs trademark. It was also received quite favourably by the international aid community and judging by how many donors and NGO have taken it up, GTZ is probably the most widespread 2nd generation version of LFA today. Despite this positive response, however, the appropriateness and delivery modes of the method, particularly for cross-cultural work, have attracted increasing discussion. These criticisms lead to substantial modifications and eventually to a downgrading of ZOPP. From the mid-90s it was treated as just one set of tools among many others, which required flexible and selective use. From 1995 onwards GTZ relaunched LFA within the broader perspective of PCM. Later on, other formats were tested, which conceptually departed from LFA and used a systemic perspective. This notably involved the work of Schwaninger (see section 3) and SINFONIE, a systemic planning method for strategy development in complex situations.

24

Beyond logframe: Critique, Variations and Alternatives

4.1. Managing for Results (Orientation on Impacts) With the formulation of the Millennium Development Goals and the Paris Declaration, the discussion on quality and effectiveness of development cooperation has taken on new importance. New approaches for measuring and attributing results were introduced. This greater emphasis on results concerned not only the procedures and instruments used but also their correct application, calling for a change in attitude and modes of conduct. The paradigm had shifted to an understanding that development cannot be precisely planned and risks are not always foreseeable. In the past GTZs quality assurance was based on detailed, goal oriented and logical project planning (quality at entry) combined with performance monitoring based on comparisons of the actual situation with targets. The assumption was that detailed situation analysis and goal-oriented planning automatically led to quality and success. But this assumption was increasingly contradicted by project practice. And it became evident that this kind of rather input-oriented project planning and implementation did not necessarily correlate with the economic and social development processes in the partner countries. This insight has led to a stronger focus on development results (quality at exit). Consequently managing for development results has been enhanced at all levels of project work (GTZ 2004).
Fig 6: Managing for development results at GTZ
Before: Input/Output Orientation
focus on output Input/Activities Outputs less on impact

Outcome/Results

Impact/Results

Today: Managing for Development Results (Outcome, Impact)


focus on outcome and on impact

Input/Activities Outputs Use of Outputs Outcome/Results Impact/Results


Source: Reuber, M. and Haas, O. (2009)

All GTZ monitoring & evaluation activities are therefore now geared towards results. The project environment is examined from two perspectives:
25

CHAPTER 1

first, the effect the environment has on the project, and second, the extent to which the project contributes to the changes in the environment. Typical questions in results-based monitoring & evaluation are: Which areas may be affected by the projects activities and outputs? Is the project about to achieve its intended results? Are there any unintended (positive or negative) results that need to be addressed? What changes can be observed in the project environment and can be plausibly attributed to the project?
Fig. 7: The GTZ results model (results chain taking the example of vocational training)
Impact / overarching indirect results

Poverty reduction

graduates find employment and income

Attribution gap
Outcome (direct results) Intermediate outcome (use of outputs) Outputs Activities Inputs materials and equipment, personnel, funding
Source: Reuber, M. and Haas, O. (2009)

graduated have the knowledge and skills damanded by the labour market

the application of new curricula at vocational training institutes

demand-orientated curricula are available development and proving of new curricula

The planning of a project is usually based on cause/effect hypotheses, i.e. assumptions of the relationship between intervention and result. The GTZ results model (orientation on impacts) represents the results of project outputs on different levels: Use of outputs: This level describes whether and how the outputs of a project are used by the beneficiaries (e.g. the application of new curricula at vocational training institutes). The direct results describe the achievements through the use of outputs (e.g. graduates have the knowledge and skills demanded by the labour market). The overall objective of the project is located at this level (out26

Beyond logframe: Critique, Variations and Alternatives

come). The overarching indirect results describe the changes of higher aggregated development progress (impact) to which GTZs work aims to contribute via direct results (e.g. graduates find employment and income possibly leading to poverty reduction). It is generally difficult to attribute development progress at higher aggregated levels because this concerns comprehensive and complex changes in the environment of the project, in which several actors are often involved. It is methodologically challenging to clearly attribute the contribution of an individual project to changes at higher levels. GTZ therefore acknowledges the existence of an attribution gap. Although it is difficult to establish an unambiguous causal relationship between the project and the changes in its environment, a plausible attribution should be attempted. After all the real purpose of project work lies in achieving development progress beyond the attribution gap (e.g. poverty reduction). In 2002 a new contractual framework (named AURA after its German acronym) has incorporated this orientation on impacts also in the contractual relations between GTZ and the Federal Ministry for Cooperation (BMZ). These contracts (and the corresponding documents) need to specify the development objectives and results to be achieved. They must be jointly agreed with the partners, who are together with the external agent (e.g. GTZ) responsible and accountable for their achievement. But the operational details are deliberately left out of these contracts and delegated to the level of the supported project or programme (in line with a recursive management philosophy). In practice this has lead to much shorter documents, less prescriptive designs and more flexibility in implementation. I have incidentally taken the basic ideas of this orientation on impacts, blended them with elements from Outcome Mapping and adapted them to suit the needs and requirements of EU Structural Fund programmes. I have labelled this impact-led monitoring approach Process Monitoring of Impacts, as it is essentially about identifying processes considered relevant for the achievement of effects, and monitoring whether these processes are valid and actually take place. It is a theory-based approach that makes use of logic models, transforms them into circular ones and deals with them in a systemic manner (Hummelbrunner 2007). 4.2. Capacity WORKS The orientation on impacts approach specifies what should be achieved, but
27

CHAPTER 1

leaves open how this is to be done. To this end GTZ has developed a new management model called Capacity WORKS 6. This model was also intended as a response to changes in the aid architecture and stakeholder landscapes, i.e. the delivery via programmes (instead of projects) and the need to steer them in supra-organisational cooperation systems (beyond managing the GTZ contributions). In short, to better handle the increasing complexity of development work. An important role of GTZ is to assist in steering development projects, i.e. establish and secure functioning cooperation frameworks. Projects and programmes are conceived as cooperation systems involving many partners. But in supra-organisational cooperation systems none of the participating entities is legitimated to lead these co-operations in the traditional sense, since decisions are not based on hierarchies. Rather these systems are based on negotiation processes and take place within political processes. Therefore Capacity WORKS should help to identify, focus and work through the relevant processes to achieve negotiated, agreed and measurable results. In elaborating this new model, GTZ could build on its long tradition of methodological development, using the experience gained and incorporating elements from its past repertoire of methods. In addition, a survey was carried out on a series of programmes that were particularly successful in the past. In-depth interviews with key staff identified several success factors that renowned experts and researchers further validated. These empirical success factors were to form the core of the model, which were then complemented by a set of tools to assist in working with them in practice. Capacity WORKS is a management model to support contract and cooperation management especially during the implementation phase, though it is also suitable for the project appraisal and preparation phases, as well as the concluding phase. The model is designed to guarantee the necessary coherence with existing procedures and tools such as project preparation, approval of the offer concept, project progress review, etc. Moreover, Capacity WORKS is GTZs management model for sustainable development and translates this corporate policy concept into practical questions. The key points of reference for Capacity WORKS are the objectives and results jointly agreed with partners. The agreement of objectives and results should therefore be understood as a fundamental process of political negotiation, and structured accordingly. Results chains are vital tools in this process.
6 Capacity WORKS is protected as a word mark both in Germany and internationally.

28

Beyond logframe: Critique, Variations and Alternatives

They describe how activities based on inputs provided by the partners, including the German ones, are to generate outputs and, ultimately, sustainable results. Subsequently all decisions should be measured against these objectives, and monitored accordingly during implementation. Capacity WORKS operates with five success factors. These are central to the approach and serve as a methodological guide for contract and cooperation management. The success factors are also instrumental in project and programme design. They provide the framework for negotiating the project or programme with partners. The success factors applied in Capacity WORKS are based on the European Foundation for Quality Management (EFQM) model, adapted to meet the specific demands inherent in steering complex projects and programmes: The five success factors of Capacity WORKS and their key messages are (GTZ 2009): 1. Strategy: Negotiate and agree on the strategic orientation. A clear and plausible strategic orientation leads to positive results. 2. Cooperation: Network people and organisations to facilitate change. A clear definition of who the project/programme will be cooperating with and how, leads to positive results. 3. Steering Structure: Negotiate the optimal structure. An effective steering structure leads to positive results. 4. Processes: Manage processes for social innovation. A clear understanding of the key strategic processes leads to positive results. 5. Learning and Innovation: Focus on learning capacity from the outset. Individual and organisational learning capacities in all success factors lead to positive results. The concept and action in each success factor are guided by key questions. And the model includes an extensive management toolbox, which is described in detail in an internal Manual. Each of the 40 tools contained in the toolbox is assigned to one of the five success factors and their key questions. Incidentally, the logframe (called Project Overview Plan in the ZOPP tradition) is also part of this toolbox. The logframe has been retained because it is still used quite widely within the GTZ community and continues to be a methodological anchor for many (older) staff members. And because it is useful for planning, provided it is applied with care and embedded in a range of complementary tools. It is therefore included as a planning tool, intended for translating strategic orientations into implementation plans.
29

CHAPTER 1

Most of the tools originate from practice in organisational development, process or (systemic) management consulting. Although all of the tools have been tried and tested, they need to be adapted to the context and the stakeholders concerned. Thus Capacity WORKS provides a structuring framework, but does not relieve the user of the need to perform some conceptual and creative work. The model does not provide blueprints, but is intended to create scope for action and provides a menu which is proven and based on experience. One of the expected benefits is to promote communication on the level of how the GTZ works, both internally (e.g. in teams, with partners) and externally (e.g. with other donors, in cooperation forums). It is expected that each officer responsible for contracts and cooperation will be able to provide soundly articulated and differentiated responses to the questions raised in the context of the respective success factors. This should support and guide the dialogue on the quality of GTZ services. The experience gained in a pilot phase (2007 and 2008) in which Capacity WORKS was tested in more than 60 projects, was used to further refine the model into its present form. Since 2009 Capacity WORKS has been gradually introduced in all GTZ projects and programmes worldwide. Initially by training all expatriate staff and selected national counterparts and then by training external experts/consultants who can then use Capacity WORKS only after they have been licensed to do so. Capacity WORKS will eventually become the standard management and steering model of GTZ, at least for all BMZfunded projects and programmes. OEAR Regionalberatung has been involved in the pilot phase and is also part of the training team for the roll-out phase of Capacity WORKS, which is still going on. Having worked on this myself as a trainer, I was able to experience the enormous interest of the trainees and their overwhelmingly positive response to the new model. To my knowledge this is, to date, the most thorough move of any major international aid organisation beyond the LFA tradition and certainly a massive investment for retooling GTZ towards a management style which is expected to cope more effectively with the complexities of future aid delivery. However, in order to unleash its full potential it will be important that partners as well as other donors will also buy into this model, and thus resonate and reinforce GTZs pilot efforts. Although Capacity WORKS has been tailor-made to suit the needs of GTZ and German development aid, its underlying principles, success factors and tools could be transformed and
30

Beyond logframe: Critique, Variations and Alternatives

adapted to fit with other contexts or needs, as well.

Concluding remarks
LFA is an analytical and management tool which is now used (in one form or another) by most donors, aid agencies and NGOs. Despite formidable arguments which have been raised against its use, when intelligently applied (i.e. taking into account its conceptual limits and conducive context conditions), LFA can be an effective tool and the logframe remains the most common format for planning interventions. Apparently LFA continues to be useful for many people and organisations in development aid, especially those under pressure to demonstrate performance, assure effectiveness and accountability. Although deeply flawed, LFA offers a middle path, as it is a component of results based management but also allows for intense stakeholder participation, at least at the planning stage. And it seems to provide some degree of certainty amid the messy realities and insecurity of aid delivery. Despite numerous attempts no other approaches have been sufficiently convincing to aid managers to replace LFA as alternatives for planning and monitoring. Until now few organisations (GTZ being the most thorough and promising exception) have managed to make a radical departure from LFA. The use of LFA is least controversial in simple situations, where interventions can rely on proven tracks and can be modelled in a linear, mechanistic manner although even under such conditions it should not be handled in a blue-print style. In complicated situations, involving multiple actors or dynamic contexts, however, LFA can only be applied with much caution and should be complemented with other methods that are better suited for these conditions. Furthermore, in complex situations, displaying recursive causality or emerging outcomes, LFA is best not used at all, as its fundamental assumptions are not appropriate and LFA will not work under such conditions. Therefore it should be possible to base the utilization of LFA on rational choice and sound evidence. And nowadays sufficient variations - and even alternatives are available to avoid inappropriate application. In other words: a frame exists for a more logical use of logframe provided this frame is noticed and used.

31

CHAPTER 1

References German
GTZ (2004) Wirkungsorientiertes Monitoring: Leitfaden fr Vorhaben der Technischen Zusammenarbeit. Eschborn: GTZ Stabsstelle 04, Unternehmensentwicklung. GTZ/BMZ (2003) Handreichung fr die Bearbeitung von AURA Angeboten. Eschborn. Hummelbrunner, R. (1997) Die Logik hinter dem Logical Framework Strken und Schwchen des Planungs- und Evaluierungskonzepts der EU. Paper presented at IR Plattform Evaluation der Evaluationen, Feb. 26 1997, Vienna. Schwaninger, M. and Krner, M. (2001a) Systemisches Projektmanagement: Ein Instrumentarium fr komplexe Vernderungs- und Entwicklungsprojekte. St. Gallen: Universitt St. Gallen Institut fr Betriebswirtschaftslehre. (2001b) Komplexittsbewltigung in Entwicklungsprojekten: Ein systemisches Instrumentarium auf der Grundlage des St. Gallener Management Konzepts. Unpublished manuscript.

English
Ambrose, K. and Roduner, D. (2009) A conceptual fusion of the Logical Framework Approach and Outcome Mapping, OM Ideas Paper No. 1. Bakewell, O. and A. Garbutt (2005) The use and abuse of the Logical Framework Approach A Review of International Development NGOs Experiences. Sida. Baumfeld, L., Hummelbrunner, R., Lukesch R. (2002) Reforming Structural Funds: Towards a New Model of Policy Making based on Systems Thinking. Vienna: AR Regionalberatung. Beer, S. (1979) The Heart of Enterprise. New York: John Wiley. Chambers, R. (1997) Whose Reality Counts? London: Intermediate Technology Publications. Davies, R. (2005) Moving from Logical to Network Frameworks: A modular matrix approach to representing and evaluating complex programmes (2009) The Use of Social Network Analysis Tools in the Evaluation of Social Change Communications. Communication for Social Change Consortium. Earl, S., Carden, F., Smutylo, T. (2001) Outcome Mapping: Building Learning and Reflection into Development Programs, Ottawa: IDRC.
32

Beyond logframe: Critique, Variations and Alternatives

Eggers, H. (1998) Project Cycle Management Revisited. The Courier, MayJune 1998, Brussels: European Commission. European Commission (2004) Project Cycle Management Guidelines. Aid Delivery Methods, Vol. 1. EuropeAid Cooperation Office, Development DG Gasper, D. (1997) Logical Frameworks: A Critical Assessment. Working Paper no.278, The Hague: Institute of Social Studies. (2000) Evaluating the logical framework approach towards learning-oriented development evaluation. Public Administration and Development, vol. 20, 1728. GTZ (2009) Capacity WORKS: The Management Model for Sustainable Development. Eschborn. Hummelbrunner, R. (2001) Use and misuse of the LFA in PHARE and PreAccession Programmes. Unpublished manuscript. (2007) Process Monitoring of Impacts Applied Study for the European Territorial Cooperation Programmes, Vienna: INTERACT (http://www.interact-eu.net/interact_studies/63) Reuber, M. and Haas, O. (2009) Evaluations at the GTZ. Eschborn. Schwaninger, M. (2006): Intelligent Organizations: Powerful Models for Systemic Management. Springer Verlag.

33

Systems Thinking and Capacity Development in the International Arena

2
Systems Thinking and Capacity Development in the International Arena
Bob Williams
Drawing on an earlier OECD/DAC paper, the EuropeAid discussion document, Institutional Assessment and Capacity Development: Why, what and how? proposes the following definitions of capacity and capacity development 1: Broadly taken, capacity can be defined as the ability to perform tasks and produce outputs, to define and solve problems, and make informed choices. Capacity development is the process by which people and organizations create and strengthen their capacity over time. An uncontroversial and common sense statement, you might think. In contrast, here is something I wrote several years ago: Think of a water tank whose purpose is to store water for irrigation purposes. The capacity for irrigating the land is stored in that tank. The fuller it is the more capacity it has for providing sustenance. However, the capability to irrigate successfully depends on more than the capacity to store water. It depends on climatic conditions, historical rainfall, the presence of something or someone to turn the tap on and off at the right moment, the quality of the tap and water distribution networks. Thats what affects the capability of that amount of water to achieve its objectives. All the skills, knowledge, technical expertise and experience in the world
1 Institutional Assessment and Capacity Development Why, what and how? (2007) Office for Official Publications of the European Communities, Luxembourg. http://ec.europa.eu/europeaid/multimedia/publications/documents/tools/europeaid_adm_concept_ paper_en.pdf, accessed 2 June 2009

35

CHAPTER 2

wont help an evaluation if the capability of the program, community, organisation or environment cannot sustain and nurture those skills and abilities. What I generally see at the moment, is capacity building (i.e. building large storage containers) without a great deal of capability building (i.e. the complex and strategic business of getting adequate water on the ground at the appropriate time). If this continues, I believe we are in danger of being all dressed up and nowhere to go. Not everyone distinguishes between capacity and capability as above. Some use the distinction but switch the definitions and some use completely different words for capacity and capability. My point is, however, that distinguishing between the theoretical ability to do something and the actual ability to do something is useful.

What is thinking systemically?


Although a single, widely agreed upon definition of thinking systemically does not exist, a working definition could be, using systems concepts in addressing situations. This definition, however, raises the question of what constitutes systems concepts? and although no single understanding exists this paper provides a framework that accommodates most of the accepted systems concepts. In some ways identifying what systems thinking is not or not entirely is easier. It is not, for example, a wiring diagram with lots of boxes with even more arrows and lines drawn between those boxes. It is also not holistic in the sense that it includes everything. Most systems thinkers acknowledge that nobody can think about everything and even if they could it would be of little practical use. You simply cant take everything into account, so dont begin to try.

Systems thinking
So if systems thinking is not primarily about boxes, lines and arrows and not about thinking about everything then what is it about? And how can it help us think about capacity and capability?

36

Systems Thinking and Capacity Development in the International Arena

From systems to thinking systemically


Some history will help. The systems field as we know it today developed in and around the Second World War. That war posed some very tricky, seemingly intractable problems at the individual, team, organisational and institutional levels. Most of these had to do with capacity and capability the achievement of complex strategies with very limited resources. The pressure for increased capacity and capability not only promoted the development of systems ideas, it also spurred important innovations in organisational development, group dynamics, and action research. Indeed these ideas often overlapped and intermingled with the development of the systems field; this is nicely illustrated in a recent compendium of system thinkers (Ramage & Shipp, 2009). The history of the systems field is thus rooted in addressing complicated and complex problems with limited time and with restricted information. Over the past 50 years the systems field has expanded from its relatively modest beginnings into a suite of perhaps 1,000 or more methods and methodologies. Its orientation towards addressing capacity and capability issues however, has remained throughout. So what, at core, might thinking systemically be about? This article identifies one way based on three core concepts. All three concepts have deep roots that stretch back many years.

Three core systems concepts: Relationships, Perspectives and Boundaries


During the 1960s and 1970s the focus of the systems field was very much on inter-relationships. In many ways this was the wiring diagram stage of thinking systemically. By the mid-1970s it was clear that the inter-relationships, whilst important, were not neutral concepts. The relative importance of particular inter-relationships depended on the different purposes you could ascribe to any single situation. Thus thinking systemically began to address the implications of applying different perspectives to the same situation. However, by the mid-1980s it was clear that these perspectives were also not neutral. Perspectives determined what was seen to be relevant or not relevant and what lay in the system and what lay outside it. Whoever defined the dominant perspective controlled the systems boundary. Thus the importance of studying boundaries and critiquing boundary decisions (and those
37

CHAPTER 2

who made them) became the third core concept underpinning systems approaches. These three concepts are essential both for understanding systemic interventions and for distinguishing systemic interventions from other approaches for dealing with complex situations. These three concepts underpin all the models, metaphors, methodologies and methods used in the systems field.

Inter-relationships
Although the systems field draws on many methods that focus on interrelationships it tends to address five main questions: What is the nature of the inter-relationships within a situation? What is the structure of these inter-relationships? What are the processes between them? What are the patterns that emerge from those processes, with what consequences and for whom? Why does this matter? To whom? In what context? Partly because inter-relationships is the oldest systems concept it is the most familiar. Questioning how things are connected and with what consequence stems from the earliest thinking about systems. Inter-relationships is also the concept most strongly embedded in the popular imagination. For example, when we talk about the filing system, or the healthcare system, the image we have in our minds is of a set of objects and processes that are interconnected in some way. The popularity of system dynamics, with its boxes and arrows, further cements the notion that inter-connection is an important systems concept. The study of inter-relationships is central to any systemic inquiry. In particular, systems approaches look at the following aspects of inter-relationships: dynamic aspects (where the way the inter-relationships affect behaviour of a situation over a period of time) non-linear aspects (where the scale of effect is apparently unrelated to the scale of the cause; often but not always caused by feedback) the sensitivity of inter-relationships to context (where the same intervention in different areas has varying results, making it unreliable to translate a best practice from one area to another) massively entangled inter-relationships (distinguishing the behaviour of
38

Systems Thinking and Capacity Development in the International Arena

simple, complicated and complex inter-relationships)

Two systems methods that highlight inter-relationships


System Dynamics is a method that seeks to explore the consequences of nonlinear relationships and delay. It is usually, although not always, used in conjunction with computer simulation. System Dynamics addresses the following questions: How does delay impact the performance of the situation? How do patterns of feedback affect the behaviour of a situation? What controls the way in which resources flow through the situation? How does this affect performance? Results chains and process models often assume cause and effect relationships that are relatively sequential, linear and non-recursive (i.e. A leads to B leads to C). For example, we might understand capacity development as being the consequence of training (A), which leads to increased knowledge (B), which leads to employment (C).
A B C

In contrast, System Dynamics acknowledges that the relationship is circular and the arrows may go in either direction.
A

B ?

So training (A) might increase knowledge (B), and this knowledge may increase the demand for further training (A) which leads to greater knowledge (B). In this relationship capacity will increase rapidly. However, it also allows for the possibility that knowledge (B) may lead to people gaining employment in the field (C), which might reduce their ability to engage in further training (A) because they are now out in the field. On top of this behaviour there may be delays in response that further complicate the picture. So whilst A might affect B quite quickly, B may affect the demand for A
39

CHAPTER 2

more slowly than C affects the demand for A. Thus in these scenarios, whilst the capacity of the situation may be enhanced at least initially (i.e. more training, knowledge, employment), over time the capability of the situation reduces. Viable System Modelling (VSM) is based on the idea that different levels within a situation need different kinds of information to function effectively. Viable System Modelling addresses the following questions: What are the operational, co-ordination, management, strategy and governance needs of the situation in order to deliver on its purpose? What information is needed at each level of the situations to achieve the purpose? How does information flow through the situation? Is the right information available at the most appropriate level of a situations hierarchy of tasks? The model has five levels or systems:
System 1 System 2 System 3 System 4 System 5 The entire collection of interacting operational units within an organisation. The system responsible for stability/resolving conflict between operational units. The systems responsible for optimisation/generating synergy between operational units. Future plans and strategies. Adaptation to a changing environment. The overall systems link to the environment. Policy. Sets system rules and direction.

In diagram form:

5 4 3 2 1 1 1 1 1 1 2 E n v i r o n m e n t

Source: Williams & Hummelbrunner (2010)

40

Systems Thinking and Capacity Development in the International Arena

Jon Walker (see references) describes each of these categories in the following way: First of all you need the working bits. This is System 1, the operational units. System 1 is the bit which actually does something. Its the muscles, the engine room, the machines, the producers. Secondly you must ensure that there are ways of dealing with conflicting interests which are inevitable in the interactions which occur as the parts of System 1 interact. Conflict resolution is the job of System 2. System 2 is also given the job of ensuring stability. Once the interactions of the System 1 units are rendered stable, it becomes essential to look at ways of optimising these interactions. This is the job of System 3. System 3 works with an overview of the entire complex of interacting System 1 units and thinks If this one does this and that one does that, then the whole thing will work more effectively. The extra efficiency is called synergy. System 3 is there to regulate System 1 its function is optimisation. Once you have a stable, optimised set of operational units, then you must ensure that it can survive in a changing environment. This is the job of System 4. System 4 looks at the outside world [environment], considers what it sees, and looks for threats, opportunities and schemes. System 4 is there to produce plans to ensure long term viability. And finally, the whole thing must function within some sort of overall context. Everyone must be pulling in the same direction. This is System 5s job. It provides the ground rules and the means of enforcing them to ensure that the system in complete. System 5 provides the ultimate authority. VSM is commonly used to understand why monitoring and information management procedures are ineffective. An organisations capacity is the sum total of all information possessed by an organisation. Each of the five systems needs information to function. However not all of the information required by one system is generated inside that system it makes demands on other systems for information. So each system has to handle its own information needs as well as the information needs of other systems. While people in each system will understand the information needs of their own system they may not understand the information needs of the other systems. Furthermore they will often lack the motivation to ensure that information required by other systems is relevant and timely. For instance, take a refugee agency. Workers in a refugee camp providing emergency food supplies (System 1), will be committed to keeping records of the amount of rice in store, but may not see the relevance of information requests for the ethnicity of food recipients. Yet at a strategic level
41

CHAPTER 2

this information may be useful to the refugee agency as a means of assessing the political conditions that will determine their approach to refugee camp placement planning (i.e. System 4) Bad information sloshes around the organisation and thus threatens its viability. In other words, capacity does not match capability. VSM is thus used as a tool to help organisations balance capability and capacity and to ensure that the right information is generated in the right place, supplied to the right people at the right time to allow them to take the right decisions. For instance, once upon a time there was a small development agency dealing with local health issues. The staff, clients and funding agencies were in close contact with each other. Generally speaking everyone was part of all five systems, although it was probable that System 4 was largely ignored. Suddenly, along comes HIV/AIDS and large sums of money from an international foundation. The organisation expands from 5 people to 50 people, the demands from the donors become more complex, the response from the environment: less predictable. The old way of management by discussion and consensus breaks down, everyone no longer can know everything that is going on, generalist tasks become specialist. At this point organisations often collapse, unable to respond to the new dynamics and needs. VSM provides a means of addressing these issues, by exploring, for instance, who needs to be trained to do what; how do you meet the competing information demands without overwhelming the task?

Perspectives
The systems field draws on a number of approaches for exposing and exploring perspectives: What are the different ways in which this situation can be understood? How are these different understandings going to affect the way in which people judge the success of an endeavour? How will it affect their behaviour, and thus the behaviour of the system, especially when things go wrong from their perspective? With what result and significance? A systemic approach is more than a study of how boxes and arrows fit together or how information networks operate. Just looking at interconnections does not make an inquiry systemic. What makes it systemic is how
42

Systems Thinking and Capacity Development in the International Arena

you look at the picture, big or small, and explore interconnections. When people observe inter-relationships they will see and interpret those interrelationships in different ways. People participate in a project for many different reasons. Think of your own involvement in the international capacity development field. How many different ways of seeing your involvement are there, and how do they affect the kinds of decisions you make? What you may regard as a situation able to deliver adequate locally resourced economic initiatives (i.e. large capacity, medium capability for economic development), someone else may regard as completely ignoring womens rights (i.e. large capacity, low capability for social development). These interpretations, these motivations and the behaviours that flow from these different perspectives may have little or nothing to do with the formal goals or objectives of a project or program. The program may have indeed been primarily about economic development. Yet the expectation by some key players for social development will affect how the program performs and ultimately affect the results. Thus we cannot comprehend the behaviour of a program, in this case capacity development, without identifying and understanding a wider range of perspectives. Perspectives help to explain and predict unanticipated behaviours because they give us a window into motivations. They also draw our attention to consequences: unplanned and unintended. Towering above this is the need to acknowledge that people make programs work, not some imagined logic like a logframe dreamed up by a funding agency. The introduction of perspectives as a core systems concept was profound. First, it highlighted the notion that a situation can be seen in different ways, and this will affect how you understand the system. Whilst that idea isnt inherently radical, most of the attention outside the systems field has been how different stakeholders hold perspectives that are different from each other. However, the systems field holds that different stakeholders may share the same perspective and most importantly any one stakeholder will hold several different perspectives, not all of which will be compatible with each other. For instance, Ive never held a single unified view on any project Ive been involved in. How I handle a situation shall I give my $5 donation to this person on the street will be the result of a complex set of internal arguments and trade-offs that can change in the time it takes for me to reach into my pocket. In other words, understanding perceptions is important at a behavioural level. Perceptions are more than how people look at a situation; these perceptions generate behaviours that impact on the way the situation
43

CHAPTER 2

operates. Yet the management theories that dominate the international development world tend to force us to pick one and pretend that its the one that should motivate everyone. And then we wonder why things dont quite work out as we planned. In terms of capacity and capability, the key question that flows from this discussion on perspectives is not whether there is capacity within a situation, but what that capacity has the capability of doing. Second, perspectives importantly drew the focus away from the system as it supposedly exists in real life and allowed us to consider alternatives: what it might be like, could be like, or even should be like. Or how different people imagine how it might be like. This opened up the systems world, because not only could you draw conclusions based on a study of the world as it is, but you could also compare alternative perceptions of what they think it is with what actually is, or with perceptions of what is or with what might be. The similarities and differences between what is and what might be create puzzles and contradictions which can stimulate deeper learning.

Two systems methods that highlight perspectives


Soft Systems is a methodology that first forces you to consider alternative perspectives (e.g. development as aid, development as patronage, development as a tool of foreign policy, development as empowerment). It then asks a series of questions that help you work out what the structure, function and logical consequences are of each perspective. Soft Systems poses the following questions of a situation: What are the different ways in which the situation can be viewed? How does each of these ways express and give meaning to what some people within the situation value? How are these perspectives reflected in peoples motivations and behaviours within the situation? How does each of these ways affect the performance of the situation? For instance, take an HIV/AIDS program in India. If you perceived development primarily as aid then you may well focus primarily on ensuring the availability of skilled medical care sourced from wherever that skill may be. Your understanding of capacity will be how well that skill is coordinated with the local situation. However, if you perceived development from an empow44

Systems Thinking and Capacity Development in the International Arena

erment perspective, you would be more concerned about progressively building up the skill from the local base. Your use of externally sourced expertise would be on a different basis; your notion of capacity would be quite different. Of course reality is always a mixture of perspectives. Soft Systems does not force you to pick one perspective over the other, or even prioritise them. Instead what it encourages you to do is separate out these perspectives, work out the implications of each and then assess how best to integrate them in a way that they work for rather than against each other. Activity Systems is an approach closely associated with the idea of communities of practice. A community of practice is a set of mutually agreed upon and shared activities directed towards a common purpose. Activity Systems add two other factors: the tools we use and the motivations that drive us towards a goal. It seeks to address problems associated with different motivations towards goals and how to constructively handle shocks to the system. Activity Systems addresses the following questions: What fundamentally are the motivations underpinning the achievement of a goal? What tools, rules and roles are necessary for that motivation to be translated into goal directed activities? How does the system handle contradictions in tools, rules, roles and motivations so that the goal is achieved? For instance, both of us could be providing HIV/AIDS counselling to fulfil a goal about reducing the incidence of HIV/AIDS. However, your motivation for providing counselling may be womens empowerment, whereas my motivation may be to complete my counselling qualification accreditation. Most of the time these two motivations will not contradict each other. However, something might occur perhaps a tricky ethical issue that bring us into conflict. How does one handle such a situation constructively or creatively rather than destructively? Activity Systems provides a framework for assisting that process. Activity Systems approaches enable people to engage constructively in resolving the tensions that arise when circumstances expose situations where people are engaged in the same activities but for different reasons. So whilst a conventional capacity development perspective might focus on the quality of the actual counselling, an Activity Systems approach would be in developing the capability to handle conflicts construc45

CHAPTER 2

tively.

Boundaries
A systemic approach to boundaries will help you navigate through complexity by posing the following general questions: How is a situation being framed? What does this imply about the way in which the situation can be managed or investigated? Boundary setting is not optional. We make situations manageable by setting boundaries. Every endeavour has to make a choice between what it includes and what it excludes. A boundary differentiates between what is in and what is out, what is deemed relevant and irrelevant, what is important and what is unimportant, what is worthwhile and what is not, who benefits and who is disadvantaged. Boundary setting also determines: how we approach a situation; what we expect from it and what methods we might use to manage it; what is deemed relevant and what is not; and which perspectives are honoured and which perspectives are marginalised. Boundaries drive how we frame situations. We manage something framed as a simple situation quite differently from a complex one; our definitions of what constitutes a simple situation and what constitutes a complex situation are core boundary decisions. In evaluation terms, boundaries are the sites where values get played out and disagreements are highlighted. A lot of power issues are wrapped up in boundaries just as the person with the magic marker controls what goes on the whiteboard, the person whose perspective dominates a project decides the boundaries. Capacity development in the international arena is full of boundary decisions: who gets what kind of resources for what purpose and whose interests are marginalised. Indeed, as we shall see in a couple of paragraphs even the notions of capacity and capability are themselves boundary choices. As suggested earlier, during the mid-1980s, questions were being asked within the systems field about how boundaries are set, who sets them and what the consequences of setting boundary are. Whilst its fine to map relationships and it may be fine to acknowledge that there will be different perspectives on those relationships, those relationships and perspectives are not neutral someone somewhere decides which relationships and perspectives
46

Systems Thinking and Capacity Development in the International Arena

are the most important. Once it was fully acknowledged that thinking systemically about perspectives and inter-relationships involved boundary choices, many in the systems field started taking a deliberate and often debated approach to boundary identification and selection.

Two systems methods that highlight boundaries


Critical Systems Heuristics (CSH) is a set of questions that helps draw out debates on boundaries. Critical Systems Heuristics addresses the following questions: Who or what is being excluded, marginalised or made a victim by the way in which this situation is being viewed or is operating? What does this say about what is valued, by whom, in this situation? What are the consequences of boundary setting decisions? How can negative effects of such decisions be mitigated? Critical Systems Heuristics involves exploring four aspects of boundaries: Purpose/Values: Whose interests are being served and whose interests should be served? Resources/Control: Who controls what resources and who should control what resources? Knowledge/Expertise: What expertise is required? Who do we trust as experts and what expertise should be required; whats the risk of assuming this is all the expertise needed? Legitimacy: Whose interests are being excluded, marginalized or harmed by the way we are framing the situation and whose interests should be excluded, marginalized or harmed? Lets consider an HIV/AIDS prevention project that used peer education methods to increase the degree of control sex workers have over their clients by improving their knowledge about HIV/AIDS and its transmission. In parallel, the project used microfinance methods to allow sex workers to develop alternative income streams, and to become less economically dependent on clients (and thus able to be more assertive with their knowledge). The evaluation of this project documented some deep systemic problems that could be more clearly understood by considering boundary issues. In
47

CHAPTER 2

terms of purpose/values, the project seemed in many ways to be driven more by the interests of the religious organisation that funded the project than by the interests of the sex workers. In terms of resources/control the project was dominated by an external agency that controlled most of the resources but did not have the internal capacity to process and monitor the loans and repayments. Thus loans were slow to be made and largely unrepaid violating two of the basic tenets of microfinance. In terms of knowledge/expertise the project critically misunderstood the expertise needed to negotiate with clients. The sex workers actually knew much more than the educational model assumed, in fact the key issue was a resource issue (lack of condoms) not an educational issue. Finally, in terms of legitimacy, key interests were excluded (e.g. hotel owners and traditional money lenders), who actively worked against the project. Although capacity development touches on all four of these categories (purpose, resources, expertise, legitimacy), issues of capability focus particularly on expertise. Capacity and capability are intricately bound up with notions of expertise and Critical Systems Heuristics poses some very challenging notions about what assumptions are being made about expertise; what expertise is regarded as relevant (or irrelevant) and who should have that expertise. Cynefin (a Welsh word roughly meaning place of multiple belongings) is a systems-based framework that distinguishes (i.e. draws a boundary) between four aspects of a situation. These four aspects are: simple situations that deal with the known (i.e. the aspect of the situation is self- evident and uncontested) complicated aspects that are about the knowable (the aspect of the situation is knowable with sufficient investigation and expertise) complex (the aspect of the situation is unknowable in advance but knowable in the present) chaotic aspects of situations (the aspect of the situation is completely unknowable) Each situation implies a different set of knowledge and management capacities, and suggests different ways to make those capacities available.

48

Systems Thinking and Capacity Development in the International Arena

COMPLEX Probe Sense Respond

COMPLICATED Sense Analyse Respond

CHAOTIC Act Sense Respond

SIMPLE Sense Categorise Respond

Source: Williams & Hummelbrunner (2010)

Simple aspects of a situation imply very little capacity development things can just be borrowed from elsewhere and applied with very little modification (i.e. sense, categorise, respond). Complicated aspects of a situation need capacities that involve research (to find out where an intervention has been used before) and adaptation (i.e. ability to understand how context will affect application). In Cynefin shorthand sense, analyse, respond. Complex aspects require capacities that are close to action research an ability to reflect critically on practice, identify patterns and modify processes (i.e. probe, sense, respond). Chaotic aspects need immediate action (e.g. triage) to stop the patterns and then consider what to do (i.e. act, sense, respond) Thus the capacity required to manage a simple aspect of a situation is likely to be inappropriate for managing a complex aspect of a situation. In terms of capacity development, Cynefin challenges the idea that there is a list of capacities and capabilities that can enable all aspects of a situation. Cynefin thus addresses the following questions: How are we framing the situation? What are the implications of this framing for how we manage a situation? What are appropriate ways of managing a situation on the basis of this framing? Referring again to HIV/AIDS projects, a simple aspect of a project may be the HIV testing process itself, whereas a more complicated aspect may be the treatment of those who have moved to an AIDS diagnosis. A complex aspect may be the medium to long term management of those with AIDS within their own communities, or the process of encouraging people to come
49

CHAPTER 2

forward for tests. Activities in the complex zone are likely to be highly sensitive to local conditions and relatively unpredictable in outcome.

Thinking systemically about capacity development


The reassessment of international development over the past decade, the struggles over the Millennium Development Goals (MDGs), the acknowledged failure of results-based management methods (United Nations, 2008) 2 and the shift toward capacity development at a national rather than at project levels poses many challenges to existing capacity development concepts. In particular it highlights the need to distinguish carefully between capacity and capability. This distinction raises many inter-relationship, perspective and boundary questions. These include at what scale (national, trans-national, local) should capacity development interventions be conceived and assessed, who should be the primary beneficiaries, who or what could be harmed by that choice, can tension be resolved, what expertise is considered necessary to provide adequate capability and who should control what resources. These are questions that the systems field is well equipped to address.

Systems thinking and systems methods


Choosing systems methods ... There will be times when the high level of abstraction in the three basic dimensions of systems thinking is inadequate. You need to dig deeper. Thats when selecting from the wide range of systems methods becomes important. In which case, how do you decide between the methods? In many ways it is personal judgement or deciding which approach feels right. In many ways it is a pragmatic call, which methods do I know best, do I feel I can learn the quickest or can I find someone with expertise. However, there are more systematic ways. One way is to consider the kinds of questions that a systems method addresses. For each of the six examples in this document there are a set of questions most closely associated with that method. Which of these questions most closely relate to the situation that concerns you? Another way relates to the nature of the situation. For instance, this
2 United Nations (2008) Review of results-based management at the United Nations. Report of the Office of Internal Oversight Services. 22 September 2008. [http://www.un.org/ga/search/view_doc.asp?symbol=A/63/268].

50

Systems Thinking and Capacity Development in the International Arena

(much debated) framework by Mike Jackson, Robert Flood and Paul Keys (for a summary see Midgley, 2000) focuses on the structure, dynamics and purpose of the situation. The dynamics of the situation and its context Is the situation unitary?: There is agreement between key stakeholders. In which case methods that are essentially technical and oriented towards inter-relationships are appropriate. These would include System Dynamics and Viable System Modelling. Is the situation pluralist?: There are disagreements between stakeholders. In which case methods that are focused on perspectival issues are appropriate. These would include Soft Systems, Activity Systems and Cynefin. Is the situation coercive?: There are power dynamics that seek to control the way in which the situation is perceived and handled. In which case methods that with care can be used to expose that dynamic. Critical Systems Heuristics is claimed by some to be an appropriate approach. Activity Systems might also be a contender as well as Cynefin. The purpose of the intervention If you wish to predict or seek to bring a situation under control: use approaches such as System Dynamics. If you wish to involve people in reflective debate so that learning may be facilitated: use approaches such as Soft Systems or Cynefin. If you wish to expose underlying assumptions to ethical critique or other forms of non-technical critique: use methods such as Activity Systems or Critical Systems Heuristics. Midgley (2000) and colleagues developed a mixed methodology approach. Rather than picking one method or methodology to address a situation, this approach uses a range of different systems approaches. For instance, at the start of an intervention you may wish to understand the dynamics of a situation, so use System Dynamics, or you may want to know how information is flowing around a system, so use Viable System Modelling, or how the situation has handled sudden shock and contradictions (i.e. Activity Systems). Once you have got this you may want to set some boundaries around the intervention you are intending for this situation and thus use Critical Systems Heuristics as a means of working out how best to frame that intervention. This may result in a set of different framings that Soft Systems may help form the platform for discussion around possible
51

CHAPTER 2

resolutions problems associated with the situation. Finally in terms of intervention you may want to determine whether to treat some aspects of the situation as complex or complicated or simple. At this point you are beginning to close the gap between thinking about systems methods and thinking systemically. ... or thinking systemically Thinking systemically is a matter of capacity development itself. There is knowledge to be acquired, skills to be gained, learnings to be acquired and opportunities to be sought to apply all of these. However, it is also critically about capability. Over the years running workshops on systems thinking, a common (and realistic) comment has been that the ideas are great, but the organisation people work with would never tolerate the kind of questions systems thinking poses. Under such conditions the question arises about where to start. Generally speaking the best place to start with is where you are right now (Midgley in Williams and Imam, 2007). Do the notions of focusing on interrelationships, perspectives and boundaries help you improve your own understanding of capacity development? Will addressing the questions under each of the main headings of this paper help you puzzle your way through a problem, or bring constructive light onto an issue? If you have answered yes to any of these questions, then you have probably found the best place to start.

This article is based on a shorter article in Capacity.Org (http://www.capacity.org/en/journal/feature/thinking_systemically) as well as previous writings by the author and contributions from Gerald Midgley, Richard Hummelbrunner, Amy La Goy, Iraj Imam, Martin Reynolds and Glenda Eoyang. In various forms it has formed the basis of workshops, lectures and articles, including the American Evaluation Association, the University of Wageningen (http://portals.wi.wur.nl/navigatingcomplexity/?Report) and The Broker (http://www.thebrokeronline.eu/en/articles/Bucking-the-system).

Useful References
Checkland, P. and Poulter, J. (2006) Learning For Action: A Short Definitive Account of Soft Systems Methodology, and its use for Practitioners, Teachers and Students. Wiley.
52

Systems Thinking and Capacity Development in the International Arena

Flood, R. and Jackson, M. (1991) Creative Problem Solving: Total Systems Intervention. Wiley. Irwin, J.S. (2000) Business Dynamics: Systems Thinking and Modelling in a Complex World. McGraw-Hill. Kurtz, C. F. and Snowden, D. J. (2003) The new dynamics of strategy: Sensemaking in a complex and complicated world. Cognitive Edge: http://www. cognitive-edge.com/ Midgley, G. (2000) Systemic Intervention: Philosophy, Methodology, and Practice. New York: Kluwer Academic/Plenum Publishers. Ramage, M. and Shipp, K (2009) Systems Thinkers, Springer. Reynolds, M. (2008) Reframing expert support for development management. Journal of International Development. 20. Williams, B. and Hummelbrunner, R. (2010) Systems Concepts in Action : A Practitioners Toolkit. Stanford University Press. Williams, B. and Imam, I. (eds) (2007) Systems Concepts in Evaluation. An Expert Anthology. EdgePress.

Some On-line Resources


Flood, R. Rethinking the Fifth Discipline: Learning Within the Unknowable. http://www.ebookmall.com/ebook/132488-ebook.htm Systems Dynamics : http://wwwu.uni-klu.ac.at/gossimit/linklist.php VSM : Walker, J. An introduction to the Viable System Model as a diagnostic & design tool for co-operatives & federations, see http://www.esrad.org.uk/resources/vsmg_3/screen.php?page=1qguide (checked 28/6/09) Soft Systems Methodology http://www.bobwilliams.co.nz. Go to the section titled systems stuff in the sidebar menu. Centre for Activity Theory and Developmental Work Research; University of Helsinki http://www.edu.helsinki.fi/activity/pages/chatanddwr/activitysystem Critical Systems Heuristics: Werner Ulrichs homepage: http://www.geocities.com/csh_home/ Cynefin http://www.cognitive-edge.com

53

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

3
Using Systems Concepts in Evaluation
A Dialogue with Patricia Rogers and Bob Williams
Edited by FASID (1)
As Mr. Williams explained in the previous article, more than 1,000 systems tools exist which can make it even more difficult for newcomers in this field to understand how systems concepts can actually be applied to development evaluation. Therefore, in this section we have asked veterans Dr. Patricia Rogers and Mr. Bob Williams to explain how we can use systems approaches in evaluation.

1. Thinking about Systems and Evaluation


FASID: Thank you both very much for taking your time to participate in this dialogue. I read Dr. Rogers fascinating article (2) about using program theory to evaluate complicated and complex interventions which also suggested possibilities for using systems thinking to tackle some difficulties in development evaluation. I have also learned from Dr. Rogers that Mr. Williams is now writing a book on this topic with Dr. Richard Hummelbrunner. I am sure it will be a great resource for evaluators and we are very pleased that we are getting a preview of it today. But, before getting into the details, would you mind telling us how you encountered this field of systems concepts and evaluation? Bob: I was originally trained as a biologist, and worked as a research ecologist for a number of years. I became mixed up in the environmental movement in the early 1970s, and ended up as a community worker. During that time I became associated with the Systems Group at the Open University in the UK. So I went from an environmental focus to a social focus via a systems focus. After a number of years working as a community worker, I began asking myself the question, Is this doing any good? What I did not know at the
55

CHAPTER 3

time was that I had wandered into an area that I did not even know existed evaluation. So here I was in the mid 1990s posing evaluation questions on the basis of my physical and social science backgrounds with a systems orientation. Patricia: My background was in political science, with a Ph.D. in evaluation. I started teaching in RMIT (3) Universitys Evaluation Program in 1989. It was an extremely unusual evaluation program set up by Jerome Winston whose disciplinary background was in physics. He had studied general systems theory at MIT and had found that the theory provided a useful framework for evaluating human service programs. He developed a graduate program in evaluation that was, I believe, unique in having several compulsory subjects in systems approaches and systems thinking. Although the systems field was an area I had not formally studied, I discovered that a key feature of systems approaches was developing and using conceptual models appropriately, remembering that the map is not the territory. This resonated with my training in political science.
Bob Patricia Systems Political Science Approaches Evalution Models to Quantitative & Qualitative Evaluation Research Models

Community Work Environment Ecology

FASID: So, you are working together in this combined field of systems and evaluation. Why? Bob: We are not doing this just because the systems approach is interesting, but because we believe that systems concepts provide something that can make evaluation better. FASID: When we talk about systems, what exactly are we talking about? Bob: Im afraid you will not find an agreed definitive definition of systems in the literature. Its a bit like evaluation. In the evaluation field, you will never find a single agreed definition of what is and what is not evaluation. People argue all the time over what it is; its the same in the systems field. Thats not to say that people havent tried to define what a system is, but in more recent times the emphasis is less on what a system is and more on what taking a systemic viewpoint implies. Derek Cabrera distinguishes between thinking about systems and thinking systemically. I think that is a useful distinction.
56

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

FASID: Would you explain the difference between thinking about systems and thinking systemically? Bob: OK, heres an example. Peter Checkland defines a system in the following way. It is a situation that displays the following properties: An ongoing purpose (that may be determined in advance purposeful, or assigned through observation purposively) A means of assessing performance A decision-taking process Components that are also systems (i.e., the notion of sub-systems) Components that interact An environment (with which the system may or may not interact) A boundary between the system and the environment (that may be closed or open) Resources Continuity So thinking about systems, would be exploring only situations that display all of these properties. The problem is that relatively few situations display all these properties. And even when they do, you find that people often draw different conclusions about them. In contrast, thinking systemically doesnt require agreement about whether or not a situation is a system, but explores a situation using principles that underpin systems ideas, revealing insights that would otherwise be hidden from view. FASID: Then, what are those principles? Bob: Many people have attempted to identify these. What we conclude from our different experiences within the systems field and within evaluation is that the principles of systems thinking can be expressed by three concepts. One of them addresses inter-relationships. The second is the concept that different people see different things when you explore inter-relationships this concept is known as perspectives. Finally, the third concept, which I personally think has enormous implications in terms of evaluation, is that it is fine to have multiple perspectives, but isnt it interesting that somebodys perspective always seems to win out that there is always a dominant perspective? This leads to a focus on boundaries the third distinction that can be made in the systems field. FASID: So, you are saying that by applying these three concepts inter-relationships, perspectives and boundaries we can improve evaluation from its planning to its final analysis? Bob: Using these three concepts can improve evaluations and for any sin57

CHAPTER 3

gle evaluation they can also help to make its limitations clear. These systems concepts highlight certain aspects of a situation which an evaluation could then examine. Taking a systems approach to evaluation is not about looking at everything, its about being aware of what to look at, what can be ignored and what are the implications of looking at one aspect and ignoring another. Systems thinking provides a powerful framework for making those decisions. Patricia: These systems principles dont only have implications for a single evaluation. They raise questions about how we organize evaluations within and across organizations. There are real limits to what you can do in any single evaluation. For example, requiring lots of small project evaluations can be a tremendous waste of resources especially if they are done too soon or too quickly to adequately report on results. It would be better, in my opinion, to reduce the scale of the evaluations required of all projects or programs, and put more evaluation resources into better understanding a few key projects or programs. People working in development evaluation need to think more strategically about the protocols and requirements. Systems thinking can help with this as well.

2. Three Core Concepts (1) Inter-relationships


FASID: You mentioned that the systems approach has three core concepts; would you please explain them briefly one by one? Patricia: Sure, Ill start with Inter-relationships. Evaluators love inter-relationships. But while we draw lots of inter-relationship diagrams, many of our methods such as program logic and logframe dont actually explore the real depth of those inter-relationships and what affects them. For example, if we were evaluating a program based around the proverb an apple a day keeps the doctor away, many evaluators would develop a logic model that went something like this:
INPUTS Apples

PROCESSES Deliver apples

OUTPUTS Apples eaten

OUTCOME/IMPACT Improved health

If we measured each inter-relationship, and addressed the issue of causal


58

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

attribution, we might be able to conclude whether or not the program had worked, but we would have no idea how it worked, or why it didnt work, and we would not have enough information to be able to transfer this evidence to a new situation or improve a program that wasnt working. If we were taking a systems approach, we would explore the relationships between eating and better health for whom, in what circumstances and how does it work? For example, the program might work very well for young children who dont have access to fresh fruit, by increasing their vitamin and fibre intake, or it might work well for overweight people by encouraging them to substitute an apple for a high-calorie snack, or it might increase the intake of Quercetin which may reduce inflammation, including influenza and heart disease. Each of these possible explanations would lead to impacts of different types of health improvements, and would have implications about other situations to which it could be generalized, and possible ways of adapting it. If it works through vitamin C, then oranges would be effective if apples were not available. Alternatively, if it works through Querecetin, red onions would be a better substitute than oranges. FASID: I see. That means inter-relationships are something we should think about at the stage of framing the evaluation? Patricia: Yes, and we can also use it at the analysis stage. Thinking about inter-relationships should be considered an analytical tool to make sense of existing data, and to work out in advance what data can be collected in the time available for adequate analysis. FASID: Do you differentiate between intended and unintended paths from input to goal? And what about unintended goals? Patricia: I think you should at least consider or investigate these different paths, as this may also help you to understand any unintended outcomes, especially negative ones. Bob: This actually leads onto the next concept about perspectives. If I pose your question from a perspectives viewpoint, I would ask Unintended by whom? So even the language we use like intended and unintended consequences, reveals something about the perspective taken about a particular intervention. An unintended consequence from your perspective may very well be a specifically intended consequence for somebody else. I sometimes say there is no such thing as an unintended consequence someone, somewhere, intended it. I know my rhetorical sleight of hand is not universally true, but it does challenge us to think about perspectives and especially those we use in the design and analysis of evaluations.
59

CHAPTER 3

3. Three Core Concepts (2) Perspectives


FASID: So, the next core concept is perspectives. Bob: Understanding a variety of perspectives is useful for both practical reasons and ethical reasons. Often its not practical to understand what is really going on by seeing things from only one point of view. A soccer match ending in a riot cannot be explained if you see a soccer match solely from the perspective of a game of skill between two teams. An ethical reason for understanding perspectives is the possibility that someone or a certain group could get harmed if you dont see things through an alternative perspective. Seeing a particular development issue solely through the lens of womens empowerment, for instance, may mean the negative consequences on social cohesion are ignored. Therefore, it is important to see things not just through one pair of glasses. FASID: We all understand the importance of considering different perspectives. But, isnt it a challenge to come up with all the possible perspectives within a given time span? Patricia: Yes, that is a very good point. We cant identify or investigate all possible perspectives, but we can identify some important perspectives. Then we can use these different perspectives when deciding what data to collect, or at least when interpreting the data we have. I know it is sometimes hard when we are very constrained not just by time but also in terms of what are seen as legitimate perspectives. But if I am looking at something with different perspectives, I can get a better understanding of why something works or doesnt work and how it can be fixed. Bob: I would also say this is not about getting things right. You cant look at everything and we will get to boundary setting in a minute to further address that. What we are trying to do is to introduce you to a new approach that, given the resources you have, can enable you to make more powerful judgements of worth. Perspectives and worth are very closely linked. You might judge the worth of a soccer match differently if you perceive soccer as a game of skill or a means of entertainment. A soccer game played by incompetents could be judged hugely entertaining, whereas a game played skillfully could be judged very dull. So, it is not about going through an enormous process to get things right. It is about being skilled on knowing the right questions to ask. FASID: How should we find different perspectives?
60

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

Bob: Well, there are several strategies to think about multiple perspectives. One is to think about the motivations of those involved in a project. This implies a focus on stakes as well as stakeholders in an evaluation. Clearly different stakeholders have different perspectives. However, one single stakeholder can have multiple perspectives depending on their < <stakes> and > their motivations. If we were evaluating the workshop weve just finished from a multi-perspective point of view, most evaluators would put Patricia and I in the < <facilitators> stakeholder grouping. However looking at things from > a< <stake> point of view opens up other ways of understanding how the work> shop went. Our stakes in the workshop included our reputation as evaluators and the workshop as a source of income, as an opportunity to visit Tokyo, as an opportunity to work together, and as a means of rolling out the idea of systems ideas into the international development arena. How we managed those different stakes influenced the seminar in a range of ways. And those stakes were driven fundamentally by our motivations. So a good place to start choosing perspectives to explore is to ask < <what are the motivations that stakeholders are bringing to the project and which motivations will play a key role in how the project performs?> > Patricia: Another strategy is to work backwards from a result, whether good or bad. And the third strategy is to think about something from a formal philosophical viewpoint, such as a Marxist view, a feminist view, a Buddhist view, an environmentalist view, and a social capital view. For example, an HIV/AIDS prevention program in Africa could be seen from a human rights perspective, a public health perspective or a poverty reduction perspective. Suppose the improvement of economic conditions (by providing alternative sources of income) is considered to lead to the empowerment of commercial sex workers (who can then negotiate with clients over condom use) and which can lead to the reduction of HIV transfer in the community. If poverty reduction is achieved but the program does not lead to a reduction of HIV transfer, the project might be considered a success from an economic perspective but not from a public health perspective. Bob: This I think also helps us understand that perspectives are a way in which we < <frame> things, and how we frame things will influence our ability > to resolve issues. For instance, an American manufacturing company that builds machinery for large infrastructure projects was in constant industrial relations strife because of the regular fluctuations in their workforce. They often had to make large numbers of people redundant when orders dried up and then cope with issues of hiring new workers when orders were high.
61

CHAPTER 3

This meant they were constantly in conflict with their trade unions over redundancy payments and training issues. Framed as an industrial relationship problem it seemed unsolvable. They tried hard for a number of years and failed. Finally they saw the issue through a different lens as an economic cycle issue. As we know, the world economy goes up and down, and infrastructural projects tend to go up and down with local economic conditions. By framing the problem as an economic cycle issue, rather than a labour relations issue, the company came to realize that their product range was too closely associated with one phase of the economic cycle. Framing the problem as an economic cycle issues allowed them to ask an important question with the current workforce skills and the plant can we manufacture something that is counter-cyclical, so that when demand for infrastructure construction is low, we can switch to an alternative product? They did find an answer to this question; creating a stable workforce on the way and resolving their industrial relations problems. Patricia: Thats a nice example of how reframing a problem can help to solve it. It may be interesting to think about this case study when you are involved in evaluation; those who appear to be negatively affected may not necessarily be so depending on how the project is framed. FASID: It is always useful to look at things from a different perspective. How can we incorporate this in our work? Bob: For example, you could include groups of people who have different viewpoints in your interview list. And when you include them, some question techniques are very helpful. For example, seek an overall judgement of whether or not the project worked, and then explore what they mean by worked. You will discover that they might have different ideas about what the project was trying to do. Ask them directly about their motivation. Why are you involved in this? What are you hoping to achieve? You are enlarging your interview group a little bit, and you are making sure there are open-ended questions in there. That would be great. Patricia: Another point to think about is, what are the consequences of not considering a certain perspective? You cannot take into account every single perspective. It is just not possible. You are going to have to draw a line somewhere. That boundary may be drawn by your client or it may be drawn by you. When you draw that boundary, when you choose to have only one or two perspectives or perhaps just the clients perspective, you need to think through the possible consequences of that decision. If you think the conse62

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

quences of leaving out other perspectives are significant, then you should see if it is possible to include them. If they are not, why bother to write them in a report? But one thing I have learned is that I take personal responsibility if I was aware of possible consequences but did not mention them. It is a professional judgement call you take on the basis of having at least explored or begun to explore some of the possible consequences.

4. Three Core Concepts (3) Boundaries


FASID: This brings us nicely to the third core concept boundaries. Bob: Right. Boundaries determine what is in and what is out, what is considered important or unimportant, or relevant or irrelevant. As I said earlier, it is good to explore different perspectives but it is not possible to look at everything. There is no such thing as holism; every human endeavor has to have a boundary around it. Now many in the systems field say very strongly that simply setting boundaries is not enough; when you set boundaries, you have an obligation, an ethical obligation, to consider the impact of setting that boundary. Patricia: This is particularly important in the development field where there is a constant struggle over setting boundaries and determining and identifying any negative consequences. For example, if you only include women in a micro-credit scheme, men who have been excluded from the project might be unsupportive and seek to destroy it. FASID: Indeed. By the way, in terms of the boundary of an evaluation, it is often set by the client so dont we have to work within that boundary? Bob: Well yes and no. I tend to get clients to think about the boundaries theyve set for the evaluation and the practical consequences (positive and negative) of that decision. I generally find that although clients have often thought about these things, the conversation and decisions could have been better thought through. Once thats been done then the key is doing the best we can within whatever boundary has been set. Of course this process is just as relevant for setting project boundaries as for setting evaluation boundaries. FASID: How do you set the boundaries? If, for example, we are thinking about a certain group of people who are negatively affected by a development program, from whose perspective do we set a boundary? From one perspective, we may say they are negatively affected, but that may not be true from another viewpoint...
63

CHAPTER 3

Bob: Absolutely. So, the argument is that if you are trying to understand how a program works, you have to understand that tension. There will be people in the program who may not even consider, for instance, girls or the environment an important aspect of a rural economic development program. Thinking about boundaries is a way of getting a bit deeper into the understanding of how a particular project works. Going back to your question about how you < <set> boundaries, I think it is > also important to understand how you consider boundaries. What are you looking for and what do you do when you find one? At this point I have to mention Critical Systems and its role in identifying and exploring boundaries. I think Critical Systems deals with issues constantly confronting evaluation but which evaluators have few ways of addressing and therefore Id like it to be much better known in the evaluation field. Churchman who developed Critical Systems firmly believed three things. Firstly, that every world-view is terribly restricted, thus systems thinking starts when you see the world through the eyes of another. Secondly that any judgement of human activity erects a boundary between worthwhile (i.e. in) and marginalised (i.e. out). Finally because, as systems thinkers, we are obliged to see the world through the eyes of others, then with every boundary we set we have to consider the implications of setting that boundary for others. This has enormous implications for evaluation in the developing world, since evaluators often take the judgements of worth of donors as the only basis for judgement. At least formally they do. The reality is different. When I talk to evaluators working in the development field I find they are acutely aware of these boundary judgements but there are few established approaches in evaluation to help them work this through. The systems field, which has a strong bias towards the resolution of opposites, has quite a lot of approaches to address this. Critical System is one of them. A guy called Werner Ulrich took these ideas and developed them into a tool Critical Systems Heuristic which dives very deeply into these issues. Its a tool that many evaluators and their clients could benefit from using perhaps more than any other tool from the systems field. Patricia: The important point that Bob just made is that you have to draw the boundaries. We cant look at everything, but what we must do is to consider the consequences of drawing boundaries and whether we need to mitigate the consequences of the boundaries we have chosen. Bob: In the paper on capacity building (4), I explore several important boundary issues. These include the purpose of a project and who is likely to benefit
64

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

from it, the amount of resources available to a project, the way in which necessary expertise is framed and what you do about the inevitable decision to exclude certain interests (such as girls or the environment from a rural economic development project). These are all extremely important boundary decisions that profoundly affect whether a project ends up being worthwhile or not (and to whom). By surfacing these boundary decisions you are able to submit them to critical analysis. If they are submerged you never do. For instance, how many projects have you been involved in where assumptions about the necessary expertise required within the project proved to be woefully wrong? Or the fact that excluding certain interests created resistance that in turn reduced the impact of the project?

5. Applying Systems Approaches to Development Evaluation


FASID: So, we can start thinking systemically using these three concepts. Patricia: Yes, and for application to development evaluation, different systems tools and methods can be utilized. For instance, Critical Systems Heuristics is oriented towards thinking about boundaries, Soft Systems about perspectives, Systems Dynamics about inter-relationships, and Cynefin which blends all three core concepts. FASID: How do you know which systems tools to use for what kind of evaluation? Patricia: In the paper on capacity development, Bob has laid out a very helpful guide indicating which tool can be used under what circumstances. This guide includes some specific questions that relate to how to determine the relevant systems tool to use. For example, specific questions for the systems method Cynefin would be the following; Does the situation display simple, complicated, complex or chaotic inter-relationships? From whose perspective? With what consequences? What patterns have developed within the situation? What have been the likely generators of those patterns? Bob: So identifying the questions you want to ask may be a way of identifying which tool to use. You may think Here is a list of questions which tool will allow me to address those questions best? The capacity paper lists those questions. It also suggests a couple of other approaches to selecting methods, but in the end it is really about which approach you feel you understand
65

CHAPTER 3

best. Even though we have said that Soft Systems is strongest when dealing with perspectival issues, it doesnt mean that it is only good at that. Remember in order for a method to be considered a systems method it has to address all three components; inter-relationships, perspectives and boundaries. The questions Ive linked with the systems methods came out of my experience in using these methods. Patricia: And the questions may be specific to a particular situation. Bob: Absolutely. So reading about and understanding the tools may give you the set of questions. Or you can work at it the other way around; being confronted with the set of questions will help you select the specific kinds of tools that are the most relevant. For instance, I think questions about legitimacy and expertise highlight two areas where evaluation is relatively weak. Historically these are areas that evaluation does not go, they are just not part of the evaluation tradition. Deeper exploration of some of these boundary issues (e.g. what expertise is considered relevant and what is not) may lead evaluation into new and tough areas. This idea of exploring the consequences of ignoring those who may be disadvantaged by a project is something that Im increasingly hearing in the evaluation field. As evaluators become more accountable for what they do to a wider range of people, accountability becomes absolutely evident. And boundary questions raise all kinds of accountability issues such as an evaluations (or projects) legitimacy. Legitimacy to whom? FASID: What do you mean by legitimacy in association with evaluation? Bob: It is really asking, in whose eyes is this project (or indeed its evaluation) a legitimate thing to do? Who may not be convinced about that and what are the consequences of ignoring those viewpoints? That is the sort of pragmatic practical side of asking those questions. But there is also the ethical side; is this project or this evaluation the right thing to do? Patricia: In evaluation, we often focus on intended outcomes, whether we were able to accomplish the objectives. As you can see from the diagram below, unintended outcomes and negative outcomes are two separate concepts. Sometimes we can have unintended positive outcomes. Unanticipated outcomes are a subset of unintended outcomes, and I think an important part of evaluation is to try to have enough openness to be able to see these outcomes that were not expected, as well as the intended but negative from some perspectives outcomes.

66

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

Unintended (Positive) Unanticipated (Negative)

Intended (Positive)

(Negative)

FASID: So, evaluators are there to set appropriate boundaries under time and resources limitations. Is it a judgement call? Bob: An informed judgement call. As an evaluator I do the best I can within those boundaries. Im informed by an awareness of the consequences of setting those boundaries, and attempt to mitigate any negative consequences of those boundaries. So the issue is, What are the consequences of those boundaries and how can we make the best within those boundaries? Take time for instance. There are also various techniques that we can use to work around a time limit. For example, many of our jobs take place in highly complex and contested areas. In such cases getting adequate representation of those issues in our data set is a big challenge even before we start thinking systemically. For instance, traditional unstructured interview-based representative sample surveys are hugely resource and time intensive. So to ensure that I am able to gain as many relevant perspectives but in a short space of time, I may consider using interviewing methods that allow respondents to get to core issues very quickly. For instance, there is a method called Convergent interviewing, developed by an Australian called Bob Dick (5) that can generate and assess a wide range of perspectives very quickly and with limited resources. Patricia: Now, the traditional way of doing evaluation is that we spend a long time planning, then we collect data, do the analysis and then we report. However, very often, the planning part, the data collection part and the analysis part take longer time than we think, and the most important parts of analysis and reporting get squeezed in at the end. So, another way of getting around this time boundary, introduced in Michael Pattons UtilizationFocused book (6), is Reality Testing, where the process of planning-data collection-analysis-reporting is done in a series of very short cycles. This is another useful method in conducting evaluation time-efficiently. FASID: When you are deeply involved in one particular perspective, then evaluation seems to be extremely difficult. Especially in the case of internal evaluation by donor agencies.
67

CHAPTER 3

Bob: Thats a risk, but again just keep with the mantra inter-relationships, perspectives and boundaries, just keep posing those dozen or so core questions and you shouldnt go too wrong. FASID: You say that in evaluation, evaluators have to draw boundaries and they should look into different perspectives as much as they can, thinking systematically and sometimes using systems tools. Could participatory evaluation play some role in it? Bob: Many in the systems field would argue that systems approaches are inherently participative. How else can you ensure that you are dealing with the appropriate and most important inter-relationships, perspectives and boundaries? Participatory evaluation raises the issue of stakeholders, and Id like to remind you of the comments made earlier about the distinction between stakes and stakeholders. As I said earlier, every person has multiple stakes in a project no matter what stakeholder grouping they fall into. Evaluation generally tends to assume that stakeholders have only one stake, a very substantial over-generalisation and often an erroneous one. A focus on stakes as well as stakeholders is a core part of the Critical Systems Heuristic tool I described earlier. FASID: The Japanese translations for both complex and complicated are the same. Could you elaborate on the difference in meaning as they relate to Cynefin? Bob: I know this will sound familiar by now, but different people will give you different answers. For me the key difference relates to knowability of outcomes. You know you are dealing with a complicated situation if the result of an intervention is difficult to predict but fundamentally knowable in advance. The situation is a complex one if, no matter how much research you do and no matter how much you try to follow good practice, the result isnt knowable in advance. In a complex setting you can only really understand the result of an intervention after the intervention. One of the big mistakes that evaluators make is to try to and understand and evaluate complex settings in which the results of the intervention are not knowable using tools that are more suited for complicated settings, in which the results of the intervention are knowable at least in a general way. Obviously context matters. Whats knowable in one context may not be knowable in another. Patricia: My take is similar. If an intervention can be adequately represented by a logic model, showing a predictable sequence of cause and effect, then the intervention is being understood as either simple or just complicated. If an intervention is truly complex, we cant predict in advance the cause and
68

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

effect relationships, but need ongoing processes of learning and adaptation. Some people use the example of raising a child as a complex activity because even when you think you have done it successfully before, you might have a child that is really different to your first child, and you are going to have to see what works with them and adapt and respond. These are specific uses of the word complex it doesnt just mean difficult. Sometimes people use the word complex as a fancy way of saying that something is difficult: this can be confusing. Bob: The reality is that most projects that we work on have aspects in both zones (complicated and complex) and they are very much intertwined. Think of Patricias child raising example. Some aspects of raising a child are complicated and some are complex and as a parent you handle these aspects in different ways. So one of the things that I often find useful to do is to untangle the two often using a framework known as Cynefin. Then in evaluation terms we can say how well did a project handle the complex parts of the intervention and how well did it handle the complicated aspects. FASID: Using systems tools, how do you strive to make a good evaluation? Patricia: I guess one way I do this is to use some of the ideas from Critical Systems to examine who is making the key decisions about evaluation and what are the consequences of those decisions. You cant just end up representing your evaluation client, and excluding other perspectives that have not been looked at. Another way is to consider simple, complicated and complex aspects of evaluation practice. There are some simple rules in evaluation that are true everywhere like keeping commitments to your client, keeping them informed, being truthful. There are some complicated aspects, where certain methods are appropriate in certain contexts, and where expertise is needed to be able to match the methods to the situation. And there are sometimes complex aspects, which cannot be totally planned in advance, where there needs to be a process of doing something (in the Cynefin framework that Bob mentioned this process is called a probe) and seeing the response. There needs to be a process for documenting and reflecting on the evaluation itself and making adjustments if needed. This requires considerable trust and partnership with the evaluation client, and is inconsistent with a contractual process where you have to set out the details of your evaluation plan in advance and cannot change them later.

69

CHAPTER 3

6. Systems Thinking and the DAC Evaluation Criteria


FASID: As you know, Japanese ODA adopted the specific evaluation criteria set by the OECD Development Assistance Committee (DAC) relevance, effectiveness, efficiency, impact and sustainability. And under each of these five criteria, more detailed evaluation criteria are set. For example, policy relevance is the foremost important issue in regards to the relevance criteria. Bob: So, lets unpick the boundary issue about policy relevance. A purely policy orientation can exclude other factors that might help us understand how a project works. I recall an evaluation where the focus of the evaluation was supposed to be about the impact of a new funding policy on the management of community organisations. What we found was that whilst the new policy improved the management of community organizations, the greatest impact was on the relationship between the funding body and the organizations. The relationship got a lot worse. We couldnt get the policy people to be interested in this, even though down the track it was likely to be far more important to the sustainability of the projects than dodgy management. The policy people were only interested in the intended change in management practice; they had set a policy boundary and had no interest in either looking over the boundary or even considering the implications of setting that boundary. Patricia: We have to think also about non-participants in terms of relevance. One of the issues is obviously that there are participants and intended participants, and one of the reasons intended beneficiaries do not participate in projects is that they do not see the project as being relevant to them. Bob: Certainly. And thats one of the areas that I have become interested in. If Im working in a project then clearly that project is terribly important to me. Its what I care about, its what I get paid to do, and it may even define my role in the community. Because my focus is on that project, I care about people benefiting from that project, and I look for them benefiting from the project. Implicitly Im assuming the project is important because people are benefiting from it, so therefore it must be relevant to them. But we must ask the question whether I have things out of proportion. Whilst I may think the benefits are important and indeed people may be benefiting, I still need to consider just how important they are in relation to all the other things that are going on in beneficiaries lives. We have a very project-centric way of
70

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

looking at the world. Going back to Churchmans comment just stepping out and looking back at the situation from the participants point of view is critical. Just how worthwhile is the project to them in the grand scheme of things? Could we have done something more worthwhile with our money and their time? Patricia: Yes, so the participants might see the project as being relevant but not in the way that you expected it to be; what they are getting out of the project may not have been what you had anticipated. Asking some of the systemic perspectives questions may help you a lot for this criterion. FASID: Another DAC criterion is sustainability. How can systems ideas help here? Bob: Sustainability is a concept that crops up frequently in the systems literature, although it is often framed as adaptation or emergence. Im very interested in sustainability, how would you assess that within your evaluation framework? FASID: We look at how the effects of the project continue to be sustained after the end of the project, usually looking at different aspects of sustainability such as policy aspects, organizational aspects, financial aspects, technical aspects, socio-cultural aspects and environmental aspects. Bob: I see. So here are some things to think about from a systemic perspective. One of the things that concerns a number of people in the systems field is that sustainability is often framed as meaning staying the same. The underpinning assumptions are that the context is both known or is knowable and is relatively stable. In other words once youve got the project right you keep going and know enough about the environment to keep things on track. However, especially in developing countries, not all environments are either knowable or stable. In which case sustainability in terms of keeping things the same is neither feasible nor arguably desirable. How many times in your evaluations have you felt people were addressing last years context? The systems world tends to think of things more in terms of adaptability rather than sustainability this may be what your evaluation criteria of sustainability actually means. There is an entire field of systems called complex adaptive systems, which asks how do we understand how situations evolve over time? and how do they duck and weave through the confusing and complicated environments that they work in? Patricia: I think one of the big issues for us in the work we do, is for us to be very clear about what it is that we want to sustain. Sometimes it is the project but that is not always the case. Sometimes a strategy to sustain the effects of
71

CHAPTER 3

the project would be to institutionalize its activities, but that will of course need ongoing resources. Other times you may want it to be actually incorporated within a business as an upfront investment, or as an idea that people can pick up and share easily, or some technical expertise. Bob: Another sustainability aspect is the sustainability of the capacity to do things. For example, although the original project to control malaria may have disappeared you may now have the capacity to undertake other health promotion activities related to HIV/AIDS that you didnt have before the malaria project. In this case its not the sustainability of the project, but the sustainability of the organization or the community capacity and capability to actually do something. FASID: How can we incorporate the systems approach when using the logical framework (LogFrame) or project design matrix? Lets take the impact criteria for example. The impact is measured mainly by the overall goal, so although we do look at other unexpected positive or negative impacts, the first place we need to look is the overall goal within the logical framework, so we tend not to place so much importance on other impacts. Patricia: Yes, but sometimes these unintended impacts can actually be more important than the goal, whether positive or negative. Bob: Thats right. Take my example of the policy intervention. Exploring inter-relationships, perspectives and boundaries can help in assessing whether these unintended impacts are likely to occur. And, no, you dont have to explore every possible inter-relationship, perspective and boundary. What you are looking for are the likely impacts that are sufficiently big to be of concern. Patricia: So, you are trying to include unintended outcomes, but how do you do that in your design, within the framework of your evaluation? Theres a method that Carol Weiss used called Negative Program Theory (7), where instead of saying Here is what we want to achieve and here is what we are doing how do we think it will work? she asks Here is what we are doing what is something bad that might happen and how might that happen? For example, if were implementing a policy to increase teachers salaries so that children will learn better, and if we apply negative program theory, we might ask What is a way that we could increase teachers salaries but end up with a result that children learn worse? How could that happen? So you set up a causal chain that is plausible and identify things that you might want to check on. You can redesign your intervention to reduce the risk or you might just want to track it and get early warnings. Negative program theory may
72

Using Systems Concepts in Evaluation A Dialogue with Patricia Rogers and Bob Williams

help you to get evidence about unintended outcomes, if you can anticipate them. However, it will not help you with the totally unanticipated outcomes. We therefore need to build into our evaluations some ways of collecting things we dont expect for example, by talking to different people with different perspectives, and just generally keeping our eyes open. As I said earlier, when you have a very limited viewpoint it is much harder, but just having some open-ended questions in the interview schedule may help you to catch unanticipated outcomes. Bob: It is certainly one of the traps within the project design matrix or logframe. In a logframe for instance, the means of verification and the indicators are very much framed around the desired impact. You will therefore be shining your data collection torch in a relatively narrow zone. Once youve designed a logframe, then it starts to have a life of its own, and changing a logframe as they go, I know, is very difficult. Im undecided whether logframe is inherently unsystemic, but I think the way its used often is. Patricia: Systems thinking may be effective here, when defining the outcome as well as when you are designing a logframe. Some questions to ask may be: What does success require? What standard of performance would be considered a success? What would be credible evidence? FASID: So summing up what would you say is a key < <take-away> from this > conversation? Patricia: If you genuinely ask questions, it will transform evaluation tremendously. Explicitly or implicitly, you should think about complicated and complex issues at key decision points in planning and conducting an evaluation. Bob: Explore inter-relationships, perspectives and boundaries. And do it with local people. That will increase legitimacy. FASID: Thank you Dr. Rogers and Mr. Williams.

(1) This dialogue was composed jointly by Patricia Rogers, Bob Williams, Nobuko Fujita, and Eriko Takahashi, drawing from material developed over the course of numerous conversations and e-mail exchanges; material was also taken from the Q&A sessions held in December 2009 in connection with a FASID seminar: Using Systems Concepts in Evaluation. (2) Rogers, P.J., 2008, Using Programme Theory to Evaluate Complicated and Complex Aspects of Interventions, Evaluation, 2008;14;29, Sage Publications. (3) Royal Melbourne Institute of Technology
73

CHAPTER 3

(4) Ch.2 of this issue (5) Dick, R., Convergent Interviewing ( http://www.scu.edu.au/schools/gcm/ar/arp/iview.html#a_iv_intro) (6) Patton, M. Q. 1997, Utilization-Focused Evaluation 3rd Edition. Sage Publications, CA. (7) Weiss, C. H., 1997, Evaluation 2nd Edition. Prentice-Hall, NJ.

74

Report

Two Cases of Collaborative Evaluation 1


Nobuko Fujita

Enriching Evaluation with Perspectives of Local People:

Both the Paris Declaration (2005) and its follow up in the Accra Agenda for Action (2008) address issues in the field of international development assistance evaluation and stress the importance of working through partner countries systems. Currently, however, so much evaluation work is still donor-driven and designed to meet the donors evaluation needs that the current state of evaluation falls far short of the ideals as stated in international commitments (OBrien 2009). The situation is unfortunate since, once the project is completed, responsibility lies with the partner countries to make positive impacts sustainable and to minimize any negative impacts. Providing partner countries with relevant information, therefore, should be one of the primary goals of conducting evaluations. Donor-centered evaluation tends to de-motivate those involved from the partner side (Lundgren & Kennedy 2009). Minimal partner involvement, in turn, lowers their interest in utilizing evaluation results and consequently, such evaluations become of little use to local decision-makers, implementers and beneficiaries. In addition, for partner countries accountability is an issue. The donors purpose in conducting evaluation is two-fold: to learn from the experience and apply any lessons learned to current and future projects and to show accountability and fulfill their responsibility to explain to taxpayers how effectively their tax money was spent. Likewise, partner countries should utilize evaluation to fulfill their responsibility to be accountable to their own citizens.

The author gratefully acknowledges the insightful comments made by Takako Haraguchi and Keishi Miyazaki who have been working for evaluation capacity building in Vietnam for the last three years.

75

Moreover, donor-driven evaluations might overlook important impacts because they are generally geared towards the needs of a projects funders and follow pre-set evaluation criteria and checklists in guidelines. In the end, international development evaluation comes down to determining what cooperation brought and what the people, especially local people or beneficiaries, think about it. No matter how extensive and comprehensive a donor-driven evaluation is, outsiders cannot feel the same way local people do. Conducting evaluation with local people; therefore is effective to make sure that the project is evaluated from local peoples perspectives and to avoid overlooking important aspects or groups of people who may be affected by the project. For some time now, so-called joint evaluations have been ongoing. Most joint evaluations, however, are still primarily donor-driven, and partner countries participation in those attempts is limited 2. How can we move to more collaboratively conducted evaluations which genuinely incorporate beneficiaries perspectives and which enrich evaluations so that they can truly help developing countries tackle the issues they are facing and make them useful for future cooperation? This report introduces two recent examples at more genuine collaboration in evaluation to see how such collaboration is useful for partners and donors alike. Section one presents a joint ODA evaluation in Vietnam. Section two explains an evaluation study conducted jointly with a Lao PDR local government. In both cases, local participants perspectives enriched evaluation considerably.

1. Innovative Joint Evaluation Approach in Vietnam


1. 1. Summary and process The first case is a joint ex-post evaluation of the Japanese concessional loan projects being conducted in Vietnam 3. It is different from conventional joint evaluations in that the Vietnamese side (developing country) is playing a major role in its implementation. The 34 member evaluation team consists of four JICA consultants (two
2 Typical form of developing countrys participation in joint evaluation is that either reviewing/commenting on report drafted by donor side or some government officials joining an evaluation committee established by a donor. Ex-post evaluation of the three Official Development Assistance (ODA) loan projects by Japan International Cooperation Agency (JICA). The three projects are Hanoi Drainage Project for Environment Improvement (1) (2), Da Nang (Tien Sa) Port Improvement Project, and Hai Van Pass Tunnel Construction project (1) (2) (3). Evaluation period is from September 2009 to August 2010.

76

Enriching Evaluation with Perspectives of Local People: Two Cases of Collaborative Evaluation

Vietnamese, two Japanese), three local evaluation experts, two local advisors who were active participants in joint evaluations in previous years, and 25 members form three working groups (one group for each project). Among the 34 members, only two are from the donor side (consultants). The remaining members come from various Vietnamese Ministries, local implementing agencies/companies, and local consulting firms. The Vietnamese members developed the evaluation frameworks and survey questionnaires, and conducted the field study. From time to time, the Japanese consultants accompanying in the field study gave some advice, but generally left Vietnamese members in charge of interviews and survey. The team members are now drafting the evaluation report in English: the common language chosen for the report. 1. 2. Merit of this approach This joint evaluation of Japanese ODA projects in which stakeholders from a partner country fully participate in the evaluation work is the first of its kind. The team accepted me as an observer for the field studies, therefore I joined two out of the three projects 4 and observed the following merits of this approach. First, the teams motivation in conducting evaluation was clear and strong. When trying to collect data the local team members stressed that their own government: not just the donor needs to know the result of the project. In a sense, without such motivation, the essence of collaborative evaluation may have been lost, and the evaluation might have become pro forma. Also, when the interviewees were approached in that way, a slight change of expression and attitude could be observed: interviewees typical defensiveness towards evaluators seemed to have faded. Second, interviewees could respond in a more straightforward manner because most of the team members were from their own country, and interviews were conducted in their local language. Asked what the City of Da Nang would have been like now without all these projects, city officials answered that the municipality would have invested their own funds anyway 5. Third, the interviewees frankness was reinforced by the local team mem4 5 The Hai Van Pass Tunnel Construction Project and the Da Nang Port Improvement Project. Only one Japanese consultant was present in the field study of these two projects. This would force the team to look for some particular benefits of investing in it with cooperation from Japan. For example, technology transfer, capacity building during the projects.

77

bers precise knowledge of the projects and the background of the project. In terms of knowledge of the local context, donors consultants cannot compete with their local partners: no matter how competent they may be. For example, the local team members had detailed background information concerning sensitive issues such as relocating residents to build a port access road, therefore they could more naturally direct interviewees to reveal what they knew. Gone was the tension or defensiveness experienced in typical ODA evaluation, instead, openly discussing such situations among the different stakeholders to see if they could do something better was just business as usual. Fourth, interview time was better utilized by, for example, minimizing the need for courtesy calls which can sometimes eat away precious time during interviews conducted by donor side evaluators. Fifth, none of the Vietnamese team members was passive and held almost complete ownership. On the last day of the field trip, when all the scheduled interviews were over, they knew exactly what they had to do before starting to write the report: follow up data collection, undertake additional interviews and make confirmations. Overall, this approach seemed to be quite successful in conducting evaluation in local context. 1. 3. Success Factors First, this joint evaluation followed two years of evaluation capacity building projects by JICA in Vietnam. Therefore, the Vietnamese side was ready to engage in full-fledged joint evaluation. The two Japanese consultants who were in charge of the first two years of capacity building kept working in the third year and by that time they had built-up a network and trust among stakeholders. Second, in Vietnam, institution building for monitoring and evaluation had been enhanced through, among others, a two-year capacity building project as mentioned above and therefore Ministries and implementing institutions have generally been more and more conscious of the necessities of conducting project evaluations. Third, the composition of the team was well thought out, possibly due to the continuity of the Japanese members. The Vietnamese consultants who participated in the evaluation capacity building projects in the first two years joined the team as core members (among them, there were experts in social research). All the concerned Ministries and project implementing institutions
78

Enriching Evaluation with Perspectives of Local People: Two Cases of Collaborative Evaluation

also joined the team. They could share all sorts of background knowledge regarding the projects and their individual field of expertise. Fourth, good teamwork also contributed to the success of the study. Prior to the start of this joint evaluation, Vietnamese consultants and evaluation advisors participated in five days of evaluation training organized by the Japanese consultants. On top of that, four to five days training was offered to the core members of the team. Through these processes, the purpose and the significance of evaluation, evaluation methods and evaluation tools were understood and ownership of evaluation was developed. Also, by working in a group throughout the training, they naturally were able to build a team which exhibited excellent teamwork. (Luckily, this team even included a team-building expert.) 1. 4. Challenges in Implementation Of course, not everything was rosy and challenges abounded: both practical and substantial. First, the cost of sharing information is higher in these types of joint evaluations. An increase in the number of the team members increases the time needed to share the information collected every day as, ideally, all the data gathered should be shared before proceeding to the next day. That means increasing the number of internal meetings required during the already-compressed field study schedule. However, this happens to any large-scale evaluation team. And as far as this particular case was concerned, the merits of a larger team composition seemed to have outweighed the demerits since data collection work was efficiently divided among team members. Second, language matters. Luckily, almost all the members of this particular team were English speaking. Had it been otherwise, the process of evaluation planning, making, using, and analyzing questionnaires, and drafting the report, translating and revising it back and forth each time would have been time consuming and would have tested everyones patience. Actually, last year (during the second year of capacity building), according to the team members who conducted joint evaluations in two languages, too much time and energy was spent translating everything back and forth. Third, cost maters. The evaluation team had members from both Hanoi and the project sites in Da Nang which is about 750km away from Hanoi. Some members had to travel back and forth for the field study and meetings necessary for finalizing the draft report. For the Vietnamese entities these costs had to be justified legally and also practically.
79

Fourth, difficulties exist fulfilling donors requirements. For donor-funded development projects, numerous requirements exist in regards to its evaluation. Explaining each requirement and ensuring all the team members pay attention to them probably improves the quality of evaluation. On the other hand, some of the detailed requirements might not even be relevant to the Vietnamese side. 1. 5. Implications The last point actually leads to a fundamental issue: how much deviation from guidelines can donors, or clients of ODA evaluations accept? Different perspectives could change the direction of the study, the evaluation questions, and even the conclusion. Would a donor accept whatever evaluation results were turned out or would the donors consultants have to rewrite the report to strictly adhere to the donors guidelines? Compared to conventional donor-driven evaluation, the standards and guidelines of joint evaluations may be more lenient. (This could mean that the purpose of a joint evaluation may have to be slightly adjusted).

(Interview at Da Nang Peoples Committee. Vice Mayor (in front) is giving authorization to respective person in charge (on the left) to provide information required by the evaluation team (on the right)).

2. A Collaborative Evaluation Study in Lao PDR


2. 1. Summary and Process The second case is a socio-economic impact study of the Second Mekong International Bridge (hereafter FBII) in Lao PDR 6. This evaluation study was conducted not as an official ODA evaluation but as a research program by FASID 7. The purpose of the study was to gain a comprehensive understanding of the impact of the FBII by collaborating with a local partner: the Department of Planning and Investment, Savannakhet Province, Lao PDR
80

Enriching Evaluation with Perspectives of Local People: Two Cases of Collaborative Evaluation

(hereafter DPI). It was free from the donors checklists and pre-set project goals, and it tried to look at project-related issues and concerns from the local peoples perspectives. Although DPI was the only implementation partner of the study, evaluation questions were derived from the preparatory study conducted in the previous year which consisted of interviews and questionnaire surveys from central and local government officials, business leaders, and think-tanks from both sides of the FBII (Savannakhet, Lao PDR and Mukdahan, Thailand). This preparatory study found that in addition to direct economic benefits the local people were very much concerned with the social impacts of the FBII, and instead of the outcomes of a single project, they were concerned with the combined impact of other projects and government policies 8. Based on the prep-study, DPI and FASID jointly developed the evaluation questions for the impact study. Because of the language difference and DPIs time-availability, a report was drafted in English, translated into Lao, reviewed and more information was added by DPI (the report in the Annex). 2. 2. Merit of this approach In this case, evaluation was conducted collaboratively which involved both sides taking ownership in the evaluation process and jointly deciding the details of the study (such as what issues to be studied; what research methods should be used; who should be interviewed; how the questionnaire should be phrased; and how the findings would be communicated and used). When compared to a donor-centered evaluation, the following significant changes resulted. First, it made the evaluations coverage more extensive. This is natural since the local people often consider what outsiders fail to. The FBII impact study covered the farmers whose land was confiscated but who were not included in the initial evaluation framework. Second, it made the information source far richer since the local partner
6 The Second Mekong International Bridge, which opened in December 2006, was built over the Mekong River between Mukdahan, Thailand, and Savannakhet, Lao PDR. It was built to promote trade between Lao PDR and Thailand and to promote economic development in the region. It was one of a number of infrastructure projects on the East-West Corridor (a transnational road connecting Vietnam, Lao PDR, Thailand, and Myanmar.). The Japan Bank for International Cooperation provided a concessional loan for construction and consulting services to both Thailand (4,079 mil. yen) and Lao PDR (4,011 mil. yen). The evaluation team consisted of a representative from DPI, two local consultants/interpreters, and two FASID staffs. Fujita 2009

7 8

81

acted proactively in finding the best information sources which in this case included brothel owners, commercial sex workers, and a human trafficking victim. Furthermore instead of focusing on the town by the FBII, the local partner stressed the impact on the areas adjoining the national road connected to the FBII. Third, it provided a more holistic interpretation by finding various explanations for one phenomenon. For example, the causes for the sharp decline in Tuk-Tuk (bike taxi) business and in the number of Tuk-Tuk drivers using the ferry port as their base were explained by not only the ferry customers shifting to using the bridge, but also the government policy to shift short-distance transportation means from Tuk-Tuks to minibuses, and the change in lifestyle such that more Lao people were using privately-owned bikes and cars rather than Tuk-Tuks as their means of transportation. Fourth, compared to donor-taken initiative, information collection done by the partner under its own initiative and using its own networks and knowledge base is more efficient. Altogether, these factors provided greater context to the evaluation, and consequently made it more useful to the local partner. An additional result was that joint evaluation greatly motivated the local partner. Towards the end of the field study, the provincial government was quite confident in running similar evaluations and requested training for evaluators for future projects. After all, partner countries ownership is critical; public policy evaluation is important beyond international development cooperation programs as there are accountability and information needs to be met throughout the public sector in partner countries. 2. 3. Challenges in implementation Some of the challenges were similar to the Vietnam case. First, communicating within the team. The fact that communication between DPI and FASID had to go through interpreters all the time sometimes restricted mutual understanding. It may work better in this case to nominate a local team leader who can communicate well with the donor side and can assume the responsibilities of running the field study in close consultation with the donor side. Second, the trade-off between cost-bearing and ownership. Unlike the first case, no long travel was involved for DPI in this study. Still, for these kind of donor initiated studies, the donor side customarily pays local partners a per diem . Such compensation increases the risk that activities are done for
82

Enriching Evaluation with Perspectives of Local People: Two Cases of Collaborative Evaluation

financial incentives and not for the utility of a study. Third, a prior understanding of how much of the results will be made public may be necessary since some sensitive issues could come up unexpectedly during the study. Fourth, certain compromise may have to be made in terms of evaluation methods. For example, in discussions with DPI during the process of shaping up the evaluation methods, PRA and random sampling were excluded. As for PRA, DPI had prior experience in which they felt researchers intentionally directed the opinions of uneducated group of people 9. Random sampling was excluded because DPI claimed that uneducated people do not express their opinions for this kind of surveys and that obtaining the opinions of the people most positively/negatively affected is more important. Such logic makes sense to some extent but may not be persuasive to a donor.

The Second Mekong International Bridge (FBII)

Interviewing a basket vendor

3. Boundaries and Perspectives


As an official ODA evaluation, the joint evaluation of the project in Vietnam had to fulfill all of the donors requirements. The boundary of evaluation, therefore, was restricted and the initial evaluation framework may have been more or less the same as a donor-driven evaluation. Perspectives, however, naturally differed from a donor-driven evaluation since it was mostly local people who implemented the evaluation and this may have changed the evaluations boundaries accordingly (the report will be made public by the end of 2010). In the Lao PDR evaluation study, no donor guideline was used and the team set the evaluation questions entirely under their own discretion. The
9 Therefore, in this evaluation study in Lao PDR, instead of targeting particular group of people, each respondents educational background was asked and later responses were compared according to their education level.

83

evaluations perspectives changed from that of the donors to that of the local government and the local people. The boundary of the evaluation was changed from whether or not the pre-set goals of the bridge project were achieved to what actually happened after the FBII was constructed. The inter-relationships of various factors were considered instead of the direct cause-effect relationships. In both cases, the evaluations perspectives and boundaries clearly changed from the donor-driven evaluation and for this reason the evaluations could help bring out the real impact of the project, and shed light upon truly important aspects. Compared to 20 years ago, huge progress has been made ensuring that the evaluation of ODA projects has become standard practice. However, in the process, too much emphasis may have been placed on the process of standardization and on meeting the formal requirements of the donors. The goal of ODA is to help development and the goal of evaluation is to help ODA. Evaluation of ODA projects can be used to help development in partner countries in a number of important ways in addition to improving a projects management or providing accountability to taxpayers.

References
Bryk, A., Stakeholder-based Evaluation, New Directions for Program Evaluation (17), 1983. Cousins, J.B.and Whitmore, E., Framing Participatory Evaluation, New Directions for Evaluation (80), 1998, pp.5-23. Dabelstein, N., Kliest, T., and Wood, B., Evaluating the Paris Declaration: A Joint Cross National Evaluation, Presented at Annual Conference of the American Evaluation Association, 2008. Daigneault, P., Jacob, S., Toward Accurate Measurement of ParticipationRethinking the Conceptualization and Operationalization of Participatory Evaluation, American Journal of Evaluation 30(3), 2009. Fujita, N., Evaluating a Project from Beneficiaries Perspectives A Case Study of the Second Mekong International Bridge , Review of Development and Cooperation, 2009. Haraguchi, T., and Miyazaki, K., A Practical Approach of Assistance in Evaluation Capacity Development Experience through Evaluation of ODA Loan Projects, Trends in Development Assistance 5, pp.53-95, 2009. http://www.fasid.or.jp/english/publication/research/evaluating.html King, J., Making Sense of Participatory Evaluation Practice, New Directions
84

Enriching Evaluation with Perspectives of Local People: Two Cases of Collaborative Evaluation

for Evaluation (80), 1998, pp.56-68. Lundgren, H., and Kennedy, M., Supporting Partner Country Ownership and Capacity in Development Evaluation. The OECD DAC Evaluation Network, in UNICEF, 2009, Country-led monitoring and evaluation systems, pp.77-95. OBrien F., Preface, in UNICEF, ibid., pp.2-3, 2009. OECD Development Assistance Committee, Paris Declaration on Aid Effectiveness http://www.oecd.org/dataoecd/11/41/34428351.pdf,2005a. Accessed on 2008. 11. 23rd . Joint Evaluations: Recent Experiences, Lessons Learned and Options for the Future, DAC Evaluation Network Working Paper, 2005b. Guidance for Managing Joint Evaluations, DAC Evaluation Series, 2006. Segone, M., Enhancing evidence-based policy-making through country-led monitoring and evaluation systems, in UNICEF, 2009, op.cit., pp.17-31. Stake, R.E., Standards-Based and Responsive Evaluation, presented at an IPEN conference on Development and Evaluation, Tbilisi, Georgia, 29th September, 2006. Weaver, L., Cousins, J.B., Unpacking the participatory process. Journal of Multidisciplinary Evaluation, I, pp.19-40. Retrieved 21st September, 2009, from http://survey.ate.wmich.edu/jmde/index.php/jmde_1/article/view/ 144/159. World Bank, The Accra Agenda for Action, 2008. Retrieved 23rd November, 2008, from http://siteresources.worldbank.org/ACCRAEXT/Resources/47007901217425866038/AAA-4-SEPTEMBER-FINAL-16h00.pdf Worthen, R., Sanders, J.R., Fitzpatrick, J.L., Program Evaluation: Alternative approaches and practical guidelines, Longman, White Plains. NY, 1997, pp.153-169.

85

Annex

The Socio-Economic Impacts of the Second Mekong International Bridge


Evaluation Study Conducted by: The Department of Planning and Investment, Savannakhet Province, Lao PDR (DPI) and The Foundation for Advanced Studies on International Development (FASID)1

Nobuko Fujita, Eriko Takahashi

1. Introduction
The Second Mekong International Bridge (FBII 2, Figure 1) opened in December 2006. Three years after its opening, in order to comprehensively understand its socio-economic impact, the Department of Planning and Investment (DPI), Savannakhet Province, and Foundation for Advanced Studies on International Development (FASID) jointly conducted the evaluation study which follows.

The evaluation team consisted of the following members: Keokhonsy Thonchankhame (DPI), Vinata Sayavong and Vannasao Soumpholphakdy (Civilise Consulting Co. Ltd.), Nobuko Fujita and Eriko Takahashi (FASID). The report was compiled by FASID (sections 1-5, 7: Fujita, section 6: Takahashi, section 8:Fujita and Takahashi) and reviewed by DPI who added more detailed information. FBII stands for the Second Friendship Bridge.

87

Figure 1. Map of the Site


Mekong River Vietnam Myanmar
Fourth Intl Bridge Naypyidaw Vientiane Third Intl Bridge First Friendship Bridge Yangon Second Intl Bridge Mawlamyine East-West Economic Corridor Savannakhet Da Nang Mukdahan Hanoi

China

Lao PDR

Thailand
Bangkok

Cambodia
Phnom Penh

(Source: Author)

2. Criteria of Evaluation
The criteria of evaluation were initially derived from the preparatory study (evaluation needs study) conducted in August 2008 3. Evaluation needs studies are conducted to determine what developing countries expect from evaluations of development assistance projects. For the FBII needs study a total of 60 interviews were conducted between July 9th and 16th, 2008 4. Interviewees were asked: If an evaluation of the construction of the FBII were conducted, what aspects would you like to look at? or What results or effects of the bridge construction would you like to know more about, or are you concerned about? Responses were grouped into six categories: actual economic benefits, social impacts, impacts on other countries, immediate issues (such as immigration services), cost-benefit analysis, and prescriptions. Among responses interviewees showed a pronounced interest in the social impacts of the bridge. Interviewees interests covered a wide range of issues: from labor migration, prostitution, HIV/AIDS infection, human traf-

3 4

Fujita 2009 Interviewees included high-ranking central and local government officials; leaders of business associations; researchers at a University and two think-tanks; residents and business people (living or running businesses near the bridge, ferry terminal, in the markets, alongside National Highway No.9, or inside the town of Kayson Phomvihane).

88

The Socio-Economic Impacts of the Second Mekong International Bridge

ficking, traffic accidents, environmental issues, to an increase in crime (murder, robbery, drug trafficking, and smuggling). These issues are prioritized according to the frequency of responses. Some of the issues were already examined to some extent by the Asian Development Banks evaluations and were therefore excluded 5. After discussing the rest of the issues with the DPI, they chose the issues to study according to their sense of urgency. In the end, they selected almost all the social impacts along with the cultural, economic impacts (Table 1).
Table 1. Criteria of Evaluation
Category Social Impact Labor immigration Remittances Traffic accidents HIV/AIDS Drugs Human trafficking Crimes Smuggling Prostitution Social exchange with Thai-side Cultural Impact Change in fashion, music, food, language. Environmental changes along the corridor Economic Impact Feeling of affluence Economic growth, trade, investment, tourism Change of occupation Others Benefit of the bridge Other changes Environmental Impact 6 Environmental changes in town Issues

3. Methods used
Key informant interviews and semi-structured interview-style questionnaire surveys were conducted during the period September 4 th-11 th, 2009, in
5 Criteria discussed in the 2008 preparatory study and determined to have been covered in evaluations by the Asian Development Bank and therefore excluded in this study included immigration services and the impact on minority villagers near the Vietnamese border. Environmental issues (such as deforestation, noise from vehicles and people, dust, garbage, and exhaust gas from increased traffic, dump trucks with toxic substances travelling at high speeds,etc.) were later excluded from the evaluation criteria of the study since it was deemed to be too much to cover in the short period of time and given the expertise needed in that area was lacking.

89

Kayson Phomvihane, Outhoumphone, and Champhone Districts in Savannakhet Province (Table 2, Figure 2, and Figure 3). Official statistics were collected wherever available.
Table 2. List of Interviewees
Key Informant Interviews Interviewees Savannakhet Province Dept. of Health Savannakhet Province Dept. of Labor & Social Welfare Savannakhet Province Dept. of Public Work and Transport Savannakhet Province, Womens Union Residents who moved to Savannakhet Province within three years Truck drivers from Thailand Service women (Kayson Phomvihane) Human trafficking victim Total Interviewees Residents in Nong Dern Village Residents in Phonsavan Neu Village Udom Viley Village Champhone District Near km 35 Service women (Beer stand along Route 9) Xay Ya Poom District (Kayson Phomvihane) Residents whose land was confiscated for the FBII construction TukTuk drivers Near the Ferry Terminal The FBII Terminal Residents north side of the FBII (within 1 km of the FBII) Residents near That Inghang Total Grand Total Number 3 1 1 1 4 2 5 1 18 Number 3 1 5 7 2 4 7 6 Hotel, restaurants and shops Notes Village where the casino is located Near the FBII, along the by-pass Near the FBII, along the highway 56 km from central Kayson Phomvihane 35 km from central Kayson Phomvihane Interview-style Questionnaire Survey 7 Notes

4 1 7 5 50 70 The most popular tourist spot in Savannakhet

Respondents were chosen from residents who had been living in Savannakhet for more than three years.

90

The Socio-Economic Impacts of the Second Mekong International Bridge

Figure 2. Area of Survey (1) Kayson Phomvihane


Udon Vilay Casino Dinosaur Junction Nong Dern Phonsavan Neu By-pass Provincial Administration Offices Immigration Office Airport Bus Terminal Market Temple Church That IngHang Kayson Phomvihane City Km35 N

Second Friendship Bridge Mekong River

XayYaPoom Ferry Terminal

500 1000 1500m

(Source: Author)

Figure 3. Area of Survey (2) Savannakhet Province


Route 13

Thailand Outhoumphone District Champhone Distirict

Vietnam

Kayson Phomvihane District FBII

Route 9

Surveyed area

(Source: Author)

4. Major Changes in the Year 2008-2009


Major changes in the town of Kayson Phomvihane in 2008-2009, included, the opening of a casino near the FBII, the reopening of Savannakhet airport, and an investment boom associated with both of these new developments. The casino (Savan Vegas) opened in August 2008 8, and as of September
8 An adjacent hotel opened in April 2009.

91

2009, has been attracting more than 1,000 tourists (mostly Thai) every day 9. The majority of Thai visitors are picked up by vans provided by the casino and offered a free lunch buffet; some tourists opt to visit local restaurants taking Tuk-Tuks, or private cars. According to the Department of Labor and Social Welfare, the casino has also generated a total of 1,200 jobs 10. Savannakhet airport reopened in August 2008. Three flights per week are now available from Vientiane and travel time from Vientiane has been shortened from an eight-hour car drive to a one-hour flight. The Bangkok route started in October 2008 has also improved access to Savannakhet tremendously for international businessmen and tourists. The FBII and related factors as mentioned above have helped activate the Savannakhet economy. New resort hotels and modern shops opened and the town has become cleaner than it was before. Other investments such as sugar cane factories and manufacturing have been on the increase as well. Despite an improved economy which has helped create a labor shortage in Savannakhet, local workers still go to Thailand for higher wages, and the numbers are increasing.

5. Social Impact
5.1. Results of the Questionnaire Survey A summary of the questionnaire survey results are shown in Figure 4 (n=50). The respondents were asked about the changes that have occurred since the opening of the FBII. According to the results, the largest change has been seen in labor migration. To address the concerns of the DPI, the results of the questionnaire survey were broken down into respondents with a middle school education or above (Figure5, n=36), and respondents with a high school education or above (Figure 6, n=27). The results show that differences in respondents educational levels did not affect the answers. 5.2 through 5.9 below explain the changes in each item.

On September 9, 2009, the so-called 999 (lucky) day, 3,000 Thai crossed the FBII to come to the casino, according to a local newspaper. 10 Out of the 1,200 workers, 200 were foreigners.

92

The Socio-Economic Impacts of the Second Mekong International Bridge

Figure 4. Perception of the Changes after the Opening of the FBII (All Respondents)
Perception of the changes after FBII (All respondents. n=50) labor migration prostitution drugs crimes traffic accidents HIV / AIDS human trafficking smuggling 0% 20% 40% 60% 80% 100%
(Source: Questionnaire Survey, 2009)

More often than before unchanged less than before dont know other

Figure 5. Perception of the Changes after the Opening of the FBII (Respondents with Middle School Education)
Perception of the changes after FBII (Respondents with education level: middle school or above. n=36) labor migration prostitution drugs crimes traffic accidents HIV / AIDS human trafficking smuggling 0% 20% 40% 60% 80% 100%
(Source : Questionnaire Survey, 2009)

More often than before unchanged less than before dont know other

Figure 6. Perception of the Changes after the Opening of the FBII (Respondents with High School Education)
Perception of the changes after FBII (Respondents with education level high-school or above, n=27) labor migration prostitution drugs crimes traffic accidents HIV / AIDS human trafficking smuggling 0% 20% 40% 60% 80% 100%
(Source : Questionnaire Survey, 2009)

More often than before unchanged less than before dont know other

93

5.2. Labor Migration Although the exact numbers are not available, since the completion of the FBII, labor migration to Thailand has been increasing 11. The questionnaire survey result shows that 56% of the respondents think that more people from Savannakhet work in Thailand since the construction of the FBII, 30% think that there has been no change, and 7% think that fewer people work in Thailand as compared to before. This seems to be a combined effect of the FBII and visa exemption. Many Lao people used to enter Thailand with a border pass (valid for three days) and work illegally in Thailand. Now, many hold a passport which costs 30US$ per applicant and allows them to stay in Thailand for up to three months. The Ministry of Foreign Affairs service to process passport applications at the FBII facilitated this move. Outgoing Lao travelers from Savannakhet by passport increased from 35,749 in 2006 to 92,283 in 2008, while those with a border pass increased in 2007 possibly due to the opening of the bridge, but then decreased again in 2008 to 2006 standards (Figure7). Although carrying a passport does not allow people to work legally in Thailand, many Lao workers understand that just carrying a passport makes it legal. Although Savannakhet now has many employment opportunities such as working in gold and copper mines, sugarcane and textile factories, the low minimum wage of 25,000 kip/day makes attracting workers difficult 12.
Figure 7. Out-going Lao Travelers from Savannakhet Province
100 90 80 70 60 50 40 30 20 10 0

travelers (1000s)

Passport Border Pass

2006

2007

2008

(Source: Lao National Tourism Administration, 2008)

11 ILO reports that 27,000 Savannakhet people worked in Thailand in 2004. In 2008, the number available was 42,069; however it is not accurate according to the Department of Labor and Social Welfare (Interviewed on Sept.11, 2009). 12 1US$= about 8,500kip as of Sept.10, 2009

94

The Socio-Economic Impacts of the Second Mekong International Bridge

5.3. Prostitution The next biggest change since opening of the FBII perceived by the interviewees regarded the issue of prostitution. Although data on the number of brothels 13 or prostitutes is unavailable, 43.8% of the respondents answered that prostitution has been increasing since the FBII opened (33.3% answered unchanged, 0% answered decreased, and 18.8% said dont know. See Figure 4). Among key informant interviews, respondents perceptions showed mixed results. At a beer house located more than 30 kilometers away from the FBII on Route 9, an interviewee stated that her business has declined since the casino opened. However, more beer houses have been established along Route 13 in Champhone District recently. Of the twenty-three respondents who said that prostitution had increased: fourteen said that the increase in prostitution came from the demand side due to the increase in the number of international tourists (including casino guests); and nine people suggested that the increase came from the supply side citing that it is a quick way to make money, especially for local people who are having difficulty finding a job or who simply want to purchase fashionable Thai clothes and cell phones. One of the authorities mentioned that because of the advocacy campaign, many of the service women 14 changed their jobs by learning various skills through training provided by Womens Union, international organizations and NGOs. One source claimed that many small beer restaurants shut down due to this fact, but available statistics could not confirm this claim. On the other hand, another official suggested that one negative impact of the bridge was that young girls living in the countryside (Louangphabang, Vangviang, Xiangkhouang) who are attracted to Savannakhet end up working as service women 15. Their main customers are Thai nationals who can now cross the FBII anytime between 6:00 A.M. to 10:00 P.M. 5.4. Human Trafficking Not all women moved into Savannakhet on their own free will. Officials in charge of public security, labor and social welfare, and womens issues, all

13 Since beer restaurants are all registered as restaurants, the official number of brothel-type beer restaurants is not available. 14 Commercial sex workers 15 Out of the 50 questionnaire survey interviewees, 6 moved in to Savannakhet in the last 3 years; 2 came to go to college, 1 came to open a store, and the remaining 3 were service women.

95

said that human trafficking has been increasing since the FBII opened. Lao immigrant workers in Thailand return home and take friends and relatives back with them to Thailand, exposing them to human traffickers. According to Savannakhet Womens Union, the number of cases of human trafficking reported in Savannakhet Province was four in 2007 and two in 2008. Data was not collected in 2006 according to the Womens Union 16. In 2009, although none of the cases reported involved Savannakhet residents, reported cases did involve two Vietnamese women, four women from Attapeu Province and eight women from Khammouane Province. Many of the traffickers use small boats and cell phones to cross the Mekong River at night. Additional factors such as an increase in labor migration and a desire for high income also had a combined impact on the increase in human trafficking. The anti-trafficking campaign might also have made people acknowledge cases of human trafficking that previously would not have been recognized as such. 5.5. HIV/AIDS & Other Communicable Diseases Cases of people with HIV reported in Savannakhet Province decreased in 2008. However, as of the first half of 2009, 90 cases have already been reported, which indicates that an increase is expected in 2009 (Figure 8). In 2008, 86% of all cases reported were in Kayson Phomvihane; 94% were along Route 9 and all the cases reported in 2008 were along either Route 9 or 13 (Figure 9). Immigrant workers and truck drivers accounted for 55 percent of all reported HIV/AIDS cases, and housewives who were most likely infected by their immigrant worker husbands accounted for 22 percent (Figure 10). The 2008 study revealed that the lower HIV infection rates in Savannakhet, Lao PDR (0.8%) compared to Mukdahan, Thailand (23%) was considered a likely reason that Thai go to Laos to obtain services. The difference in infection rates was about the same in 2009 (Mukdahan: 23%, Savannakhet: 0.9%) according to Savannakhet Department of Health (DOH) 17.

16 Interviewed on September 9, 2009. 17 Interviewed on September 7, 2009.

96

The Socio-Economic Impacts of the Second Mekong International Bridge

Figure 8. HIV Positive Cases Reported in Savannakhet Province


200 Number of report 150 100 50 0 2006 2007 2008 2009 (Jan.-June)
(Source: Savannakhet Province Department of Health, 2009)

Figure 9. HIV Positive Cases Reported in 2008 (by Districts 18)


(Savannakhet Province, Jan. to Dec.2008)

Route 13
THAILAND XBL

Khammouane Province VIETNAM

5 6
OUT

3
APT

1
PHA

Route 9

113
KAY

1
CHA

FBII

2
SON

Saravane Province
(Source: Savannakhet Province Department of Health, 2009)

18 XBL: Xayboury Province, KAY: Kayson Phomvihane, OUT: Outhoumphone, CHA: Champhone, SON: Songkhone, APT: Atsaphangthong, PHA: Phalanxay Districts.

97

Figure 10. HIV Positive Cases Reported in 2008 (by Occupation)


(Savannakhet Province, Jan. to Dec. 2008) 60 50 40 Numbers 30 20 10 0
nw ork er en n nt ist lice po tr u se wif e tw ork er plo ye e dre rch a rm wo m sty l ck d ch il t fa riv er er

48 Blood Tested: 2.986; HIV +: 131; Female: 62 29 21

5 1

Ho u

gra n

em

me

se rvi ce

gra n

cti o

ns tr u

mi

(Source: Savannakhet Province Department of Health, 2009)

The H1N1 flu is a recent concern in the Savannakhet Province where more than 20 cases have been reported as of September 7, 2009. The first patient from Savannakhet was infected in Thailand, where he worked (reported on June 26, 2009). On July 12, the first patient in the casino was reported. The casino, which was crowded with gamblers, their families, spectators, and free lunch takers, became a hot spot and 16 staff members were diagnosed as contracting H1N1 19. DOH and the casino immediately took counter-measures such as introducing thermographs and masks, and according to DOH, no further patients have been reported since August 15, 2009 (as of September 7, 2009) 20. Regional surveillance which was conducted for 10 infectious diseases in 2008 increased to 19 diseases in 2009. In the case of H1N1, coordination among hospitals in other provinces and Thailand was done smoothly. DOH constantly updates information regarding infections and makes such information readily available. 5.6. Drugs, Crimes and Smuggling Official data recording increases/decreases in drug confiscations, the num-

19 Many of the casino guests come in families, including grandparents and small children. 20 DOH interview on September 7, 2009.

98

co

mi

be

au

ty

The Socio-Economic Impacts of the Second Mekong International Bridge

ber of crimes, and the incidences of smuggling was not available. Therefore, the following information has been drawn from key person interviews and the questionnaire surveys. Eighteen out of 50 respondents (36%) of the questionnaire survey responded that drugs have been on the rise because of easier access by the FBII and increased traffic. Ten respondents (20%) said unchanged, and the same number of respondents (20%) said drug use has decreased due to tighter control. Either way, drug use is a two-way issue, i.e., drugs coming in from Thailand and going out from Lao PDR. As for general crimes, 15 respondents (30%) said, increased. Four of them were either victims of burglaries or neighbors of victims; items stolen included bikes, helmets, cows, shoes and canopies. Some people mentioned that crimes have increased because Thai criminals have joined forces with local criminals. Casino customers in particular have been made prime targets of robbery and snatching since it first opened. Furthermore, the losers of gambling can become potential robbers. Twenty-two respondents (44%) answered that the number of crimes has not changed, and four (8%) said that crimes have decreased due to tighter security (initiated by the government campaign Village without crimes). One said the newly built road (access road to the casino) also contributed to a decrease in the number of crimes. As for smuggling, only three respondents (6%) said, increased, they explained that they had witnessed gasoline and food being smuggled by passenger cars and trucks. Twenty-one (42%) said unchanged. Eight respondents (16%) said, decreased, due to tighter control and regulations. One of them said that easier transportation by the FBII made smuggling unnecessary. The respondents who answered unchanged said that smuggled goods (in this case, small daily items such as detergents and candy) are brought in from Thailand by small boats, therefore the construction of the FBII is not related in any way to cases of smuggling. 5.7. Traffic Accidents According to the Provincial Police Department, safety campaigns conducted over the last three years have helped decrease traffic accidents in Kayson Phomvihane (Figure 11). However, according to the questionnaire survey, 16 people (32%) responded that traffic accidents have increased as opposed to two respondents (4%) who said decreased. Respondents that believe traffic accidents have increased mentioned that There have been many motorbike accidents recently (18%), The number of cars on the road have increased
99

(12%), Many second-hand Korean cars are imported (2%), and Drunk drivers coming out of the casino get involved in accidents (2%). According to the respondents, the causes of accidents include drunk driving, the absence of a license, and higher accessibility to motor bikes especially for young people due to cheap bikes imported from China. Fifty-eight percent said unchanged implying that the FBII is not a direct cause of the increase in accidents since most of the accidents involve motorbikes, and motorbikes are prohibited from crossing the bridge.
Figure 11. The Number and Cause of Traffic Accidents in Kayson Phomvihane
Speeding Drunk driving Change lanes No waiting Wrong lane Pass Fall over Others 0 100 200 300 400 500 number of accidents 600 700

2004-2005

2007-2008

(Source: Kayson Phomvihane Provincial Police, 2009. Figures are for fiscal year, October to September)

5.8. Access to Health/Medical Services in Thailand According to the questionnaire survey, 22 respondents (45.8%) said they would seek health/medical care in Thailand, of which 16 (33.3%) said they would use the FBII to do so. The same number of respondents (22 or 45.8%) said they would not seek such services. Respondents who answered that they would seek medical attention in Thailand offered the quality of available services as a reason. Respondents who said that they would not go to Thailand for medical care gave reasons which included, My family is very healthy (7 respondents), No money (6), It is more convenient to use local services (ID or passport is not required) (2), and Local hospitals are getting better (2). Consequently, the people who seek Thai medical services may be limited to those people who can afford it, and the FBII may facilitate such people to obtain medical services. In terms of emergency care, the FBII benefited the whole population of Savannakhet since Category I patients 21 are transferred to hospitals in Mukdahan, and the DOH stated that the FBII has undoubtedly facilitated emergency medical care.
21 Patients with life-threatening conditions

100

The Socio-Economic Impacts of the Second Mekong International Bridge

5.9. Negatively-affected People


5.9.1. Farmers:

The questionnaire survey showed that the respondents think farmers (whose land was expropriated for the FBII) benefited least by the FBII. For some of these people, the process for receiving expropriation compensations is not over, even though their land was expropriated as early as 2000. According to the interviews of six farmers who use to own paddy, fruit and vegetable fields and farmhouses at the construction site, alternative land was offered which was too rocky to make suitable for farming, and was one hour away by car. (Back then, their farms were in walking distance.) All six farmers therefore chose monetary compensation instead. The first payment was made in 2005 and small-scale farmers were all compensated. Large-scale farmers were to be paid in several payments but as of September 2009, the final payment had not yet been made 22. It is noteworthy that, this issue seems to have some influence in land expropriation for the Savan-Seno Economic Zone (SEZ). A farmer whose paddy field is located in one of the SEZ sites said that he would not stop planting rice until compensation is completed and that his neighbors have the same view.
Table 3. Who Benefited the Most and the Least from the FBII
Who Benefited the Most Transportation companies Traders/Merchants Government officials/Immigration officers Travel agencies Hotels Restaurants Casino Markets Tuk-Tuk drivers Investors Ordinary people
(Source: Questionnaire Survey, 2009)

Responses

Who Benefited the Least

Responses 12 3 3 2 1 1 1

13 Farmers (whose land was expropriated) 13 Ferry company owners and employees 6 Restaurants near the ferry terminal 4 Tuk-Tuk drivers 3 Small store owners 3 Students (*because they dont use the FBII) 3 Government officials 3 3 2 1

(n=50)

5.9.2. Ferry Companies:

According to the study conducted in 2008, ferry companies lost half of their
22 Note: Time constraints precluded conducting interviews with government officials to clarify this issue.

101

customers after the FBII opened. This is despite the fact that many people still prefer to use the ferry due to the ferry terminals more convenient location downtown and because it is closer to the market or their residence. Restaurants by the ferry terminal also experienced a downfall in customers.
5.9.3. Tuk-Tuk Drivers who Used to Work for Ferry Passengers:

Tuk-Tuk drivers who operate near the ferry terminal have seen a decline in their customers; some say half of the pre-FBII customers have gone. According to the remaining active drivers, many of their colleagues lost their jobs and now work in construction and other industries. According to the Department of Public Work and Transport, the number of registered TukTuks is as follows 23.
Figure 12. Number of Tuk-Tuk in Kayson Phomvihane
450 400 350 # of vehicles 300 250 200 150 100 50 0 2005 2006 2007 2008 2009
(Source: Savannakhet Province Department of Public Work and Transport, 2009)

Tuk-Tuk Sky-lab type Tuk-Tuk

Tuk-Tuk drivers have not left for other jobs simply due to the decrease in ferry passengers. Another factor is the Department of Public Work and Transports policy to shift from Tuk-Tuks to mini-buses as a means of safe, short distance transportation. New registrations of the Sky Lab-type TukTuks (Figure 13) are no longer admitted, due to termination of the production and difficulties in proper maintenance. In addition, in line with a booming economy, the mode of transportation is shifting from Tuk-Tuks to private motor bikes or cars. Therefore, in addition to the FBII, a number of other factors can be attributed to the decrease of Tuk-Tuks and drivers.

23 There was no data available for the number of drivers. However, number of drivers and vehicles are more or less the same, according to the Department of Public Work and Transport.

102

The Socio-Economic Impacts of the Second Mekong International Bridge

Figure 13. Sky-Lab and Tuk-Tuk


Sky-Lab Tuk-Tuk

(Source: Savannakhet Province Department of Public Work and Transport, 2009)

6. Economic Impact
6.1. Gross Provincial Product Savannakhet Province has maintained a high Gross Provincial Product (GPP) rate for the past couple of years. As the graph below (Figure 14) demonstrates, Savannakhets real GPP growth rate (the provincial growth rate) greatly exceeded the real GDP growth rate of Lao PDR (the countrys growth rate). Particularly in the years prior to the FBII construction, Savannakhets GPP growth rate was increasing at a rapid pace. Although anticipation of the bridges completion may have stimulated the provinces economy (for example, by increasing investment) the growth can more likely be attributed to the development of the provinces gold and copper mining industry (IDE-JETRO, 2008). Full-fledged development of the mining industry began in 2003 with Australian company Oxianas 30 million dollar investment in the Sepon Mines (IDE-JETRO, 2008). Since then, an increase in gold and copper exports has been evident; 2005 exports of gold and copper doubled in comparison to the year before (IDE-JETRO, 2008). Consequently, although the Lao PDRs real GDP growth rate has decreased slightly since 2006, Savannakhet has maintained a high real GPP growth rate.

103

Figure 14. Real Growth Rates of Savannakhet GPP and Lao PDR GDP
14 12 Growth Rate (%) 10 8 6 4 2 0 2004 2005 2006 Year 2007 Laos PDR GDP 2008 7.8 8.1 7.1 6.4 7.9 7.5 10.0 11.9 11.9 12.5

Savannakhet GPP

(Source for Savannakhet GPP Real Growth Rate: Savannakhet Province Department for Planning and Investment, 2007, 2008, 2009. Source for Lao PDR GDP Real Growth Rate: World Bank, 2009)

Breaking down Savannakhet Provinces GPP by sectors (agriculture, industry and services), Figure 15 illustrates that, for the years 2004 through 2008 (at current market prices), all three sectors appear to be enjoying a positive trend. Table 4 breaks down the growth rate of the provinces GPP by sector. The GPP of the agriculture sector throughout the observed years has maintained a growth rate of 7-9%. The services sector, likewise, has been growing at a rate of approximately 13% (both before and after the completion of the bridge), although growth slowed slightly in 2008. As for the industry sector, although the growth rate was lower than that of the services sector at approximately 12.6% in 2005, for 2006 (the year the FBII was completed) and for the years following, the growth rate has exceeded 18%, which largely surpasses the growth rates of the other sectors. Although drawing any conclusions is premature, it is possible that the FBII has had a positive effect on Savannakhets sectoral GPP growth rates particularly on the industry sector.

104

The Socio-Economic Impacts of the Second Mekong International Bridge

Figure 15. Savannakhet Current GPP by Sector


6,000 5,000 Billion Kips 4,000 3,000 2,000 1,000 0 2004 2005 2006 Year 2007 2008

AgricultureIndustryServices
(Source: Savannakhet Province Department for Planning and Investment, 2007, 2008, 2009)

Table 4. Savannakhet Current GPP Growth Rate by Sector


2005 Agriculture Growth Rate (%) Industry Growth Rate (%) Services Growth Rate (%) 7.73 12.62 13.22 2006 7.31 19.94 13.77 2007 7.53 18.73 13.71 2008 8.92 19.80 11.76

(Source: Savannakhet Province Department for Planning and Investment, 2007, 2008, 2009)

6.2. Trade and Investment Savannakhets principal imported goods for the period 2006 to 2008, are shown in Table 5 below (expressed in US$ thousands). In general, most of Savannakhets principal imported goods increased after 2007 although data is insufficient to draw any definite conclusions that those increases were brought about by the completion of the FBII. The large increase in imported fuels in 2008, for example, probably reflects the soaring oil prices that year.
Table 5. Principal Imported Goods (US$ thousands)
2006 Food Clothes Agricultural tools and machines Medicine Construction materials Fuel Electric appliances Vehicles and small parts Spirits and tobacco 4,853 1,249 528 373 1,452 29,646 167 8,897 25.4 2007 6,376 1,409 369 369 1,413 26,783 426 12,271 48 2008 6,366 N/A 686 686 3,992 47,897 971 17,322 3,597

(Source: Savannakhet Province Department for Planning and Investment, 2007, 2008, 2009)

105

Figure 16. Inbound/Outbound Commercial Trucks and Freight Volume at the Border between Savannakhet (Lao PDR) and Mukdahan (Thailand)
25 20 19.17 19.46 195 158 126 6.02 6.80 173 7.10 7.41 20.32 225 190 157 21.21 260 300 250 200 150 100 50 0

Vehicles (1000s)

17.85

15 146 10 5 0 132 6.24

2004

2005

2006 Year

2007

2008

Commercial Truck (Inbound) Freight volume (Inbound)

Commercial Truck (Outbound) Freight volume (Outbound)

(Source: Data from the Savannakhet Province Department of Public Work and Transport, 2009)

Figure 16 illustrates the number of inbound and outbound commercial trucks taken at the border between Savannakhet (Lao PDR) and Mukdahan (Thailand), also represented in terms of freight volume. The graph shows the steady inflow of both the number of commercial trucks as well as freight volume both prior to and subsequent to the inauguration of the bridge, hence the effects of the bridge remain unclear. Furthermore, although the bridge is considered advantageous especially for land transportation and freight forwarding businesses, at present, the bridge is being used merely as a transshipment point for Thai and Vietnamese cargo trucks (IDE-JETRO, 2008). To improve the current situation, investment in key industries in Savannakhet is crucial and, therefore, further improvements need to be made concerning the investment climate of the province (IDE-JETRO, 2008). Trends are also not so apparent in regards to Savannakhets principal exported goods. Table 6 below illustrates Savannakhets principal exports in terms of US$ thousands for the period 2006 to 2008. Apart from the plunge in gold and copper exports in 2007, the probable result of a large decrease in the production growth rates of the two minerals, no particular trends concerning exported goods can be seen overall (ADB, 2009). Moreover, Figure 16 clearly shows that both the number of inbound commercial trucks and the total freight volume largely exceeds that of the outbound number and volume; most likely due to the fact that a considerable portion of Savannakhets GPP is accounted for by its primary (agriculture, mining) industry.
106

Freight Volume (1,000 tons)

The Socio-Economic Impacts of the Second Mekong International Bridge

Consequently, the economic impact of the FBII on Savannakhets exports is likely to be a mid- to long-term effect, largely depending on the development of the provinces secondary and tertiary industries.
Table 6. Principal Exported Goods (US$ thousands)
2006 Wood and wooden articles Agricultural products Manufactured products Gypsum Gold Copper Clothes Forest products 17,317 4,708 524 708 92,603 385,404 1,193 324 2007 19,972 1,900 1,711 N/A 19,467 118,412 N/A 42 2008 10,238 4,175 1,487 N/A 82,346 479,198 N/A 353

(Source: Savannakhet Province Department for Planning and Investment, 2007, 2008, 2009)

The following graph (Figure 17) presents the inflow of foreign investment in Savannakhet Province, in terms of capital costs in US$ millions, and the number of approved investment projects, for the years 2006 through 2009. It can be seen from the graph that in 2006, the year the FBII was completed, large-scale investments exceeding a total of US$400 million were made in Savannakhet. Investments approved that year included a US$250 million multinational project conducted by an Indian company that made investments in a Eucalyptus tree plantation and the construction of paper factories. This project, as well as many other investment projects that have been implemented in anticipation of the completion of the bridge are long-term projects (e.g. tree-planting projects) and therefore will require a relatively long amount of time before the effects are observable (IDE-JETRO, 2008). Consequently, although investment may have been enhanced due to improved access to the province and increased investment opportunities, supporting evidence is insufficient to draw any definite conclusions.

107

Figure 17. Foreign Investment Capital Costs and Number of Approved Projects
450 400 350 US$ million 300 250 200 150 100 50 0 2006 2007 Year Capital Cost Projects
(Source: Savannakhet Province Department for Planning and Investment, 2007, 2008, 2009)

25 20 15 10 5 0 Number of Approved Projects

2008

2009

Initially planned by the government in 2002-2003, the SEZ (a special economic zone located in Savannakhet Province) is currently undergoing development. It consist of four zones covering an area of approximately 800 hectares and is expected to include an export processing zone, a free trade zone and a logistics center. It is anticipated that establishment of the SEZ will attract additional foreign investment. 6.3. Tourism The construction of the FBII has had a significant impact on the tourism sector of Savannakhet Province and thus on the tourism sector of Lao PDR as a whole. Figure 18 describes the number of international tourist arrivals (in thousands) in Lao PDR by different ports of entry Savannakhet Province, the First Friendship Bridge in Vientiane, Vientiane Wattay International Airport and other entry points. In 2007 and 2008, a sharp rise in tourist arrivals via Savannakhet can be seen. As a port of entry, Savannakhet accounted for almost 25% of all Lao PDR tourist arrivals in 2007 and 2008 a potential indication of the positive impact of the FBII.

108

The Socio-Economic Impacts of the Second Mekong International Bridge

Figure 18. Lao PDR Tourist Arrivals by Port of Entry


2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0

Tourists (1000s)

Other Wattay Airport Friendship Bridge I Savannakhet

2003 174.08 72.22 357.04 33.02

2004 242.99 86.33 451.10 114.39

2005 313.08 105.53 541.02 135.69

2006 361.34 106.23 582.18 165.36

2007 462.13 105.67 656.47 399.67

2008 628.46 151.94 547.59 408.80

Year
(Source: Lao National Tourism Administration, 2008, 2009)

Figure 19. Inbound/Outbound Passenger Cars at the Border Between Savannakhet (Lao PDR) and Mukdahan (Thailand)
16,000 14,000 12,000 Vehicles 10,000 8,000 6,000 4,000 2,000 0 350 277 2004 352 190 2005 2006 Year 2007 2008 2,161 1,760 5,371 4,438 13,354 11,192

Passenger Car (Inbound)

Passenger Car (Outbound)

(Source: Savannakhet Province Department of Public Work and Transport, 2009)

109

Figure 20. Inbound/Outbound Passengers at the Border Between Savannakhet (Lao PDR) and Mukdahan (Thailand)
120 109.04 100 Passengers (1000s) 82.49 80 62.49 60 45.12 40 35.87 20 0 38.95 48.12 60.09 69.35 80.15

2004

2005

2006 Year

2007

2008

Passengers (Inbound)

Passengers (Outbound)

(Source: Savannakhet Province Department of Public Work and Transport, 2009)

The number of passenger cars and passengers respectively that have entered or exited Savannakhet Province also support this assumption as evident in Figures 19 and 20. From Figure 19, it can be seen that the number of both inbound and outbound passenger cars have increased dramatically at around the same rate. Figure 20 shows that the number of inbound and outbound passengers has been on the rise since completion of the bridge in 2006. Since completion, the number of incoming passengers has exceeded the number of outgoing passengers, with passenger inflow growing at a noticeably higher rate than passenger outflow 24. It can be concluded that the construction of the bridge has played a positive role in attracting tourists to the province. The development of Savannakhet Provinces tourism sector, which has been further enhanced with the inauguration of the FBII, can be observed from Figure 21. The graph shows recent trends (2003 through 2008) in the number of accommodation facilities located in Savannakhet Province (including hotels and guesthouses) and the number of available rooms in those facilities. Since completion of the bridge in 2006, approximately 10 new hotels and guestrooms were constructed, and nearly 400 more rooms were made available, to help accommodate the increase in incoming travelers to the

24 It should be noted that Figure 20 accounts for passengers traveling in regular passenger cars only; passengers who cross the bridge by bus are not represented in the number of inbound/outbound passengers.

110

The Socio-Economic Impacts of the Second Mekong International Bridge

province. Figure 22 presents the room occupancy rates for all accommodation facilities in Savannakhet Province for the years 2003 to 2008. The room occupancy rates have shown a gradual increase throughout the period; after the inauguration of the bridge, the occupancy rates have remained in the high 60s, supporting the assumption that there has been a boost in tourism and traveler arrivals.
Figure 21. Number of Accommodation Facilities & Rooms in Savannakhet Province
90 80 Accomodation Facilities 70 60 60 50 40 30 20 10 0 2003 2004 2005 2006 2007 Rooms 2008 979 995 53 1,257 1,236 67 72 1,627 81 83 1,758 2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 Rooms

Accomodation Facilities
(Source: Lao National Tourism Administration, 2008, 2009)

Figure 22. Room Occupancy Rates for Accommodation Facilities in Savannakhet Province
80 Room Occupancy Rate (%) 70 60 50 40 30 20 10 0 2003 2004 2005 Year Room Occupancy Rate
(Source: Lao National Tourism Administration, 2008, 2009)

68 58 49 60 61

65

2006

2007

2008

111

Recreational facilities, including Lao PDRs first casino hotel, which opened in 2009, have also been developed. The casino (a US$56 million joint venture project Laos (40%) and a Macau-based gaming investment company (60%)), specifically targets the Thai population living in the vicinity of Mukdahan Province which is the region in Thailand linked via the FBII to Savannakhet. Since its opening, the casino has been attracting many Thai customers and has provided employment opportunities 25. Consequently, the revitalization of Savannakhets tourism sector has contributed to the enhancement of the local economy and the creation of employment opportunities.

7. Cultural Impact
As for fashion, food, music and language, interviewees were asked if the influence from Thailand has become stronger since the FBII opened. The answers are as follows (Figure 23). Only for fashion did respondents that answered Yes exceed those that answered No. Changes in style of fashion include young girls wearing revealing clothes such as mini-skirts and short pants instead of the traditional wrap skirt (Sin). This type of change in fashion was considered favorable by some people (7 respondents) and unfavorable by others (7 respondents). On the other hand, as for the influx of Thai cuisine, only positive opinions were given, such as the increase of a variety of different tastes. For music and language, more influence comes from TV programs and other media, and physical improvement of access does not seem to contribute much.
Figure 23. Increase in Cultural Influence from Thailand since the Opening of the FBII
Fashion Yes Cuisine No Music Language 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% dont know

(Source: Questionnaire Survey, 2009)

25 Please refer to section 4, Major Changes in the Year 2008-2009.

112

The Socio-Economic Impacts of the Second Mekong International Bridge

According to some respondents, the FBII has facilitated visiting Thailand for cultural events (attending festivals, visiting temples, etc). Thirty % of the survey respondents said that thanks to the FBII they cross the Mekong River for such occasions more often than before. However, 46% answered unchanged, because they use small boats, rather than ferries or the FBII, to cross the river depending on the accessibility (location, means of the transportation to the access point) 26. Thirteen % said they have never visited Thailand for cultural events (Figure 24).
Figure 24. Cultural Visits to Thailand
other 10% Never gone for cultural events 13% dont know 2% less than before 4% More often than before 28%

n=45
unchanged 43%
(Source: Questionnaire Survey, 2009)

8. Conclusion
8.1. Benefits from the Bridge for the People of Savannakhet The completion of the FBII has brought about evident changes to Savannakhet. The people of Savannakhet have clearly benefited from the bridge, in terms of the positive impacts it has made on the local economy. As stated in section 6, the FBII has had a significant impact particularly on the tourism sector of the province. The inauguration of the FBII has resulted in an enormous influx in the number of travelers coming into Savannakhet, an increase in the number of accommodation establishments, and an improvement in the availability of employment opportunities. The FBII has also brought convenience to the people of Savannakhet in terms of shopping, accessing Thai health/medical services, and many other
26 A small boat to cross the river costs 100 Baht/person (round trip), a ferry costs 14,000 kip/ person (oneway) and the cross border bus to cross the FBII costs 50 Baht (one-way). Cost-wise, there is no difference. (1Baht=250 kip as of Sept.9th, 2009)

113

cross-border activities. On the other hand, not all the people use the FBII and benefit from it. Some of the locals that live right by the FBII have never used it, and instead cross the Mekong River by ferry or small boats from nearby cross border points as they did before. Savannakhet residents who directly benefit from the bridge are limited to people with a degree of financial status and can afford Thai medical/health care services and enjoy shopping, dining, and sightseeing in Mukdahan and beyond. The questionnaire survey asked if respondents have benefited from the FBII; 66% answered Yes, 18% answered No, 16% answered dont know.
Table 7. Have you Benefited from the FBII? Questionnaire Survey Results (n=50)
Yes To cross the Mekong River For shopping in Thailand Good business due to customer increase Convenient to go to hospitals on Thai side Convenient to go to festivals Family member lives in Thailand and convenient to come and go ATM of Thai banks are now accessible TOTAL
(Source: Questionnaire Survey, 2009)

No

# 4

dont know

# 8

17 Never crossed the FBII 8 Increase of competition, 3 loss of customers (Tuk-Tuk drivers, 2 restaurant owners) 1 1 1 33 TOTAL

5 dont know

9 TOTAL

8.2. Relationships of Various Social Issues Economic activities in Savannakhet have been activated by the FBII and the airport. Investments, which include the casino, hotels, restaurants and sugar cane factories, have contributed to greater employment opportunities. Nevertheless, it appears that the establishment of the FBII and the visa exemption agreement has become a push factor for labor migration to Thailand as higher wages in Thailand entice more local workers. The FBII has also promoted the movement of people, particularly young women from other regions in Lao PDR who are attracted to Kayson Phomvihane and towns along National Routes 9 and 13. In some cases, they have fallen susceptible to human trafficking. As stated before, there may be no direct causal relationship between the construction of the FBII and the increase in human trafficking; the influx of people and the activation of the local economy may have acted as a pull factor attracting young women to towns nearby the FBII. The decision to open the casino in Savannakhet was inseparable from the opening of the FBII and the casino has undoubtedly generated employment and income for Savannakhet. At the same time, however, it has created an
114

The Socio-Economic Impacts of the Second Mekong International Bridge

extremely crowded spot in a once sparsely populated province which has increased the risk of spreading globally infectious diseases such as H1N1. In response, a global surveillance system was reinforced and the Department of Health of Savannakhet Province is now ready to take actions against such outbreaks of disease and infections.
Figure 25. Changes after the FBII
Casino open Airport reopens Convenience in transportation

Increase in investment

Activation of economy

Increased employment

Improved access to Thai medical services Increased risk of infectious diseases

Increase in human trafficking

FBII open Convenience for traveling Visa exemptions


(Source: Author)

Increase in tourists Increase in labor migration Longing for high income Increase in service women

8.3. Implications In conclusion, since the inauguration of the FBII, some signs of economic benefits to Savannakhet are evident but before the province can reap fullfledged benefits such as the development of local industry, more time is needed. Nonetheless, Savannakhet is evidently changing at a rapid pace and further changes are anticipated as the SEZ further develops and as other economic activities increase. In light of such rapid change, continuously assessing the social impacts of economic development and being able to take swift measures will be important. As illustrated in a previous section, when the H1N1 flu struck the casino, the close cooperation and sharing of information between the relevant parties successfully prevented the virus from spreading, demonstrating the benefits of quick and efficient cooperation and transparency between the public sector and the private sector. If Savannakhet can better foster such partnerships not just within Savannakhet but also domestically with other regions and even internationally, the people of Savannakhet will almost certainly enjoy
115

further benefits of the FBII as the provinces economy improves.

References
Asian Development Bank, Reviewing the Poverty Impact of Regional Economic Integration in the Greater Mekong Sub-Region, Regional Synthesis Report, 2006. , Guidelines for Preparing Performance Evaluation Reports for Public Sector Operations, 2006. , Lao PDR and Viet Nam: Greater Mekong Subregion: East-West Corridor Project, Completion Report, November 2008a. , Lao Peoples Democratic Republic and Socialist Republic of Viet Nam: Greater Mekong Subregion: East-West Corridor Project, Performance Evaluation Report, December 2008b. , Greater Mekong Subregion: Maturing and Moving Forward, Regional Cooperation Assistance Program Evaluation, December, 2008c. , HIV Prevention and Infrastructure: Mitigating Risk in the Greater Mekong Subregion, May 2008d. Department for Planning and Investment of Savannakhet Province, Statistical Yearbook 2006 Savannakhet Province, 2007. , Statistical Yearbook 2007 Savannakhet Province, 2008. , Statistical Yearbook 2008 Savannakhet Province, 2009. Department of Public Work and Transport in Savannakhet, Statistic of Freight, Passenger, and Vehicle for crossing the border between Savannakhet (Laos) & Mukdahan (Thailand), 2009. Foundation for Advanced Studies on International Development, Public Sector and Private Sector Perception towards Tourism in Savannakhet, 2006. Fujita, N., Evaluating a Project from Beneficiaries Perspectives A Case Study of the Second Mekong International Bridge , Review of Development and Cooperation, 2009. IDE-JETRO. 2008 Yearbook of Asian Affairs, 2008. Lao National Tourism Administration, 2007 Statistical Report on Tourism in Laos, 2008. , 2008 Statistical Report on Tourism in Laos, 2009. Sanum Investments Ltd., Executive Summary. Retrieved 17th December, 2009, from http://www.sanuminvestment.com-a.googlepages.com/Sanum_ Investments_Summary.pdf Savannakhet Province, Savannakhet Investment & Tourism Magazine, No.01,
116

The Socio-Economic Impacts of the Second Mekong International Bridge

October 2007. Suzuki, M., Keola, S., Perspective of Savan-Seno Special Economic Zone and Vientiane Industrial Park in the Lao PDR. In The Formation of Industrial Clusters in Asia and Regional Integration (Chapter 4). Retrieved 17th December, 2009, from http://www.ide.go.jp/Japanese/Publish/Download/Report/pdf/2008_ 0111_ch4.pdf World Bank, World Development Indicators 2009 Database. Retrieved 6th January, 2010.

117

About the Contributors (at the time of writing)

Richard Hummelbrunner Richard Hummelbrunner is a freelance consultant in the field of local and regional development. He has more than 30 years of professional experience both in his native Austria and internationally, particularly in Europe and SubSaharan Africa. His most recent specialization is in policy consultancy as well as evaluation and he has been active in promoting the use of systems concepts in these fields as a practitioner, trainer and author. Bob Williams Bob Williams is a freelance consultant, specializing in empirical field research and evaluation technique, systems theory and practice, organizational development and action research. He works with evaluators of various organizations in New Zealand, Australia, Asia, USA and Europe, and is internationally well-known for his contributions in promoting systems approaches and concepts in the evaluation field. Patricia Rogers Patricia Rogers is Professor of Public Sector Evaluation at RMIT University (Royal Melbourne Institute of Technology), Australia. For more than 25 years she has worked in public sector evaluation and research across a wide range of programs and levels of government. She was the inaugural Chair of the Systems in Evaluation Topical Interest Group of the American Evaluation Association and her particular field of interest is in evaluation methods for complicated and complex programs and policies. Nobuko Fujita Nobuko Fujita is the Deputy Director of the International Development Research Institute (IDRI) at Foundation for Advanced Studies on International Development (FASID), where she has been involved in evaluation-related research, training, and consulting. Eriko Takahashi Eriko Takahashi is a Junior Program Officer of the IDRI at FASID, where she has been involved in evaluation-related projects and developmental research.
119

You might also like