UNIVERSIDAD NACIONAL DE LA PLATA

Ctedra de Gestin Ambiental (Ecologa)

ACTUALIZACIONES AL LIBRO DE CONSULTA PARA EVALUACIN AMBIENTAL DEL BANCO MUNDIAL

Material seleccionado por: Ing.Marcelo Gavio Novillo

La Plata Noviembre 2.000

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

April 1993 Number 1

The World Bank and Environmental Assessment: An Overview

The sustainability of economic development has become one of the most important challenges facing the World Bank in recent years. In response to growing awareness regarding environmentally sustainable development, the Bank is promoting a variety of initiatives ranging from environmental assessment (EA) for specific projects or sectors to national environmental strategies and action plans. This first issue of EA SOURCEBOOK UPDATE provides a brief review of Bank efforts to integrate environmental concerns into the mainstream of its lending activities during the last two decades. It also gives a summary of the main findings and conclusions of the first EA Annual Review, and institutional responsibilities for implementing the EA process in Bank operations. This Update belongs to Chapter 1 in the Update Binder and does not replace any section of the EA Sourcebook.

Background A brief review of policy statements and directives during the past decade reveals the World Banks increasing efforts to broaden its traditional approach to project development. Environmental concerns first became an explicit part of Bank activities when the position of Environmental Advisor was established in 1970. The Bank played an active role in this area by becoming the first multilateral development agency to screen projects for their environmental consequences and to adopt environmental guidelines for the evaluation of future lending operations. One of the first significant policy statements issued was Operational Manual Statement 2.36: Environmental Aspects of Bank Work, in May 1984, requiring that environmental considerations be introduced at the time of project identification and preparation, and recognizing that modification could also occur at the time of appraisal, negotiations, and implementation. For all projects, staff were instructed to use prudent judgement when assessing environmental effects that were potentially irreversible (for example, the extinc-

tion of species or ecosystems). The new policy further stipulated that projects with severe environmental impacts would not be financed without mitigatory measures acceptable to the Bank (para. 9[b]). By the mid-1980s, the Bank was financing projects containing environmental components, including several free-standing environmental projects, which had specific environmental objectives, such as reforestation, pollution control, and water resource management. Although these measures were designed to help both borrowers and the Bank improve the environmental quality of projects, problems remained. In developing countries, serious environmental degradation accelerated and began to constrain economic development. It became apparent to Bank management that the degree of effort devoted to environmental issues and the approaches actually used were insufficient to ensure full consideration of adverse environmental impacts during project identification, design, and implementation. This, combined with a few well-publicized cases in which Bank-financed projects were found to have

negative environmental consequences such as contributing to the destruction of tropical rain forests and posing threats to wildlife, indigenous people, and established human settlements prompted the institution to adjust its policies toward environmental management and to systematically bring environmental issues into the mainstream of its lending activities. In 1987, the Bank implemented a series of structural changes that included strengthening of environmental policies, procedures, and staff resources. A central Environment Department and Environment Divisions in each of the four regional Technical Departments were created, leading the way to a significant expansion of Bank staff assigned to execute, monitor, and support environment-related activities. The new Environment Department would help set the direction of Bank environmental policy, planning, and research and take the lead in developing strategies to integrate environmental concerns into Bank lending and policy work. The regional divisions would review and give clearance to Bank-financed projects and provide technical support on individual projects, sector programs, and country or regional strategies, thus helping borrowing countries to design and implement more sustainable approaches to development. EA Guidelines for the World Bank Since the 1987 reorganization, the Bank has moved decisively in institutional strengthening and policy reform. A notable achievement was the introduction of an Operational Directive on Environmental Assessment (OD 4.00, Annex A) in October 1989. This comprehensive and detailed new policy mandated an environmental assessment for all projects that may have significant impacts on the environment. Bank staff were required to screen and categorize all prospective loans (Category A through D) for potential adverse environmental problems at the time of project identification. Although the borrower had the responsibility for carrying out the environmental assessment, the Bank would play a review and follow-up role to ensure that the project had been adequately prepared; for example, assisting the borrower in preparing the terms of reference (TOR) for the environmental assessment. With careful planning, generally in the form of a mitigation plan, adverse environmental effects could be avoided or mitigated. The directive recommended regional and sectoral assessments to reduce the work subsequently needed on specific investment projects. It also required that the borrower undertake appropriate consultation with and disclosure of information to affected groups and local nongovernmental organizations. After two years of Bank experience with environmental assessments, the operational directive was revised to broaden its scope and applicability. The

Bank recognized that projects aimed at achieving environmental objectives could sometimes have negative and unanticipated effects. In October 1991, OD 4.00, Annex A was replaced by OD 4.01 (see summary of revisions in Box 1). The revised directive, which incorporated the guidelines contained in OD 4.00, Annex A, introduced a new system for classifying projects according to the nature and extent of their environmental impact. It provided more specific instructions to staff on public consultation and disclosure of information. It also recommended that the Environmental Assessment Sourcebook, published in the World Banks Technical Papers Series during fiscal 1991-92, be used for guidance throughout the environmental assessment process. First Environmental Assessment Annual Review The first annual EA review, initiated in late 1991, covered the period from October 1989 to October 1991 and extended through the end of fiscal 1992. It was prepared in close collaboration with the Regional Environment Divisions and reviewed by the Environmental Assessment Steering Committee, as well as staff throughout the Bank. The main findings and conclusions were based on interviews with Bank regional environmental staff and selected task managers, a survey of twenty projects for which environmental assessments had been completed and evaluated by the end of fiscal 1992 under the requirements of OD 4.00, Annex A, and detailed case studies of seven of these projects. The Banks EA procedures were proven to be realistic, workable, and instrumental in helping to improve development planning and environmental management. Even in countries with minimal or no prior experience, the EA process was a valuable tool for identifying project problems and the means to avoid or mitigate them. Although the review documented considerable progress, it also revealed that the new policy required additional human and financial resources for Bank EA work. The review identified the following areas where Bank efforts should be continued and/or strengthened to make the EA process more effective: Borrower capacity building, including technical assistance, institutional support and in-country training should be given greater attention to help ensure that environmental assessment becomes an effective planning tool during project preparation and that the EA report presented to the Bank adequately covers all relevant issues and are of acceptable quality. Borrower requirements should be explicitly taken into account in defining the EA preparation schedule to ensure that the EA process is initiated at an early enough stage to accommodate in-country review procedures.

Box 1. Operational Directive 4.01: Environmental Assessment

The revised OD 4.01: Environmental Assessment includes the following principal changes from O.D. 4.00, Annex A: Screening. Projects with multiple components are to be classified according to the component with the most significant adverse impact, i.e., if there is an A component, the entire project is A (para. 17 and Annex E). The D Category has been eliminated as well as combined categories (A/D, or B/D) to reduce screening differences. Consultation. Consultation with affected groups and NGOs should include: (i) making available to the groups being consulted a summary description of the project, its objectives and potential adverse impacts, shortly after assignment of the EA category; (ii) providing a summary of the conclusions of the draft EA report, in a form and language that are meaningful to the groups being consulted; and (iii) taking consulted groups views fully into account in the design of the EA study and the project, as appropriate (paras. 19-21). Release of the Environmental Assessment to the Executive Directors (EDs). The Bank requests the borrowers advance permission to release the EA report to the Executive Directors (EDs) because the report is the borrowers property. On receipt of a copy of the EA report from the borrower, an English-language summary is made available to the EDs, and a copy of the report is deposited in the project file (para. 22). Environmental Mitigation Plan. Annex C outlines items to be covered in the environmental mitigation plan and incorporated into the project. Global Environment Facility. GEF projects and components are subject to EA requirements under OD 4.01 (para. 1 and footnote 1). Environmental Advisory Panels. The borrower should normally engage advisory panels of environmental specialists for major, risky or contentious projects; this section tightens the original OD 4.00, Annex A, which recommended that establishment of such panels for projects with major environmental concerns be explored with Government (para. 13). Relation Between Preparers of EA and Feasibility Study. In the case of projects with potentially major adverse impacts, the preparer of the EA should not be affiliated with the project, but should liaise closely with project preparation/feasibility teams (para. 16). Grandfathering. OD 4.01 applies to projects with IEPS issued after October 1, 1991

EA training within the Bank should continue, as should the use of case studies for this purpose. However, training should become more sectorspecific. Bank case studies should be developed to provide greater orientation to staff on methods of local consultation and regional and sectoral environmental assessments. Additional guidance on project screening should be provided to Bank task managers to ensure that classification for purposes of environmental assessment is consistent across regions. Field visits, including consultation with affected populations and local NGOs, should be a consistent part of scoping activities for projects requiring a full EA. The Checklist of Potential Issues for an Environmental Assessment (OD 4.00, Annex A-2) should be used for Terms of Reference (TOR) preparation for EA work and associated scoping so that key aspects and possible impacts are not missed. The potential benefits of regional and sectoral EAs should be more widely disseminated to borrowers and Bank staff and utilized to a greater extent.

Work is currently underway to follow up on these recommendations, particularly in the area of EA training. Institutional EA Responsibilities Country Departments in the operational vice presidencies, working closely with the Regional Environment Divisions, are directly responsible for overseeing and appraising project preparation. The project task manager, in consultation with the Regional Environment Divisions, is responsible for environmental screening, preparation of terms of reference for EA work when an environmental analysis is required, monitoring of the EA process, and review of the report in draft and final form. A summary of the findings is prepared as an annex to the project staff appraisal report. The Regional Environment Divisions oversee the work in their respective regions and offer technical expertise and policy advice to TMs and borrower governments. The Environment Department works closely with operational vice presidencies, particularly the Regional Environment Divisions, to ensure that the environmental assessment process is effectively integrated into Bank activities. To this end, the role of the Department, now housed in the Environmentally Sustainable

Development Vice Presidency (ESDVP), is to give clearer focus to analytical, policy, and operational support to activities related to brown (e.g., industrial pollution and urban infrastructure) as well as green (e.g., agriculture, forest management and biodiversity) issues in its work program. Recently, the Department has also become a focal point for addressing social and resettlement issues in Bank lending operations. The green unit of the Environment Department, the Land, Water and Natural Habitats Division (ENVLW), oversees and coordinates the Banks work on environmental assessment. The Division provides operational support for analytical work, project preparation, and supervision on issues related to biodiversity, forestry, coastal zone management, and environmental information management. The Divisions EA team systematically monitors the Banks follow-up of the EA Operational Directive and is responsible for preparing the EA Annual Review, which is submitted to the President. The team is also responsible for continuously updating the EA Sourcebook as a living reference manual for Bank staff and external users through the series of EA Sourcebook Updates. Finally, the team offers project-specific operational support on EA-related issues, especially during the preparation and supervision stages of the project cycle. Special attention is given to sectoral and re-

gional environmental assessments, and environmental aspects of privatization activities, financial intermediary and adjustment operations. The brown unit, the Pollution and Environmental Economics Division (ENVPE), provides technical and economic analytical support with respect to issues relating to land and water pollution, atmospheric emissions and air quality, and global climate change. It is also responsible for work on the economic valuation of environmental damage and the impacts of economic policies on the environment. The Division offers guidance to Bank staff on how to internalize environmental costs and benefits in economic analysis, which is particularly important when considering alternative project options. The Social Policy and Resettlement Division (ENVSP) addresses the social dimensions of sustainable development, particularly the social organization of resource management. It provides leadership in social impact assessment, and support work on the social strategies in project design and implementation. The division supports the EA process by improving social analysis in specific areas of Bank activity, e.g., involuntary resettlement, treatment and rights of indigenous peoples and other ethnic minorities, public participation and consultation, NGO intermediation and institutional capacity building.

EA SOURCEBOOK UPDATE is designed to provide the most up-to-date information on the Banks policy and procedures for conducting environmental assessments of proposed projects. This publication should be used as a supplement to the Environmental Assessment Sourcebook, which provides guidance on the subjects covered in Operational Directive 4.01. Please address comments and inquiries to Olav Kjrven, Managing Editor, EA Sourcebook Update, ENVLW, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. S-5123, (202) 473- 1297.

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

April 1993 Number 2

Environmental Screening
Environmental screening is intended to ensure that proposed projects are subject to the appropriate extent and type of environmental assessment (EA). This EA Sourcebook Update provides guidance to Bank staff on environmental screening, based on the Banks EA policy and evolving EA experience. This Update replaces the screening section in chapter 1 of the Environmental Assessment Sourcebook (pp. 4-5).

Requirements for screening The World Banks EA process generally begins with screening at the time of project identification. At this stage, the project is classified into one of three categories prior to issuance of the Project Concept Document. The chosen category signals the appropriate level of EA required. Environmental screening also helps determine the choice of EA instruments, depending on the needs of the project. Determining the level of EA The Bank uses the following three categories to signal the appropriate level of EA for any given project: Category A: A full EA is needed in accordance with the specific requirements of the Banks EA policy and procedure for Category A projects, including in areas such as public disclosure, public consultation, and the timing for submitting the EA report to the Banksee Operational Directive(OD) 4.01, soon to be reissued as Operational Policy/Bank Procedure/ Good Practice (OP/BP/GP) 4.01. Category B: EA is required, but its scope corresponds to the limited environmental impacts of the project (again, the Banks EA policy and procedure provide specific guidance). Category C: No EA is required. The selection of the category should be based on professional judgment and information available at the time of project identification. If the project is modified or new information becomes available, Bank EA policy permits the TM to reclassify a project with the concurrence of the RED. For example, a Category B project might become Category A if new information reveals that it may have diverse and
Reissued with revisions: November 1996

significant environmental impacts when they were originally thought to be limited to one aspect of the environment. Conversely, a Category A project might be reclassified as B if a component with significant impacts is dropped or altered. The option to reclassify projects relieves some of the pressure to make the initial decision the correct and final one. However, reclassification is not free of cost. For example, if a Category B project is later changed to Category A, additional resources will be required for environmental studies, public consultation, and report preparation. The schedule for project preparation will almost certainly be adversely affected. Selecting appropriate EA instruments In parallel with determining the appropriate project classification, the screening process should also identify the types of EA instruments that may be suitable for the project. Box 1 lists different project categories and EA instruments that are often useful in conjunction with such projects. This list is illustrative and the final decision on the use of these instruments has to involve the borrower. Opportunities for undertaking strategic forms of EA (sectoral or regional EA) should be considered earlyif possible before the identification of concrete projects in order to optimize the influence of the EA process on the selection and design of projects (Updates nos. 4 and 15 provide guidance on sectoral and regional EA). Criteria for making the classification decision Initially, it is useful to look at key words in OP 4.01 (forthcoming) to describe each screening category. Projects are
Insert into Update Binder chapter 1

Box 1. Project categories and corresponding EA instruments

This list provides an overview of EA instruments that are typically relevant for different categories of projects. For any one project, the choice of instrument(s) should be tailored to the environmental issues at hand. For some situations, the use of just one instrument is appropriate, other times it makes most sense to combine two or more instruments (for example EIA and risk assessment) or to sequence them (for example, a sectoral EA followed by a number of EIAs for subprojects). Investment projects Sector adjustment loans Sector investment and time-slice operations Urban development program/project Rural development program/project Riverbasin or watershed program/project Rehabilitation and maintenance projects Industrial expansion projects Privatization projects Financial intermediary loans Social investment funds EIA, risk assessment Sectoral EA Sectoral EA, EIA (for subprojects) Regional EA, EIA (for subprojects) Regional EA, EIA (for subprojects) Regional EA, EIA (for subprojects) Environmental audit, hazard/risk assessment EIA, environmental audit, hazard/risk assessment Environmental audit, hazard/risk assessment EIA, environmental audit, hazard/risk assessment, checklists, screening and review criteria EIA, checklists, screening and review criteria

classified into Category A if they are likely to have significant adverse impacts that are sensitive, diverse, or unprecedented, or that affect an area broader than the sites or facilities subject to physical works. The impacts of Category B projects are site-specific in nature and do not significantly affect human populations or alter environmentally important areas, including wetlands, native forests, grasslands, and other major natural habitats. Few if any of the impacts are irreversible, and in most cases mitigatory measures can be designed more readily than for Category A projects. In order for a project to classified into Category C, it must be considered likely to have no adverse impacts at all, or the impacts would be negligible. In practice, the significance of impacts, and the selection of screening category accordingly, depends on the type and scale of the project, the location and sensitivity of environmental issues, and the nature and magnitude of the potential impacts. These dimensions are discussed below. Project type and scale Boxes 24 provide illustrative lists of projects assigned to each of the three categories based upon prior Bank and international experience. These lists provide a good starting point and framework for the screening decision. However, because of other factors involved such as project siting, the nature of impacts, and the need for the EA process to be flexible enough to accommodate them, the lists should not be used as the sole basis for screening. Category A. In general, certain types of projects such as those listed in box 2, are likely to have adverse impacts of a Category A nature. Category A includes projects which have one or more of the following attributes that make the potential impacts significant: direct pollutant discharges that are large enough to cause degradation of air, water or soil; large-scale physical disturbance of the site and/or surroundings; extraction, consumption, or conversion of
2

substantial amounts of forest and other natural resources; measurable modification of hydrologic cycle; hazardous materials in more than incidental quantities; and involuntary displacement of people and other significant social disturbances. Category B. Projects in Category B often differ from A projects of the same type only in scale (see box 3). Large irrigation and drainage projects are usually Category A; however, small-scale projects of the same type may fall into Category B. Similarly, a 50-meter hydroelectric dam is clearly large in scale and will usually require Category A classification, while low-head power dams may be Category B. Construction of a 50-km expressway would also require Category A due to scale, while rural road rehabilitation will tend to raise only minor environmental issues (Category B). Projects entailing rehabilitation, maintenance or upgrading rather than new construction will usually be in Category B. A project with any of these characteristics may have impacts, but they are less likely to be significant. However, each case must be judged on its own merits. Many rehabilitation, maintenance and upgrading projectsas well as privatization projectsmay require attention to existing environmental problems at the site rather than potential new impacts. Therefore, an environmental audit may be more useful than an impact assessment in fulfilling the EA needs for such projects (see Update no. 11: Environmental Auditing). Category C. Typical Category C projects are listed in box 4. However, before classifying a project in this category it is important to consider potential issues, some of which may not immediately spring to mind. For example, disposal of medical wastes may be an issue in many health projects. Likewise, while most technical assistance (TA) projects should fall into Category C since

Box 2. Category A projects

The projects or components included in this list are likely to have adverse impacts that normally warrant classification in Category A. Aquaculture and mariculture (large-scale) Dams and reservoirs Forestry production projects Hazardous waste management and disposal Industrial plants (large-scale) and industrial estates, including major expansion, rehabilitation, or modification Irrigation, drainage, and flood control (large-scale) Land clearance and leveling Manufacture, transportation, and use of pesticides or other hazardous and/or toxic materials Mineral development (including oil and gas) New construction or major upgrading of highways or rural roads Port and harbor development Reclamation and new land development Resettlement River basin development Thermal power and hydropower development or expansion Water supply and wasterwater collection, treatment and disposal projects (large-scale)

Box 3. Category B projects

The following projects and components may have environmental impacts for which more limited EA is appropriate. Agro-industries (small-scale) Electrical transmission Energy efficiency and energy conservation Irrigation and drainage (small-scale) Protected areas and biodiversity conservation Rehabilitation or maintenance of highways or rural roads Rehabilitation or modification of existing industrial facilities (small-scale) Renewable energy (other than hydroelectric dams) Rural electrification Rural water supply and sanitation Tourism Watershed projects (management or rehabilitation)

along watercourses, in aquifer recharge areas or in reservoir catchments used for potable water supply; and on lands or waters containing valuable resources (such as fisheries, minerals, medicinal plants, prime agricultural soils). Experience to date shows that precise identification of the projects geographical setting at the screening stage greatly enhances the quality of the screening decision and helps focus the EA on the important environmental issues. A map of the project area that includes key environmental features (including cultural heritage sites) is invaluable for this purpose. Information on the project setting may be available to the TM from the RED, from colleagues in country departments, or in-country environmental profiles or Bank reports on other projects in the vicinity. Local institutions and NGOs are also valuable sources. In the absence of any such information, the TM should consider sending a reconnaissance mission to provide the basis for proper screening. Often a product of this mission is a draft of the Terms of Reference (TOR) for the EA. Sensitivity of issues Environmental issues that are sensitive within the Bank or the borrowing country require special attention during the EA process. The project may involve activities or environmental features that are always of particular concern to the Bank as well as to many borrowers. These issues may include (but are not limited to) disturbance of tropical forests, conversion of wetlands, potential adverse effects on protected areas or sites, encroachment on lands or rights of indigenous peoples or other vulnerable minorities, involuntary resettlement, impacts on international waterways and other transboundary issues, and toxic waste disposal. The best way to ensure proper treatment of such issues is to classify the project as Category A, so
3

they involve no physical works, certain TA operations are designed to pave the way for major investments or privatization (often in a particular sector). In such cases, it is appropriate to undertake a limited review of the environmental institutional and regulatory framework for the sector and recommend improvements (as needed). Category B is normally the correct classification for such projects. Project location The selection of a screening category often depends substantially on the project setting, while the significance of potential impacts is partly a function of the natural and sociocultural surroundings. There are a number of locations which should cause the TM to seriously consider an A classification: In or near sensitive and valuable ecosystems wetlands, wildlands, coral reefs and habitat of endangered species; in or near areas with archaeological and/or historical sites or existing cultural and social institutions; in densely populated areas, where resettlement may be required or potential pollution impacts and other disturbances may significantly affect communities; in regions subject to heavy development activities or where there are conflicts in natural resource allocation;

Box 4. Category C projects

These projects are likely to have negligible or no environmental impacts. EA is normally not required. Education Family planning Health Institution development Most human resources projects Nutrition

tions, while a decrease of 3.0 mg/l will in many circumstances. The effect of either decrease on the aquatic ecosystem will be different depending on its duration and frequency of occurrencecontinuous or permanent, seasonal, intermittent or accidental. Where it is possible to assign probabilities to potential impacts, which often cannot be done without detailed analysis, the risk of occurrence becomes an aspect of magnitude. One of the requirements of a full EA is that other current and proposed development activities within the project area and more spontaneous activities spurred by a project (such as migration of people into an area opened up by a road project) must be taken into account. Such cumulative or induced impact may sometimes be the primary determinant of the appropriate level of EA. Screening of operations with multiple subprojects Financial intermediary, sector investment, and social investment fund operations present unique problems during screening because the details of the subloans or subprojects are usually not known at the time of project identification. One of the TMs responsibilities is to see that the loan includes a mechanism for conducting environmental screening of subprojects and, as appropriate, carrying out and reviewing EAs. However, the entire loan must also be assigned to a category for the purpose of preparation and appraisal. If it becomes evident that one or more subprojects will require full EA, the entire loan should be classified as Category A (see Chapter 6 of Volume I, Environmental Assessment Sourcebook). When screening a sector investment loan, the need and opportunity for undertaking a sectoral EA should be considered (see Update no. 4: Sectoral Environmental Assessment). Future Updates will discuss EA for financial intermediary lending and social investment funds in more detail. Outputs of screening The screening results are recorded and explained in the Project Concept Document and the Environmental Data Sheet. The Monthly Operational Summary also records the screening decision. The Bank reviews the results with the borrower, especially with regard to the type of EA instruments required, the general scope of the EA, public disclosure and consultation requirements, schedule, and implementation arrangements. As soon as possible after screening, the borrower should prepare the Terms of Reference (TORs) for any EA required. The Bank assists as necessary in preparing the TORs and always reviews their contents.

that the level of effort will be adequate in terms of analytical expertise, decision-making, interagency coordination, and public involvement. Nature of impacts It is difficult to describe the nature of impacts without having some overlap with the concepts of sensitivity and project type. The TM should take into consideration the following examples of impacts that warrant Category A attention: Irreversible destruction or degradation of natural habitat and loss of biodiversity or environmental services provided by a natural system; risk to human health or safety (for example, from generation, storage or disposal of hazardous wastes, or violation of ambient air quality standards); and absence of effective mitigatory or compensatory measures. Magnitude of impacts There are a number of ways in which magnitude can be measured, such as the absolute amount of a resource or ecosystem affected, the amount affected relative to the existing stock of the resource or ecosystem, the intensity of the impact and its timing and duration. In addition, the probability of occurrence for a specific impact and the cumulative impact of the proposed action and other planned or ongoing actions may need to be considered. For example, the resettlement of 5,000 families is a large impact, in absolute terms. Conversion of 50 hectares of wetland, however, differs markedly in significance depending on its size relative to the total area of wetlands in the country or region. An average decrease in dissolved oxygen concentration of 0.05 mg/l in a receiving water is unlikely to have serious biological or chemical implica-

This Update was prepared by Thomas E. Walton and Olav Kjrven. The EA Sourcebook Updates provide up-to-date guidance for conducting EAs of proposed projects and should be used as a supplement to the Environmental Assessment Sourcebook. The Bank is thankful to the Government of Norway for financing the production of the Updates. Please address comments and inquiries to Olav Kjrven and Aidan Davy, Managing Editors, EA Sourcebook Update, ENVLW, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. S-5139, (202) 473-1297. E-mail: eaupdates@worldbank.org.

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

June 1996 Number 14

Environmental Performance Monitoring and Supervision

One of the purposes of Environmental Assessment (EA) is to minimize potentially adverse impacts and enhance the overall quality of the project. The effectiveness of an EA in meeting these objectives is determined during project implementation through performance monitoring by the borrower and Bank supervision. Environmental performance monitoring should be directed towards measuring and evaluating changes brought about by projects and assessing the effectiveness of proposed mitigation measures. Bank environmental supervision activities ensure diligence of Borrowers in implementing mitigation measures. A recent review of the Banks experience with EA concluded that arrangements for monitoring and supervision could be made more effective. The 1992 Wapenhans Task Force Report Effective Implementation: Key to Development Impact emphasized the importance of managing the quality of projects under implementation. This resulted in a Bank-wide effort to develop practical performance monitoring indicators. The Environment Department recently issued a note on Environmental Performance Indicators (EPIs) which complements this Update. While the EPI note presents examples of measurable indicators to monitor the environmental impact of projects, this Update discusses the process of implementing effective environmental performance monitoring and supervision of Bank projects. Definitions and objectives Environmental performance monitoring may be defined as technical and institutional activities that are implemented by a Borrower to measure and evaluate environmental (including health and socioeconomic) changes induced by a project. The overall objective of performance monitoring is to identify predicted and unanticipated changes to the physical, biological and social environment brought about by the project. This requires baseline information on predevelopment environmental and social conditions, against which development and post-development impacts and mitigation measures can be measured and compared. Deviations from the baseline beyond predetermined limits should trigger corrective actions. In this respect monitoring is a dynamic activity as opposed to passive collection of data. During the construction phase, for example, monitoring may involve checking the performance of contractors or government institutions against commitments expressed in formal documents, such as contract specifications or loan agreements. During the post-construction phase, it may involve measuring physical, biological or social parameters against required limits (e.g. measuring air and water discharges against Borrower country or World Bank guidelines). As the degree of uncertainty in impact prediction or effectiveness of mitigation measures increases, monitoring becomes more critical. Performance monitoring is therefore concerned with the immediate outcome and longer term impact of development projects (whether positive or negative). The EPI note provides examples of measurable indicators drawn from different environmental sectors. Supervision, in contrast, relates to the implementation process. A well-designed and executed monitoring program ensures that information is provided in the correct form and at the right time to trigger the appropriate supervisory response. Environmental spervision may be defined as any Bank activity directed towards ensuring that Borrowers implement projects responsibly, regarding

Box 1.

Environmental Management Plan

Environmental Management Plans (EMPs) outline the measures to be taken during project implementation and operation to control adverse environmental impacts and the actions needed to implement these measures. Such plans are essential elements of Category A projects; for many Category B projects, mitigation plans alone suffice. A mitigation or management plan should include the following items: Identification and summary of all anticipated adverse environmental impacts; Description of each mitigation measure, including the type of impact to which it relates and the conditions under which it is required (for example, continuously or in the event of contingencies), together with designs, equipment descriptions, and operating procedures, as appropriate; Description of the elements of the monitoring program (as outlined in the main text); Institutional arrangements (responsibilities for mitigation and monitoring), which may include recommendations for strengthening existing institutional capacity; Implementation schedule for measures that must be carried out as part of the project, showing links with overall project implementation plans; Monitoring and reporting procedures that are designed to ensure early detection of conditions that necessitate corrective actions, and provide information on the progress and results of mitigation and institutional strengthening measures; and Cost estimates and sources of funds for both the initial investment and recurring expenses for implementing the EMP, integrated into the total project costs.

the environmental requirements (of which performance monitoring is but one element) within an Environmental Management Plan (EMP) (see box 1). EMPs should demonstrate that proposed monitoring activities will encompass all major impacts and will be integrated into project supervision. The Task Manager should ensure that the Borrower drafts Terms of Reference (TORs) for the EA which adequately reflect the importance of preparing a comprehensive EMP. Developing a performance monitoring program An effective environmental performance monitoring program should consist of the following elements: Monitoring objectives; description of performance indicators which provide linkages to impacts and mitigation measures identified in the EA; description of parameters to be measured, methods to be employed, sampling locations, frequency of measurements, detection limits (where appropriate) and definition of thresholds that will signal the need for remedial actions; institutional responsibilities, timing and timescales for monitoring; reporting arrangements (to the regulatory authorities and the Bank); and costs and financing provisions. Monitoring objectives must clearly spell out the questions to be answered by measurement activities. By way of illustration, suppose particulate emissions from a proposed project are a concern. If the monitoring objectives are simply to determine whether these emissions will cause a public nuisance, citizen complaints would be a suitable indicator. If the objectives are to ensure that respiratory risks attributed to particulate matter are reduced, an ambient monitoring program for particulate materials with a diameter of less than 10 microns (PM 10) would be appropriate. However, if the objective is to control the health risks from toxic constituents, a more extensive monitoring program focusing on the fate, transport and health effects of these constituents might be necessary. Once the monitoring objectives have been established, both the immediate outcome of the project (for example, reduced particulate emissions from the smoke stack) and the longer term environmental impact (ambient concentrations of particulates) should be monitored. An example of the use of measurable indicators in the Lithuania Siauliai Environment Project is provided in Annex 2 of the EPI note. The monitoring program provided in the EMP should clearly show the linkages between specific impacts identified in the EA and indicators to be measured. These linkages should be demonstrated in

agreed environmental safeguards and the need to address unanticipated environmental problems. This typically involves visiting project sites and meeting with Borrower representatives, and reviewing environmental monitoring reports and correspondence with the Borrower. It may also involve a sectoral or regional implementation review covering several projects. Where environmental or socioeconomic problems arise during project implementation, the Bank works with the Borrower to resolve them. This may require corrective actions by the Borrower, such as modifying the design of the project or mitigation program. Project preparation and appraisal considerations Opportunities for preventing, minimizing, mitigating or compensating for adverse impacts can only be realized if they are integrated into overall project implementation. This is supported by clearly defining
2

a way that can be readily understood by decisionmakers. It is not essential to have complete details of monitoring in the EA Report. In some cases further collection of baseline data may be necessary. Nevertheless, the EMP should describe the process through which final monitoring arrangements will be agreed. Associated costs, funding and institutional needs (training, legislative or regulatory) required to complete the plan should be identified. The costs (including personnel, sampling and analytical charges) are integral to the project, and therefore will need to be factored into loan negotiations. Verification of mitigation measures prior to final approval of the EMP is important. A formal environmental sign-off should cover both detailed mitigation and monitoring proposals (as well as associated management and training) prior to loan negotiations. It is necessary to ensure that appropriate skills are available to design and verify the mitigation measures agreed during project appraisal. Similarly, commitments to environmental protection during construction need to be checked against bid specifications and tenders. Preparing a supervision plan In preparing a supervision plan for a project (as required by OD 13.05 on Project Supervision) the Task Manager should consider : the magnitude and significance of the projects potential environmental impacts; uncertainties associated with the predicted environmental effects; past experience of similar projects within the sector, region or with the same implementing agency; institutional complexities and capacities of the various agencies that may be involved in project monitoring and management; and availability of specialist skills within the Bank, implementing agency, Non-governmental organizations (NGOs) or consultants. Projects with potentially large, significant and uncertain environmental impacts will normally require more intensive supervision. This is especially valid where past experience shows an unwillingness or inability of the implementing agency to monitor or undertake corrective actions effectively (see box 2). Supervision is based in part on project conditionality. Project legal documentation (which encompasses the loan, credit or grant agreements per se, subsidiary agreements, bidding documents and related contracts) provide much of the framework to support and enforce supervision. It is therefore critical that such documents adequately reflect the project

Box 2.

Benefits of intense supervision

The Category A Yacyreta II Hydro-electric project on the Rio Parana between Argentina and Paraguay has been closely supervised due to the scale and sensitivity of its potential social and environmental impacts. A Bank team was assembled to supervise the project and provide technical assistance to the Borrower on environmental aspects. While supervision on this scale is atypical, the Yacyreta experience demonstrates the benefits of intense supervision for complex and sensitive projects. Critical elements of the Resettlement and Environmental Management Plan (REMP) include water quality monitoring, establishment of compensatory protected areas and wildlife rescue, archaeological salvage, and housing and economic rehabilitation for resettled families. By providing technical assistance and emphasizing the importance of the REMP, the Bank helped ensure compliance with REMP requirements prior to reservoir filling. A Bank mid-term review recommended institutional improvements prior to the second phase of reservoir construction and filling. It also highlighted the importance of intensive supervision in ensuring compliance with the REMP. Reliance on intensive Bank supervisory presence may be essential where institutional capacity or political commitment to managing environmental and social aspects is insufficient or questionable.

environmental requirements and implementation mechanisms (see box 3). It is good practice to ensure that addressing major environmental components be linked to disbursement conditions. Two vital links with the measures contained in the EMP are the Project Implementation Plan (PIP) (see Bank Procedure 10.00) and the environmental covenants agreed to by the Borrower as a condition for receiving the loan. The Task Manager and Borrower should agree on the most important environmental performance (and other) indicators specified in the PIP. These help to reinforce project supervision. To reinforce the legal documentation it is recommended that a summary schedule of performance monitoring be included as an annex to the environmental covenants. The annex should be as precise as possible. For Category A projects, the Task Manager should ensure annual participation of an environmental and/or social sciences specialist (depending on the importance of environmental or social issues) in project supervision. For category B projects,
3

Box 3. Translating EA recommendations into project documentation

EA requirements are not always well-reflected in Staff Appraisal Reports (SARs) and other project documents. Implementation schedules and budgets for mitigation and monitoring measures are often ill-defined. The Second Ertan Hydroelectric Project in China provides an excellent example of effective incorporation of EAderived mitigation and monitoring measures into the SAR, in the form of a detailed chronological schedule. The environmental mitigation and monitoring plan summary defines the agencies and individuals responsible for the programs. It includes details on coordination of sub-components, agencies contracted for programs, agency representatives, the supervisory agency and the supervisor. The summary plan also outlines the timing of program sub-components, the location of studies to be carried out, the data collection methods and type of data analysis. In addition, budgetary provisions for the seven year monitoring period are outlined. The detailed environmental measures defined in this SAR institutionalize the projects environmental aspects within the design and implementation plans, thereby providing a sound basis for monitoring and supervision. One aspect rarely addressed in detail in project SARs is inclusion of environmental requirements in bidding and construction contracts. During project preparation, environmental and social provisions should be reflected in bidding documents for major project components. These might include restrictions on location or restoration of borrow areas and construction camps. Where construction camps are essential, issues to be addressed should include recruitment of local labor, controlled access to natural resources (such as fuelwood), healthcare provisions, and treatment or disposal of solid and effluent wastes (primarily domestic sewage).

Box 4. Evaluating implementation of a monitoring plan

The Guangdong and Henan Provincial Highway projects in China included construction of four lane highways in Guangdong, a two lane expressway in Henan and improvement programs for a few thousand kilometers of existing roads. Monitoring requirements for the new highway components were specified in the EMPs of the project EAs and fully reflected in the implementation and budgetary arrangements specified in the Staff Appraisal Reports (SARs). Submission of EMPs for road improvement sub-projects were a condition of disbursement. The Bank also required that each local project supervision team had at least one environmental specialist. A 1993 supervision mission found implementation of monitoring measures to be variable, often lagging behind committed timescales. Dust and noise monitoring of the expressway projects and the training component required strengthening. Inter-project communication between monitoring staff was also inadequate, and some agreed monitoring programs had not been initiated. In addition, staffing of the environmental unit with implementation responsibility was insufficient. However, following the signing of contracts with local environmental institutes to undertake monitoring, significant progress was made with implementing agreed monitoring measures. Fully equipped monitoring stations are due to be operational before the expressways are opened to traffic. The inclusion of environmental monitoring requirements within the EMP and project legal documentation provided a sound basis for measuring progress with implementation during supervision and for Borrower compliance with monitoring commitments.

participation in mid-term reviews should be the minimum requirement. Implementation considerations During project implementation, the Borrower is responsible for undertaking performance monitoring in accordance with the EMP, and for reporting the results to the Banks Task Manager. In addition, there may be a requirement to report monitoring results to national or local regulatory authorities. Where results violate pre-agreed limits the implementing agency has the responsibility to take corrective action to achieve the projects environmental requirements.
4

Any monitoring program requires allocation of responsibilities. The task of assigning roles can be aided by the use of the matrices developed to show linkages between impacts and monitoring. This can help establish the appropriate level of expertise for particular tasks, and in assigning functions to different organizations. For example, inspection of construction might be assigned to a consultant engineers staff, while monitoring might be performed by staff from the implementing agency. Cross-checking can be facilitated by comparison of results from different monitoring sources, including local communities. For example, the effectiveness of bans on timber extraction by contractors can be verified using aerial photography supplemented by field observations. The recipients, structure and intended usage of monitoring reports should be clearly identified. It is

Figure 1: Monitoring and supervision during project implementation

Project Implementation

Implementing Entity

Implementing entity takes corrective action

Inadequate

Implementing entity undertakes monitoring per EMP requirements

Reporting

Regulatory entity assesses compliance & recommends action

Reporting

Monitoring Supervision
Banks Task Manager informs Implementing entity No

Bank confirms adequacy of monitoring information Adequate

World Bank

Bank assess compliance with EA requirements

imperative that the reporting structure ensures that non-compliance is rapidly brought to the attention of the appropriate decision makers, to facilitate timely corrective actions. In addition, the structure, content and timing of reporting to the Bank should be designed to facilitate supervision and provide background for Bank missions. The Task Manager should carefully consider arrangements for tracking receipt (and subsequent dissemination) of monitoring reports. The importance of missions in project supervision is highlighted in box 4. The Task Manager, who has the pivotal role in project supervision, confirms that progress reports submitted by the Borrower include the requisite performance monitoring information. It is advisable that the Banks Regional Environment Divisions (REDs) review these reports and help Task Managers to assess compliance with EA requirements. If unsat-

isfactory, the RED should agree on an appropriate course of action with the Banks Country Department staff. The Country Department notifies the Borrower of this action and any necessary follow up and consults with the RED on significant environmental changes during implementation. The likely interfaces between the Borrower, implementing agency and Bank staff during project implementation are outlined in figure 1. Adherence to the monitoring and supervision schedule should ensure that the necessary mitigation measures are completed in a timely and satisfactory manner. It should also assist personnel involved in monitoring and supervision (particularly where responsibilities change), by providing a succinct summary of the agreed environmental requirements. To reinforce effective meshing of environmental and other project requirements, financial disincen5

Project status reports (Form 590) Box 5. Towards improved monitoring and supervision strategies in Sri Lanka
Under the National Environmental Act, the Government of Sri Lanka requires EAs for major projects. The Act and its implementing regulations define explicit monitoring and supervision requirements as part of the project planning and implementation cycle. However, as in many other countries, in Sri Lanka monitoring is often the weakest link in the project implementation cycle. The Central Environmental Authority (CEA) and the United States Agency for International Development convened an intensive workshop in 1995 to identify and address critical issues impeding environmental monitoring and supervision programs. Attendees included representatives from government agencies, universities, research institutions and NGOs. Consensus was reached on a number of important issues including improving: Regulatory programs and policies by clearly defining monitoring and supervisory responsibilities, development of sectoral guidelines and establishing time frames for legal action in response to non-compliances; capacity of agencies with monitoring responsibilities through provision of additional resources, staff and specific budgetary allocations; proficiency of laboratories and environmental professionals by establishing certification procedures; and inter-agency coordination of monitoring activities by developing legal provisions under memoranda of understanding; establishing a central database within the CEA to improve the quality of and accessibility to baseline data, and maintaining an up-to-date register of data sources.

In recognition of the need for more realistic and transparent performance assessments of projects, the Banks Operations Policy Department (OPR) has recently revised the project status reporting format (form 590). This report is completed by Bank Task Managers following project supervision missions. The change in reporting arrangements should be fully operational by the end of FY97. Two important changes include identifying key performance indicators linked to development objectives (which may include environmental objectives), and ratings of implementation progress. Guidance on specific rating criteria which apply to the performance of environmental, and resettlement components has recently been issued by OPR. The implementation progress rating criteria for the environmental components (often specified in the EMP) are as follows: Highly satisfactory (HS) Environmental components are being implemented in a timely and satisfactory manner. Implementation of the critical aspects of the environmental components is proceeding in a timely and satisfactory manner. There are some problems with certain other aspects, but these are not undermining the progress of implementing the components and are being addressed by the borrower/ implementing agency. There are major problems in implementing some or all of the environmental components of the project, including aspects critical to the success of the components and the project as a whole. However, appropriate measures are being taken by the borrower/ implementing agency to address the problems. As in U, but problems are not being satisfactory (HU) addressed appropriately.

Satisfactory (S)

Unsatisfactory (U)

tives can be utilized. Refusal to disburse funds unless work is done in compliance with environmental requirements is a standard construction contract condition that could also be elevated to the loan agreement level. In addition to assessing predicted impacts, monitoring may also reveal unanticipated impacts. Effective supervision ensures that corrective action is taken commensurate with the scale of such impacts. This can significantly influence project costs, which underlines the importance of accurate and timely reporting. This redesign or mobilization of contingency funds may be required.
6

Highly UnSatisfactory (HU) Institutional issues

Many IBRD projects require the strengthening of environmental management capacity within agencies responsible for overseeing the implementa-

tion of mitigation measures and monitoring. This should result in a better understanding of performance monitoring and better implementation of future projects (see box 5). Monitoring may be shared between an implementating entity and an environmental regulatory entity, to help develop management capacity. It is important that some funding for monitoring be allocated to the relevant agencies, otherwise the necessary environmental actions may not be effectively implemented. Opportunities should also be sought to develop links with national monitoring networks and to ensure appropriate roles for environment management institutions. Public consultation and involvement Public involvement is crucial to a sound decisionmaking process. The EMP should therefore clearly describe and justify the proposed performance monitoring approach. Local residents will often be in the best position to observe and report environmental and social impacts. A first step is to involve them in establishing baseline conditions before a project proceeds, typically through the use of local NGOs or universities with the requisite skills. In addition, there should be one clearly identified primary point of on-site office contact with the public on compliance issues and complaints. This office should keep complete records and provide input to resolution of issues. There should also be provision for appeals by interested parties. Many countries, especially in Latin America, are establishing positions of environmental omsbudman, to hear public concerns and bring them to the attention of the authorities. Consultation with affected people and NGOs (as specified within OD 14.70 on Involvement of NGOs in Bank Supported Activities) should be an integral part of the EMP for all Category A and are advisable for many Category B projects. In some instances NGOs have been invited to be part of the monitoring process. Bank supervision should ensure that local concerns are adequately addresed during implementation. Disclosure of information is an important element of the consultation process. Information on progress with mitigation implementation and monitoring activities should be shared with the public. Panel of experts In some instances, there may be disagreement as to the significance of the impacts of a project or the effectiveness of mitigation measures. The option of an independent panel of experts is provided for under Bank EA procedures and the involvement of such a group in the design and implementation of a monitoring program can be useful.

Box 6: Involvement of Panel of Experts

The Ghazi-Barotha Hydropower project in Pakistan is designed to meet the acute power shortage in Pakistan. The main project elements include a barrage located on the Indus River, a power channel (designed to convey water from the barrage to the power complex) and a power complex. The SAR for the project includes detailed TORs for an independent environmental and resettlement panel (ERP) for the implementation phase of the project. A separate ERP was actively involved in project preparation. During implementation the ERP (which will comprise an environmental specialist, an agricultural specialist and a social scientist) will work closely with the agencies involved in project implementation. The chairman of the ERP will have responsibility to appoint short-term specialists if necessary. The ERP will be required to undertake semiannual onsite reviews of the following: implementation of the environmental mitigation, monitoring, and resettlement plans; activities of the Project Information Center; actions or studies which should be undertaken to support the projects environmental, resettlement and public information objectives; TORs for environmental and resettlement activities at the request of the implementing agency; and appropriate advice or guidance required to address unanticipated impacts during project implementation.

The findings of each review will be reported and detailed briefings will be given to the implementing agency and other concerned parties.

The principal advantage is that an arms-length institutional arrangement can provide a forum for discussion of technical issues and recommendations on future actions, without the appearance of a vested interest that Borrowers or the Bank may incur (see box 6). To maintain its independence it is important to establish clear procedures for disclosure of the Panels deliberations. Sectoral and regional EAs The use of sectoral EAs (SEAs) and regional EAs (REAs) is increasing within the Bank, which offers certain advantages for monitoring and supervision, compared with a project-specific approach (see Update no. 4: Sectoral EA and forthcoming Update no. 15 on Regional EA). They provide a framework for assessing institutional needs and project-specific proposals.
7

They can also help to rationalize monitoring and supervision efforts where similar projects are funded from different sources. REAs are compatible with development of monitoring protocols that are contiguous with the boundaries of natural or man-made systems such as watersheds or large metropolitan areas, and can also facilitate inter-agency coordination. This approach can result in the systematic strengthening of overall environmental management capability in a region. Conclusions It is essential that monitoring and supervision be guided by clear objectives. Specific requirements should be linked to impacts identified during the EA process. The incorporation of detailed monitoring and supervision arrangements into project legal documents is essential to ensure effective implementation. Where supervision identifies inadequacies in the approach to monitoring, project legal documents should provide a sound basis for agreeing to and implementing remedial actions. Institutional capacity to implement or oversee monitoring should be evaluated, and reinforced as appropriate. The associated costs should be an integral part of the project costs. It is important to maintain effective public involvement throughout project implementation. This should include provision of a forum for submitting and resolving complaints, access to monitoring information, and involvement in monitoring to the extent feasible.

References AGRAF. 1995a (draft). Land Quality Indicators for Agricultural and Resource Management Projects. Washington: the World Bank, Agriculture and Natural Resources Department. AGRAF. 1995b (draft). Performance Indicators in BankFinanced Agricultural Projects. Washington: the World Bank, Agriculture and Natural Resources Department. Dixon, J.A., Kunte, A. and Pagiola, S. (1996). Environmental performance Indicators. A First Edition Note. World Bank, Washington, D.C. ENVGC. 1995a (draft). Monitoring and Evaluation Guidelines for ODS Phaseout Investment Projects, Washington: The World Bank, Global Coordination Division. ENVGC. 1995b (draft). Monitoring and Evaluation Guidelines for GEF Global Warming Investment Projects, Washington: The World Bank, Global Coordination Division. ESD. 1995. Monitoring Environmental Progress: A Report on Work in Progress. Washington: The World Bank, Environment Department. Global Environment Facility. 1992. Guidelines for Monitoring and Evaluation of GEF Biodiversity Projects. Washington: Global Environment Facility. OECD. 1994. Environmental Indicators. Paris: OECD. Pieri, Christian, J. Dumanski, A. Hamblin, A. Young. 1995. Land Quality Indicators. World Bank Discussion Paper No. 315.

This Update was prepared by Aidan Davy and Phil Paradine, an independent consultant. Extensive review and comments have been provided by John Dixon, Arundhati Kunte (ENVPE) and John Butler of IRG consultants. Based on Bank policy and procedures on Environmental Assessment (EA) (Operational Directive 4.01), the EA SOURCEBOOK UPDATE provides up-to-date guidance for conducting EAs of proposed projects. This publication should be used as a supplement to the Environmental Assessment Sourcebook. Please address comments and inquiries to Olav Kjrven and Aidan Davy, Managing Editors, EA Sourcebook Update, ENVLW, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. S-5139, (202) 473- 1297. The Bank is thankful for the Government of Norway for financing the production of the EA Sourcebook Update.

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

June 1996 Number 15

Regional Environmental Assessment

Regional environmental assessment (REA) is a tool to help development planners design investment strategies, programs and projects that are environmentally sustainable for a region as a whole. REAs take into account the opportunities and limitations represented by the environment of a region and assesses on-going and planned activities from a regional perspective. This EA Sourcebook Update describes REA in terms of its nature and purpose, advantages, operational context, selection criteria, and key components. It also discusses challenges associated with REA preparation and offers examples from Bank experience. The Update expands upon existing information in Chapter 1, pp. 1214 in the EA Sourcebook. Background Generic guidance on regional EAs was introduced in the World Bank in 1989 with the adoption of an Operational Directive on environmental assessment (amended in 1996 as OP/BP/GP 4.01). The Environmental Assessment Sourcebook (1991) provided additional advice on the nature and purpose of REAs and possible operational contexts. REAs have had limited use to date, but emerging experience in the Bank and other development institutions has revealed that this tool can improve regional development planning by helping to formulate more sustainable investment strategies and to build environmental management capacity at the regional level. This Update draws on this experience. The term region is discussed in box 1. Nature and purpose of REA The main purpose of REA is to improve investment decisions by bringing environmental opportunities and constraints into development planning at the regional level. REA is similar to sectoral EA (see Update no. 4: Sectoral Environmental Assessment) in that it is used during the early stages of development planning, before decisions about specific projects have been made and with the purpose of influencing such decisions. Both EA types allow for comprehensive assessment of environmental issues (one within a sector and the other a region), and can be used to establish environmentally sound development policy. Regional EA differs from other forms of EA because it assesses environmental issues and impacts in a disREAs are more comprehensive undertakings than project-specific EAs being broader in terms of the physical area to be assessed, the time frame to be considered, and the analytical content. REAs are also more openended in terms of impact predictions and recommendations. They do not substitute for project-specific EAs of individual investments, but can limit the need and scope of project-specific EAs downstream. Some regional EAs are used in a proactive manner as a development planning tool for a region. They examine a given region in terms of its natural resources, ecological and socioeconomic characteristics and identify investment projects that are environmentally sustainable for the region as a whole. The end result may be a comprehensive regional development plan. However, most tinctly spatial setting. Ongoing activities, plans and potential projects are assessed by how they may cumulatively affect the ecology and human living conditions within a larger area. The spatial area to be investigated can be delimited based on ecological, socio-economic, administrative or other boundaries (see Box 1). In the first case, the geographic area may be a river basin, a coastal zone, a highland area or other areas that can be viewed as naturally bounded. In the second case, a regional EA may, for example, focus on a province, a group of counties or a municipality. The spatial area can also be delineated by demographic factors; this applies most commonly to urban areas. Sometimes a defined region extends to more than one administrative area, e.g., more than one municipality, county, province or even nation state. In such cases, cross-jurisdictional issues often create a need for innovative institutional arrangements.

Insert in Update Binder chapter 1

Box 1.

Definition of region

One definition of a region is any subnational area that a country calls a region for purposes of planning or development (Organization of American States OAS, 1984). Such an area is usually demarcated along administrative boundaries and may be composed of one or more municipalities, provinces or states. Other times, a region is defined as the locus of a specific problem (e.g., poverty, social tension, population pressure) or according to ethnic makeup. Socioeconomic characteristics may also define regions, such as a generally poor rural area, or a major industrial area. For purposes of integrated regional planning and REA, the ideal approach is normally to define the region in natural-spatial terms. Common geographically defined units are river basins, mountain plateaus, forested areas, coastal zones, airsheds and island configurations. An urban area can also be a very useful unit of analysis, often providing a degree of consistency across natural-spacial, socio-economic and administrative boundaries. Given this wide variety of definitions and characteristics, regions have no general distinguishing characteristics as study areas. However, methodology for regional development planning in general is relatively well defined and advanced.

A range of countries have experience with REAs and other environmental planning tools with a regional application. This experience suggests that, effectively used, REA can provide a number of additional benefits: Provide a baseline overview of environmental conditions within the study area (a regional state of the environment), which is key to making reliable impact assessments and monitoring environmental changes over time; Assist governments in forming a long-term view of regional planning and increase the transparency of the planning process (that is, show the reasoning behind development plans), thereby modifying or eliminating decisions that might be environmentally harmful; Analyze the institutional and legal framework relevant to the particular region, identify institutional and jurisdictional gaps, and make recommendations regarding, for example, environmental standards and law enforcement appropriate for the region (thus reducing the need for similar analysis in downstream EA work); Suitably collect and organize regional environmental data and, in the process, identify data gaps and needs at an early stage, and outline methods, schedules and responsibilities for data collection and management during program or project implementation; Allow for comprehensive planning of region-wide environmental management and monitoring, and identify broad institutional, resource and technological needs at an early stage, including potential funding problems; Provide a basis for collaboration and coordination across administrative boundaries and between sector-specific authorities and help avoid contradictions in policy and planning while enhancing efficiencies; Strengthen preparation and implementation of individual projects within the region, by recommending criteria for environmental screening, analysis and review of such projects and setting standards and guidelines for project implementation; and Provide a vehicle for public participation in shaping the future development of a region, thereby building public support for the process. The operational context The World Bank may support REAs in the context of: (1) a region-specific investment program involving multiple subprojects, such as the Argentina Flood Protection Project featured in box 3; (2) a series of independent investments within a region where the Bank may be involved in one or several of them; (3) a large, single project with complex implications for other activities in the region, such as the Lebanon Solid Waste and Envi-

regional EAs are more in the nature of a cumulative impact assessment of multiple projects and activities that may be on-going, planned, or simply expected. Such an REA may be carried out in conjunction with a regional development plan, and may also help shape investment priorities and activities downstream. However, the emphasis is on influencing an evolving strategy or plan (including projects) and assessing cumulative impacts rather than designing a full-fledged development plan. Boxes 2 and 3 discuss two REAs of the second type. Advantages of REA REAs can influence investment planning in a large area where project-specific EAs can only address the design and management of individual projects. REAs can assess the cumulative and interactive environmental impacts of several projects where the projectspecific variant looks at site-specific impacts. Like sectoral EAs, REAs move environmental analysis upstream in the planning process into the policy arena, at a stage where major strategic decisions have yet to be made. Thus, REAs offer opportunities for more comprehensive and realistic assessment of investment alternatives and can help eliminate at an early stage those investments that might generate particularly adverse environmental impacts.
2

Box 2.

Biodiversity Conservation Project in Indonesia: The regional impacts of conservation

recommendations in a number of areas. For example, the REA found that the park boundaries were inadequate for biodiversity conservation, and that a clear commitment from the Government of Indonesia to rationalize the boundaries progressively over a five-year period would be needed. Perhaps the most important finding of the REA was that current rates of deforestation and land degradation are so high in the area that for the project to be viable, immediate and strong environmental management measures were required. Logging, agricultural encroachment, mining and road development are activities where immediate control and enforcement measures are needed. The REA outlined a number of such measures and also discussed how the proposed project might influence regional development opportunities. It concluded that the project has a low opportunity cost because remaining land areas with biodiversity value have major use limitations. Suitable land for agriculture has already been converted. Further logging and conversion to agriculture and other uses are already causing major erosion problems as remaining forested areas are typically in steep hillside areas. While the project is not likely to make a significant regional economic contribution in the short-term, it might lead to a significant slowing in the degradation of land, water and biological resources. Over the medium to long term, park-based tourism could become an important source of income.

This project, financed in part by the Global Environment Facility (GEF), aims to conserve biological diversity and improve natural resource management and economic opportunities for local communities in and around the Kerinci-Seblat National Park on the island of Sumatra. A REA was prepared to assess the environmental, land use and socioeconomic development implications of this project and help design appropriate policy, administrative and fiscal responses from national and local governments. The REA was an integral part of project development from the earliest stages and evaluated the following: the extent to which the design of the project met GEF criteria and objectives for natural resource and biodiversity conservation; the potential impacts of adjacent development activities on biodiversity conservation and integrity of the national park; and the impacts (positive and negative) of the project on adjacent human communities and on opportunities for future economic development.

The REA included detailed baseline surveys of the natural environment and of socioeconomic conditions in the project area, enabling the REA consultants to evaluate the evolving project design and give concrete

ronmental Management Project discussed in box 4 ; and (4) regional planning that may go ahead independently of any Bank-supported project preparation in the region. In the first three instances, the REA would normally be tied to one or several projects. These may have potential environmental impacts that are either significant or moderate (classified by the Bank in categories A or Bsee Update no. 2: Environmental Screening). Thus, the REA should be carried out in accordance with the Banks EA policy and procedure (OD 4.01: Environmental Assessment). However, in the fourth context the REA is employed without any direct link to lending activities that normally require EA work. In this case, the REA can be tied to Bank economic and sector analysis for any given country, or to technical assistance or adjustment operations. The Banks EA policy does not apply although it, as well as this Update, may provide valuable information. Criteria for undertaking REA The following criteria help indicate if a REA is appropriate to a given investment project or program (where OD 4.01 applies):

Is the Bank considering supporting an investment program or project(s) in a region with existing, significant and inter-related environmental problems or major uncertainties about ecological functions and relationships? Could the proposed program or project(s) have significant region-wide environmental impacts (including environment-related social impacts and impacts on cultural heritage) that need a comprehensive treatment to be understood and addressed? If the answer is yes to one or both of these questions, a REA should be considered. In addition, there are conditions that, when met, increase the value and feasibility of a REA: Is the borrower at an early planning stage or at a new major investment phase, where important strategic decisions have not yet been made concerning development of the region? Are the economic and social conditions in the region relatively stable and predictable (as opposed to rapid and unpredictable change), to allow for a medium to long-term planning horizon and enhance the long-term value of the REA?
3

Box 3.

A proactive regional EA: A flood protection project in Argentina

were road construction, followed by poorly planned urban expansion, and effluent from the meat packing industry. Another significant finding was that poor urban sanitation services were directly undermining existing flood protection works. For example, many communities disposed garbage along protective dikes, attracting rodents which weakened the dikes by digging tunnels making them ineffective against floods. The REA helped design four key project components to help improve the environmental and economic benefits of the project. These included (a) a component to strengthen EA procedures in key institutions within the seven provinces; (b) technical assistance for urban environmental management; (c) environmental education and awareness programs in communities benefiting from protection works; and (d) support to protection and management initiatives for wetlands and other ecosystems. However, perhaps the most important contribution of the REA was its direct contribution to screening potential investments under the project and assessing the cumulative impacts of selected sub-projects. From a total of some 150 possible investments, 51 subprojects with clear economic, social and environmental benefits were short-listed. Projectspecific EAs were prepared for all sub-projects, on the basis of which the REA team examined the likely cumulative impacts of all the 51 subprojects, ensuring that such impacts would be minimized. Public consultation was an important part of the entire process and, in one case, a subproject was significantly redesigned following community input. A highly experienced team was used and the total cost of the REA did not exceed US$300,000a small amount compared to the cost of the project and the expected economic, social and environmental benefits.

The central objective of the Argentina Flood Protection Project is to improve flood protection for the human communities inhabiting the flood plains of the Parana, Paraguay and Uruguay rivers in northern Argentina. It would finance implementation of a comprehensive program of investments within the flood plains lying within the boundaries of seven provinces. This region has suffered enormous economic and human loss from floods, the last occurring in 1983 and 1992. At the same time, the periodic flooding sustains ecological systems and many forms of productive activities. The project has therefore adopted a living with floods strategy. The investments include both structural and non-structural measures to protect important economic and social infrastructure and enhance the provincial capacity to deal with periodic flooding. At the Banks suggestion, a REA was initiated during the earliest stages of project preparation, to serve as an input to its design. The REA studied the interaction of natural and manmade systems within the flood plains, including the ecological functions of the periodic floods and the current state of critical ecosystems such as wetlands and gallery forests. The study, undertaken by an Argentine team led by a Colombian specialist, found that to a surprising extent many ecosystems and human activities depended on the floods. This had a direct impact on the way the project was designed. Criteria for the selection of investments were modified to ensure that flooding would continue but not threaten human well-being and economic infrastructure. The study also documented the extent to which wetlands, gallery forests and aquatic ecosystems of the tributaries to the three rivers are threatened by ongoing human activities. The REA found that the most disruptive activities

Is the borrower likely to give due consideration to the findings and recommendations of the REA? Undertaking a REA: The process Like any other form of EA, the process of undertaking a REA can be divided into designing the study and executing the study. The design stage essentially consists of determining the scope of the REA (scoping), preparation of TOR and selection of the REA team. The execution stage involves the preparation of the REA. What makes the REA process different from other EA processes is that the regional/spatial perspective needs to be present throughout, and that the scope poses some particular challenges in design and execution. Designing the study Since a REA by its very nature is a complex undertaking and can have significant impacts on the develop4

ment of a region, it is strongly recommended that a comprehensive scoping exercise be undertaken to help lay a solid foundation for the study. The scoping should encompass the following activities: Understanding the regional planning framework. There must be a compelling connection between the REA process and the existing development planning framework for the region. In some cases the situation is clear-cut: the boundaries of an existing regional institution, such as a river basin authority, provides both an organizational home and the geographical boundaries for the REA. At other times, a region is characterized by overlapping responsibilities across institutions and even overlapping jurisdictions, making it more complicated to institutionally position the REA and determine the appropriate boundaries. The key is to understand the existing institutional setting and link the REA in a functional way.

Box 4.

A solid waste and environmental management project on Lebanons coastal zone

Task 3: Prediction and evaluation of the cumulative impacts of the scenarios What are the significant adverse and beneficial impacts/resource use conflicts by the year 2010? What are the relative probabilities of the impacts occurring? Which of these expected adverse impacts are irreversible? Which adverse impacts are reversible and how might this be achieved? How might expected beneficial impacts be enhanced? What is the temporal and spatial distribution of the impacts and resource use conflicts? Which protected areas, non-designated but important sites or resources might be affected? Which protected, rare or endangered species might be threatened? Which international or national standards, criteria or guidelines, and policies for pollutant levels, human health/activities, protected areas/species will be contravened?

This project has two main objectives: Rehabilitation of Lebanons solid waste management system, and reversing the current trend of environmental degradation in coastal areas of the country. The second objective is addressed through a coastal zone management (CZM) component aimed at strengthening development planning and environmental management in coastal areas. A regional EA for Lebanons coastal zone is the first step in the process of designing this component. It is intended to identify the cumulative pressures and impacts of coastal developments under different investment scenarios, identify the hot spots in terms of environmental sensitivity and potential hazard, and propose potential mitigation actions. It will also provide key information that will feed into (a) the establishment of a Geographic Information System (GIS) for physical planning and monitoring; (b) preparation of a CZM plan to be approved and legally binding for all future developments on the coast; and (c) implementation of emergency actions to protect and rehabilitate coastal resources. Below are exerts from the Terms of Reference (TOR), in the form of questions that the REA needs to address: Task 1: Description of the development scenarios What are the expected developments/trends at the micro and macro levels that will affect the coast? What are the assumptions behind the scenario(s) identified and which are the main uncertainties? What is the most likely scenario? What are the main concerns related to expected coastal impacts, 19952010? What are the expected resource use conflicts, 1995 2010? What are the institutional responsibilities and functions in control and management of the coast?

Task 4: Comparative analysis of the development alternatives What are the main differences between the scenarios in terms of environmental impacts and resource use conflicts? Are these differences of type or scale? What are the major differences in type? What are the major differences in scale? What are the implications of these differences for institutional arrangements and the nature/type of Coastal Zone Management Plan needed?

Task 5: Recommendations/input to Coastal Zone Management Plans What are the priority hot spots which need remedial action? What sectors are contributing most to these hot spots? What types/mix of measures are needed to improve the hot spots in the short and long term? What beneficial impacts could be enhanced? What measures could achieve this objective? What might be the costs and institutional implications of the recommended measures? What capacity building measures might be needed?

Task 2: Description of the environment What are the main features and natural resource uses of the coast? What and where are the main, current environmental impacts and resource use conflicts? What are current trends in coastal environmental quality? What and where are the relative contributions of the various economic sectors to these trends? What and where will the net effect be of current infrastructure rehabilitation?

Defining the spatial context. Defining the geographic and environmental context for a REA is essential. Some steps that will help define context include the following: (a) identify the major ecosystems of the region to understand the broader ecological context and the rational boundaries for natural resources management (e.g., a river basin or a coastal zone); (b) determine the boundaries of economic and market systems; (c) define the pri-

mary, secondary and tertiary urban centers of the region; and (d) assess demographic patterns and migration trends to understand the population dynamics likely to influence the regions development. Determining the optimal multi-sectoral focus. Development problems are frequently defined in a sectoral context that often obscures casual relationships. Because
5

sectoral problems frequently require multi-sectoral solutions, the challenge is to design studies with a sufficiently broad technical focus. Consequently, before embarking on a REA, the following steps are needed: (a) determine which sectors have the potential to cause environmental problems in the region so that the study can be designed to include all relevant aspects; (b) analyze sectoral cause-and-effect relationships so that the REA is directed toward addressing main causes rather than symptoms; and (c) encourage sectoral institutions to supply information, feedback and political support. Involving sectoral interests early may defuse potential conflicts and reveal false assumptions that can misdirect the study. Limiting the study goals while retaining an integral focus. One common problem in designing regional studies involves stretching limited financial resources to cover too many study objectives. The challenge is often to cut the problem down to fit the resources available, without losing the integral focus of the study or producing plans and proposals too general to be implemented. Ways to balance these concerns may include: (a) limiting the overall size of the geographical area; (b) studying the region at different levels of detail, phasing the investigation from the general to the specific; (c) limiting the time horizon of the proposed development plan and projects; (d) limiting the sectoral focus of the study after a rapid integrated overview of the various sectors; or (e) focusing the study on a particular target population within a designated geographical area. Setting up appropriate institutional arrangements. This point is particularly important for REAs where the main objective is to shape regional development planning. The most common management structure is a technical unit composed of national and international professionals and support personnel who jointly execute the study. It is important for the effectiveness of the study that the degree of interaction and teamwork be very high. In terms of institutional arrangements at the national and subnational levels, several options are available, including (a) setting up a task force of agencies to execute the study; (b) working with a sectoral executing agency under the aegis of a national or provincial planning agency; (c) working with a natural resource or environmental agency; (d) working with a regional development corporation or similar agency; and (e) helping establish an independently funded study team that can evolve into a regional development institution. All these options have their strengths and weaknesses and specific country circumstances should guide the final selection. Developing a detailed TOR. The TOR needs to address the following: (a) define the objectives of the study and tasks to be performed to achieve the objectives, including clear allocation of responsibilities among those involved; (b) identify the specific technical products to be
6

delivered (reports, maps, training, development project proposals, etc.); (c) define available information resources; (d) allocate the available human and financial resources to the various tasks; (e) establish a time frame for the delivery of different products; (f) design a system for continually integrating information and forcing the interaction of specialists throughout the study; and (g) provide for appropriate levels of public consultation. Planning appropriate public consultation. Public consultation is an integral part of the EA process (see Update no. 5: Public Involvement in Environmental Assessment: Requirements, Opportunities and Issues). Since a REA is invariably initiated before concrete investment decisions are made, it may not always be possible to systematically consult representatives of potentially affected people during REA preparation. The exception is if activities with potential impacts on specific locations are already being planned. Consequently, an appropriate approach may be to target consultations with those NGOs that are active in the region, scientific experts, relevant agencies from all levels of government and perhaps industrial, commercial and labor interests. The scope of the consultations would normally be at the level of strategic choices for regional development. The REA can also be an effective vehicle for designing a public consultation plan to be carried out within the context of the implementation of a regional development program. In the case of a proposed urban infrastructure program for the Indonesian island of Bali, the regional EA will design detailed provisions for consulting with stakeholders at local (banjar), county (kabupaten) and provincial levels. At the local level, the need to consult women has been particularly emphasized, as they are the principal end users of water, solid waste and other services. A well-planned and successfully implemented consultation process will help ensure public support for the final regional plan or program, and increase overall public awareness of major environmental issues and their interrelatedness in the region. Defining a review process. It is important to agree on a review process in advance, including scope and process for report distribution, allocation of review responsibilities and definition of review stages. In establishing the review process public consultation needs and requirements should be fully taken into account. Executing the study REAs vary in scope and content according to the region of concern, types and significance of environmental issues and the operational context. Primary areas requiring attention in a full (Category A) REA are presented below. It is important to keep in mind that flexibility is needed to adjust the REA process and methodology according to the particular context (geographical, administrative and operational). Being a relatively new ap-

proach, REAs require innovative thinking, careful judgment and a structured learning-by-doing process. Policy, legal and administrative framework. This section should be considered a fundamental part of a REA report. It is important to take into account those aspects of the national policy, legal and institutional framework likely to influence environmental management in the region, in addition to the regional framework itself. This should include looking at existing political priorities and how they might constrain or facilitiate implementation of environmental policy and activities in the region. If other, recent studies have analyzed these dimensions adequately, the REA should use this work rather than duplicate it. The national framework. The relevant national environmental policies, laws and regulations should be assessed for completeness and appropriateness in light of the particular conditions and problems of the region, and gaps and weaknesses noted. Non-environmental laws and policies of significance to the regions utilization of resources, production processes, or pollution should also be identified. Similarly, the national regulatory framework for EA preparation and review should be assessed. The REA should look closely at the institutional capacity of the main environmental or natural resources ministry or agency, in terms of effectiveness and capacity for providing guidelines, setting and enforcing standards, and reviewing EAs. The capacity and performance of agencies responsible for specific environmental services such as nature conservation and cultural heritage should also be reviewed when relevant. The regional framework. The REA should analyze regionally-based policies and regulations that have environmental implications. It should also identify how responsibilities are distributed and assess institutional capacity for environmental management (including enforcement). The regional investment planning process should be carefully reviewed in terms of objectives, methodology and procedures for review and approval of plans and projects. The relationship between timing of project review, issuance of licenses and permits, and the regional planning process more generally should be clearly indicated. Baseline conditions. A central component of the REA study is the assessment of the existing environmental conditions in the region, including the development constraints and opportunities that the environment and natural resource base pose. The baseline survey should be clearly focused on those aspects of direct relevance to the key environmental issues the REA is intended to address, rather than attempting to cover all possible environmental aspects. The survey would normally cover the following main themes, reviewing the subthemes on a selective basis:

The physical environment: Geology; topography; soils and land capabilities; meteorology; surface and groundwater hydrology; water quality and quantity; air quality; potential natural hazards. The biological environment: Flora and fauna (particularly rare or endangered species); critical habitats and ecosystems; parks and reserves; significant natural sites; and species of commercial importance. The socioeconomic and -cultural environment: Population; land use and patterns of land ownership and tenure; planned development activities; community structure; employment; distribution of income, goods and services; public health; cultural heritage; indigenous peoples; and customs, aspirations and attitudes. The emphasis given to each of the issues listed above would depend on the given regional setting, particularly on whether the setting is urban or rural but also whether it is coastal or inland, densely or scarcely populated, largely pristine or highly developed, and so on. Thus, those who prepare REA TORs should provide clear direction on these matters on a case-by-case basis. Box 5 describes criteria for baseline data collection and summarizes some lessons of experience. Where a project-specific EA describes conditions around a proposed project site, the REA should concentrate on the issues and problems that are characteristic of the region as a whole. For example, deforestation may be a dominant problem in one region, leading to associated problems such as loss of biodiversity, soil erosion and sedimentation of river systems. Or, if rapid industrialization and urbanization is taking place, the main concerns may be management of water supply, air and water pollution, or congestion. Important regional resources and activities should be given particular emphasis. For example, if the region of concern is a rural coastal flood plain, major sectors are likely to be agriculture/irrigation and fisheries. The interrelationship of these sectors and the impacts on the natural resources of the flood plain and coastal zone could be major issues. Water quality and floods could be other areas of concern. Geographic Information Systems (GIS) may be an effective instrument for gathering, organizing and analyzing baseline environmental data (see Updates nos. 3 and 9), particularly at the regional and national levels. If GIS capability is not in place, a regional EA may provide opportunity and rationale for introducing it immediately or in the future. It is always useful to cross-reference the pertinent environmental regulations and standards when describing baseline conditions. If regulations and standards are absent, this should be explicitly noted. The
7

Box 5.

Conducting baseline surveys

Updates nos. 3 and 9, on GIS and EA). In the Pilcomayo River Basin (Argentina, Bolivia, Paraguay), satellite imagery was the key to the rapid analysis of land capability and the delineation of development zones worthy of more detailed study. Use resource survey specialists who are experienced development practitioners as well as good scientists. Such specialists can direct and train local professionals along practical and efficient lines of investigation as well as interpretation of data from several disciplines for analytical purposes. Where possible, a single national agency should be responsible for resource surveys. This greatly facilitates the compatibility and integration of data. Use maps to synthesize the final products of sectoral investigations. Integrating mapped information is one way to merge the work of different disciplines. As devices for data integration, map-overlay techniques and composite maps are particularly useful, although availability of base maps is limited in many countries. In the Eastern Cibao valley of the Dominican Republic, for example, the map-overlay technique was used very effectively for agricultural zoning. The combination of information about geomorphology, vegetation and land capability in specific areas facilitated the identification of viable agricultural and agro-industry projects.

General criteria for data collection: Information contributed by the main sectors covered by the REA should be at approximately the same level of detail, with any bias reflecting the emphasis of the REA. Setting common data standards and formats in advance can be useful. Data should prove or disprove specific hypotheses related to the studys objectives and answer specific questions about the environmental situation in the region. The TOR should specify who will use the data being developed and how. Where possible, data collection should be undertaken in conjunction with domestic institutions. The use of international consultants to do basic data collection can be very costly and needs special justification. Local populations, NGOs, research institutions and universities should be used as sources of information. This saves time and resources and gives access to unique information. Keep description to a minimum and emphasize analysis.

Lessons of experience Use modern satellite imagery and remote-sensing techniques for resource surveys of large sparsely settled or unpopulated regions about which little is known. These tools facilitate rapid and relatively accurate mapping and analysis of geology, geomorphology, soils, natural vegetation, land use and so forth when coupled with ground verification (see also

Source: Integrated Regional Development Planning: Guidelines and Case Studies from OAS Experience. Organization of American States, 1984.

REA report should also critically take into account the quality of available data and note major data gaps. Description of development plan and associated projects. To the extent that the REA is intended to assess an existing development plan, including associated project proposals, the REA needs to provide a detailed description of the plan and the proposed projects. If several variations of a plan are being considered, or if one or more plans are under conceptual development, the REA should describe these in sufficient detail for a useful analysis of impacts and consequences. Inventory of other plans and projects. The development context of a region is as important as its resources and population dynamics. The quantity of existing plans and proposed projectsprepared with or without Bank or other international development assistancewill normally be high. (In the Chapare region in Bolivia, for example, 54 agencies supported development activities during the same period, many of them conflicting.) Making an inventory of all relevant plans and projects is exceedingly important, but can be overwhelming.
8

Guidelines from the Organization of American States (OAS, 1984), on integrated regional development plans, suggest some ways to avoid problems: Identify all significant development plans and projects in the region, no matter what agency is involved, but collect detailed information only about those that serve or contradict the proposed development objectives. Efforts at being too comprehensive will only delay the study. Organize planned projects in a time sequence, and avoid including projects that fall outside the time horizon of the study. Be sure to identify projects with high-level political support that have already gained momentum. Cumulative impact assessment. The main objective of a REA in terms of impact assessment is normally to estimate (or forecast) the potential cumulative impacts of planned activities on a regions environment, natural resource base, and socio-economic conditions, taking into account the baseline situation, and activities included in the inventory of plans and projects, and expected spontaneous developments.

The purpose of this assessment is first and foremost to be able to analyze the environmental impacts of the planned activitiesusually formulated as a development planagainst other options, before the opportunity for realistic evaluation of these alternatives is closed (see below under Analysis of Alternatives). The underlying objective should be to determine which options present a framework for environmentally sustainable development. Step one. The first step is to undertake limited, preliminary impact assessments of the major individual investments under consideration, covering potential direct and indirect, positive and negative impacts. These assessments cannot be as rigorous and thorough as for project-specific EAs, due to the fact that REA is an upstream planning instrument normally undertaken at a pre-feasibility stage where investments have not yet been outlined sufficiently to make detailed analysis feasible or economically justified. If any proposed subproject is expected to cause significant impacts, the REA should recommend a course of action to address them, including carrying out a project-specific EA. Step two. The second step is to look at the sum total of the individual activities and estimate their cumulative effects. For example, regional authorities may be contemplating a plan within a largely forested watershed which includes agricultural expansion, hydropower development and rural road construction. All these activities would result in some deforestation as a direct impact. There would most likely be induced impacts from these developments, such as migration and land settlement along roads. This, in turn, could lead to increased deforestation as an indirect impact. Improved access for loggers could lead to additional deforestation. The REA needs to take into account these impacts, first separately and then in terms of their added and cumulative impact. The cumulative effect has often been found to be bigger than the sum of its parts. It may be, for example, that the total deforestation from the three developments described above would reduce the forest cover of the river basin to such an extent that remaining tracts would be too small as to be viable habitats for wildlife. Loss of wildlife would then be a cumulative effect on top of the deforestation. Or, the cumulative impact might surface in the form of rapidly increasing soil erosion and siltation of rivers. This in turn could negatively effect the hydropower development as well as other economic activities. Where project-specific EAs could effectively assess the impacts of each individual investment and develop good mitigation plans to limit and manage these impacts, they would not be well placed to analyze such cumulative and interactive effects. Only a REA type assessment can determine how multiple projects, in a variety of sectors, may cumulatively affect the environment and each other, including

the risk that the economic viability of one or several projects might be undermined. In many cases, it may also be important to include economic policies in the cumulative impact assessment. For example, tax and subsidy rates on extraction and use of natural resources may greatly influence extraction rates and patterns of consumption. Sometimes, changing such variables may alter the significance of environmental impacts more profoundly than technical or other changes in plans and programs. Economic analysis is needed to address such issues, a topic covered in a forthcoming Update. Analysis of alternatives. The REA should always compare the results of the impact assessment against the impacts of realistic alternative plans, if they exist, and the no plan (business as usual) option. If there are no plans under consideration other than the one the REA is related to, the REA should not develop elaborate alternatives just for the sake of analysis, but rather compare the plan with other broad strategic options for the region. Perhaps the most important purpose of a REA is to analyze the environmental costs and benefits of major alternative strategy and investment options, and recommend a course of action that will best achieve environmental sustainability. Comparison of alternatives becomes particularly important when a region is under environmental stress, where severe competition for scarce natural resources is predicted, or when considering plans for a relatively pristine area. Options should be evaluated carefully for extent of irreversible impacts and effects on long-term productivity of the regions natural resources. Alternative investment options should be considered in terms of their individual and cumulative impacts. Private as well as public sector activities may be included in the analysis, as appropriate. Where several donors are involved in the region, the REA should review their existing and/or planned activities and, if necessary, suggest ways to improve coherence and complimentarity. Alternative mitigation options may also be included in the analysis, where preparation of individual investments is relatively advanced. Experts responsible for REA preparation may propose mitigation measures and compare them with those proposed by the government or third parties. In other cases, this analysis is better included in the mitigation plan. It is useful to prepare a matrix summarizing the project-specific and cumulative impacts of the various options under consideration. A forthcoming Update will provide further guidance on analysis of alternatives in EA.
9

Recommendations towards an optimal regional investment plan. The REA can be a useful vehicle to design an optimal regional investment plan based on the previous analysis, by feeding into strategy formulation, the identification of potential projects, development of environmental guidelines for certain sectors or activities (e.g., land zoning), and the formulation of a detailed and coherent investment or action plan. More often, however, the mandate for the REA will only allow it to give recommendations for improving an existing plan. At any rate, REA provides unique opportunities to internalize environmental factors into regional development planning to minimize future environmental costs and ensure long-term economic and environmental sustainability. Environmental management strategy. Based on its findings, the REA should propose a strategy for strengthening environmental management within the region. The plan would normally include: (a) measures to ensure that environmental impacts of proposed projects are adequately mitigated; (b) general guidelines for long-term environmental monitoring; and (c) a plan for institutional strengthening. Mitigation. Mitigation measures are generally of a detailed, technical nature, and therefore normally developed within project-specific EAs. However, REA can be an effective tool for identifying, at an early stage, projects that will require special mitigation measures. A REA may also suggest broad solutions for reducing negative impacts on important regional environments and natural resources, or develop mitigation guidelines for specific activities. For example, construction of secondary sewage treatment plants may be recommended to contain the effects on coastal ecosystems of anticipated rapid urban growth. Or, requiring catalytic converters on automobiles might be a solution for containing the growth of urban air pollution. Reforestation along important waterways to reduce soil erosion is another example of a mitigation measure that requires a comprehensive, strategic approach. REA may be an effective vehicle for recommending mitigation measures that can only be implemented at the regional level for regulatory or economic reasons. For example, a regional EA may fall within an urban jurisdiction with mandate to set its own air quality standards. Similarly, costly mitigation solutions that require economies of scale (such as a hazardous waste incinerator) are normally best introduced in conjunction with development planning at a regional, sectoral or even national level.

Monitoring. The REA should provide general guidelines for long-term environmental monitoring to ensure adequate implementation of the regional program or set of projects and evaluate progress. The findings of the baseline data section should be used to measure progress over the course of implementation. The REA should also recommend measures needed to collect and organize needed data. Geographic Information Systems (GIS) should be considered, if not already in place, to support regional environmental monitoring activities. Institutional strengthening. The REA might recommend training or additional hiring, or more sweeping changes such as reorganization of units or agencies, and redefinition of roles and responsibilities. Under certain circumstances, a REA may recommend that a new institution be created to manage a regions natural resources base, e.g., a river-basin authority. This section might also include recommendations on policy and regulatory instruments for environmental management in the region. If not already in place, an environmental screening and review process for individual projects should be designed and instituted. For further reading Little has been written about regional environmental assessment. However, there is a body of literature on regional development and environmental planning, such as those indicated below. In addition, the REA prepared for the Argentina Flood Protection Project is highly recommended reading and is available (in Spanish) from the Banks Public Information Center. The Environment Department also has available TORs for REAs. An Update is forthcoming on the application of REA for development planning within river basins. The existing Updates referenced in this document also provide guidance on aspects relevant to REA (e.g., coastal zone management, Geographic Information Systems, and public involvement in EA). Other relevant sources of guidance are: Asian Development Bank. 1988. Guidelines for Integrated Regional Economic-cum-Environmental Development Planning. Environment Paper No. 3 (two volumes). Manila. Organization of American States. 1984. Integrated Regional Development Planning: Guidelines and Case Studies from OAS Experience. Washington, D.C.

This Update was prepared by Olav Kjrven. Based on Bank policy and procedures on Environmental Assessment (EA) (Operational Directive 4.01), the EA SOURCEBOOK UPDATE provides up-to-date guidance for conducting EAs of proposed projects. This publication should be used as a supplement to the Environmental Assessment Sourcebook. Please address comments and inquiries to Olav Kjrven and Aidan Davy, Managing Editors, EA Sourcebook Update, ENVLW, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. S-5139, (202) 473- 1297. The Bank is thankful for the Government of Norway for financing the production of the EA Sourcebook Update.

10

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

December 1996 Number 16

Challenges of Managing the EA Process

Successful implementation of environmental assessments (EAs) must address a number of challenges for those managing EAs and preparing EA reports, particularly in countries with limited EA experience. This Update attempts to identify the most important of these challenges (including selection and management of EA teams, impact identification and prediction, evaluation of impact significance, and information presentation) and suggests approaches to addressing them. It supplements information in chapter 1 of the EA Sourcebook. Background EA is a tool to manage the development process and not just a technical aid in project appraisal. Consequently, managing the EA process requires a combination of scientific judgment and management skills. EAs are not scientific studies; instead they use the results of scientific studies, and techniques based on scientific principles, to provide defensible and justifiable conclusions which form a partial basis for decision making. They are usually undertaken within strict budgetary and time constraints and in situations where data may be unavailable or of dubious reliability. EAs are also often undertaken in situations where social and environmental conditions can change quickly. The key objectives in undertaking EAs are to implement an effective EA process and produce a useful EA report. Both are important, but in different ways. The EA process is related to the project conceptualization, preparation and implementation phases and should influence the production of an environmentally sound project. The EA report (and preliminary or interim versions) synthesizes results and is a formal demonstration to key decision-makers, NGOs, and the public that the EA has been done according to good professional practice. Key technical and managerial challenges to implementing an effective EA process include: Selection of an appropriate EA team, in terms of technical and managerial capabilities; identifying the likely environmental impacts and determining their anticipated relative significance in the early stages of the process; determining the range and type of baseline data needed to make defensible and robust impact predictions, collecting these data and making the predictions; evaluating the significance of predicted social and environmental impacts; and effectively presenting the information obtained at relevant decision making stages. Collecting, evaluating and presenting relevant environmental information for use in project planning and decision making are important aspects of the EA process. For most projects, the output consists of predictions on how the environment may change if specified development alternatives were to occur and how best to manage the anticipated environmental changes. Decisions occur throughout the development process from initial concept to decommissioning or abandonment. In the context of Bank projects, there is no single big decision (with the exception of the posssible decision not to proceed with a project), but a sequence of linked decisions often involving a variety of decision-makers. EA team structure and management EAs require a variety of specialist inputs depending on the potential impacts to be addressed; for major developments these can be extensive. To predict environmental impacts it is necessary to firstly evaluate the elements of the existing environment (air, land, water and social) and interfaces between these elements. The results are related to alternative development scenarios to assess environmental impacts. Thus, specialists may be needed in topics as diverse as air emissions dispersion modeling, health risk assessment, prediction of pollutant migration within aquifers or ecological assessment, deInsert in Update Binder chapter 1

Box 1. Role of EA manager for a regional EA of Victoria Falls, Zimbabwe & Zambia
This regional EA identified environmental implications of cross-sectoral development scenarios (high, medium and low growth) for an area within a 30 km radius of the Victoria Falls World Heritage site on the border between Zimbabwe and Zambia, up to the year 2010. The predicted impacts of the scenarios show the expected state of the environment in 2010. Based on the EA results a management plan was prepared to help avoid, or reduce, the severity of significant adverse changes and to assist realization of potential benefits. The EA manager performed the following key managerial functions: Established the characteristics of the scenarios investigated (they were updated and revised throughout most of the EA work); issued detailed guidance notes to each specialist instructing them on the scope of their work, the baseline information needed, the consultations required, and the type of data needed on impacts for inclusion in the EA Report; organized and managed face-to-face discussion with individual team members to deal with specific issues and periodic team meetings to discuss results obtained and further work; managed the continuing program of consultations with stakeholders; prepared interim and mid-term reports on progress of the EA for an EA Steering Group; and produced the draft and final EA reports.

always involve consultations with potentially affected people and relevant local organizations (see Update no. 5), is often called scoping, environmental reconnaissance or initial environmental examination. Category A projects need such scoping in order to focus the EA process on the key environmental issues. It can also be a cost-effective approach for many Category B projects since the results of the scoping may be documented as the main written output of the EA process. When the scoping determines a need for further EA work (which would be the normal situation for Category A projects and also some Category B projects), it is important to translate the scoping results into coherent terms of reference (TORs) and schedules for undertaking this work. Scoping should therefore normally precede development of detailed TORs for the EA or, alternatively, be an integral part of preparing the TORs. Experience shows that TORs ground truthed through scoping are more focused on the key environmental issues and risks than desk based TORs, which tend to demand coverage of all potential issues. Too often, such TORs result in production of voluminous and unfocused EA reports. Since the outcome of scoping (for example, a short report or a TOR) may significantly influence the focus and cost of any further EA work, it should be subject to review by the Bank and Borrower prior to proceeding with any such work. Impact identification is a continuing process which occurs during screening and scoping and continues through impact prediction as new information becomes available and insights are obtained. A systematic and rigorous approach to identifying impacts (as an aid and supplement to scoping activities and for providing a framework to guide EA implementation) can be based on the following methods: Checklists; interaction matrices; networks; and overlay mapping and GIS.

The Environmental Management Plan will be jointly implemented by the Governments of Zimbabwe and Zambia.

pending on the scope of the project and characteristics of the existing environment. Each EA team needs an effective manager, experienced and familiar with handling the impacts to be analyzed. Only someone with this expertise and experience can ensure the quality of individual impact studies and integration of results into an overall picture of environmental consequences. The appointment of a competent EA manager makes, perhaps, the greatest contribution towards effective EA. Identifying impacts The initial identification of potential environmental impacts takes place during environmental screening (see Update no. 2). Following screening, a fieldbased examination is often required to identify more precisely the range of relevant impacts and indicate their relative importance. This examination, which should
2

Each has advantages, drawbacks and potential application in other EA tasks. They can be used in combination as well as singly; for example, a matrix can be used to identify direct impacts which in turn can be used as a basis for constructing a network. Checklists There are many different types of checklists. Some are lists of environmental factors while others list environmental factors and developmental actions likely to cause impacts. By systematically comparing factors and actions likely impacts are identified. Also, there are checklists, listing the typical impacts of specific project types which are easily available. The Banks Operational Directive onEA (OD 4.01, soon to be reissued as OP/BP/GP 4.01)

Box 2. Example of a network diagram

Soil erosion into streams Change in slopes Alteration of stream turbidity Decreased light penetration

technology and expertise allows overlay mapping to be incorporated within a Geographic Information System (GIS) (see Updates nos. 3 and 9). Direct and indirect impacts can be identified, broadly and generally, by superimposing a map showing the proposed development (with required infrastructure such as roads and transmission lines) and associated projects (for example, a new quarry or one which will expand operations) onto the composite map. Overlay maps are useful for identifying impacts and comparing alternatives for all types of development, but achieve most usefulness for EAs of linear developments (pipelines, roads and transmission lines) and multiple investments or activities resulting in cumulative impacts. Impact prediction

Cut & fill operation

PROJECT ACTIVITY INITIAL CONDITION PRIMARY IMPACT SECONDARY IMPACT TERTIARY IMPACT

contains a checklist. All such checklists can be useful sources of information to guide and structure scoping and other impact identification activities. Interaction matrices A matrix is a diagram which links environmental features, or potential environmental impacts on these features, with actions associated with a proposal. Matrices may be constructed before scoping, and used to guide scoping sessions with participants discussing the significance of proposed actions for environmental features. The decision is recorded by marking the box representing the intersection between an action and an environmental factor (if no impact is expected, the box is left blank). The completed matrix forms the basis for determining EA work and can be updated as necessary. One difficulty with matrix methods is the inability to clearly identify the links between impacts. The systematic consideration of indirect impacts, and their linkages, is vital in all EAs. Networks The network, or flow diagram, was developed to identify the links between different impacts and the ways in which aspects of the environment might be affected by more than one impact pathway. A generic structure for a network is shown in box 2. This model can be adapted to meet particular needs. Networks may be partially constructed in advance of scoping sessions or they can be built-up as part of the session. Once constructed they provide a framework to guide EA work and can be updated or amended as work progresses. Overlay mapping/GIS To use overlay maps it is necessary to prepare maps that show the position, nature and extent of natural and human attributes of an area. Attributes which may be mapped include surface water bodies, agricultural land, wetlands, settlements, and cultural resources. The features mapped are those which are expected to be sensitive to the project. Individual transparency maps are overlaid to provide a composite picture of the environment in terms of its basic components (see box 3). If available, computer

Prediction is the technical heart of EA and is an attempt to assist decision making by isolating and reducing uncertainty with respect to anticipated environmental changes. Prediction is a complex activity and the following techniques may be used to quantify the nature and extent of environmental changes: Mathematical models (such as noise propagation models, air or water dispersion models, income multipliers); physical models (such as wind tunnels and hydraulic models of, for example, estuaries); field experiments; structured or semi-structured approaches to produce a mix of qualitative and quantitative predictions (for example, landscape change and social impacts); and scientific experience and judgment. Most EAs use a mix of these techniques with many relying heavily on the latter two. EA team members must determine the range and type of baseline data needed to make defensible and robust impact predictions. These requirements dictate the technique to be used and not the reverse a common misconception in predictive modeling. A risk- based approach can be useful in determining the appropriate degree of detail for data collection. In general terms, where uncertainties regarding the occurrence of potential impacts are large, and the consequences of the impact occurring are significant (for example, deteriorating air quality affecting the health of people or crops), detailed data collection is appropriate. If the potential consequences are not significant, detailed data collection is inappropriate regardless of the level of uncertainty. Given the lack of standardized guidance it is unsurprising that the quality of impact predictions is a common technical weakness in EA work.
3

Box 3. Illustrative example of overlay mapping

Composite constraints map

Ground water regime

+
Natural habitats

+
Land capability

+
Population distribution

Impact prediction must not only concern itself with estimating the magnitude or scale of change, but also provide information on the following aspects of impacts: Duration (time period over which they will occur); likelihood or probability of occurrence (very likely or unlikely); reversibility (natural recovery or aided by human intervention); area affected (size and whether near or far from the project); number (and characteristics) of people likely to be affected and their locations; and transboundary aspectsdo impacts cross national borders? There is also the issue of determining distribution of impacts. For example, impacts may be identified individually (noise, ecological and health effects) and sensitive receptors determined. It is only following the prediction of each impact that the geographic overlap of impacts and their relative spatial and temporal distribution may be determined. If specific human settlements or natural habitats are subject to a range of impacts, the cumulative impact has to be identified and evaluated. The overlay mapping method is useful in this regard. It should not be forgotten that the sustainability of a project can be influenced by the impacts it causes. EA practitioners sometimes incorrectly view the project as active and the environment as passive. A project and its environment form a dynamic system with interacting
4

components. Often, there are feedback loops through which the sequence of environmental changes caused by a project can ultimately affect it. A well-known example concerns the reduction in the efficiency of hydro-power projects in the tropics due to increased sedimentation (see box 4). Such threats to project sustainability will not occur in every case, but it is important that the EA evaluate such possibilities. Early warning of possible threats, and initiation of mitigation measures to prevent or reduce their severity, can make important savings to projects and regional or even national economies. Baseline studies Where there is a lack of information, many EA teams have to revert to gathering baseline data before proceeding to the stage of impact prediction. This is done to establish an overview of the environment; its main features and key natural processes, and any identifiable trends (for example, whether air quality is stable or declining). The role of baseline studies in EA is frequently misunderstood. Traditionally, obtaining and interpreting information to describe existing environmental conditions was needed to assist identification of impacts. This is now the major objective of scoping and does not require large expenditures of time and resources. Locally based participants in scoping bring their environmental knowledge with them to the scoping sessions. Participants from central or local government agencies and other non-local organizations may provide secondary sources of information.

Environmental constraints

Under certain circumstances it may be necessary for EA teams to collect information prior to scoping. Usually, little time will be available for detailed surveys. Experienced EA practitioners need little guidance on how to obtain sufficient information for scoping. Others can use checklists. However, most checklists are generic and strict adherence may result in time and resources being wasted on unnecessary data collection. The best approach to baseline studies is to use the impacts identified in scoping to guide data collection. This does not preclude subsequent gathering of other data, but does help to avoid non-directed baseline data collection. The key to cost-effective baseline studies is to strike a balance between obtaining sufficient information to describe existing features, their inter-relationships, and overall environmental status or quality (realizing that impact data will be synthesized at the end of the EA) while obtaining sufficiently detailed data on current status and trends to enable specific impacts to be predicted. The danger lies in collecting too much detailed information of limited practical value to EA. In baseline studies, the relevance of the data is much more important than the amount. EA managers should resist deriving false comfort from collection and presentation of large amounts of baseline data. There are a number of approaches to do this quickly and cost-effectively. Site visits are useful, but may be difficult because of lack of access, support facilities and the nature of the terrain. In such circumstances it is useful to begin with maps, aerial photographs and satellite images. Considerable information can be obtained rapidly from such sources, especially if there are images or photographs taken over a period of time. Where practicable, a well-focused site visit should be undertaken to ground truth data. If time and budget allows, a GIS may be initiated particularly for EAs of large-scale or linear projects, such as roads or pipelines, and for sectoral and regional EAs (see Updates, nos. 4 and 15). EAs always face problems of data availability and reliability. Some techniques for acquiring data have been outlined above. Other sources of data include: universities and research institutes; local and national government departments and agencies; NGOs; and local people. Such information sources should be used initially and primary data collection undertaken only if there are serious data gaps, or doubts regarding reliability, as it can be expensive and time-consuming. The issue of time and resources for data collection is a constant challenge in EA work. For many ecosystems, habitats and species, little information exists on their behavior, variability and trends. How is it possible to predict changes when so little is known about what exists? One approach is for the Task Manager and implementing agency to allocate sufficient time to obtaining data spanning climatic cycles (wet/dry seasons; summer/winter).

For example, in India the EA law requires that certain EAs are allocated at least one year to achieve this objective. Many projects have long lead times and EAs should be scheduled to tie in with the key points in the project cycle. If, for example, ecological impacts in an area with limited data available are identified as potentially significant, work should begin early enough to collect information for the main seasons. Those implementing EAs have a responsibility to ensure that they exercise best professional judgment as to the minimum data needed to describe the environment and to make defensible predictions. If essential data cannot be gathered, additional time or financing might be requested. Alternatively, data gaps or weaknesses should be identified and evaluated in the EA report. One final issue deserves attention. Almost all EA guidelines state that the impact of alternative proposals should be compared, at a minimum, with the no development or no-action situation. Unfortunately, many consider the no-action scenario to be identical to the current environmental conditions (determined by baseline studies). Rather, the most useful comparison is the situation as it would be if the proposal did not occur. For example, the no-action alternative to a regional strategy to shift energy supply from coal to natural gas would need to consider the implications of additional coal fired capacity to supply projected increases in demand. The environment is not a static entity; it changes due to natural processes and human activities. Thus, in EA it is important to consider, explicitly, the moving baseline. This is not easy and attempting to do so makes EA work more complex. Nevertheless, an attempt should be made to identify the most significant causes of current environmental change and to project their effects into the future. Also, it is necessary to identify other projects, either underway or planned (with a high probability of implementation), which may affect the environment and perhaps even the sustainability of the project subject to the EA. The condition of the environment, resulting from this analysis, should be used as the baseline. In many situations, it may only be possible to make a best guess in relation to specific environmental features. Evaluating impact significance Quantifying impacts is an objective, technical task where-as evaluating significance is subjective and political. There are two aspects to assigning significance. First, there is the importance of individual impacts. For example, if a proposed route alignment for a new road will increase night-time noise levels by 10 dB(A) and ground-level concentrations of NOx by 5 g/m3, are these significant in terms of sleep disturbance and respiratory effects? Secondly, if an alternative alignment would increase noise levels by 5 dB(A) and NOx by 10
5

Box 4.

Saguling hydropower, Indonesia

The EA of the Saguling hydropower project in Indonesia considered the likelihood of feedback loops affecting the viability of the project. Using a network, an attempt was made to identify the ways in which the impacts of the project might interact, through time, and reduce the likelihood of it achieving some of its objectives. The EA identified the strong possibility that local people, resettled from the area to be inundated, would return and attempt to establish farms on the slopes above the reservoir. This would involve deforestation to create areas for cultivation. Over time, erosion would increase and the reservoir would receive enhanced sediment loads which would reduce the capacity of the reservoir and hence its usefulness. Also, run-off from such cultivated land might reduce the potential benefits from the planned aquaculture projects in the reservoir. The EA identified a number of ways in which the viability of the project would be threatened and suggested mitigation measures to prevent them from occurring.

components such as noise levels or water quality also have inherent significance criteria (sometimes referred to as Thresholds of Concern). For example, the threshold of concern for drinking water quality may be based on national or World Health Organization (WHO) standards. Thresholds can represent an objective to be achieved or a limit not to be exceeded. Any impact which exceeds a limit or does not achieve an objective should be considered significant to decision makers. Secondly, where environmental aspects are not covered by predetermined criteria, significance must be determined in the context of the project. During scoping, stakeholders may assign significance to both environmental features and likely impacts. Prior to preparing a draft EA report, the significance of predicted impacts should be discussed with key stakeholders with a view to building consensus. The more representative the stakeholders, the greater the likelihood of the consensus holding. Assigning significance, therefore, should be a joint effort on the part of the EA team and stake-holders. Values, special interest considerations, and best professional judgment can be applied jointly to assign or re-assign significance. Much of this work will form a political process with discussion, argument, negotiation and compromise. There are a number of approaches to help groups reach a compromise or consensus, as outlined in the Banks Participation Sourcebook. Comparative evaluation of alternatives As a summary display format, checklists can be adapted or expanded to comparatively illustrate alternatives. Predicted impacts (for example, amount of agricultural land or natural habitat lost) of development alternatives may be converted into simple scales such as low, medium and high. A reader can readily see how each alternative compares in terms of impacts. Unfortunately, where more than two or three alternatives are being evaluated, it is usually not easy to identify one preferred option. This problem has encouraged the development of methods that place all impacts on a single scale (scaling) and assign a numerical expression of relative importance to each impact (weighting). Once done, it is possible to manipulate, mathematically, the EA results to form a total score (or index) for each alternative option. This score includes all beneficial and adverse impacts and enables, easily, a preferred option to be identified. Box 6 shows the basic structure and operating principles of these methods. A variation on this method is provided by an EA of water supply options for Kathmandu (see box 7) which incorporated a probabilistic dimension. There are different views of the validity and usefulness of such methods. Many criticize them for being:

g/m3, what is the relative importance of these alternative impact scenarios in relation to each other and to other predicted impacts? Explicitly addressing such tradeoffs is necessary if a number of alternatives are being compared. The concept of significance in EA presents theoretical and practical difficulties. The most common approach is to assume that significance, once assigned, is constant. It is, however, not appropriate to maintain a constant view of significance if additional data and experience indicate that an impact scenario needs revision. There is a growing realization that EA should incorporate a more fluid concept of significance depending on knowledge at the various EA stages and on the views of important stakeholders. The role of EA throughout the project cycle (particularly during implementation) makes a rigid assignment of significance unjustifiable. Significance should be determined on the basis of biophysical context and sensitivity of receptors; socioeconomic and cultural context; characteristics of the impacts such as magnitude, duration, and reversibility; and applicable environmental laws and regulations. Box 5 illustrates how these factors may be combined, in a local context, to produce differing views of significance. Currently, there are two basic approaches to assigning significance (they are not exclusive and should normally be used in combination). Firstly, any formal predetermined criteria must be respected. For example, international treaties and conventions, national legislation, government policies and regional/local plans will often have established that certain natural resources or environmental and cultural features are important (sometimes through formal protection). Ambient standards or environmental quality objectives for specific environmental
6

Box 5. Differing perceptions of environmental benefits: Kathmandu water supply options, Nepal
An EA was undertaken for three options to supply water to the KathmanduLalitpur urban area. Two of the options required a balancing reservoir to be located near the urban area. Kathmandu is built on the Bagmati river which is sacred to Hindus. The Pashtupatinath temple complex, used for ritual bathing and cremation, is sited on the banks of the Bagmati. The Bagmati has periods of low flow in which the effects of existing water pollution are exacerbated. Those undertaking the EA saw a possibility of augmenting the low flow of the Bagmati with water from the balancing reservoir to reduce the adverse effects of water pollution during natural low flow conditions. This was seen as a positive impact and potential benefit of the water supply scheme. However, this proposal was viewed differently by many devout Hindus who considered the addition of water, from a source external to the Bagmati river system, as a type of ritual pollution and therefore an adverse impact despite the water quality advantages which all agreed would occur.

tation of these methods and in critical review of their results is limited because of the complexity which characterizes their operations); and reductionist (it is difficult to deal with indirect impacts and feedback loops). There are also advantages to using these methods in EA, as they provide answers to complex questions within the restricted timescales and budgets which tend to characterize the development process. In recent years, attempts have been made to improve them by incorporating a wider public input into the weighting schemes, estimating probabilities and using sensitivity analysis to test the results. Unfortunately, the improvements can make the methods more complex and less amenable to critical scrutiny not only by the public, but also by experts. These methods are perhaps most useful where there is a large number of alternatives to be assessed and a need to discard some early in the EA process. In such circumstances a simple and basic use of this type of method might be appropriate. Additional guidance on comparative evaluation of alternatives is given in Update no. 17: Analysis of Alternatives in Environmental Assessment. Information presentation

Simplistic (reducing a complex, multi-dimensional environmental reality to uni-dimensional scales and indices); falsely objective (the numbers imply some form of scientific credibility, whereas the scales and weights are often the product of the subjective views of the EA team only); technocratic (using these methods may force specific, single decisions on decision makers); non-participatory (the extent to which the public and other stakeholders can participate in implemen-

The role of EA reports is to inform all stakeholders of the expected impacts of alternative development actions and the mitigation measures which will be needed. Readers of EA reports will include experts in environmental issues, interested parties such as NGOs, and members of the public. As a result, EA teams have a responsibility to communicate effectively with a very diverse audience. Maximum use should be made of presentation techniques which facilitate effective transfer of information.

Box 6. A comparative analysis of three alternatives using scaling-weighting checklists

(Relative importance)
Environmental factors 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Flora Fauna Water quality Air quality Noise levels Wildlife Land use Topography Soils Micro-climate Relative importance 7 6 9 10 7 5 8 3 6 5

(Magnitude of impact)
Magnitude of impact on each alternative A B C 7 10 5 6 6 2 5 5 4 7 4 7 6 8 4 2 7 2 8 6 10 8 3 1 7 2 2 9 9 7

Grand index

Grand index A 49 36 45 70 42 10 64 24 42 45 Total 427 B 70 36 45 40 56 35 48 9 12 45 386 C 35 12 36 70 28 10 80 3 12 35 321

Step 1

Step 2

Step 3

Step 4

It can be useful to determine, in advance, the desired length of the various sections of an EA report and use these as a guide in report preparation. It also is helpful to decide which information can be appended on the basis that it provides background to the main findings of the EA. All EA reports should contain an Executive Summary providing information on the key issues pertinent Box 7. Comparative evaluation of water supply options
A comparative assessment method was used for assessing the biophysical environmental impacts of three water supply options for KathmanduLalitpur, Nepal. Each impact was described in terms of magnitude (major, moderate, minor); extent (regional, local, site only); and duration (long-term, medium-term, short-term). A simple scale was devised for each of these characteristics and specific impacts were assigned a value from the scale as follows: Magnitude Major 60 Moderate 20 Minor 10 Extent Regional 60 Local 20 Site 10 Duration Long-term 20 Medium-term 10 Short-term 05

to the approval decision. It should not be a summary of all the contents of the EA report and should be restricted to 1020 pages. The EA report itself should normally not exceed 150 pages (excluding technical annexes). The EA report should be written in a consistent, simple and clear style. Technical terms, acronyms and jargon should only be used when essential and a glossary should be supplied to explain the meaning of such terms. It is not acceptable to produce an EA report which has been compiled by binding together a series of specialist reports produced by different experts at various times in the EA work. A coherent text with maximum use of cross-referencing is needed. The use of visual aids is strongly encouraged to help clarify locations of places or geographic features referred to in the text, the extent of environmental resources, locations of people or aspects of the natural environment affected by the project, and sampling locations. However, too many EA reports contain maps and other diagrams which are poorly prepared and cannot be reproduced easily and clearly. For location maps, it is recommended that a limited number of representative base-mapping scales and aerial coverages be used consistently throughout the text. Methods used to help identify impacts, such as checklists or matrices, can also be used to present results in a visual summary. For further reading Biswas, A.K. and Q. Geping (eds). 1987. Environmental Impact Assessment for Developing Countries. Tycooly Publishing. London. Canter, L. 1995. Environmental Impact Assessment. 2nd Edition. McGraw-Hill. New York. ENVLW. 1996. The Impact of Environmental Assessment: The World Banks Experience. The World Bank. Washington, D.C. Morris, P. and Thrivel, R. (eds). 1995. Methods of Environmental Impact Assessment. UCL Press Ltd. London. Wathern, P. (ed). 1988. Environmental Impact Assessment: Theory and Practice. Unwin Hyman. London. UNEP. 1996. Environmental Impact Assessment : Issues, Trends and Practice. United National Environment Programme. Nairobi.

Assigning these numerical values to an impact incorporates an element of weighting as a site-specific, shortterm impact of major magnitude is less significant than one of minor magnitude, but of regional extent and longterm duration (75 and 90 points respectively). The following impact scores were derived for reduced downstream water quality and increased erosion: Impact Magnitude 10 60 Extent 20 20 Duration 20 20 Total 50 100

Water quality Erosion

Probabilities were estimated on a scale of 0.1 to 1.0. The probability for each impact was multiplied by the total scores. When multiplied by the probabilities for the three options (a) Melamchi, (b) Modified Melamchi and (c) Lower Rosi Khola, the following scores were derived (probabilities are in parentheses): Impact Option 1 Option 2 Option 3

Water quality 50(0.1) = 5 50(0.1) = 5 50(0.2) = 10 Erosion 100(0.4) = 40 100(0.4) = 40 100(0.2) = 20 Such scores were obtained for all impacts of the three options and an aggregate total obtained. The option with the smallest total was the least environmentally damaging.

This Update was prepared by Safei El-Deen Hamed, Ron Bisset of Scott-Wilson Resource Consultants, and Aidan Davy. The EA Sourcebook Updates provide guidance for conducting environmental assessments (EAs) of proposed projects and should be used as a supplement to the Environmental Assessment Sourcebook. The Bank is thankful to the Government of Norway for financing the production of the Updates. Please address comments and inquiries to Olav Kjrven and Aidan Davy, Managing Editors, EA Sourcebook Update, ENVLW, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. S-5139, (202) 473-1297. E-mail: eaupdates@worldbank.org.

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

December 1996 Number 17

Analysis of Alternatives in Environmental Assessment

The Banks Operational Directive OD 4.01 on environmental assessment (EA) calls for, inter alia, systematic comparison of the proposed investment design, site, technology, and operational alternatives in terms of their potential environmental impact. Despite this emphasis on evaluating alternatives, the recently completed Second EA Review of Bank-financed projects found that analysis of alternatives is often inadequately addressed. Reasons include the timing of key decisions in relation to EA and lack of methodological guidance. As a result, many EAs focus only on minimizing the adverse impacts of a given project proposal. This EA Sourcebook Update builds upon the lessons learned from the Second EA Review, and outlines how analysis of alternatives can be undertaken at different levels of development planning, through project-specific, sectoral and regional EAs. Although it does not attempt to provide a thorough review of decision methods for comparative assessment of alternatives, this Update aims to provide broad guidance on comparative assessment and an introduction to systematic methods for comparing alternatives. This Update expands on existing material in chapter 1 of the EA Sourcebook. Background Since the introduction of the EA process and subsequent development of EA methodologies and legislative provisions, the analysis of alternatives has been one of the main tenets of EA policy and procedures. Indeed, a thorough, unbiased and transparent assessment of investment alternatives from an environmental and social perspective (as well as a technical and economic standpoint) is one of the most important contributions EA can make to improving decision-making. Alternatives analysis in EA is designed to bring environmental and social considerations into the upstream stages of development planningproject identification and earlieras well as the later stages of site selection, design and implementation. In the absence of such consideration, those steps in the project cycle are taken solely on the basis of technical feasibility, economics, and political preferences, and the EA for such a project tends to be directed to supporting or affirming a project proposal. At best, EA becomes a damage limitation exercise, with the benefits restricted to identification of mitigation measures. Whereas environmental and social analysis at an earlier stage might have revealed another cost-effective way of achieving the same project objectives at lower environmental or social cost (measured either by the severity of the impacts or the costs of measures to mitigate them), the likelihood of finding it late in the process is small. Furthermore, even if such an option were to be found in the project EA, it often cannot be implemented without disrupting project preparation in a manner that is so time-consuming and expensive as to be impractical. Alternatives that differ in environmental and social impact may be found at several levels in development planning: alternative policies and strategies at the national and sectoral levels; alternative patterns of economic growth, land development and resource use for regions; and alternative sites, technologies, designs and operating procedures for individual projects. Within the Bank, too, the environmental and social dimensions of alternatives can be considered in economic and sector work (ESW) and formulation of a Country Assistance Strategy (CAS). Ideally, the environmental impacts of the alternatives at the higher levels would be evaluated and compared as an integral element of the planning process, in parallel with the economic analysis, and hence

Insert in Update Binder chapter 1

would have been taken into account prior to project identification. In some instances, environmental considerations are partially factored into development planning, but the process is often neither transparent nor systematic. However, the EA process at the Bank and in many countries is evolving in that direction, with increasing efforts to introduce environmental concerns further upstream and to combine or at least closely link EA with feasibility studies. However, the evolution is not complete, and it is still common to encounter proposed projects in which the decisions up to and including selection of site and technology have been made with little or no environmental consideration. Avoidance of the disruptive and sometimes artificial process of examining alternatives once the site and technology have been chosen is one of the main reasons the Bank encourages the use of strategic EA (encompassing both sectoral and regional EA), which provides a framework for systematic analysis of alternatives. This Update therefore describes a tiered approach, in which analysis of alternatives occurs at strategic and projectspecific levels. Not all of the elements described here

will apply in every situation. Bank staff and borrowers should apply this guidance selectively to take advantage of opportunities to examine available and relevant alternatives, and to enhance alternatives analysis in future development planning. Process of alternatives analysis in EA Sectoral EA (see Update no. 4) should be used for distinguishing among alternative strategies and investment programs within a sector (such as power), and for reviewing the effects of sectoral policy changes. Regional EA (see Update no. 15) should be used to compare alternative development scenarios and to recommend sustainable policies and development patterns at a regional level. Both sectoral and regional EA may be used to screen project alternatives based on limited data, prior to more detailed studythey facilitate development of an overarching framework within which individual project proposals can be examined. The subregional energy sector study for the Mekong Basin (box 1) illustrates the benefits of such upstream assessments.

Box 1. Analysis of alternatives in a strategic power & water sector study for the Mekong region
In 1993/94 the Asian Development Bank (ADB) undertook an energy sector study to evaluate supply options over a 25 year period for the Mekong region, an area of rapidly growing energy demand. The objective was to identify the scope, opportunities and means for enhancing cooperation in the fields of water resources, electric power and natural gas between Cambodia, Lao PDR, Myanmar, Thailand, Vietnam, and Yunnan Province of the Peoples Republic of China. Two basic power development strategies were comparednational self-sufficiency versus regional cooperation. In the self-sufficiency scenario, Thailand would have to generate significant new thermal generating capacity based on its high power demand. A regional cooperation scenario however, taking full advantage of the Mekongs hydropower resources, would reduce the need for additional thermal capacity. The two scenarios were compared on the basis of a number of technical, economic, environmental, and social criteria. Some 54 hydropower alternatives were evaluated based on criteria which included: installed capacity; ease of access; dam type and height; degree of flow regulation; area of agricultural and forest land inundated; number of people resettled; multi-purpose use of impounded water; and impacts on ethnic minorities. Options were ranked as having low to severe impacts using comparative indices where possible, such as area inundated or persons resettled per kWh. Environmental and social aspects were combined with technical and economic evaluations to identify potential least-cost projects. The environmental aspects of fuels used in thermal power stations were assessed with regard to impacts on the atmospheric environment acid deposition causing pollutants (sulfur dioxide and oxides of nitrogen) and the greenhouse gas carbon dioxideand human health impacts of particulate matter. On balance, the most sustainable hydropower projects would be a more benign solution to meeting future energy demands than new thermal power options. If thermal capacity is developed without power sharing, the region will experience a serious decline in air quality. Substituting new thermal capacity with renewable hydro through power sharing would reduce atmospheric emissions within importing countries like Thailand, Vietnam, and Yunnan. For example, CO2 emissions are projected to increase from 50 million tonnes (in 1995) to 357 tonnes by 2020 assuming power sharing, or to 432 million tonnes with no power-sharing. Over the same period, SO2 emissions would increase from 0.18 million tonnes to 1.4 million tonnes if power sharing takes place, or otherwise to 1.8 million tonnes. The power sharing scenario would also reduce regional investment in generating capacity by 15-20 percent. Finally, it is worth noting that the evaluation criteria for hydropower alternatives did not encompass ecological concerns. Therefore in ranking the Nam Theun 2 alternative in Laos (see box 2) as one of the most promising options, the conservation value of the Nakai plateau where the reservoir is located, was not considered.

Figure 1. Stages in analysis of alternatives

PROJECT-SPECIFIC EA Define project objectives Produce development proposal
*For natural resource projects, substitute resource management strategies for technologies

Identify None alternative technologies* Define sectoral and regional devlopment objectives Identify alternative development strategies Define resource requirements

Identify alternative locations Screen alternative technologies*

Screen alternative locations

Evaluate selected alternatives Operational & mitigation alternatives Implementation alternatives Design configuration alternatives

Comparatively assess alternatives

Proceed with preferred alternative(s)

STRATEGIC EA

When the framework that strategic EA can provide is absent, as is frequently the case for projects proposed to the Bank, it may be necessary to examine some sectoral or regional alternatives such as sites and technologies within the project-specific EA. In any case, there will be design and implementation alternatives to be examined at the project level. Because it is easier to introduce the alternatives analysis procedure in the context of a project, steps for generation and analysis of alternatives illustrated in figure 1 are discussed for the project-specific case first, then extended to strategic EA. Stages in the analysis process The starting point is the overall project objective. Alternatives that will meet the objective should be identified with as much freedom from limiting conditions as possible, consistent with maintaining reasonableness and practicality. This is fundamental; it fosters the kind of creative planning and engineering needed to reveal options that are truly different, not only in terms of environmental impact, but also cost and ease of implementation. It also sends a message to affected communities and other interest groups that decisions still remain open in the areas usually of most concern to themlocation, size and technologyin contrast to cases in which the nature of the project and its location have already been decided. Box 2 illustrates the point. When first proposed by the Lao Peoples Democratic Republic (PDR) to the Bank and IFC, Nam Theun 2 was a hydropower project with established dam height, reservoir surface area and generating capacity. The 600 MW of power was to be sold to Thailand. Within the project, alterna-

Box 2. Example of a tiered approach to analysis of alternatives: Nam Theun 2

The Nam Theun 2 Hydroelectric Project in the Peoples Democratic Republic of Laos is intended to increase revenues and strengthen the base for economic development by exporting power to Thailand. An alternatives analysis study was designed to ensure that the project complied with the Banks EA requirements. An important aspect of this was public participation appropriate to each stage of the study. The key stages are as follows: Evaluation of the potential for demand side management (DSM); identification and screening of alternative energy sources to hydropower; evaluation of realistic alternative energy sources; comparative assessment of alternatives; identification of hydroelectric alternatives; evaluation of hydroelectric alternatives; comparative assessment of hydroelectric alternatives; and comparison of conceptual and design alternatives for the proposed project.

The final product is designed to be useful in national power sector planning by the Lao PDR; in planning by development finance institutions for their activities in the region and the sector; for planning by private investors; in identifying stakeholder concerns and building consensus; and as an input to preparation and environmental assessment of individual proposed projects such as Nam Theun 2.

Box 3. Screening alternative technologies in EA

The Environmental Improvement and Clean Fuels Project in Thailand is designed to help meet the Royal Thai Governments (RTG) clean fuel specifications introduced in 1993. The project finances upgrading of the equipment and processing facilities at the Bangchak refinery in Bangkok. In addition, Bangchak refinery safety and environmental standards are to be upgraded to meet expected standards to the year 2000. The primary specification changes affecting the refinery products involve removal of lead; reduction of aromatics and benzene levels in gasoline; and reduction of the sulfur content and boiling range of diesel oil. Four alternative processing configurations were considered including deep gas oil hydrotreating (DGHT); hydrocracking; fluid catalytic cracking (FCC) combined with a deep gas oil hydrotreater; and a reduced crude catalytic (RCC) cracker combined with a deep gas oil hydrotreater. The first two stand-alone options require the importation of FCC quality high octane, low aromatics and benzene gasoline which are unavailable on a long term basis in large quantities. The RCC would produce 5 tons per day of spent catalyst compared to one ton per day from the FCC. The comparative assessment was therefore limited to the DGHT, hydrocracker combined with DGHT and FCC combined with DGHT. The preferred option, on the basis of environmental and economic considerations (plus security of supply) was the FCC combined with DGHT.

tions. This might include production, abatement or treatment technologies for industrial processes, or alternative transport modes for transportation projects. The identification of alternative technologies may relate to the entire project or specific components. For example, energy production alternatives might include combined cycle gas turbines, hydropower, coal-fired plants or wind power. Within each of these options, a variety of alternative technologies could be considered, such as flue gas desulfurization alternatives for coal-fired plants. A similar approach should be followed for natural resource management projects (for example agricultural, forestry or water resources development proposals). However, in this context alternative strategies should be considered (as opposed to technologies). For example, if a proposed agricultural program supports conversion of wetlands to rice production, alternative strategies such as resorting to higher yield varieties in existing fields or conversion of other land types should be considered. The potential of wetlands to sustain fish yields or other edible resources should also be assessed (the no-project option). It may also be appropriate to resort to other crop types in alternative locations. At this stage it is important to consult with key stakeholders, including relevant government institutions, agencies and non-governmental organizations (NGOs), on whether the range of technologies being considered is complete. A workshop provides a suitable forum, which would also facilitate development of a public consultation program. Having defined a range of technologies or strategies, resource requirements should be determined for each alternative. This includes energy types and quantities, water, land areas, associated infrastructure, staffing, raw materials/fuel, solid waste and effluent disposal and other requirements plus associated costs. All phases of the project should be consideredsite preparation, construction, operation and, if applicable, decommissioning or closure. Screening of alternative technologies helps to limit the efforts and costs associated with data collection and processing. Screening should be based on factors such as ability of the technology to meet the project objectives, availability of resource requirements (at a macro level), suitability in a particular situation, and the broad environmental and economic acceptability (see box 3). The lead-times associated with bringing projects on-line are also important in determining the suitability of alternatives. The screening process should define a realistic range of alternatives for further consideration. At this stage, a consultation exercise involving key stakeholders should take place in order to seek consensus on the short-listed technologies.

tive dam heights and tailrace alignments had already been considered, and alternative transmission line rights-of-way were being evaluated. Before it would agree to finance the project, the Bank requested that the project proponents return to the overall objective, a 600 MW increase in generating capacity for Thailand, and analyze alternative ways to achieve it. Identifying the alternatives. For energy or water supply projects, an evaluation of the potential for demand-side and supply-side efficiencies should be incorporated at an early stage, which may lead to a refinement of project objectives and consequently the development proposal. However, while demand-sidemanagement and supply-side-management measures complement power supply expansion programs by attenuating electricity demand, they usually are not a substitute for generation capacity expansion in developing countries with rapidly increasing demands. For industrial, infrastructure and energy projects, it is recommended that alternative technologies be identified generically, without reference to project loca4

Box 4. Siting alternatives in an EA for a hydropower project (Pakistan)

The Ghazi-Barotha Hydropower Project is a major run-of-river power project designed to meet the acute power shortage in Pakistan. The main project elements include a barrage located on the Indus River, a power channel (designed to convey water from the barrage to the power complex) and a power complex. Alternative locations for these elements were evaluated based on technical, economic, environmental and social constraints by an interdisciplinary project team and subject to review by an external environmental and resettlement panel. Initial screening of five barrage sites identified by the project consultants resulted in two options being selected for detailed evaluation. The preferred option had less storage capacity than the main alternative, but was preferable in terms of environmental impact. The most economical alignment for the power channel would have necessitated resettlement of an estimated 40,000 people. Movement of the alignment to less densely populated areas, although technically more complex and financially less attractive, reduced the resettlement requirement to approximately 900 people. Additional modifications further reduced the impact on archaeological sites and graveyards. Five power complex sites were initially studied, and three remained for detailed evaluation following screening. Topographical factors determined the preferred option as the environmental implications were broadly similar in each case. Sub-elements of the power complex such as access roads, headpond capacity and embankments were chosen based on environmental and technical considerations. Finally, four alternative alignments were evaluated for the 500 kV transmission line connections to the Peshwar grid station. The selected routes had minimal environmental and socio-cultural impacts. Detailed design of this component will focus on choosing alignment and tower locations with minimal impacts on dwellings, agricultural land and archaeological sites.

lines. Identification of suitable alternative locations should take into consideration the resource requirements identified for the short-listed technologies. The basis for screening alternative locations is similar to that used for screening technologies and includes ability to meet project objectives, resource requirements for short-listed technologies, and broad environmental planning and economic considerations (including the ability to meet Bank economic rate-ofreturn criteria). For example, reasons for rejection of alternative locations could include conflict with existing planning policies or settlements, encroachment into conservation areas or habitat of endangered species, disturbance of archaeologically important sites, opportunity cost of inundating high quality agricultural land, seismic hazard, and risks to groundwater. Significant social concerns, such as involuntary resettlement, often form the basis for rejection of locations. During the initial screening of alternative locations, the concerns of the wider public may be represented by government agencies, institutions, community organizations or NGOs. Once the short-list of alternative project proposals (or project element proposals) is finalized an evaluation of each alternative should be undertaken. Environmental, social and health impacts of the shortlisted alternatives should be determined in sufficient detail to facilitate their comparative assessment. Engineering feasibility and institutional issues should be addressed concurrently, and factored into the evaluation. Where possible, external environmental costs which have not previously been accounted for should be evaluated and internalized within the overall economic analyses to reflect the effects of environmental costs on the rates-of-return of alternatives. Integration of externalities can either be achieved by direct monetary valuation (see forthcoming Update on Economic Analysis in EA) or by the use of comparative assessment techniques described below. The latter can be used to account for environmental, health or social impacts that do not readily lend themselves to monetary valuation, such as loss of biodiversity or cultural heritage. In many cases, the evaluation can be carried out with little fieldwork other than site reconnaissance and review of existing information sources, such as documentation on performance of technologies or methods, aerial photographs and satellite imagery, geological and soil surveys, and hydrologic records. However, the EA team should have resources available for fieldwork to obtain missing information that it determines will be critical in discriminating among alternatives. Typical examples are site visits to estimate the extent of resettlement that would be required or to verify soil survey data on quality of
5

Having identified a shortlist of alternative technologies, the next stage is to identify a range of alternative locations. It may be appropriate to identify alternative locations for the entire project or selected components of the project (see box 4). In many instances, some elements of the project may be fixed, such as the ore body in a mining project, terminal points of road and rail projects, and location of oil reserves. However, this still leaves scope for analysis of alternative sites for facilities such as tailings dams or alternative alignments for roads, rail lines or pipe-

Box 5. Strategic EA of flood protection in Argentina

The central objective in the Argentina Flood Protection Project is to improve flood protection for communities inhabiting the flood plains of the Parana, Paraguay and Uruguay Rivers in northern Argentina. This region has suffered enormous human and economic losses as a result of flooding, most recently in 1992. However, periodic flooding sustains ecological systems and many productive agricultural activities. The project has therefore adopted a living with floods strategy. A comprehensive investment program in structural and non-structural measures has been designed to enhance the provincial capacity to deal with periodic flooding. At the Banks suggestion, a regional EA initiated in the early stages of project preparation determined the importance of flooding for natural and manmade systems within the flood plains. These included the ecological importance of floods in sustaining critical natural habitats such as wetlands and gallery forests. Accordingly, initial criteria for the selection of investments were modified to ensure that flooding would continue without threats to human well-being or economic infrastructure. All 150 investments initially identified within the project were screened based on these selection criteria. The regional EA helped select 51 subprojects with clear economic, social and environmental justification. Brief project-specific environmental appraisals were prepared for all subprojects by the regional EA team. Upon completion of these appraisals, the cumulative impacts of all 51 subprojects were evaluated, and mitigation measures designed to minimize overall environmental impacts. Public consultation was an important component of the selection and refinement process. For example, in one instance a subproject was significantly redesigned to reflect social concerns.

The final stage is to compare alternatives based on the output from the evaluation. There are a variety of tools which may be employed for this purpose. In all cases, the basis for selection of the preferred alternative(s) should be transparent and clearly described. Where alternatives have been selected that are sub-optimal from an environmental perspective, the justification for their selection should also be documented. Additional guidance on comparative methods is given below. Application to sectoral and regional EA When analysis of alternatives is conducted in strategic EA, the sectoral or regional development objectives are a key component of the framework for screening strategic development options. Demand side management and supply side efficiencies are particularly relevant to energy or water supply strategies in a sectoral and regional context. The stages in figure 1 which relate to generation of realistic alternatives (technology or strategy and location identification and screening) at a project-specific level are arguably more applicable at a sectoral and regional level. For example, a regional EA for a water supply project with the objective of developing optimal resource allocation strategies should first evaluate supply side efficiencies. A broad range of supply options should be identified which might include increased exploitation of surface and groundwater sources, construction of additional storage capacity to harness and exploit peak flows (run-of-river or bank-side), inter-basin transfers or re-use of treated wastewater. Resource requirements in this context would include water volumes, land areas for reservoir construction, and infrastructure such as pipelines, pumping stations, and water treatment plants. Screening of alternative strategies should be based on suitability to supply projected uses (domestic, industrial, agricultural, civic, recreational or ecological) qualitatively and quantitatively, and broad environmental and economic acceptability. In this example, alternative locations might include specific water bodies or aquifers, alternative sites for dam or reservoir construction, pipeline routes or points for transfer of water between rivers. Screening of such alternatives would need to consider conflicting uses of the water bodies (for example effluent disposal and recreation), current abstractions and end uses, water quality criteria vis--vis the intended uses, and the implications for aquatic ecology (including commercial fisheries). Terrestrial constraints would include land use, planning policy designations, sociocultural and ecological conflicts. Economic factors should focus on opportunity costs, and approximate rates-of-return. A similar approach could be applied to sectoral and regional development investments in

agricultural land that would be converted, and plant visits to see alternative technologies in operation and discuss operating experience with owners. During evaluation, the process of public consultation should be continued to ensure that decision makers and stakeholders (including those at the individual sites) have confidence in the process. As a first step, stakeholders should be identified based on a review of the institutions or agencies that may become involved in implementation of project activities, NGOs and community groups local to the short-listed sites. Consultation should entail clearly presenting alternatives to all parties, in the local language(s), in a forum that encourages discussion.
6

transport, energy, agriculture, sanitation, flood protection and other sectors. The regional EA prepared for the Argentina Flood Protection Project is an example (box 5). Linkages to the project cycle It is essential to integrate the identification of alternatives into the project identification process (prior to production of concept paper) to ensure a comprehensive analysis of alternatives (see figure 1). This is usually the pre-feasibility stage of a project, which may involve reconnaissance visits and preliminary investigations. At the project identification stage, the onus is on borrowers to generate realistic alternatives (supported by the Bank) that can be carried through to project preparation. The evaluation and comparative assessment of realistic alternatives should be an integral part of the EA and pre-feasibility studies, and should be described in the EA report prior to appraisal. It is imperative that Task Managers ensure that EA TORs adequately reflect the need to consider alternatives. No-action alternative The no-action or no-project alternative should routinely be included in analysis of alternatives in EA. (Only in rare cases is it not relevantfor example when an investment is necessary to respond to legislative requirements.) This involves projecting what is likely to occur if proposed investment projects are not undertaken. It provides the means to compare the environmental, social, and economic impacts of various project alternatives with those of a scenario in which the project is not implemented. In evaluating the no-action alternative, it is important to take into account all probable public and private actions which are likely to occur in the absence of the project. For example, if the development proposal is to construct a rail link between an industrial area and a port facility to alleviate road congestion, the no-action alternative should consider: the implications for increased traffic and related air pollution and noise as industrial output increases, and the associated effects on adjoining communities; the potential disincentives for further investment in the industrial zone; requisite road improvements to accommodate traffic increases and the effects on adjoining properties; and the economic costs of delays in transport and shipment. Conducting a truly objective evaluation of the no-action alternative requires extra care, since various interest groups have historically used it to support positions for and against projects. Environmental groups that favor preservation over development

have used it to highlight the negative impacts while downplaying project benefits. At the other extreme, advocates of development within the sector concerned tend to emphasize the economic benefits that will be foregone, using the no-action option as a vehicle for providing support for a project proposal. A balanced evaluation can provide objective guidance to support informed decision making. Data requirements An analysis of alternatives is dependent upon the availability of sufficient data. The data base must be designed so that the data describe the characteristics of the variables to be compared and allow data to be transformed and aggregated satisfactorily at the different stages of the analysis process. Ideally, the data should be as homogenous as possiblecollected in a methodologically consistent manner, representative for the time of project planning and implementation, and collected to comparable standards of accuracy. In general, the investment in collecting and processing data must be relative to the benefit of their application. Existing data sources should be used wherever possible, particularly in the earlier stages of analysis, subject to their efficacy. Baseline studies are usually only appropriate in the evaluation stage of selected alternatives. Public involvement Providing opportunities for stakeholders to express their views during alternatives analysis can be beneficial in two waysto obtain information and to build consensus. First, some stakeholders will be sources of valuable local knowledge, others may be experts in the sector, and stakeholders in general are the main source of information on acceptability of certain alternatives. Second, participation throughout identification of the alternatives that will be considered, as well as during their evaluation and comparison, helps to build consensus for the preferred alternative. Consensusbuilding is particularly important in operations like integrated conservation and development projects that depend on stakeholders for successful implementation. It is also critical where controversy is likely, most notably in selecting sites for dams, thermal power plants, or waste disposal facilities. One of the best ways to counter the not in my backyard reaction is to conduct an analysis of alternatives that is perceived as transparent, balanced, and responsive to stakeholder views. In a straightforward, non-controversial project, the general public consultation process for the EA may be sufficient. When a project is potentially controversial, however, as in the case of hydropower projects or highways through populous or environmentally sensitive
7

areas, it is advisable to focus additional consultation efforts on the analysis of alternatives, primarily for consensus-building. Stages in the process where consultation may be worthwhile include: Development of analytical methodology and TORs; selection of alternatives to be analyzed; determination of weights or importance values for evaluation parameters (discussed further below); comparison of alternatives; and formulation of recommendations. There is ample guidance available on consultation and decision-making techniques that provide for public involvement, such as The Participation Sourcebook and Update no. 5: Public Involvement in EA. In applying it to alternatives analysis, however, it is important to remember that different levels of public involvement are usually appropriate at each stage of the process, and this often dictates involving different stakeholders at each stage. For example, in the case of the EA for Nam Theun 2 (box 2), involving communities adjacent to possible dam sites in Lao PDR in discussions of analytical techniques or demand side management and alternative energy sources in Thailand would neither be helpful to the process nor meaningful to the communities. However, environmental NGOs (Thai, Lao, and international) as well as technical experts and institutions, consumer groups, private industry, and representatives of the power and energy sectors would be intensely interested in this phase and could usefully contribute. Conversely, potentially affected communities and local NGOs interested in environmental and social issues would be the primary stakeholders in analysis of alternative sites for power generating stations. The design of the consultation elements can be facilitated by social assessment, which can help to identify key stakeholders and establish an appropriate framework for their involvement. Application to constrained project scenarios Where project identification has largely been completed prior to Bank Group involvement, elements of the project may be fixed, including the location. This is particularly true of some private sector projects where a borrower or sponsor may be responding to a site or technology specific project proposalexamples could include mining concessions or privately financed toll roads (in response to a specific alignment). It is also true of many situations where Bank assistance is being sought where project planning is almost complete. In such instances, to what
8

extent should project identification be revisited or a more thorough analysis of alternatives be undertaken retroactively? Where the Bank (IBRD and IDA) is directly involved (as opposed to IFC and MIGA), the opportunity exists for constructive dialogue with governments. Revisiting project identification or requiring a more thorough analysis retroactively should be considered where there are potentially significant environmental and social impacts associated with the project as proposedas was the case with the Nam Theun 2 Hydropower project (see box 2). In addition, the option of examining sectoral or institutional issues within a project-specific EA should be pursued when it appears advisable and the proponent has not taken the opportunity to explore them in strategic EAs. The opportunities to revisit project identification are more limited with IFC and MIGA. However, in a situation like Nam Theun 2, assuming the Bank were not involved, an alternatives analysis would need to consider alternative dam heights, locations for the power complex, alternative approaches to handling water from the turbines (which cant be routed back to the Nam Theun River), and alternative alignments for access routes. In such circumstances, and where no strategic analysis has been undertaken, the project level EA should determine the acceptability or sustainability of the project as proposed with appropriate mitigation measures. In addition, consideration should be given to the cumulative impacts of other cross-sectoral developments. For example, the economic sustainability of the project might be adversely affected by sedimentation resulting from deforestation and poor land management practices within the river catchment. Comparative assessment of alternatives The objective of comparative analysis is to sharply define the merits and demerits of realistic alternatives, thereby providing decision makers and the public with a clear basis for choosing between options. The key challenge to EA practitioners in comparative assessment is to show distinctions objectively, and as simply as possible. The adoption of unnecessarily complicated techniques can confuse decision-makers and exclude the public from effective participation. As a general rule, the following principles should be adopted in determining an appropriate comparative assessment methodology: In every case, a table or matrix should be prepared summarizing qualitative or quantitative information for each option with decision criteria (economic, technical, environmental and social) on one axis and options on the other.

Box 6. Comparative assessment of alternative toll road alignments, Indonesia

To alleviate congestion and meet projected traffic increases between Cirebon and Batang in northern Java, the Indonesian Ministry of Public Works is planning for construction of a new four-lane dual-carriageway toll road. The project feasibility study included an analysis of three alternative route alignments, which had been selected based on broad technical, social and environmental criteria. The northern Java coast runs approximately East to West in the project area, and the three route options follow a broadly similar alignment and comprise a coastal route (option 1), a direct route (option 2), and an inland route (option 3). The comparative assessment of route options was based on criteria which included: Construction and maintenance costs (including land acquisition and resettlement costs); savings to road users arising from reduced travel times and vehicle operating costs; support for economic development policies and consistency with mational and regional plans; impacts on protected areas, ecology and aesthetics; and direct and indirect impacts on households, and on the integrity of communities. likely impact of the route on the decision criterion. Importance weightings of 0.5 to 4 were assigned to impact criteria based on their relative importance. For example, disturbance of nature reserves was weighted more highly than changes in landscape, whereas people indirectly affected were weighted below people directly affected or community severance. Importance weightings were multiplied by scaling scores, to derive impact criterion scores for each route. These scores were summed to give an overall environmental impact rating for each alternative, and similar approaches were followed for the traffic, planning and engineering aspects. The coastal route had the best overall environmental impact scores, but fared badly based on traffic and engineering criteria. The inland route had the highest overall environmental impact, but scored best based on engineering criteria. Overall, the preferred option was the direct route, route 2, based on an aggregate score of all factors. Assessment Environment Negative impacts Positive impacts Traffic Planning Engineering Total score Priority rating Rating for economic & financial assessment Summary of weighted scores Route 1 -48 6 30 49.5 31 68.5 3 3 Route 2 -57.5 8.5 73 71.5 48 143.5 1 1 Route 3 -72.5 5 47 34 71 84.5 2 2

No single route had clear advantages over the other two options based on all criteriafor example, the inland route would have the greatest ecological and aesthetic impacts, whereas the direct route would involve resettlement of twice the number of households as the inland routeand the choice of an optimal route entailed a trade-off between the various factors. A tabular summary of potential environmental impacts (21 negative and 6 positive) was constructed. For the three route options, each impact criterion was scaled as 1 to 3, based on the

In many cases, particularly where only a few alternatives have been generated, a preferred alternative will become apparent by inspection of the matrix. Where the environmental or social impacts are broadly similar for each option, technical or economic factors should determine the preferred alternative. Where a larger number of realistic alternatives has been generated or where options have varying levels of impact, it may not be possible to identify a preferred alternative from the matrix. The matrix should still be prepared, since it enhances transparency of the process and provides the information that other reviewers of the analysis will need if they wish to

check its conclusions or apply their own methods to compare alternatives. However, a more systematic approach may be needed, involving the use of multiattribute decision making techniques. More complex techniques and associated sensitivity analyses should only be used if straightforward methods fail to provide a clear basis for decision making. Systematic approaches to comparative assessment of alternatives involve the application of scaling, rating or ranking checklists. These are used in conjunction with the results derived from the comparative evaluation of selected alternatives in the EA process (based on decision criteria such as effects on air quality, ecology, and human health). Importance
9

weighting of decision criteria may also be used, either in isolation from or in combination with scaling, rating or ranking methods. Ranking entails ordering alternatives from best to worst in terms of potential impacts on decision criteria. Rating refers to the use of a pre-defined rating scheme to rate the significance of decision criteria for each option. Scaling involves the assignment of numeric or algebraic scales to the impact of each alternative on each decision criterion (see box 6). Importance weighting involves assigning a weighting factor to each decision criterion relative to the other decision criteria (see box 6). Explanations on the various techniques, and their limitations, may be obtained from the growing body of literature on EA methods. Additional guidance may also be obtained from Update no. 17: Challenges of Managing the EA Process.

For further information Canter, L.W. 1994. Environmental Impact Assessment. McGraw-Hill. New York. Dixon, J. et al. 1994. Economic Analysis of Environmental Impacts. 2nd Edition. Earthscan Publications. London. ENVLW. 1996. The Impact of Environmental Assessment: The World Banks Experience. The World Bank. Washington, D.C. ENVSP. 1995. Social Assessment. Environment Department Dissemination Note No. 36. The World Bank. Washington, D.C. ENVSP. 1996. The Participation Sourcebook. The World Bank. Washington, D.C.

This Update was prepared by Aidan Davy, Thomas E. Walton and Tor Ziegler. The EA Sourcebook Updates provide guidance for conducting environmental assessments (EAs) of proposed projects and should be used as a supplement to the Environmental Assessment Sourcebook. The Bank is thankful to the Government of Norway for financing the production of the Updates. Please address comments and inquiries to Olav Kjrven and Aidan Davy, Managing Editors, EA Sourcebook Update, ENVLW, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. S-5139, (202) 473-1297. E-mail: eaupdates@worldbank.org.

10

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

April 1998 Number 23

Economic Analysis and Environmental Assessment

Successful economic development depends on the rational use of natural resources and on reducing as far as possible the adverse environmental impacts of development projects. Environmental assessment (EA) is a primary tool for achieving this objective, by inserting critical environmental information into the process of project identification, preparation, and implementation. Economic analysis, by comparison, is employed to determine if the overall economic benefits of a proposed project exceed its costs, and to help design the project in a way that produces a solid economic rate of return. Adverse environmental impacts are part of the costs of a project, and positive environmental impacts are part of its benefits. Consideration of environmental impacts, therefore, should be integrated with the other aspects of the project in the economic analysis to the extent possible. This EA Sourcebook Update discusses the relationship of EA and economic analysis and gives guidance on how economic analysis might incorporate environmental costs and benefits. This Update replaces guidance provided in Chapter 4 of the EA Sourcebook. Bank requirements The World Banks Operational Policy on EA (OP 4.01) states that environmental costs and benefits should be quantified to the extent possible, and economic values should be attached where feasible. This should be done for both alternative project designs and alternative mitigation options. Moreover, the Operational Policy on Economic Evaluation of Investment Operations (OP 10.04) states that EA findings and recommendations should be taken into account in project appraisal and supporting benefit-cost analysis. EA, economic analysis, and the project cycle EA is an information-gathering and analytical process that helps avoid environmentally unsound development. It focuses on environmental externalities: unintended adverse effects of development projects on the environment. For example, land clearing for an aquaculture project could convert wetland areas, resulting in reduced bird habitat and water quality. EA seeks to identify and evaluate these environmental effects in qualitative terms, and to quantify them when feasible (for example, air pollution in parts-per-million, or tons of topsoil lost to erosion). The impacts identified in the EA process have not often been converted into monetary terms, however. A major reason behind the generally weak link between EA and economic analysis has been the lack of useful guidance on converting physical impacts into monetary terms. Recently, however, the science of environmental economics has moved to a point where it can be more readily applied in the project context. This Update seeks to explain how this might be done. The role of EA and environmental economic analysis in the Banks project cycle is illustrated in simplified terms in table 1. Environmental economic analysis can play an important role at three main stages: (i) in the assessment of the impacts of a proposed project and its various alternatives; (ii) in the analysis of preventive or mitigative options; and (iii) in project appraisal, once a specific alternative has been selected. In the case of both economic analysis and environmental assessment, the important distinction is between what would happen with the project and without the project, not other changes that may be happening over time. This point is sometimes lost as there may be important long term trends that occur irrespective of the project itself. In the first stage, the economic analysis will normally consist of estimating monetary costs and benefits (valuation) of the various environmental impacts identified in the EA, using a range of
Insert in Update Binder chapter 1 4

Table 1. EA, economic analysis, and the project cycle

Project stage Preparation EA activity Environmental screening Associated economic analysis activity Potential environmental costs and benefits are considered on a preliminary basis Requirement to quantify environmental impacts and assign monetary values spelled out EA team includes resource or health economist, as appropriate EA team analyses the impact of project alternatives and compares them, using monetary values on their costs and benefits, where feasible The Bank reviews the EA report, including the economic analysis EA findings, including the environmental costs and benefits, are incorporated into the project economic analysis and the estimation of the economic rate of return

Preparation of EA TORs

EA team selection EA preparation

Review of EA

Appraisal

Incorporation of EA into project design and documentation Agreements reached on actions to be taken, based on the findings of the EA Environmental supervision

Negotiations

Implementation

Supervision includes monitoring the projects actual environmental costs and benefits

valuation techniques. In the second stage, the analysis is extended to consider the costs and benefits of preventive and mitigative measures, so that comparisons can be made with the original project impacts. At the third stage, the monetary values for the selected alternative are integrated into the overall economic evaluation of the proposed project. These evaluation techniques, which are generic, are discussed briefly towards the end of this Update. For the integration of EA and economic analysis to be successful, both need to be designed and undertaken with the needs of the other in mind. All indices of environmental damage are not equally helpful for economic analysis. For example, a measure of soil loss in areas affected by erosion will be less useful than a measure of the resulting change in agricultural productivity. Similarly, consideration of the economic benefits at stake can help target EA resources to the areas of greatest interest. These possible inter-relationships should be incorporated from the beginning in the development of TORs, the selection of the EA team, and other stages of the EA and project preparation process (see table 1). Needless to say, the services of a trained economist will be required. Valuing environmental impacts For a projects environmental impacts to be valued, they must first be identified and measured. This is
2

generally far from straightforward. Environmental impacts are often dislocated in time and space, making cause and effect difficult to establish. The severity of environmental impacts often depends on the accumulation of problems (over time, over space, or both). Many environmental goods and services do not enter markets, or do so only imperfectly. The difficulties this causes for valuation are compounded by the empirical limitation that available data are often scarce or of poor quality. Total economic value. Economic valuation is still an evolving science. For some goods and services (for example, a kilo of rice or fish, or a cubic meter of timber), the market provides prices that are good reflections of the values society places on that good or service. For other goods and services, market prices either do not exist or only capture a small part of the total value. Examples of such goods and services include endangered species and scenic vistas. To ease in the task of analysis, therefore, it is often useful to disaggregate any environmental impact into individual components of value. One approach to doing this is called the Total Economic Value (TEV) approach, whereby an impact is decomposed into a number of categories of value (figure 1). The idea behind the TEV approach is that any good or service is composed of various attributes, some of which are concrete and easily measured, while others may be more difficult to quantify. The total value, however, is

the sum of all of these components, not just those that can be easily measured. The breakdown and terminology for the components of TEV vary slightly from analyst to analyst, but generally include (i) direct use value; (ii) indirect use value; and (iii) non-use value. The former two are generally referred to together as use value. Each is often further subdivided into additional categories. Direct use value. Direct use value, also known as extractive, consumptive, or structural use value, derives from goods which can be extracted, consumed, or directly enjoyed. In the context of a forest, for example, extractive use value would be derived from timber, from harvest of minor forest products such as fruit, herbs, or mushrooms, and from hunting and fishing. In addition to these directly consumed goods, direct use values can also be nonconsumptive. For example, people who enjoy hiking or camping in the same forest receive a direct use value, but do not actually consume any of the forest resource. Similarly, in a coral reef direct use values can include the harvesting of shells and catching of fish, or the non-consumptive use of the reef by scuba divers. All of these benefits are real, can be measured, and have values, even if the consumption by one individual does not reduce the consumption by another (economists call this non-rival consumption, and these goods are classified as public goods). Consumptive use is generally the easiest to value, since it usually involves observable quantities of products whose prices can usually also be observed. Non-consumptive use is often more difficult to value since both quantities and prices may not be observed.

Indirect use value. Indirect use value, also known as non-extractive use value or functional value, derives from the services the environment provides. For example, wetlands often filter water, improving water quality for downstream users, and national parks provide opportunities for recreation. These services have value but do not require any good to be harvested, although they may require someones physical presence. Measuring indirect use value is often considerably more difficult than measuring direct use value. The quantities of the service being provided are often hard to measure. Moreover, many of these services often do not enter markets at all, so that their price is also extremely difficult to establish. The visual aesthetic benefits provided by a landscape, for example, are non-rival in consumption, meaning that they can be enjoyed by many people without detracting from the enjoyment of others. Option value. Option value is the value obtained from maintaining the option of taking advantage of somethings use value (whether extractive or nonextractive) at a later date. It is, therefore, a special case of use value, akin to an insurance policy. (Quasioption value, which derives from the possibility that even though something appears unimportant now, information received later might lead us to reevaluate it, is a related concept.) Existence and bequest value. In contrast to use value, non-use value derives from the benefits the environment may provide which do not involve using it in any way, whether directly or indirectly. In many cases, the most important such benefit is existence value: the value that people derive from the knowledge that something exists, even if they never plan to use it. Thus, people place a value on the

Figure 1. Total economic value and selected valuation techniques Total economic value Non-use value

Use value Direct (extractive) use value

Change in output of marketable goods Cost-based approaches Hedonic princes Contingent valuation Travel cost

Indirect (non-extractive) use value

Cost-based approaches Contingent valuation Change in output of marketable goods

Option value
Contingent valuation Hedonic princes

Existence value
Contingent valuation

Bequest value
Contingent valuation

existence of blue whales, or of the panda, even if they have never seen one and probably never will; if blue whales became extinct, many people would feel a definite sense of loss. Bequest value is the value derived from the desire to pass on values to future generations. Non-use value is the most difficult type of value to estimate, since in most cases it is not, by definition, reflected in peoples behavior and is thus wholly unobservable. Benefit-cost vs cost-effectiveness. Two approaches are possible to the economic analysis of environmental impacts. The first is to use standard benefit-cost criteria, in which the benefits of an action are compared to its costs to determine whether the action is worth undertaking. This approach is commonly used to compare alternative options and requires that the environmental impacts be identified and that monetary values be placed on the outcomes. An example is the analysis of different air pollution control measures and the expected health benefits associated with each alternative. In some cases, however, a traditional benefit-cost analysis may not be feasible or desirable. It may not be possible to make monetary estimates of benefits. For example, some natural areas may be so unique that it might be felt they should be conserved at all costs. In other cases, there might be substantial uncertainty about the benefits provided by environmental goods and services, either now or in the future, or great problems in determining appropriate values in monetary terms. When loss of these goods and services would be irreversible, it may be desirable to choose the strategy that minimizes maximum possible losses due to environmental damage, unless the social cost to do so is unacceptably large; this is known as the safe minimum standard approach. In such cases, the appropriate approach to the analysis is one of cost-effectiveness rather than costbenefit; that is, the issue becomes one of finding the cheapest and most effective way of achieving the conservation objective or some other goal. Note that the cost-effectiveness approach does identify the most efficient way of reaching a goal, but does not tell you if the expected benefits justify the costs. Answers to the latter question must rely on informed judgment and common sense. Valuation techniques Incorporating environmental impacts identified in the EA into the project analysis is a two-step process. First, one has to understand what are the impacts. This information is provided by a traditional EA. Second, one has to estimate the value of the impacts (where feasible and appropriate) in monetary terms to determine their relative economic importance, and assess the benefits and costs of various alternatives.
4

This section focuses on valuation techniques, and their use in project analysis.1 In most cases, the techniques have two parts: measuring the physical impact, and then assigning a value to that impact. As can be seen in figure 1, a number of valuation techniques are potentially applicable to each category of value. Figure 2 provides a simplified guide to choosing an appropriate technique for a given situation. The flow chart begins with an environmental impact and asks if there is a measurable change in production, or a change in environmental quality. Depending on the answer, it traces out different possible scenarios and their possible impacts. It shows the most commonly-used techniques used to estimate monetary values for each kind of impact. As an example, consider the case of an aquaculture development project which will reduce the area of mangrove forest. The EA might identify reduced water quality due to loss of the mangrove forests water-filtering services and loss of habitat as adverse impacts. Unless the mangrove forests are directly harvested, techniques such as change-in-productivity will clearly not be very useful. Several techniques might be used to value the reduction in water quality; some are based on the cost of obtaining clean water by other means (for example, replacement or relocation cost), while some are based on the consequences of reduced water quality (increased sickness or death). The specific choice of technique will depend on the situation and on data availability. Likewise, the loss of habitat could be valued in a number of ways, depending on the specific nature of the situation. Figure 2 is only intended as an indicative guide; depending on the specific conditions encountered and on the data available, other techniques may be preferable in a given situation. I. Valuing Changes in Outputs and Direct Costs Change in output of marketable goods In many cases, the environmental effects of projects manifest themselves (at least in part) in changes in output of marketable goods: loss of forest, for example, results in the loss of timber products, of fuelwood, of fodder (whether collected or eaten on site by livestock grazed in the forest), and a variety of non-timber products such as fruit, herbs, and mushrooms. In cases such as these, the value of the unintended benefits and costs can be estimated by using the simple technique of valuing the change in output caused by the project. This approach is often referred to as the change-in-productivity approach. In Croatia, for example, reforestation activities under the Coastal Forest Reconstruction and Protection Project were estimated to result in increased wood production, which would be harvested at various intervals in the future. Using the increased wood

Figure 2. Choice of valuation technique

Environmental impact

Measurable change in production

Change in environmental quality

Yes

No

Habitat

Air and water quality

Health effects

Recreation Aesthetics, Biodiversity, Cultural, Historical Assets

Nondistorted market prices available?

Opportunitycost approach

Sickness Costeffectiveness of prevention

Death

Travel cost

Yes

No

Replacementcost approach Preventive expenditures

Loss of earnings

CEA of prevention

Contingent valuation Contingent valuation

Use change-inproductivity approach

Use surrogate market approaches, apply shadow prices to change in production

Land-value approach

Medical costs

Human capital

Contingent valuation

Replacement/ relocation costs

Source: Dixon and Boj in Dixon et al., 1994.

output (in terms of both quantity and quality), the expected prices at time of harvest, and a discount rate of 10%, the present value of increased wood production was estimated at between 2.5US$/ha and 82US$/ha, depending on the site. Box 1 below provides an additional example of the use of this technique, to value the damage to agriculture resulting from flooding and damage to irrigation caused by watershed degradation in Haiti. Even when prices cannot be observed (for example, products harvested for home consumption), there are generally-accepted and reliable ways to estimate the value of the products (for example, by using the value of close substitutes or the cost of collection). The biggest difficulty in valuing such impacts generally arise from measuring the amounts of goods being produced and in predicting how these amounts will change with and without the project. The EA can be very helpful in arriving at estimates of these changes. Once these estimates are in hand, valuing the changes is usually relatively simple. Cost of illness and human capital Many environmental impacts, such as air and water pollution, have repercussions for human health. Valuing the cost of pollution-related morbidity (sickness) requires information on the underlying damage function (usually some form of a doseresponse relationship) which relates the level of

pollution (exposure) to the degree of health effect as well as information on how the project will affect the level of pollution. The costs of an increase in morbidity due to increased pollution levels can then be estimated using information on various costs associated with the increase in morbidity: any loss of earnings resulting from illness, medical costs such as for doctors, hospital visits or stays, medication, and any other related out-of-pocket expenses. This approach is symmetric: the benefits of actions that reduce the level of pollution and hence of morbidity are estimated in the same way. This approach was applied in Santiago, Chile, where a package of air pollution control investments was evaluated in terms of the health benefits from reduced levels of particulates, volatile organic Table 2. Annualized benefits and costs of air pollution control strategy in Santiago, Chile
(US$ millions)
Program component Fixed sources Gasoline vehicles Buses Trucks Control strategy
Source: World Bank, 1994.

Benefits 27 33 37 8 108

Costs 11 14 30 4 60

Net benefits 16 19 7 7 48

compounds (VOC), and nitrous oxides (NOx) in the city. As shown in table 2, the analysis estimated that the benefits of the overall strategy exceeded costs by about 70 percent, with the benefits of individual components of the strategy exceeding their costs by between 20 percent and 140 percent (World Bank, 1994; Ostro and others, 1996). The costs of doing this type of study, both in money and time, can be considerable. In the case of Santiago, however, since data were available for many variables, it was possible to produce good results in a timely manner. The details of the analysis are presented in the papers cited earlier. The estimates obtained in this manner are interpreted as lower-bound estimates of the presumed costs or benefits of actions which result in changes in the level of morbidity, since this method disregards the affected individuals preference for health versus illness, for which they may be willing to pay. Also, the method assumes individuals treat health as exogenous and does not recognize that individuals may undertake defensive actions (such as using special air or water filtration systems to reduce exposure to pollution) and incur costs to reduce health risks. In addition, the method excludes non-market losses associated with sickness, such as the pain and suffering to the individual and to others concerned, and restrictions on non-work activities. Also, the approach ignores other, non-health benefits from reducing pollution levels such as amenity values (better views), productivity losses, and ecosystem impacts. When this approach is extended to estimate the costs associated with pollution-related mortality (death), it is referred to as the human-capital approach. It is similar to the change-in-productivity approach in that it is based on a damage function relating pollution to productivity, except that in this case the loss in productivity of human beings is measured. The human-capital approach is an extension of the more standard human capital theory which relates the demand for education to its potential payoff in terms of expected life-time earnings. Because it reduces the value of life to the present value of an individuals future income stream, the human-capital approach is extremely controversial when applied to mortality. We recommend, therefore, that this approach not be used. In many cases, the costs and benefits of activities that affect mortality can be expressed in terms of changes in the number of deaths (without monetary values) and a cost-effectiveness approach used. Alternatively, the US/OECD type estimates of the value of a statistical life based on willingness to pay estimates (which includes much more that just lost productivity and is often 5 to 10 or more times larger than the straight human-capital estimates) might be used, adjusted using relative per capita GNP
6

(see World Bank, 1996c, for more on these estimates). In general, estimating monetary values for mortality is a complicated, quite subjective process that has to be used with great caution and transparency. Cost-based approaches When the benefits of a given environmental impact cannot be estimated directly, information on costs can be used to produce valuable information. For example, an order of magnitude estimate of the potential costs (or savings) to society from a change in an environmental problem, can be obtained by using the cost of reducing or avoiding the impact, or the cost of replacing the services provided by the environmental resource. The major underlying assumptions of these approaches are (i) that the nature and extent of physical damage expected is predictable (there is an accurate damage function available), and (ii) that the costs to replace or restore damaged assets can be estimated with a reasonable degree of accuracy. It is further assumed that these costs can be used as a valid proxy for the cost of environmental damage. That is, the replacement or restoration costs are assumed not to exceed the economic value of the asset. These are strong assumptions and may not be valid in all cases. It simply may cost more to replace or restore an asset than it was worth in the first place. For example, cultivated hillsides may be eroding and there may be methods available (terracing, changes in cropping patterns) to reduce or prevent the erosion. Each of these preventive measures has a cost, however, and it is the responsibility of the analyst to determine if the total costs of prevention are greater or less than the benefits of preventing erosion. In some case, the costs of erosion control may be so high (and/or the benefits from controlling erosion may be so low) that erosion control measures would be an inappropriate use of scarce resources. In some cases, there may also be more cost-effective ways to compensate for environmental damage than to replace the original asset or restore it to its original condition, and these substitution possibilities are ignored with the use of this technique. If substitutes are available, the method will likely overestimate the value of the damaged or destroyed asset. Because of this, these methods are generally thought to provide an upper-bound estimate of the benefits of measures taken to prevent the damage from occurring. Replacement cost. The replacement cost approach is often used as an estimate of the cost of pollution. This approach focuses on potential damage costs as measured by ex ante engineering or accounting estimates of the costs of replacement or restoration if damage from pollution were to occur. For example, the costs of air pollution-related acid deposition in urban areas could be approximated by the restoration

and replacement costs from damaged infrastructure. Similarly, the cost of restoring a river or a wetland could be used as an estimate of the costs of environmental damage to these natural assets. Note that this approach provides only a partial measure of damages caused by pollution. For example, the replacement cost technique can be used to estimate the costs of pollution of potable water supplies. Pollution of water resources by agrochemicals is common in many countries, resulting in drinking water below acceptable health standards. Quantifying the aggregate health impacts, or estimating a damage function for this type of water pollution, is often difficult because of the complex relationship between ambient quality, exposure, and illness. However, order of magnitude estimates of the cost of providing alternative safe water supplies often are possible. The incremental investment cost of alternative water supply usually can be derived from proposed water supply investment projects containing data on total investment cost and the quantity of incremental water supply. Using the replacement cost technique, an estimate of the annual

cost of pollution of potable water can be made. For instance, assuming an estimated investment cost for alternative water supply on the order of US$0.51.0 per cubic meter, and current level of total potable water use at about 100 million cubic meters per year, the cost of pollution of potable water would be approximately US$50100 million per year at current levels of water use. Box 1 provides an example of the use of replacement cost to value the benefits obtained from reduced flooding thanks to watershed protection in Haiti. The replacement cost technique is particularly useful to assess the costs associated with damage to tangible assets, the repair and replacement costs of which are easily measurable. This information can then be used to decide if it is more efficient to allow the damage to occur and pay the replacement costs or to invest in preventing the pollution in the first place. The technique is less useful, however, for very unique assets, such as historical or cultural sites and unique natural areas, which cannot be replaced and cannot easily be restored, and about which restoration costs are uncertain.

Box 1. Costs and Benefits of the Pic Macaya National Park in Haiti
The economic analysis of the Natural Reserve Management component of the Forest and Parks Protection Technical Assistance Project in the Republic of Haiti provides an example of the use of several valuation techniques in a data-scarce environment (World Bank 1996d). The main objective of the project is to protect critical remnants of Haitis forest ecosystems, including the Pic Macaya National Park. By protecting this area, important benefits are expected in terms of (i) protecting some of the last remnants of the Hispaniolan moist forest ecosystem, which is considered regionally outstanding and as having the highest priority for conservation at the regional scale; (ii) preserving the potential for ecotourism development; (iii) protecting downstream areas, which include one of the countrys main irrigated areas, from damage from flooding and sedimentation; and (iv) helping to regulate downstream water flows. Change in productivity. Protecting the Pic Macaya watershed will help reduce damage to irrigation systems downstream, both from siltation and from reduced dry-season flow, which is forcing a reversion to rainfed cultivation and a switch to lower-value crops. Returns from rainfed production are about 200-800US$/ha/year lower than for irrigated production, depending on the crops being planted. Combined with estimates of the effects of damage to irrigation infrastructure on the area irrigated, the present value of the benefits of avoided reduction in productivity, relative to the no-project case, were estimated to be about US$27 million (the wide range reflecting the weakness of the underlying data). Replacement cost. The costs of damage to irrigation and other infrastructure were estimated using their replacement costs (2,500 to 5,000 US$/km for irrigation canals and 3,500 to 7,000 US$/km for roads). This resulted in an estimated benefit from avoided damage, relative to the no-project case, of about US$2.55 million. Opportunity Cost. Protecting the Pic Macaya area also means forgoing its use as agricultural land and the benefits of harvesting standing timber. By using crop production budgets developed for the surrounding area, the potential value of the Pic Macaya area for agriculture can be estimated. Assuming that half of the area still under forest at Pic Macaya (about 3,500ha) is suitable for agricultural use (the rest being too steep or inaccessible), the present value of forgone agricultural production is about US$175 thousand. This relatively low estimate is due to the unsustainability of agriculture under conditions such as those at Pic Macaya. The foregone wood harvesting benefits could not be estimated for lack of data on standing timber, but were expected to be relatively low due to the high transport costs of extracting timber from the area. Table 2 summarizes the results of the economic analysis of Forest and Parks Protection Technical Assistance Project, which covered Pic Macaya and two other National Parks.

Relocation cost. Similar to the replacement cost approach, the relocation cost approach uses estimated costs of a forced relocation of a natural or physical asset due to environmental damage. For example, the construction of brackish water ponds in a coastal area of Thailand resulted in the discharge of salt water into nearby freshwater streams traditionally used for irrigation and domestic water supply. Part of the environmental costs associated with this discharge is the need to relocate the intakes for both irrigation facilities and domestic water supply. The costs of relocating these intakes can then be compared to the alternative costs of redesigning and constructing drainage canals from the brackish water ponds directly to the ocean, to obviate the need for discharge into the nearby freshwater streams. Another recent example of the relocation cost approach can be found in China, where the government decided to relocate Shanghais water intake. Shanghai, a city of 14 million, was facing increasing difficulties in ensuring a safe potable water supply. The lower Huangpu River was heavily polluted by wastes from industries and ships, and by municipal sewage. The cost of moving the municipal water intake up river to take advantage of cleaner water supplies, reduced pre-use treatment costs, and lower risk of major pollution incidents was estimated. This was then compared to subjective estimates of the cost to clean up the industries and plants discharging wastes into the river that is, to clean up the existing pollution. The analysis did not consider the benefits of a clean, safe water supply, however. The costs of relocation were judged less than the costs of cleanup, so the relocation option was chosen. Opportunity cost. In some cases it is decided to protect a particular resource and forego other development options. The term opportunity cost refers to the value of these lost economic opportunities due to environmental protection. It is, therefore, a measure of the cost of environmental protection in terms of development benefits foregone. Box 1 provides an example of the use of this approach in Haiti. Since this approach gives no information on the expected benefits from protection, society must still decide if the opportunity cost of conservation is acceptable or not. This is a very powerful technique, however, since it clearly identifies the expected economic cost of protection to society. In many cases, this amount is actually very small; in other cases, this information can be used to mobilize other sources of funds to compensate individuals or society for the opportunity cost of protection. The Global Environment Facility (GEF) and other donors may be willing to provide grant funds to cover these types of costs, especially when the benefits produced are important at the global level.
8

II. Valuing Environmental Amenities: Recreation, Nature, and Biodiversity Often, the environmental good or service being valued is not traded per se in the market place. Examples of these amenity-type services include recreational sites and the preservation of biodiversity. A number of valuation techniques exist that can be used to place monetary values on these resources and this information, in turn, can be incorporated into a more conventional benefit-cost analysis. Hedonic analysis We know that environmental quality affects the price people are willing to pay for certain goods or services. Ocean front hotels, for example, charge different rates depending on the view (rooms with ocean views cost more than the same size room with a garden view usually a nice way of saying the parking lot!). Hedonic models have been widely used to examine the contribution of different attributes to prices for housing and to wage levels, including the contribution of environmental quality.2 Many observed prices for goods are prices for bundles of attributes. For example, property values depend on physical attributes of the dwelling (such as number and size of rooms, amenities such as plumbing, condition); on the convenience of access to employment, shopping, and education; and on a number of less tangible factors such as environmental quality. Since each house will differ slightly from others, the influence of the various factors on its price can be broken down statistically, provided sufficient observations are available. This approach is of interest because many environmental dimensions are likely to be embodied in property values. A home in a neighborhood with low air pollution, for example, should sell for more than a similar home in a neighborhood with high ambient air pollution. Hedonic techniques allow this effect to be measured, holding other factors such as size and amenities constant. In essence, the technique estimates the implicit prices for various attributes, which together make up the sale price. When applied to housing data, this approach is often referred to as the property value approach; when applied to wage data, it is generally referred to as the wage differential approach. In Croatia, for example, a hedonic analysis was used during preparation of the Coastal Forest Reconstruction and Protection Project to help estimate the landscape benefits of reforestation. Analysis of hotel room prices showed that rooms with views of forested landscapes cost, on average, about 36US$/day more than rooms in hotels in areas without such views. The challenge of both of these techniques is to correctly specify the relevant variables and the functional forms.

Hedonic methods require observations of the prices of goods and of the attributes of these goods. To enable the effect of the many different factors to be distinguished, large data sets are usually needed. Because of their data intensity and the need for open reporting of prices, the application of these techniques has had limited (but often successful) application in developing countries. Travel cost The travel cost (TC) method is an example of a technique that attempts to deduce value from observed behavior.3 It uses information on visitors total expenditure to visit a site to derive their demand curve for the sites services. The technique assumes that changes in total travel costs are equivalent to changes in admission fees. From this demand curve, the total benefit visitors obtain can be calculated.4 The TC method was designed for and has been used extensively to value the benefits of recreation. The TC method depends on numerous assumptions, many of which are problematic in the context of international tourism. The basic technique generally assumes that travel cost is proportional to distance from the site and that people living at the same distance from the site have identical preferences. While these assumptions are often valid in the case of national tourism (tourism within a country), neither assumption may be valid in the case of international tourism. The technique also assumes a single-purpose trip and encounters difficulties when trips have multiple purposes. It should also be borne in mind that the resulting estimates are site-specific. The main application of TC methods in developing countries is to valuing tourists willingness to pay for national parks. In Zimbabwe, a TC analysis of tourists found that they derived about US$610 per person of benefit (consumers surplus) from their trip, of which about US$275 was obtained from visiting national parks (Brown, Ward, and Jansen, 1995). In Costa Rica, the benefit obtained by tourists visiting the parks and reserves was about US$1,150 per person (Mekhaus and Lober, 1996). Contingent valuation Unlike techniques which use observed data, the Contingent Valuation (CV) technique relies on direct questioning of consumers (actual or potential) to determine their willingness-to-pay (WTP) to obtain an environmental good.5 A detailed description of the good involved is provided, along with details about how it will be provided. The actual valuation can be obtained in a number of ways, such as asking respondents to name a figure, having them chose from a number of options, or asking them whether they would pay a specific amount (in which case,

follow-up questions with higher or lower amounts are often used). CV can, in principle, be used to value any environmental benefit. Moreover, since it is not limited to deducing preferences from available data, it can be targeted quite accurately to ask about the specific changes in benefits that the proposed project would result in. This also means that, with appropriately-worded questions, CV can provide an all-encompassing estimate of the perceived costs and benefits of environmental changes, in contrast to other techniques which, as noted above, often only provide a partial estimate of environmental costs and benefits. Because of the need to describe in detail the good being valued, interviews in CV surveys are often quite time-consuming. It is also very important that the questionnaire be extensively pretested to avoid various sources of bias. CV methods have been the subject of severe criticism by some analysts (see, for example, Hausman, 1993). A blue-ribbon panel was organized by the US Department of Interior following controversy over the use of CV to value damages from the 1989 Exxon Valdez oil spill. The report of this panel (NOAA, 1993) concluded that CV can provide useful and reliable information when used carefully, and the panel provided guidance on doing so. This report is generally regarded as authoritative on appropriate use of the technique. In some cases it is possible to do both a CV and a Travel Cost analysis for the same valuation question. This allows the analyst to cross check the two estimates and get an idea of the robustness of the results. This approach has been used a number of times in determining the consumers surplus of safari visitors to game parks in East Africa, with surprisingly consistent results. It is particularly useful since one measure is based on observed behavior (the travel cost approach) while the other is based on hypothetical survey information (the CV approach). Box 2 provides some examples of the application of CV methods in the context of Bank operations. Benefits transfer Benefits transfer is not a methodology per se, but rather refers to the use of estimates obtained (by whatever method) in one context to estimate values in a different context.6 For example, an estimate of the benefit obtained by tourists viewing wildlife in one park might be used to estimate the benefit obtained from viewing wildlife in a different park. This has, in fact, been done in East Africa where estimates of the consumers surplus for safari visitors in one country have been used to estimate the benefits to new safari destinations in nearby countries. The main attraction of benefit transfer is that it provides a low-cost way of estimating values when time or resources do not allow fuller valuation studies, or when the good or
9

Box 2. Applications of contingent valuation in Bank operations

In recent years, there has been increasing use of contingent valuation (CV) techniques to value environmental goods and services, in both developed and developing countries. Recent examples carried out in the context of World Bank operations include: In Madagascar, CV was use to value the cost to local communities of refraining from using the area of the Mantadia National Park, established under the Bank-financed Forest Management and Protection Project, to gather a variety of products (Kramer and others, 1995). Local residents were asked whether they would be willing to accept specified levels of compensation (denominated in units of rice, the local staple food) to forgo access to the forests in the Park. Their responses were used to estimate a mean value per household of about $108very similar to the value obtained by other means. A separate CV survey of international tourists showed they would be willing to pay about $65 more per tourist for access to the new Park. In Croatia, CV was used to estimate tourists willingness to pay for restoration of forested landscapes in coastal areas that were destroyed during the war, as part of the economic analysis for the Coastal Forest Reconstruction and Protection Project (World Bank, 1996b). Two parallel surveys were carried out, one among tourists in Croatia itself and one among foreign tourists at similar destinations in Italy. Both showed a willingness to pay for forested landscapes of about US$3/person/day. In Morocco, CV was used to estimate tourists willingness to pay for conservation and rehabilitation of the historic Medina at Fs, a UNESCO World Heritage site that is rapidly deteriorating. The survey distinguished the willingness to pay for improvements of visitors of the site itself (who derive use value from visiting the site) from those of other visitors to Morocco (who only derive existence value and option value). Visitors to the site itself were found to be willing to pay as much as US$70 each in the form of increased tourist taxes or admission fees for improvements aimed at preserving and improving conditions in the Medina, while non-visitors would be willing to pay about US$30 each.

service to be valued has not yet been created (for example, a new safari-tourism destination national park) so that there are no users to survey. This approach also has considerable risks, however. For many reasons, estimates derived in one situation can be inappropriate in another. As a result, benefits transfer has been the subject of considerable controversy in the economics literature. A consensus seems to be emerging that benefit transfer can provide valid and reliable order-of-magnitude estimates under certain conditions: The commodity or service being valued have to be very similar at the site where the estimates were made and at the site where they are applied; and The populations affected must be very similar. Of course, the original estimates being transferred must themselves be reliable for any attempt at transfer to be meaningful. The estimates of the value of timber products produced by reforestation in Croatia cited previously indicate the limitations of benefits transfer techniques: even in a seemingly homogeneous area, environmental benefits can vary by an order of magnitude. The likelihood that benefits transferred from another area will be appropriate is, therefore, extremely low. Conversely, the use of CV to value tourists willingness to pay for forested landscapes in Croatia (see Box 2) provides an example of a situation in which benefits transfer can be used with considerable confidence. Since tourists visiting Croatia are drawn from the same pool as those visiting other Mediterranean resort areas, and since
10

forested landscapes are relatively similar, estimates of tourist willingness to pay obtained in one location can be used in another. The benefits transfer technique should be used with caution, therefore, and only when no site-specific measures are possible. Incorporating environmental costs and benefits into economic analysis The choice of technique depends on the specific problem being studied. Except in very simple situations, however, it is likely that a variety of techniques will be necessary to estimate the full range of benefits. Moreover, where substantial investments are contemplated, it might be desirable to cross-check estimates by deriving them from multiple sources. Once the various environmental impacts have been identified and the benefits and costs of various alternatives assessed, this information can be incorporated into the broader economic analysis of the project. This is usually done in a benefit-cost framework, whereby the streams of benefits and costs of a proposed project (including both direct project inputs and outputs, as well as environmental impacts to the extent that they can be identified and monetized) are compared over some period of time. The three main decision criteria used in benefit-cost analysis are: net present value (NPV), internal rate of return (IRR) and benefit-cost ratio (BCR). All of these criteria rely on the concept of discounting a stream of benefits and costs which occur at different times over

the duration of the project being evaluated. Discounting puts all of these costs and benefits into a common time frame to allow for more accurate comparison. Adding environmental costs and benefits does not change the method of analysis and guidance is available in various Bank publications, such as the Handbook on Economic Analysis of Investment Operations (World Bank, 1996a). However, several aspects of project analysis need particular attention when environmental problems are present. The impacts of many environmental changes, whether positive or negative, are often only felt in the future, long after the activity which caused the change has ceased. Similarly, effects are often felt far beyond the boundaries of the project itself. Special attention must be given, therefore, to the temporal and spatial boundaries of the analysis. Temporal Boundaries. Since environmental impacts extend long beyond the normal life of the project, it is important to extend the time horizon of the analysis so as to include all the benefits and costs associated with environmental impacts, even if they go further into the future than the normal life of a project. The effective length of the time horizon of an analysis is determined by both the number of actual years included in the analysis and the discount rate used. Using too short a time horizon effectively ignores many environmental impacts, both positive and negative. For example, an activity that results in the permanent loss of a fishery should include in the analysis the present value of the entire future loss of that resource, even if the activity itself only lasts for a few years. The choice of the appropriate discount rate is also an important decision, since a high discount rate effectively reduces to zero the present value of benefits and costs that occur many years in the future. This does not imply that a different discount rate should be use when environmental impacts are important; in fact, it is always wrong to mix discount rates within one analysis. Given the importance of the discount rate, however, it is important to do sensitivity analysis using different discount rates. This can yield useful information to the decision maker when comparing alternatives that have very different time profiles of benefits and costs (including environmental ones). Two approaches are possible to incorporating longterm environmental effects. One approach is to extend the time horizon of the entire analysis to cover a period long enough to include all environmental effects (at least to the point where, given the discount rate, any additional environmental impact has no further effect on the analysis, typically after 50-100 years). Alternatively, the present value of the entire future stream of environmental impacts (benefits and

costs) can be computed, and then incorporated in the normal project analysis framework in the same way that a residual value estimate for a long-lasting capital good would be. Spatial Boundaries. When environmental effects are present, careful thought must also be given to the appropriate spatial boundary of the analysis. The analyst often has to look far beyond the geographical boundaries of the project itself, especially when water or air pollution is involved. In other cases, global aspects may be important and require a further expansion of the accounting stance of the analysis. With both spatial and temporal externalities, the important rule is to be transparent in the assumptions being made, and explicitly state the adjustments that have been used in defining the analytical boundaries for the projectboth in space and over time. Whatever the actual techniques used to estimate the value of environmental benefits or damages, an important point that should be borne in mind is the likelihood of underestimation. Inevitably, some types of value will prove impossible to estimate using any of the available techniques, either because of lack of data or because of the difficulty of extracting the desired information from them. To this extent, any estimates of value will underestimate the total value; the estimates of project benefits will, therefore, be conservative, while estimates of costs will be optimistic. That some environmental benefits cannot be quantified, however, does not mean that they should be ignored. Rather, any unquantified benefits should be described qualitatively to the extent possible. Table 3 illustrates how a mix of quantifiable and unquantifiable benefits might be presented in a table. Several of the benefits that were not quantified in this instance are in fact potentially quantifiable, using the techniques indicated, but data and budget constraints prevented this. Since the quantifiable benefits were large enough to justify the proposed investments by themselves, devoting additional resources to quantifying the remaining benefits was judged to be unnecessary. Another potential problem which must always be considered is the risk of double-counting. The likelihood that total benefits will be underestimated because some benefits cannot be measured is well-recognized. Less well recognized is the opposite danger: that benefits (even if accurately measured) might be overestimated because some benefits are counted twice. An example will illustrate the problem. Suppose that the project aims to reduce air pollution at the site by relocating or shutting down polluting activities. The benefit of this reduction could be estimated by predicting the reduction in the prevalence of respiratory illnesses and valued using
11

Table 3. Quantified and unquantified benefits of protecting national parks in Haiti Benefits
Costs Project expenses Forgone agricultural income Forgone logging income Total On-Site benefits Biodiversity conservation Tourism potential Sustainable harvest of timber products Non-timber products Sub-total Off-site benefits Reduced damage to irrigation Reduced damage from flooding Increased water availability Sub-total Total quantifiable benefits 624 46 ? 10-30 10-30 Siltation and reduced dry-season flow reduce yields and area irrigated Flooding damages standing crops and infrastructure Reduced dry season flow endangers the populations water supply Change in productivity Replacement cost Replacement cost

Amount (US$ million)

6 2 ? 8 ? ? ? ?

Comments

Valuation technique
From PAD Opportunity cost Opportunity cost

Unlikely to be large due to high transport costs

Regionally outstanding ecosystem with many endemic species Considerable potential, but will require additional investment Limited potential due to high transport costs Considerable potential, but no data exist to estimate incremental benefits

CV CV, TC Change in productivity Change in productivity

Notes: All amounts are expressed in present value terms, discounted at 10%. Source: Adapted from World Bank, 1996d.

the reduction in treatment costs. At the same time, suppose that a hedonic technique is used to estimate the value of overall environmental quality. Since air pollution is part of environmental quality, treating these two estimates as though they described separate problems and adding the corresponding benefits together would be inaccurate. A final point. One should not lose sight of the beneficiaries and stakeholders involved in any project. Because of the existence of externalities, the costs and benefits of various activities to individual actors can vary substantially. The socially desirable outcome or action may well not be privately beneficial. Therefore it often is important to carry out an appropriate analysis of private returns (financial analysis) in order to understand the individual perspective. Only if this is done will it be possible to identify the policies or measures necessary to reconcile the individually and socially desirable actions.
12

Bibliography ADB. 1996. Economic Evaluation of Environmental Impacts: A Workbook. Manila: Asian Development Bank. Braden, J.B. and C.D. Kolstad (eds). 1991. Measuring the Demand for Environmental Quality. Contributions to Economic Analysis No.198. Amsterdam: North-Holland. Brookshire, D.S., and H.R. Nell. 1992. Benefit Transfers: Conceptual and Empirical Issues. Water Resources Research, Vol.28 No.3, March, pp.651-655. Brown, G., M. Ward, and D.J. Jansen. 1995. Capturing the Value of National Parks in Zimbabwe. ZWLMEC Project. World Bank.

Carson, R.T., 1991. Constructed Markets. In J.B. Braden and C.D. Kolstad (eds), Measuring the Demand for Environmental Quality. Contributions to Economic Analysis No.198. Amsterdam: NorthHolland. Dixon, J.A., L.F. Scura, R.A. Carpenter, and P.B. Sherman. 1994. Economic Analysis of Environmental Impacts. London: Earthscan. Dixon, J.A., and P.B. Sherman. 1990. Economics of Protected Areas: A New Look at Benefits and Costs. Washington: Island Press. Gittinger, J.P. 1982. Economic Analysis of Agricultural Projects. Second edition. Baltimore: Johns Hopkins University Press. Hanemann, W.M. 1992. Preface. In S. Navrud (ed.), Pricing the European Environment. Oslo: Scandinavian University Press. Hausman, J.A. (ed.). 1993. Contingent Valuation: A Critical Assessment. Contributions to Economic Analysis No.220. Amsterdam: North-Holland. Hufschmidt, M.M., D.E. James, A.D. Meister, B.T. Bower, and J.A. Dixon. 1983. Environment, Natural Systems, and Development: An Economic Valuation Guide. Baltimore: Johns Hopkins University Press. Kramer, R.A., N. Sharma, and M. Munasinghe. 1995. Valuing Tropical Forests: Methodology and Case Study of Madagascar. Environment Paper No.13. Washington: World Bank. Mekhaus, S., and D.J. Lober. 1996. International Ecotourism and the Valuation of Tropical Rainforests in Costa Rica. Journal of Environmental Management, Vol.47, pp.1-10. Mitchell, R.C., and R.T. Carson. 1989. Using Surveys to Value Public Goods: the Contingent Valuation Method. Washington: Resources for the Future. Navrud, S. (ed.). 1992. Pricing the European Environment. Oslo: Scandinavian University Press. NOAA, 1993. Report of the NOAA Panel on Contingent Valuation. Federal Register, Vol.58 No.10, Friday January 15, pp.4602-4614. Ostro, B., J.M. Sanchez, C. Aranda, and G.S Eskeland. 1996. Air Pollution and Mortality: Results from a Study of Santiago, Chile. Journal of Exposure Analysis and Environmental Epidemiology. Vol.6 No.1, pp.97-114.

Palmquist, R.B. 1991. Hedonic Methods. In J.B. Braden and C.D. Kolstad (eds), Measuring the Demand for Environmental Quality. Contributions to Economic Analysis No.198. Amsterdam: NorthHolland. Pearce, D.W., and J.W. Warford. 1993. World Without End: Economics, Environment, and Sustainable Development. Oxford: Oxford University Press. Randall, A. 1991. Total and Nonuse Values. In J.B. Braden and C.D. Kolstad (eds), Measuring the Demand for Environmental Quality. Contributions to Economic Analysis No.198. Amsterdam: NorthHolland. World Bank. 1994. Chile: Managing Environmental Problems: Economic Analysis of Selected Issues. Report No.13061-CH. Washington: World Bank. World Bank. 1996a. Handbook on Economic Analysis of Investment Operations. Washington: World Bank. World Bank. 1996b. Staff Appraisal Report: Republic of Croatia Coastal Forest Reconstruction and Protection Project. Report No.15518-HR. Washington: World Bank. World Bank. 1996c. Valuing the Health Effects of Air Pollution. DEC Note No.7. Washington: World Bank. World Bank. 1996d. Technical Annex: Republic of Haiti Forest and Parks Protection Technical Assistance Project. Report No.T-6948-HA. Washington: World Bank. Endnotes 1. For a general survey of techniques used in valuing environmental benefits, see Dixon and others (1995). For a more detailed exposition of the use of many of these techniques, see Hufschmidt and others (1983). For a technical discussion of the economic theory behind many of these technique, see Braden and Kolstad (1991). Hanemann (1992) provides a historical account of the development of the principal environmental valuation techniques. 2. Palmquist (1987) reviews the theory that forms the basis of hedonic estimation. 3. The theory and application of TC methods are described fully in Hufschmidt and others (1983). For numerous examples of the application of TC methods to value recreational benefits in Europe, see Navrud ( 1992).
13

4. These benefits take the form of consumers surplus, the benefit they enjoy above the costs involved in taking part in the recreational activity. A basic assumption is that the consumers surplus of the most distant visitor is zero, and that anyone more distant does not come to this site since the costs (travel costs) exceed the value of the benefits of the visit (It is important to note that the value of the site is not given by the total travel cost; this information is only used to derive the demand curve and thereby estimate the consumers surplus of visitors.) 5. A vast literature has developed on contingent valuation techniques. The standard text is Mitchell

and Carson (1989); for a more theoretical exposition, see Carson (1991). 6. A special issue of the Water Resources Research was devoted to benefits transfer, and provides the best available overview of the conceptual and empirical issues involved; see Brookshire and Nell (1992) and the following papers in that issue. A recently completed report by the Asian Development Bank relies heavily on the use of benefit-transfer (ADB, 1996), and contains many examples of the application of benefit transfer.

EA Sourcebook Updates
Number and Title 1 The World Bank and Environmental Assessment: An Overview 2 Environmental Screening 3 Geographic Information Systems for Environmental Assessment and Review 4 Sectoral Environmental Assessment 5 Public Involvement in Environmental Assessment: Requirements, Opportunities and Issues 6 Privatization and Environmental Assessment: Issues and Approaches 7 Coastal Zone Management and Environmental Assessment (also in Arabic) 8 Cultural Heritage in Environmental Assessment (also in Arabic) 9 Implementing Geographic Information Systems in Environmental Assessment 10 International Agreements on Environment and Natural Resources: Relevance and Application in Environmental Assessment (second edition ) 11 Environmental Auditing Date Apr. 93 Apr. 93 Apr. 93 Oct. 93 Oct. 93

Number and Title 12 Elimination of Ozone Depleting Substances

13 Guidelines for Marine Outfalls and Alternative Disposal and Reuse Options 14 Environmental Performance Monitoring and Supervision 15 Regional Environmental Assessment 16 Challenges of Managing the EA Process 17 Analysis of Alternatives in Environmental Assessment 18 Health Aspects of Environmental Assessment 19 Assessing the Environmental Impact of Urban Development 20 Biodiversity and Environmental Assessment 21 Environmental Hazard and Risk Assessment 22 Environmental Assessment of Mining Projects

Date Mar. 96
Mar. 96 Jun. 96 Jun. 96 Dec. 96 Dec. 96 Jul. 97 Oct. 97 Oct. 97 Dec. 97 Mar. 98 Apr. 98

Mar. 94 Mar. 94 Sep. 94 Jan. 95 Mar. 96

Aug. 95

23 Economic Analysis and Environmental Assessment

This Update was prepared by John Dixon and Stefano Pagiola of the Environmental Economics and Indicators Unit. The EA Sourcebook Updates provide guidance for conducting environmental assessments (EAs) of proposed projects and should be used as a supplement to the Environmental Assessment Sourcebook. The Bank is thankful to the Government of Norway for financing the production of the Updates. Please address comments and inquiries to Colin Rees and Aidan Davy, Managing Editors, EA Sourcebook Updates, Environment Department, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. MC-5-105, (202) 458-2715.

14

Environmental Assessment S o u r c e b o o k

UPDATE

Environment Department The World Bank

January 1999 Number 25

Environmental Management Plans

Prediction of the potential adverse environmental and social impacts arising from development interventions is at the technical heart of the environmental assessment (EA) process. An equally essential element of this process is to develop measures to eliminate, offset, or reduce impacts to acceptable levels during implementation and operation of projects. The integration of such measures into project implementation and operation is supported by clearly defining the environmental requirements within an environmental management plan (EMP). EMPs provide an essential link between the impacts predicted and mitigation measures specified within the EA report, and implementation and operational activities. They outline the anticipated environmental impacts of projects, the measures to be undertaken to mitigate these impacts, responsibilities for mitigation, timescales, costs of mitigation, and sources of funding. This Update provides an introduction to mitigation measures and EMPs, identifies the policy framework for preparing EMPs for Bank financed activities, outlines the main components of EMPs, and discusses means to ensure that commitments within the EMP are carried through to implementation and operation. It compliments information presented in Chapter 1 of the EA Sourcebook.

Introduction A development projects mitigation measures encompass all actions taken to eliminate, offset, or reduce potentially adverse environmental impacts to acceptable levels. Such measures are typically associated with the latter stages of project planning, although in practice they may occur at any stage throughout the project cycle. Normally, potential impacts are identified early during the initiation and scoping stages of EA for a project, and measures to avoid or minimize impacts are incorporated into the alternatives being considered. In this respect, some of the most important measures to protect the environment or local communities become integral to the project design, and are never reflected within a formal environmental management plan (EMP). For example, the environmental and social bases for choosing a preferred location for a project will not necessarily be evident within the EMP. Some of the primary objectives of mitigation measures are summarized in Box 1. Mitigation measures may be broadly defined as either structural or non-structural. Non-structural measures include improvements to the legal or institutional framework, economic incentives (such as realistic pricing of utility services), training, and measures to

enhance public awareness. An example is the introduction of irrigation service fees to promote efficiency of water usage and water conservation. Structural measures encompass design or location changes, engineered structures, or landscape treatments, based on the use of environmentally sound techniques and technologies. Environmental management plans (EMPs) outline the mitigation, monitoring, and institutional measures to be taken during project implementation and operation to avoid or control adverse environmental impacts, and the actions needed to implement these measures. They provide a crucial link between alternative mitigation measures evaluated and described within the EA report (see Update no. 17, Analysis of Alternatives), and ensuring that such measures are implemented. EA reports are essentially planning documents with no legal basis. In many cases, mitigation measures outlined in EAs are described in illustrative terms, or have neither been committed to by the borrower nor reflected in the project design. In this regard, the EMP is a basis for negotiation and reaching agreement with borrowers on a projects key social and environmental performance standards. The components of an EMP are described in the section below on components.
Insert in Update Binder chapter 1

Box 1. Primary objectives of mitigation measures

Mitigation measures aimed at eliminating, offsetting, or reducing adverse environmental impacts can have a range of objectives. Some of these are briefly defined below in approximate order of priority. The first priorities are avoidance or prevention of impacts, whereas the priority of the other categories is less rigid.
n

associated action plan (in effect an EMP) is often the best approach to determining the nature and extent of environmental concerns at an existing facility. The plan should identify appropriate mitigation measures, estimate the cost of proposed measures, and recommend a schedule for implementing them. For certain projects, the EA report may consist of an environmental audit; in other cases, the audit is part of the EA documentation and the EMP. World Bank Policy in relation to EMPs The Banks Operational Policy 4.01 (OP 4.01) identifies EMPs as an essential feature of category A projects; for category B projects, the EA may result in development of an EMP only, with no separate EA report. The specific requirements relating to EMPs are set out in Annex C to the Banks business procedure 4.01 (BP 4.01)these procedures are mandatory. The content of BP 4.01 is largely reflected within this Update, which also includes many elements of recommended good practice. Components of an EMP There is no standard format for EMPs. The format needs to fit the circumstances in which the EMP is being developed and the requirements which it is designed to meet. Mitigation measures and the means of ensuring their implementation for larger category A projects will often be described in some detail. Conversely, an EMP arising from an environmental audit might be summarized in a one or two page schedule. An EMP may be presented as two or three separate plans depending on borrowing country requirements and project circumstances. The EMP should be formulated in such a way that it is easy to use. References within the plan should be clearly and readily identifiable. Also, the main text of the EMP needs to be kept as clear and concise as possible, with detailed information relegated to annexes. The EMP should identify linkages to other relevant plans relating to the project, such as plans dealing with resettlement or indigenous peoples issues. The following aspects should typically be addressed within EMPs. Summary of impacts: The predicted adverse environmental and social impacts for which mitigation is required should be identified and briefly summarized. Cross-referencing to the EA report or other documentation is recommended, so that additional detail can readily be referenced. Description of mitigation measures: The EMP identifies feasible and cost effective measures to reduce

Avoidance. Avoiding projects or activities that could result in adverse impacts; avoiding certain types of resources or areas considered to be environmentally sensitive. This approach is most effective when applied in the earliest stages of project planning. Prevention. Measures aimed at impeding the occurrence of negative environmental impacts and/or preventing such an occurrence having harmful environmental and social impacts. Preservation. Preventing any future actions that might adversely affect an environmental resource or attribute. This is typically achieved by extending legal protection to selected resources beyond the immediate needs of the project. Minimization. Limiting or reducing the degree, extent, magnitude, or duration of adverse impacts. Mitigation can be achieved by scaling down, relocating, or redesigning elements of a project. Rehabilitation. Repairing or enhancing affected resources, such as natural habitats or water sources, particularly when previous development has resulted in significant resource degradation. Restoration. Restoring affected resources to an earlier (and possibly more stable and productive) state, typically background/pristine condition. Compensation. Creation, enhancement, or protection of the same type of resource at another location, to compensate for resources lost to development.

EMPs are essential elements of EAs for Category A projects, but for many Category B projects, a simple EMP alone may suffice. Typically, this applies to smaller projects not affecting environmentally sensitive areas, which present issues that are narrow in scope, welldefined, and well understood. Alternative approaches may be more effective in integrating environmental concerns into the planning process of such projects, including: environmental design criteria and emission standards for small-scale industrial plants; design criteria and construction supervision for small-scale rural works; and environmental siting criteria, construction standards, and inspection procedures for many social fund projects, such as housing or local schools projects. For industrial rehabilitation, expansion, or privatization projects, an environmental audit and
2

potentially significant adverse environmental and social impacts to acceptable levels. Each mitigation measure should be briefly described with reference to the impact to which it relates and the conditions under which it is required (for example, continuously or in the event of contingencies). These should be accompanied by, or referenced to, designs, equipment descriptions, and operating procedures which elaborate on the technical aspects of implementing the various measures. Where the mitigation measures may result in secondary impacts, their significance should be evaluated. Description of monitoring program: Environmental performance monitoring should be designed to ensure that mitigation measures are implemented, have the intended result, and that remedial measures are undertaken if mitigation measures are inadequate or the impacts have been underestimated within the EA report. It should also assess compliance with national standards and World Bank Group requirements or guidelines. The monitoring program should clearly indicate the linkages between impacts identified in the EA report, indicators to be measured, methods to be used, sampling locations, frequency of measurements, detection limits (where appropriate), and definition of thresholds that will signal the need for corrective actions, and so forth. Although not essential to have complete details of monitoring in the EMP, it should describe the means by which final monitoring arrangements will be agreed. For additional details, refer to EA Sourcebook Update no. 14: Environmental Performance Monitoring and Supervision. Institutional arrangements: Responsibilities for mitigation and monitoring should be clearly defined. The EMP should identify arrangements for coordination between the various actors responsible for mitigation. Environmental management usually involves many governmental bodies and other agencies, and links between the various actors are often complex and the hierarchy for decision making unclear. Agencies may be somewhat territorial, and reluctant to consult with or share information with others. Conflicts may also arise between institutions, particularly between those promoting development and those with a mandate for environmental protection. Shared or overlapping responsibilities within several institutions can also hinder effective enforcement of environmental control measures. It is important to account for such locationspecific circumstances and constraints in developing EMPs. Some of the key legal and institutional considerations with respect to EMPs are:

Legal framework for environmental protection: Is the framework for protection and management of resources affected by the project robust, and supported by enforcement measures? Responsibilities for environmental management: Are these clearly defined in relation to the resources affected by the project, and adequately resourced? Responsibilities for implementing mitigation measures: Are these clearly defined for environmental and social mitigation measures, and adequately resourced? Legal basis for mitigation: Do the project legal documents help to ensure that borrowers implement mitigation measures? When the borrowers institutional or technical capacity to effectively implement mitigation measures proves inadequate, a specific environmental component may be included in the project to strengthen that capacity. Institutional difficulties may sometimes be best resolved through establishing an autonomous, stable and qualified project implementation unit. If the project implementing agency lacks sufficient commitment or capacity for carrying out mitigation measures, it may be preferable to contract out operational responsibility for these measures. Implementation schedule and reporting procedures: The timing, frequency, and duration of mitigation measures should be specified in an implementation schedule, showing links with the overall project implementation plans (PIP). Where implementation of mitigation measures is tied to the project legal agreements, these linkages should be outlined. For example, some mitigation measures may be made conditions for loan effectiveness or disbursement. Procedures to provide information on the progress and results of mitigation and monitoring measures should also be clearly specified. As a minimum, the recipients of such information should include those with responsibility for ensuring timely implementation of mitigation measures, and for undertaking remedial actions in response to breaches of monitoring thresholds. In addition, the structure, content and timing of reporting to the Bank should be designed to facilitate supervision. The Task Manager should carefully consider arrangements for tracking receipt (and subsequent dissemination) of monitoring reports. Cost estimates and sources of funds: These should be specified for both the initial investment and recurring expenses for implementing all measures contained in the EMP, integrated into the total project costs, and factored into loan negotiations.
3

Where practicable, decisions regarding appropriate mitigation measures should be justified by an economic evaluation of potential environmental impacts, aimed at: Measuring the cost-effectiveness of different mitigation options where a project is required to meet a set of environmental standards or achieve specific environmental objectives Determining the appropriate level of mitigation where there is scope for a trade-off between environmental quality and the costs (and benefits) of achieving it Internalizing the economic value of residual impacts or intended environmental improvements into the final economic appraisal of the project.

Flexibility is best achieved by ensuring that response arrangements can be rapidly adapted to new and changing circumstances. Decentralization of responsibilities for EMP implementation, where appropriate, can aid flexibility, for example by empowering those responsible for monitoring to sanction and undertake remedial measures. In some cases, the project design changes following appraisal and even during implementation, which highlights the importance of periodically revising EMPs. Public involvement in developing EMPs The EMP should clearly describe and justify the proposed mitigation measures to facilitate public consultation. Consultation with affected people and NGOs should be integral to all Category A projects, and are advisable for many Category B projects in order to understand the acceptability of proposed mitigation measures to affected groups. In some situations, the development of environmental awareness amongst stakeholders is important to ensuring effective consultation on the EMP. Where projects involve socially and politically sensitive land acquisition or resettlement, these issues should be fully addressed in either the EMP or resettlement action plan (RAP) or indigenous peoples development plan (IPDP). The consultation process should help to design achievable mitigation measures. In this regard, it is recommended that the affected public participate in the design of such measures, particularly where their success depends on buy-in or actions on their part. Where appropriate, this may be supported by including formal requirements within the TOR for public participation in developing the EMP. Usually there are no formal mechanisms for the public to determine whether environmental commitments made by a proponent or government in the EMP are followed. It is recommended that information on progress with implementing mitigation and monitoring activities should be shared with the affected public. EMPs for strategic EAs Whereas the aim of project specific EAs is to ensure that projects are implemented in an environmentally acceptable manner, strategic EAs are aimed at ensuring that the choice and design of projects are inherently acceptable from an environmental perspective. They also aim to establish a sound basis for mitigation, monitoring, and management at the project level. An EMP derived from a regional EA should help to establish a sound planning and management framework that provides for addressing cumulative, direct and

The level of mitigation required may be determined by political or legal considerations, Bank requirements, international agreements (see Update no. 10: International Agreements on Environment and Natural Resources), stakeholder needs, or a combination of these factors. It is important to capture all costsincluding administrative, design and consultancy, and operational and maintenance costsresulting from meeting certain standards or modifying project design. The aim is to satisfactorily mitigate adverse impacts at least cost. The costs of preparing an EMP, which are borne by the borrower, vary depending on factors such as the complexity of potential impacts, the extent to which international consultants are used, and the need to prepare separate EMPs for sub-projects (for example with sectoral investment loans). Options for financing include the borrowers own resources, project preparation facilities (PPF), institutional development funds (IDF), or trust funds. Implementation costs may be met from the project loan, the borrowers/sponsors own resources, or from trust funds. Maintaining flexibility of EMPs EMPs should be dynamic flexible, and subject to periodic review. The extent to which EMPs should be reviewed and updated varies between and within sectors. As a rule, where the major environmental impacts are associated with the operational rather than the construction phases (particularly where operations are inherently variable), EMPs should be regularly revised. In part this is linked to the influence of changes in legislation on such operations. For example, an EMP for a road or a river crossing should be subject to review throughout the construction period, but thereafter the impacts are either static or best controlled by traffic management. Conversely, an EMP for a power plant or mining operation should be revised at regular intervals throughout the operational lifetime of the project.
4

indirect impacts of ongoing and planned investments through a spatial approach to mitigation, monitoring and management. EMPs for sectoral EAs should recommend measures for mitigating, monitoring, and managing impacts within the sector of interest, taking into account the overall environmental management capacity in the country. EMPs for strategic EAs should include the following: Recommendations for mitigation measures, tied to alternative development scenarios, to be implemented at the national, area-wide or sector level Guidelines for strengthening, adjusting, developing or harmonizing long-term environmental monitoring programs, with reference to applicable standards and international obligations Recommendations for changes to the environmental management framework to ensure adequate environmental resources management, and interagency coordination.

environmental management, to ensure that the EMP is adequately budgeted, and to determine whether the mitigation measures are properly addressed in project design and economic analysis. The Project Appraisal Document (PAD) should summarize the main measures contained in the EMP, describe environment-related loan conditionalities and covenants, and outline the program and budget for environmental supervision. During negotiations, Bank staff are responsible for translating the mitigation measures in the EMP into the project implementation plan (PIP) and loan agreement. Implementation and supervision of the EMP provisions should occur during project implementation, operation and evaluation. During project implementation, the borrower reports on compliance with environmental commitments, the status of mitigatory measures, and the findings of monitoring programs as specified in the project documents. The Bank bases supervision of the projects environmental aspects on the EMP as set out in the legal agreements for the project. Translating EMP into legal agreement It is important to translate the main findings and recommendations of the EMP into the legal documents, the principal legal instruments being the loan agreement (for the IBRD and IFC) or development credit agreement (for IDA) entered into with the borrower and the project agreement entered into with the project executing agency (where applicable). The loan/credit agreements normally include the terms of the loan or credit/guarantee, repayment obligations, provisions for the use of the proceeds of the loan and for the procurement financed by it and, in this context, obligations with respect to the carrying out of the project in question. Such obligations can include the borrowers undertakings with respect to the environment and are complemented at times by detailed requirements in project agreements concluded with the project executing agencies. A key condition for effective environmental implementation is the extent to which there are clear and specific environmental deliverables and performance indicators in the EMP and PIP, including adequate budgeting and specificity about financing of environmental measures, and associated environmental clauses in loan agreement and contract documents. When this is not the case: there is frequently less commitment on the part of the borrower and its implementing agencies; it is more difficult to supervise projects in terms of their environmental aspects; the supervision is less effective in terms of influencing the subsequent course of action;
5

Strategic EAs should highlight major impacts of concern in the sector or region, and strategic EMPs may prescribe standard approaches to subproject design and mitigation through environmental guidelines, manuals and monitoring requirements. This reduces the scope of work for individual EAs and EMPs for subprojects, and can positively influence investment activities throughout the sector or region. Ensuring implementation of EMP commitments Effective implementation of commitments set out in EMPs is aided by the borrower and Bank staff understanding their responsibilities at each stage of the project cycle, and by translating these commitments into a form that is legally binding. These aspects are discussed below. Responsibilities of Bank staff and borrowers The borrower or sponsor is responsible for preparing and implementing the EMP, and TOR for EAs should clearly reflect the importance of preparing a comprehensive EMP. In practice, when the draft EA report and EMP are submitted to the Bank prior to appraisal, many mitigation measures will be tentative or illustrative. However, the EMP needs to be finalized during appraisal and agreed upon during negotiations, together with performance monitoring indicators. The responsibilities of Bank staff vary depending on the stage of the project cycle. During appraisal, Bank staff are required to review the EMP with the borrower, to assess the adequacy of the institutions responsible for

and there is generally less environmental information in supervision reports. Lack of specific environmental clauses also makes it more difficult for implementing agencies and contractors to know what is expected. Some of the options for improving the legal basis for environmental mitigation are given in Box 2. Box 2. Improving the legal basis for mitigation

The vital link with the activities defined in the EMP is the environmental conditions and covenants the borrower and the Bank agree on. The basic factor to be considered include required activities and timing, assignment of responsibilities, and inclusion of appropriate funding and reporting arrangements. To reinforce

It is increasingly important to translate EMP commitments into appropriate language for environment-related conditions, covenants and implementation schedules in the legal agreements. For this purpose, several options are available: Conditions for loan negotiations, Board approval or effectiveness. This approach limits flexibility, but may be the best means of ensuring progress with mitigation up to loan effectiveness. Conditions for disbursement. While the Banks leverage decreases as project implementation progresses, conditioning disbursement on the achievement of key environmental milestones carries considerable weight. For example, when subprojects of sector loans that may be environmentally damaging are not appraised prior to project approval, it is important to condition disbursement on satisfactory review of EA reports. Dated covenants. NBthis needs revising; very general). These usually relate to the borrowers management decisions. As a result of these covenants, Bank supervision could effectively address problems and delays in the concerned areas whereupon implementation performance improved. Covenants have the advantage of bringing key issues to the full attention of both Bank supervision staff and the implementing agencies on the borrower side. Inclusion of mitigation measures within the Description of the Project under Schedule 2 in the legal agreement. This approach is useful insofar as it recognizes that environmental aspects are integral to the project, on an equal footing with other components. It is also a useful place for providing a detailed timetable for undertaking mitigation measures. The timetable should ideally be reinforced by covenants and conditions in the main sections of the legal agreement. General statement. This commits the borrower to following applicable national environmental regulations and/or international standards and practices, acceptable to the Bank. Where the regulations in a borrowers country and capacity for implementation and enforcement are adequate, a general statement may be sufficient.

the legal documents it is recommended that a monitoring summary be prepared that is linked to the project legal agreement. Bidding and contract documents It is important to translate EMP requirements into bidding and contract documents to ensure that obligations are clearly communicated to contractors. The bidding documents might also include environmental criteria as

part of the basis for selecting contractors. Contractors should also be obliged to follow appropriate environmental, health, and safety standards to reduce associated risks during construction and operation (see EA Sourcebook). For example, construction specifications might include clauses relating to: Preservation of the natural landscape to the extent possible, and restoration in the event of unavoidable damages

Conversion of construction camps and buildings into permanent residences, or removal to avoid deterioration into shantytowns Prevention of accidental spillage of contaminants, debris, or other pollutants, especially into streams or underground water resources Noise, dust and sediment loading Ensuring proper disposal of waste materials and rubbish.

Supervision of the EMP Bank experience confirms that systematic supervision and monitoring are fundamental to the successful

implementation of projects, and EMPs in particular. Therefore, it is vital that an appropriate supervision program be developedpreferably in collaboration with governmental implementing and environmental agencies as well as with affected populations and NGOsto monitor progress and analyze and overcome any obstacles encountered in the implementation of the EMP. Detailed guidance on environmental performance monitoring and supervision is provided in EA Sourcebook Update no. 14: Environmental Performance Monitoring and Supervision.

Previously published EA Sourcebook Updates

Number and Title 1 The World Bank and Environmental 1 The World Bank and Environmental Assessment: An Overview
2 Environmental Screening 3 Geographic Information Systems for Environmental Assessment and Review 4 Sectoral Environmental Assessment 5 Public Involvement in Environmental Assessment: Requirements, Opportunities and Issues 6 Privatization and Environmental Assessment: Issues and Approaches 7 Coastal Zone Management and Environmental Assessment (also in Arabic) 8 Cultural Heritage in Environmental Assessment (also in Arabic) 9 Implementing Geographic Information Systems in Environmental Assessment 10 International Agreements on Environment and Natural Resources: Relevance and Application in Environmental Assessment (second edition ) Auditing 11 Environmental 12 Elimination of Ozone Depleting Substances

Date Apr. 93 Apr. 93

Apr. 93 Apr. 93 Oct. 93 Oct. 93

Number and Title 1 The World Bank and Environmental 13 Guidelines for Marine Outfalls and Alternative Disposal and Reuse Options
14 Environmental Performance Monitoring and Supervision 15 Regional Environmental Assessment 16 Challenges of Managing the EA Process 17 Analysis of Alternatives in Environmental Assessment 18 Health Aspects of Environmental Assessment 19 Assessing the Environmental Impact of Urban Development 20 Biodiversity and Environmental Assessment 21 Environmental Hazard and Risk Assessment 22 Environmental Assessment of Mining Projects 23 Economic Analysis and Environmental Assessment 24 Environmental Assessment of Social Fund Projects

Date Apr. 93 Mar. 96

Jun. 96 Jun. 96 Dec. 96 Dec. 96 Jul. 97 Oct. 97 Oct. 97 Dec. 97 Mar. 98 Apr. 98 Jan. 99

Mar. 94 Mar. 94 Sep. 94 Jan. 95 Mar. 96

Aug. 95 Mar. 96

This Update was prepared by Aidan Davy, with assistance from World Bank environment specialists. The EA Sourcebook Updates provide guidance for conducting environmental assessments (EAs) of proposed projects and should be used as a supplement to the Environmental Assessment Sourcebook. The Bank is thankful to the Government of Norway for financing the production of Updates. Please address comments and inquiries to Colin Rees, Managing Editor, EA Sourcebook Updates, Environment Department, The World Bank, 1818 H St. NW, Washington, D.C., 20433, Room No. MC-5-143, (202) 458-2715.