Environmental impact assessment in a development context

Introduction

The concept of environmental impact assessment (EIA), either by legislation or by administrative control, has been in evidence in the developing countries of Southeast Asia for a considerable number of years (USAID 1979). An OECD working group on environmental assessment and development assistance found that in many ways Southeast Asia was the most advanced region in the developing world regarding the establishment of, and experience with, environmental assessments (Evers 1985). In Indonesia, for example, Article 16 of the Law con-cerning "Basic Provisions for the Management of the Living Environment" (Act No. 4 of 1982) requires: "Every plan which is considered likely to have a significant impact on the environment must be accompanied with an analysis of environmental impact, carried out according to government regulation."

In developing countries--perhaps more so than in many developed countries-environmental impacts of development are clearly recognized as encompassing the social dimensions of change, not just the physical and biological dimensions. In reviewing several EIA case studies in the Philippines, Uriarte (1985) concluded, "The human environment, not the water, nor the air, nor the trees, is the core of any environmental impact assessment." This overstates the case somewhat, ignoring the social consequences of impacts on biophysical systems, though it does make clear a requirement of an anthropocentric emphasis in environmental impact evaluation. In Indonesia, Hardjasoemantri (1983) points out that since the community, especially in the rural areas, is tradition bound, impacts on the socio-cultural environment deserve special attention, and this is unequivocally stated within the Indonesian legislation. Social impact assessment (SIA) and environmental impact assessment are synonymous in the development context.

In the Southeast Asian region, the OECD working group (Evers 1985) found that insufficient manpower--skilled or unskilled was a bigger constraint faced by developing countries in assessing environmental impacts than lack of legislative or institutional frameworks. Within Indonesia, attempts to remedy this situation have included development of an internal training program in a network of Environmental Study Centres (PPLH 1982) and aid programs such as the Canadian Environmental Manpower Development in Indonesia Project (Hainsworth 1984). This paper stems largely from the experience of the author (and colleagues) in teaching a three-months training course on environmental impact assessment in Indonesia. A major component of the training course was to plan, implement, and report on environmental assessment of a flood mitigation proposal a dam at Depok on the Ciliwung River near Jakarta. The dam proposal, while relatively small, would displace some four thousand people from the inundated area, with a possible displacement of an additional eight thousand from the shoreline areas. This training EIA simulated an EIA that could be undertaken with local resources, with constraints on time, budget, and human resources that were by no means unrealistic. More to the point, it is highly likely that the report prepared by the training program was used later as the official EIA of the project--a telling comment on the scarcity of skilled professionals to undertake assessment work in developing countries. The intention of this paper is not to examine the particular project but instead to reflect on problems with doing and teaching environmental impact assessment in a developing country. In summary, these observations are:

• reliance on impact identification "methodologies" as the beginning and end of an environmental assessment inhibits effectiveness;

• moving from impact identification in general to the identification of specific tractable impacts, and then to specific modifiable impacts, is a critical but missing component of methodology (the terms "tractable" and "modifiable" impacts are explained below);

• nonexistence or nonavailability of data requires reliance on primary sources;

• moving "beyond EIA" to an interrelated environmental planning assessment and design is a prerequisite to effective reduction of adverse effects of projects and programs.

There is a need to differentiate between two situations in which impact assessment may be carded out in a developing country. In the first, a project or program is externally funded, and the donor agency has a requirement that the impact of the project be assessed. In this situation, responsibility for any or all of the impact assessment, and modification is likely to be the province of the donor agency. In the other, the project or program is the province of the host government, and any assessment has to be carried out predominantly with local resources. It is the latter that is of interest in this paper.

Impact Assessment "Methodologies"

Effective impact assessment is, at least, a four-stage procedure: identification of potential impacts, prediction of the magnitude of impacts, evaluation of the importance of impacts individually and perhaps collectively, and project or program modification to ameliorate the impacts. The literature on environmental impact assessment over the last decade has been voluminous, and particularly dominant in this material have been so-called methodologies for impact assessment (for example, Shopley and Fuggle 1984). One feature of such methodologies has been an emphasis on impact identification techniques (checklists, matrices, networks, etc.), with prediction of impacts, evaluation, and discussion on ways to reduce impacts often receiving scant attention. To the uninitiated, and sometimes the initiated too, this imbalance in the literature can generate the impression that these "methodologies" have substance. Presumably, it is meant to be understood that after impacts have been identified, predicting the impacts and evaluating their significance would readily follow. However, this literature has in the past generally failed to convey that the essence of impact assessment and the more difficult tasks are prediction, evaluation, and mitigation.

Structured procedures for identification of potential impacts such as checklists and matrices have much to commend them in a learning situation. Porter (1985) comments that the matrix undoubtedly has value in helping the inexperienced to identify issues that need to be addressed, but also points out that the system of applying numerical values to perceived impacts and their summation as encouraged by a matrix is inherently dangerous. For the experienced, structured impact identification procedures are unlikely to be necessary.

The problem is that a misplaced acceptance of these limited "impact assessment methodologies" has led to conclusions such as "activities in tin mining had lessened the quality of the environment by nineteen percent" (Zaenal Anwar Zen and Tan Malaka 1981). Assessments of this type cannot play a useful role in a reduction of adverse effects of a project or reinforcement of positive effects. Criticism has to be made of the application of an incomplete, if not inappropriate, technology of impact assessment. Environmental impact assessment "methodologies," concentrating almost solely on impact identification, are not standalone environmental evaluation and planning tools that can be directly adopted in developing countries----they require subsequent application of specialist prediction and evaluation skills in each area of identified impact. They are not effective stand-alone tools in the developed countries in which they originated either, but there, at least, impacts identified may be subject to public scrutiny, and the administrative, judicial and, political processes themselves may play significant roles in impact evaluation and project modification. This will rarely be the case in developing countries.

Relying on impact assessment "methodologies" as the beginning and end of impact assessment not only inhibits effective assessment but can be positively damaging. It gives the impression to all concerned the proponent, the decision maker, perhaps even the assessor that the environmental impacts of a proposal are being considered and that, by some ethereal process, they will also be mitigated. Of equal concern is that these procedures are seductively easy to teach, though there is likely to be but little reward from a large number of people trained in impact identification if skills in impact prediction, evaluation, and project modification to reduce impact are not also developed.

Tractable and Modifiable Impacts

In the training program in which we were involved the study attempted to proceed beyond impact identification. The difficulty arose, as in any large-scale and multidimensional project, that a large number and range of potential impacts could be identified, but, as always, time and resources for prediction of impact magnitude and evaluation were constrained. Thorough investigation of all identified impacts can never be undertaken; this was why the procedure of "scoping" environmental assessment studies has evolved over time. Scoping is the process of circumscription of the content and depth of the assessment at a very early stage of a study. In contrast to impact identification, there is little discussion in the literature on methodologies for this essential step.

Our own solution was pragmatic and two-staged. From the list of potential impacts, students were encouraged to identify, from amongst the impacts they thought were important, those for which they could collect data to provide baseline information, and for which they could provide reliable predictions of the magnitude of impact. In other words, they were asked to identify which of the problems of impact prediction were tractable within their resources. This scoping phase, while clearly a compromise, aims to ensure that study team resources and the final report focus directly on impact and not merely on description of the existing environment. It can be argued that those impacts that were not seen as tractable, but which could be important, may be overlooked in this approach. However, this is not the case. Instead, the procedure encourages clear reporting of uncertainty--"this impact has been identified as potentially important, but the study team is unable to provide any information on the magnitude of impact."

The second stage of this scoping exercise was even more pragmatic. Of the impacts identified as tractable, most study resources were concentrated on those that were perceived as potentially modifiable through changes to the original proposal. To give an example: a detailed analysis of the sensitivity of impacts to the design water-level height is particularly valuable in evaluation of the impacts of inundation of a dam project. Effects such as number of displaced persons and lost area of agricultural production may be altered dramatically through relatively small changes to dam height, and significant study resources were devoted to water-level-related impacts. By contrast, agricultural management practices in the catchment, being in the hands of many small and traditional landholders over a large area, were seen by participants as being unalterable in the short or medium term. Thus, dam water-quality issues related to agricultural practices in the catchment, while still recognized and reported as important, received far less attention from the study team.

This same pragmatism was also extended to no-build and alternative site options. While discussion of these options sparked considerable interest amongst participants, they decided not to devote resources to their investigation. Impact assessment should not discard consideration of these options lightly, but more often than not the dynamics of the development process may be such that assumptions of this type of flexibility may not be at all realistic.

This emphasis on impacts that are perceived as modifiable injects both political and technocratic elements into the scoping process. The political elements include judgments concerning the sensitivities of the decision maker and the strength and flexibility of the proponent. Documentation of the existing environment and objectively assessing all impacts of a proposal may seem a desirable scientific goal--but that goal is utopian in a development context where resources for evaluation are scarce. Scoping the assessment through early recognition of tractable and modifiable impacts can represent an efficient use of these scarce re-sources. The ultimate measure of efficiency has to be whether environmental assessments results in any changes to projects or programs, not whether the EIA provides an accurate scientific record of the altering environment.

Teaching strategies for environmental impact assessment must explicitly work through this process. Our experience was that attempting to scope a project in this way encouraged a participant's critical thought about the function and potential of the impact assessment procedure. It provided clearer directions as to where resources had to be allocated in data collection and analysis. Of most value was that it developed an appreciation of the need for "quantification" of the impacts. Students were asked to put themselves in the position of a decision maker who had already committed significant resources to the design of the project and to ask themselves, "What environmental information would convince me of the need to abandon this project or to make modifications to it?"

Availability of Data

Emphasis in the scoping stage of assessment on the availability of data for impact prediction takes on particular importance in a developing country. Participants in the training program were strongly encouraged to consider secondary sources of data in the first instance: maps and aerial photographs, government records, parallel studies conducted elsewhere, data collected for other purposes, and so on before embarking on any field measurement exercise. In general, they had very little success in accessing information. For example, aerial photographs-an essential tool for efficiently estimating impacts on people, agriculture, and infrastructure within the inundated area of the dam--were not available, at least not within the time constraints. In another example, rudimentary traffic volume records on roads to be inundated were not accessible. It is perhaps not so much that secondary data does not exist, but that there are institutional difficulties in gaining access to it and that the knowledge and skills of participants were not highly developed in overcoming these barriers.

In Indonesia, these concerns over information sources have been articulated by the Special Assistant, Minister of State for Population and the Environment (in Hainsworth 1984). His concerns were that: data presently available were very inadequate; there were many gaps and information was often dated and its reliability questionable; information that was available was often scattered across agencies and unorganized within agencies, and mechanisms for dissemination were inadequate; sharp divisions existed between sectors and departments, each of which may apply its own methods, protocols and organizational approaches, and data were rarely organized to facilitate decision making and administration.

At least for the time being, prediction of impacts in developing countries must rely heavily on primary sources of data. EIAs and EIA training programs need to take this likely paucity of data into account (Burdge 1990).

Moving beyond EIA

Environmental impact assessment is intended as a purposeful activity, the purpose being either to discourage or modify a project/program that may have unwanted environmental or social effects, or to reinforce a project/program that has positive environmental or social effects. However, to be purposeful in this way it is essential that the assessment process is in some way linked to relevant decisionmaking processes, and preferably to project planning and design processes too.

There is no single model for these linkages. They can vary from prescription by government legislation, confrontation in the courts (as a result of NEPA in the USA), consultation and intragovernmental consensus (as in Hong Kong : Ashcroft 1985), to financial persuasion by a superior level of government (as required by the US Department of Transportation Federal Aid Highway Program).

The difficulty experienced by participants in our training course was to determine what linkages existed between their assessment and any decision-making process on the project. Further, they found it difficult to determine whether the planning and design process was complete before they began their assessment, or whether there remained opportunities for project modification. It is suggested that most impact studies in developing countries would encounter similar difficulties. Environmental impact studies may tend to exist in a vacuum--performed by one set of professionals expert in their field but with little or no interaction with sectors responsible for the individual projects or decision making.

These conditions can undermine EIA as a purposeful activity and raise the question whether the Western model, with its separate study team and most likely a stand-alone report is, by itself, the appropriate means of environmental assessment in a development context. With scarce expertise and scarce resources for training more expertise, it could be that more immediate results would be obtained by following different models or combinations of models. Possibilities include the training of management and technical professionals from the existing development sectors in philosophies and techniques of environmental assessment, "cook books" in environmental assessment specifically designed for particular development sectors within government, and perhaps a small core of professionals set up to evaluate and advise on impact assessments carried out by individual departments. While these suggestions are not without their drawbacks, in particular removing what can be regarded in some circumstances as the useful adversary nature of EIA where protagonist and evaluator are within different organizations, they do attempt to graft EIA onto the existing planning and decision-making process--moving beyond EIA to take it out of the vacuum in which it appears to operate at the moment.

Conclusions

Attempting, albeit in a teaching program, to assess the environmental effects of a development project in Indonesia was beset with difficulties. Secondary sources of data that one would rely on in a developed country were not available. The stage that the planning and design process for the project had reached and the role that the environmental assessment might have in decisions concerning the project were not explicit. Further, a prevalent view that EIA methodology began and ended with the identification of potential impacts without comparable effort in prediction, evaluation, and impact mitigation inhibited effective assessment.

A more structured approach to project scoping is required, identifying those impacts that are tractable and those that are modifiable. Stand-alone impact assessments---a common model in developed countries--may be an inappropriate use of scarce skills in developing countries. Administrative models that can effectively graft the techniques and philosophies of impact assessment to the existing development and decision-making processes need to be explored.

Tor Hundloe and Geoff McDonald, co-workers on the EIA training program, provided helpful input to the ideas discussed in this paper.