The quality of environmental impact statements and environmental impact assessment practice in Bangladesh

Abstract

This paper investigates the quality of environmental impact statements (EISs) and gives us an understanding about the performance of environmental impact assessment (EIA) practice in Bangladesh. EIA has been formally practised in Bangladesh since 1995.However, no study has yet been conducted on the quality of EISs. This study fills this gap. This empirical study shows that the quality of EISs in Bangladesh is ‘just satisfactory’ only. A significant proportion of EISs (34%) are still unsatisfactory. Finally, possible factors affecting the quality of EISs are examined and measures to improve the quality of EISs are recommended. The findings will be useful to EIA practitioners and other stakeholders in Bangladesh. This study will also provide a general guideline for other developing countries with similar socio-economic context.

Keywords:

• EIA system
• quality of EIS
• Bangladesh

1. Introduction

The overall effectiveness of environmental impact assessments (EIAs) depends on many aspects but among these the quality of environmental impact statement (EIS) is of particular importance (Lee and Colley 1991, Lee et al. 1999). An EIS is the product of an EIA process: as such the quality of the document is deeply associated with the quality of the whole EIA process (Fuller 1999). As a ‘rule of thumb’, a poor quality EIS can result from a good EIA process but it is less likely that a good quality EIS results from a poor EIA system (Fuller 1999). Indeed, it is the fundamental indicator of an effective EIA system (Barker and Wood 1999, Modak and Biswas 1999) that it presents scientific information (Pinho et al. 2007) for sustainable planning of a project.

While there have been extensive studies on the quality of EISs in developed countries (Lee et al. 1999, Weston 2000, Glasson et al. 2005), little has been done in developing countries. This may be due to the fact that developing countries are relatively late starters with the adoption and application of EIA. However, there have been some recent studies in developing countries, such as in South Africa by Sandham and Pretorius (2008), in Sri Lanka by Samarakoon and Rowan (2008), and in Egypt by Badr et al. (2011). In Bangladesh, although EIA was formally introduced in 1995, no empirical research was undertaken except a few general studies (e.g. Momtaz 2005) to understand the quality of EISs. This study aims to attempt to fill this gap.

This paper is divided into three main sections. First, it describes criteria and interview methods to review the quality of EISs and identify factors influencing the quality of EIS respectively. Second, it presents results and discussion on the quality of EISs including factors influencing the quality of EISs. Finally, conclusion was drawn and recommendations were made with a view to improving the effectiveness of EIA in Bangladesh.

2. Methodology

In this study, 30 EIA reports of development projects from various sectors were reviewed. The EIA reports were selected randomly from the EISs produced between 1994 and 2008. A set of review criteria (review package by Lee and Colley 1992) was used to assess the quality of EISs in Bangladesh. After the assessment of quality, face-to-face in-depth interviews were conducted to identify the underlying factors influencing the quality of EISs in Bangladesh.

2.1 Review criteria

Lee and Colley's (1992) approach relies upon the use of a hierarchy of detailed review criteria. This approach has been widely used by previous researchers elsewhere (e.g. Hughes and Wood 1996, Glasson et al. 1997, Barker and Wood 1999, Simpson 2001, Cashmore et al. 2002, Gray and Edward-Jones 2003, Badr et al. 2004, Pinho et al. 2007, Sandham and Pretorius 2008, Badr et al. 2011) for the last two decades. So the approach, along with good practice criteria, is internationally well recognized and applied in the context of Bangladesh as well.

However, according to Lee (2000), the quality of EISs has to be assessed taking into account the regulatory and procedural context in which they are prepared (Lee 2000, p. 138). Therefore a careful selection of criteria is necessary. As the basic requirements of EIA processes are all very similar, relatively little amendment is necessary to make a review package (such as that given in Lee and Colley 1992) suitable to the particular context of concern (Lee and Colley 1991). Previous studies (e.g. Sandham and Pretorius 2008, Badr et al. 2011) held outside the UK context have taken this concern into account. Appendix 1 shows how we have adapted the Lee and Colley (1992) review criteria to make the review package suitable in the context of Bangladesh (the few criteria added are marked with an asterisk). It should be noted that all the criteria are divided into three categories; namely area (1, 2, 3 and 4), category (1.1, 1.2 etc.) and sub-category (1.1.1, 1.1.2 etc.). All the areas, categories and sub-categories are referred to as criteria or tasks.

2.1.1 Assessment procedure

To assess the quality of an EIS, each sub-category was graded using the letters A to F (where A = Excellent, B = Good, C = Just satisfactory, D = Poor, E = Very poor, F = Fail) based on the quality of information presented under that sub-category. Where the criterion was irrelevant or not applicable, the symbol ‘N/A’ was used. The judgement of the quality of information for each criterion and subsequent grading were based on adequacy and reliability of information. Then the grades were determined for each respective category. This way the overall grade was determined for each area. Finally, from the grades given to each of the four areas, an overall grade of the EIS was determined. The grading of categories or areas was determined by the performance grade and relative weight of the respective sub-category (or category) rather than simple numerical average. Two independent reviewers perform the assessment separately since the attribution of grades is inherently subjective (Lee and Colley 1992). Significant differences found for a few criteria were carefully reviewed for the third time and the final result was accepted.

2.2 Interview method

After reviewing the quality of EISs, a face-to-face interview was conducted to identify the factors influencing the quality of EISs in Bangladesh. A total of 30 participants were purposively selected and contacted by emails and telephone. A total of 26 participants responded to attend an interview. Finally, a total of 22 participants were interviewed. The interviewees were asked two open-ended questions:

•	What are the factors affecting the quality of EISs in Bangladesh?
•	What are the recommendations to improve the quality of EISs in Bangladesh?

The interviewees belong to various organizations such as universities, government agencies, donor agencies and environmental NGOs. The two questions were asked and data were recorded using voice recorder and note taking. Then after a systematic analysis of data using NVivo software, judgement was used to select the relevant factors influencing the quality of EIS in Bangladesh.

3. Results and discussion

3.1 Overall quality of EISs in Bangladesh

Figure 1 shows the quality of 30 sampled EISs of development projects. From the graph it is seen that 66% of EISs are graded sastisfactory (A–C) and 34% of EISs are graded unsatisfactory (D–F). These findings correspond to the findings of other similar studies (Lee and Colley 1992, Glasson et al. 1997, Barker and Wood 1999, Cashmore et al. 2002, Sandham and Pretorius 2008, Jalava et al. 2010, Badr et al. 2011). However, ‘satisfactory’ does not mean that information presented in EISs is adequate; there are still deficiencies in information presented in the EISs since major tasks are poorely performed and omitted. It is notable that among the EISs graded satisfactory (66% (21)), 30% are graded as A and B, and the share of grade C is the highest (36%). This indicates that the quality of EISs in Bangladesh tends to be ‘just satisfactory’.

Figure 1 Overall quality of EISs in Bangladesh.

In addition to the overall quality of EISs, it is important to identify the strengths and weakness of the contencts of EISs. It is imperative to know which criteria are adequately addressed in the EISs and which are not. Therefore a more detailed analysis of the quality of EISs with reference to the performance of each review area is necessary.

3.2 Review results of the quality of EISs by area

Table 1 shows the performance of each review area. From the table, it is obvious that the performance (quality) of four areas in the EIA reports (see Appendix 1) varies significantly. The table shows that areas related to description of the development and baseline conditions (area#1) and the presentation of EIS and communication (area#4) are better performed than areas related to identification and evaluation of key impacts (area #2) and alternatives and environmental mitigation (area#3). Among the four areas, area#1 is the best performed area and area#2 is the worst performed area. For area#1 and area#4, only 17% and 23% EISs are unsatisfactory respectively (Table 1). On the other hand, area#2 and area#3 involve 50% and 27% unsatisfactory EISs respectively. To understand the strengths and weaknesses of the EIA reports in detail, a discussion of the performance of criteria under each area is given below.

Table 1 Comparison between areas of EISs (the contents of the EISs are divided into four major areas, as described in Appendix 1).

Grades	Area#1	Area#2	Area#3	Area#4
A = Excellent	13%	0%	7%	21%
B = Good	27%	23%	23%	40%
C = Just satisfactory	43%	27%	43%	16%
D = Poor	13%	27%	16%	16%
E = Very poor	4%	16%	7%	7%
F = Fail	0%	7%	4%	0%

3.2.1 Area#1: Description of development and baseline

This is one of the two better performed review areas in the EISs. Nevertheless, 17% of EISs are poor or unsatisfactory. The relatively straightforward and readily available tasks such as the background of the project, EIA aims, policy and legal framework for EIA, and description of the project performed best. Previous similar studies, such as Wood et al. (1996), Barker and Wood (1999), Lee et al. (1999) and Cashmore et al. (2002), also demonstrated that such descriptive tasks tend to be of higher quality. However, there are still deficiencies found in some tasks, for example related to baseline information where about 30% (9) EISs are unsatisfactory. In this study, although all of the sampled EISs contained baseline data, the coverage and the quality of baseline information tended to be poor. Also methods used for data collection, timing and duration of survey, and source of data are not clearly mentioned. These are the significant weaknesses given that the baseline data form the foundation of prediction and evaluation of impacts.

3.2.2 Area#2: Identification and evaluation of key impacts

This review area examines criteria such as methods used to identify and predict impacts, comprehensiveness of impacts, nature and magnitude of predicted impacts, assessment of the significance of predicted impacts, analysis of alternatives and selection of best options, and involvement of the public. This is the most important area of an EIS since design of projects depends on impacts identified and evaluated. However, this is the least well-performed review area with 50% (15) of EISs found to be poor and very poor. These findings correspond with previous studies (e.g. Wood et al. 1996, Fuller 1999, Glasson et al. 1997). Omissions and deficiencies occurring in this area relate to the inadequate coverage of impacts identified, inadequate explanation of methods used to predict and evaluate the impacts, and failure to adequately include the community in the EIA process.

All the studies used either experts' judgement or models or both to predict impacts but without detailed justification and clear description about the process. The EISs reviewed hardly used evaluation criteria and rarely interpreted why and how the predicted impacts are determined as significant. Earlier studies (e.g. Ross et al. 2006) also substantiate this finding. Many EISs reviewed in this study even failed to recognize that impact prediction and assessment of significant impacts are two separate stages. The overall deficiency in the prediction and evaluation of impacts thereby shows poor presentation of scientific information in EISs in Bangladesh. These results substantiate previous studies undertaken elsewhere (e.g. Warnken and Buckley 1998, Morrison-Saunders and Bailey 2003) and the studies found that the presentation of scientific information by EISs was considerably poorer than the expectation.

3.2.3 Area#3: Alternatives and environmental Mitigation

All EISs under this study have designed mitigation measures to address the significant environmental impacts of projects. However, more than 27% of EISs are poor and very poor in this area, making it one of the relatively poorly performed review areas in this study. Common omissions and deficiencies include poor description of significant adverse impacts to be mitigated, lack of justification of mitigation measures, lack of public involvement in implementing mitigation measures, lack of adequate mitigation measures for social impacts, and lack of justification of how the mitigating measures would be successful in protecting the environment. Also, monitoring mechanisms are not always properly described detailing parameters to be monitored and methods to be used. These deficiencies were found in other similar studies (e.g. Wood et al. 1996, Cashmore et al. 2002, Gray and Edward-Jones 2003, Sandham and Pretorius 2008).

3.2.4 Area#4: Presentation of EIS and communication

This is one of the best performed review areas where 80% of EISs are graded satisfactory and above and 20% EISs are unsatisfactory. Most of the EISs have used maps, pictures, tables and charts where appropriate. All EISs reviewed under this study included an executive summary or non-technical summary (NTS) and provided definitions of technical terms. Previous studies (e.g. Cashmore et al. 2002, Sandham and Pretorius 2008) also found that tasks related to general structure, layout and presentation of EIS tend to be performed well.

However, it was observed that in many cases NTSs were not written in simple language and rather often were full of jargon and technical terms. As a short version of information detailed in the EISs, the aim of the NTSs is to deliver a message using simple language to the stakeholders of all levels. Furthermore, all EISs along with the NTSs were written in English language, not in Bengali, the official language of Bangladesh, and therefore a large number of stakeholders do not have access to these documents. At least the NTSs should be published in Bengali.

3.3 Factors determining the quality of EISs

The interview results revealed a number of factors that appeared to be relevant for the variation of quality of EISs in Bangladesh. Overall, the omissions and deficiencies identified in the EISs can be attributed to a range of factors such as inadequate time and funding to conduct EIA, inadequate baseline data, lack of consultants' independence, and poor terms of reference (TOR). Also the absence of legal provisions (code of conduct and accrediting system for consultants), inadequate best practice guidelines for preparation of good EISs, and lack of EIA experts influence the quality of EISs in Bangladesh. These factors are similar to those identified by previous studies (e.g. Wood et al. 1996, Barker and Wood 1999, Morrison-Saunders et al. 2001, Morrison-Saunders and Bailey 2003, Sandham and Pretorius 2008, Badr et al. 2011).

For example, the shortage of study time is one of the important factors responsible for poor quality EISs in Bangladesh. Experiences of EIA consultants in Bangladesh reveal that often proponents want to complete the EIA study quickly and ask the consultants to complete the study within weeks rather than months (Interview#1). Similarly, the lack of available baseline data often complicates the job of consultants and hence the quality of EISs. Particularly, the scarcity of ecological baseline data is very critical (Interview#2) and initiatives to overcome this problem are not in place.

Furthermore, consultants hired by the proponent often tend to serve the commercial interest of proponents ignoring the actual objective of the preparation of EIS. This attitude is particularly acute among the proponents in the private sectors, where the proponents tend to consider financial benefits only. In this case, the consultants just touch on the points to get the environmental approval of the project (Interview#22) and thereby serve the interest of the proponent (Interview#13) playing down the most important information that is useful for effective design of mitigation measures.

4. Conclusion and recommendations

The purpose of this study is to explain to what extent the stages of EIA and related tasks are performed well in Bangladesh. With this view the quality of EISs was reviewed. The findings reveal that the quality of EISs in Bangladesh is satisfactory in general though a significant number (34%) of EISs are still poor and very poor. This means that still there are gaps and lapses in the contents of EISs. A number of factors responsible for these gaps and lapses were also identified in this study.

To improve the quality of EISs, this study recommends that it is important to change the proponent's attitude in the allocation of adequate funds and time to prepare a good EIS. This is possible only when they are well aware of ultimate benefits of EIA application. The TOR must be prepared with adequate consultation of DOE (Department of Environment) before conducting the EIA study. The provision of codes of ethics and an accreditation system can force consultants to prepare a good quality of EISs. Furthermore, given the lack of available guidelines, the development and dissemination of adequate best practice guidelines for preparation of EISs will enable both the proponents and consultants to improve the quality of EISs in Bangladesh.

Acknowledgements

The paper is part of a PhD research project funded by the Endeavour Postgraduate Award Program, the Government of Australia. The authors are thankful to Dr Angus Morrison-Saunders and two other reviewers for their valuable comments on earlier drafts of this paper.

Appendix 1: Abbreviated review criteria developed in the context of Bangladesh based on Lee and Colley's (1992) review package structure

1. Description of the development and baseline
1.1 Description of development	2.4 Prediction of impacts
1.1.1 Background and objectives of project	2.4.1 Data gaps and uncertainty
1.1.2 Design and size of the project	2.4.2 Methods used with justification
1.1.3 Some indication of physical presences	2.4.3 Quantitative expression where possible
1.1.4 Nature of production process	2.4.4* Uncertainty
1.1.5 Nature and quantities of raw materials	2.5 Assessment of impact significance
1.1.6* Policy and legal framework for EIA	2.5.1 Significance of impact on community and environment
1.1.7* EIA aim and objectives	2.5.2 Methods used for evaluation of impacts
1.1.8* Limitation of study	2.5.3 Justification of methods used
1.2 Description of site	2.6*Community involvement
1.2.1 Land area taken by the development site	2.6.1 Description of community affected
1.2.2 The use of land taken	2.6.2 Involvement of community
1.2.3 Estimated duration of project alternatives	2.6.3 Methods of community involvement
1.2.4 Number of workers and means of transport	2.6.4 Inputs from community
1.2.5 Means of transporting raw materials	3. Alternatives and environmental mitigation
1.3 Waste generation	3.1 Alternatives
1.3.1 Types and quantities	3.1.1 Alternative sites
1.3.2 Production process and waste generation	3.1.2 Alternative process, design and activities
1.3.3 Treatment of wastes generated	3.1.3 Selection of alternatives
1.4 Environmental description	3.2 Scope and effectiveness of mitigation measures
1.4.1 Environment expected to be affected	3.2.1 Description of adverse impacts to be mitigated
1.4.2 Offsite impacts	3.2.2 Mitigation measures with justification
1.5 Environmental baseline	3.2.3 Residual impacts
1.5.1 Description of important components	3.3 Commitment to mitigation
1.5.2* Natural physical environment	3.3.1 Implementation arrangements
1.5.3* Biological environment	3.3.2 Monitoring program
1.5.4* Socio-economic environment	3.3.3* Parameters to be monitored
1.5.5 Methods and sources of data with justification	3.3.4* Feedback and reporting mechanism
1.5.6 Future state of environment without project	4. Presentation of EIS and communication
2. Identification and evaluation of key impacts	4.1 Layout
2.1 Definition of impacts	4.1.1 Introduction and brief description
2.1.1 Impact types	4.1.2 Logical arrangement of information
2.1.2 Impacts with regard to humans, ecology etc.	4.1.3 List of references
2.1.3 Impacts of accidents	4.2 Presentation
2.1.4 Impacts as the deviation of baseline	4.2.1 Comprehensible to non-specialist
2.2 Identification of impacts	4.2.2 Defining technical terms
2.2.1 Methods used for impacts	4.2.3 Presented as an integrated whole
2.2.1 Justification of methods used	4.3 Emphasis
2.3 Scoping	4.3.1 Potentially severe adverse impacts
2.3.1 Arrangements for scoping	4.3.2 Unbiased statements
2.3.2 Methods of collection of opinions	4.4 Executive summary
2.3.3 Inputs from stakeholders	4.4.1 Findings presented in a non-technical way
2.3.4 Selection of key impacts to be investigated	4.4.2 Recommendations