the role and functioning of environmental assessment: theoretical reflections upon an empirical...

16
Journal of Environmental Management 88 (2008) 1233–1248 The role and functioning of environmental assessment: Theoretical reflections upon an empirical investigation of causation Matthew Cashmore , Alan Bond, Dick Cobb InteREAM, School of Environmental Sciences, University of East Anglia, Norwich, Norfolk NR4 7TJ, England Received 2 October 2006; received in revised form 27 April 2007; accepted 19 June 2007 Available online 1 August 2007 Abstract Environmental assessment is an intriguing policy phenomenon: it is employed in an ever-increasing range of contexts the world-over, yet research indicates it rarely efficiently or effectively achieves its principal purpose of promoting sustainable development. Increasingly, practical limitations in the effectiveness of environmental assessment are attributed to its theoretical shortcomings, particularly in relation to the conception of causation. This research advanced debate on environmental assessment by examining the theoretical implications of an empirical analysis of its causal operation in purposefully selected cases. The causal models derived from the research data illustrate the diversity of mechanisms through which environmental assessment can contribute to sustainability, and provide an insight into the complexity and contextuality of causation in the empirical realm. The research findings also highlight a multiplicity of societal expectations concerning environmental assessment’s contribution to sustainable development. It is concluded that the interplay of non-rational variables (such as power, agency, experiences and expectations) necessitates the pursuit of a reflexive accommodation of purposes, methods and context in environmental assessment practices. r 2007 Elsevier Ltd. All rights reserved. Keywords: Environmental assessment; Policy appraisal; Causation; Sustainable development; Governance 1. Introduction Environmental assessment—a collective term for forms of policy appraisal that address the probable environmental and social outcomes of human actions—is an intriguing policy phenomenon. Originating from vague aspirations espoused in the US National Environmental Policy Act of 1969, the scale and rapidity of its institutionalisation is remarkable. It is now routinely employed in the appraisal of trade negotiations, poverty reduction strategies, government policies, and regional and local development initiatives in more than 100 nations and by various bilateral, multilateral and interna- tional organizations (Dalal-Clayton and Sadler, 2005; Eur- opean Commission, 2005; Russel and Jordan, 2006; Wood, 2003). Yet despite the remarkable institutionalisation of environmental assessment, empirical evidence indicates that its effectiveness in promoting sustainable development— which is generally interpreted to be its primary purpose (Glasson et al., 2005)—is limited (Cashmore et al., 2004; Taylor, 1984; Wood and Jones, 1997). A further characteristic of the phenomenon is that environmental assessment is a singularly theoretically underdeveloped field of endeavour (Lawrence, 1997). Increasingly, the twin issues of limited effectiveness and theoretical impoverishment are perceived to be linked (Cashmore, 2004; Dalkmann et al., 2004; Deelstra et al., 2003). Causation in environmental assessment (put simply, the causal mechanisms and outcomes through which it contributes to sustainable development) is a central focus of contemporary debate upon theory and effectiveness, albeit often unwittingly. The causal model that has underpinned mainstream practices since its inception developed from the modernist notion of ‘speaking truth to power’. Environmental assessment was envisaged as a tool that promoted sustainable development by generating scientific data which apolitical stakeholders (usually public administrators) used, alongside other information, to make rational design and approval decisions (Nilsson and Dalkmann, 2001). This was to be achieved predominantly ARTICLE IN PRESS www.elsevier.com/locate/jenvman 0301-4797/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jenvman.2007.06.005 Corresponding author. Tel.: +44 1603 593797; fax: +44 1603 591327. E-mail address: [email protected] (M. Cashmore).

Upload: matthew-cashmore

Post on 26-Jun-2016

214 views

Category:

Documents


2 download

TRANSCRIPT

ARTICLE IN PRESS

0301-4797/$ - se

doi:10.1016/j.je

�CorrespondE-mail addr

Journal of Environmental Management 88 (2008) 1233–1248

www.elsevier.com/locate/jenvman

The role and functioning of environmental assessment: Theoreticalreflections upon an empirical investigation of causation

Matthew Cashmore�, Alan Bond, Dick Cobb

InteREAM, School of Environmental Sciences, University of East Anglia, Norwich, Norfolk NR4 7TJ, England

Received 2 October 2006; received in revised form 27 April 2007; accepted 19 June 2007

Available online 1 August 2007

Abstract

Environmental assessment is an intriguing policy phenomenon: it is employed in an ever-increasing range of contexts the world-over,

yet research indicates it rarely efficiently or effectively achieves its principal purpose of promoting sustainable development. Increasingly,

practical limitations in the effectiveness of environmental assessment are attributed to its theoretical shortcomings, particularly in

relation to the conception of causation. This research advanced debate on environmental assessment by examining the theoretical

implications of an empirical analysis of its causal operation in purposefully selected cases. The causal models derived from the research

data illustrate the diversity of mechanisms through which environmental assessment can contribute to sustainability, and provide an

insight into the complexity and contextuality of causation in the empirical realm. The research findings also highlight a multiplicity of

societal expectations concerning environmental assessment’s contribution to sustainable development. It is concluded that the interplay

of non-rational variables (such as power, agency, experiences and expectations) necessitates the pursuit of a reflexive accommodation of

purposes, methods and context in environmental assessment practices.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Environmental assessment; Policy appraisal; Causation; Sustainable development; Governance

1. Introduction

Environmental assessment—a collective term for forms ofpolicy appraisal that address the probable environmental andsocial outcomes of human actions—is an intriguing policyphenomenon. Originating from vague aspirations espoused inthe US National Environmental Policy Act of 1969, the scaleand rapidity of its institutionalisation is remarkable. It is nowroutinely employed in the appraisal of trade negotiations,poverty reduction strategies, government policies, andregional and local development initiatives in more than 100nations and by various bilateral, multilateral and interna-tional organizations (Dalal-Clayton and Sadler, 2005; Eur-opean Commission, 2005; Russel and Jordan, 2006; Wood,2003). Yet despite the remarkable institutionalisation ofenvironmental assessment, empirical evidence indicates thatits effectiveness in promoting sustainable development—which is generally interpreted to be its primary purpose

e front matter r 2007 Elsevier Ltd. All rights reserved.

nvman.2007.06.005

ing author. Tel.: +441603 593797; fax: +44 1603 591327.

ess: [email protected] (M. Cashmore).

(Glasson et al., 2005)—is limited (Cashmore et al., 2004;Taylor, 1984; Wood and Jones, 1997). A further characteristicof the phenomenon is that environmental assessment is asingularly theoretically underdeveloped field of endeavour(Lawrence, 1997). Increasingly, the twin issues of limitedeffectiveness and theoretical impoverishment are perceived tobe linked (Cashmore, 2004; Dalkmann et al., 2004; Deelstraet al., 2003).Causation in environmental assessment (put simply, the

causal mechanisms and outcomes through which itcontributes to sustainable development) is a central focusof contemporary debate upon theory and effectiveness,albeit often unwittingly. The causal model that hasunderpinned mainstream practices since its inceptiondeveloped from the modernist notion of ‘speaking truthto power’. Environmental assessment was envisaged as atool that promoted sustainable development by generatingscientific data which apolitical stakeholders (usually publicadministrators) used, alongside other information, to makerational design and approval decisions (Nilsson andDalkmann, 2001). This was to be achieved predominantly

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481234

through an analysis of the environmental consequences ofa proposed action and the passive communication of thefindings (Bartlett and Kurian, 1999; Department ofEnvironment Transport and the Regions, 1999; Glassonet al., 2005; McDonald and Brown, 1995; Weston, 2000).The conventional model of causation in environmentalassessment can thus be represented as:

Underlyingcausalstructure

Causal mechanism

Outcome

Modernistsensibility}

Analysis ofconsequences andpublication of results

-

Informedconsent anddesign decisions

A variety of variables that operate contingently upon thecausal mechanisms, thereby influencing the occurrence andcharacter of the outcomes, can be added to this simplemodel. These include such factors as: the consideration of areasonable range of feasible alternatives; the use ofaccepted scientific methods; value-neutrality; and, thetimeliness of the results (Lee et al., 1999; Morgan, 1998;Sadler, 1996; Steinemann, 2001).

The conventional model of causation has proven to haveenduring appeal, particularly amongst policy makerssearching for civil legitimacy in their actions (Hills, 2005;Owens et al., 2004). Nonetheless, the gradual accumulationof research data, practical experience and alternativetheories, in combination with the expansion of theevidence-based policy agenda, has focused scholarly atten-tion on its limitations. Significantly, a longstanding faith inthe capacity of comprehensive and rigorous proceduralimplementation to redress limitations in environmentalassessment’s influence has diminished in recent years withgrowing acceptance of the irresolubly political nature ofdecision-making (Beattie, 1995; Deelstra et al., 2003;Flyvbjerg et al., 2003; Pope et al., 2005). This has resultedin proposals for more politically astute conceptions ofcausation: from action-forcing initiatives (e.g. the intro-duction of a substantive mandate for environmentalassessment or greater integration with decision processes)to new modes of operation (e.g. the employment of anassessment framework that promotes and facilitatesadversarial deliberations) (Dalkmann et al., 2004; Flyvb-jerg, 1998; Jay et al., 2007).

The conventional causal model has also been criticizedfor equating environmental assessment’s contribution tosustainable development solely to informing consent anddesign decisions (Cashmore et al., 2004). It has long beenpostulated that environmental assessment might promoteadditional outcomes, such as institutional reform andmultiple forms of learning, or that it could play a morecreative role in policy design (Bartlett and Kurian, 1999;Caldwell, 1991; McDonald and Brown, 1995). It has alsobeen suggested that environmental assessment providesimportant apertures for societal debate on the meaning ofsustainable development (Owens and Cowell, 2002), a

significant outcome given that this is an inherently vaguemaxim with little consensus existing on its practicalramifications. Yet with few exceptions, there has been aninexplicable lack of empirical investigation of these andother ‘non-conventional’ theories.Not all criticisms of the conventional causal model relate

to inherent theoretical limitations. Clearly, the nature ofand links between science, society and technology havechanged considerably since environmental assessment’sinception, with important theoretical and practical im-plications. To name but a few of these changes, centralizedgovernment is increasingly being replaced by multilevelgovernance systems, with an emphasis on public empower-ment and devolved autonomy (Pratchett, 2004; Schout andJordan, 2005); there has been a shift in scientific philosophyaway from a realist view of knowledge to a constructivistone (Jasanoff, 2004a); and, society’s faith in technology haseroded as their awareness of uncertainty and risk hasincreased (Beck, 1998). Indeed, the variety of factorscontributing to contemporary debate on causation partlyreflects the centrality of causal considerations to environ-mental assessment theory; it is axiomatic that its contribu-tion to sustainable development cannot be articulatedadequately independently of causal theory. This makes theinvestigation of causation a complex, and invariably acontested, endeavour.This article presents the findings of research designed to

advance theory on the role of environmental assessment inpromoting sustainable development and how such contribu-tions can be achieved; hereafter, this is referred to simply astheory on its role and functioning. This is undertakenthrough an empirical investigation of causation in purpose-fully selected contexts. The aim of the research was toinform debate on causation by investigating, if as previousresearch indicates the conventional causal model is empiri-cally flawed, how is environmental assessment operating inpractice and what are the theoretical implications. Thisapproach is valuable, not just due to the paucity of rigorousempirical research undertaken specifically within the contextof environmental assessment, but because it facilitates thesimultaneous investigation of contested issues in a mannerthat respects their interdependencies. The development of arich understanding of environmental assessment’s contex-tual operation consequently has the potential to generatepowerful, grounded theoretical insights.The analysis in this article commences, after the

methodology has been introduced, with a detailed examina-tion of causation in the contexts investigated. Based on theseobservations, the theoretical implications are explored in thediscussion section, whilst an integrative summation of theprincipal findings is presented in the conclusions.

2. Methodology

The research methodology was based upon a criticalrealist philosophy of science. This is important to statebecause critical realism involves a distinctive view of

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–1248 1235

causation and a stratified ontology. Critical realism rejectsthe standard Humean ‘successionist’ view of causation onthe basis that it confuses prediction with explanation (Keatand Urry, 1982). Explanation is not believed to involveidentifying patterns of regularities because what causessomething to happen has nothing to do with the number oftimes it is observed (Sayer, 2000). Instead, explanationdepends upon identifying the underlying causal mechan-isms (which are typically not open to observation) thatinitiate an event (which may or may not be observed in theempirical realm), how they work and under what condi-tions they are activated (the empirical) (Bhaskar, 1978).The critical realist model of causation is presentedgraphically in Fig. 1.

The research focused on Verstehen (understanding),rather than statistical generalisations. A case study researchdesign (see Yin, 1994) was, therefore, employed to achievethe research goals and a judgement-based sampling protocolwas developed to actively select cases thought to possess ahigh capacity to contribute to the development of Verstehen.The criteria employed in the sampling protocol were selectedto limit the influence of recognized technical contingencies,such as evident analytical bias, inadequate resourcing, andpoor study design. They thus focused partly on suchvariables as the nature of the development (environmentalassessments of larger and more controversial developmentstend to be better resourced (Cashmore et al., 2002)) and thequality of environmental assessment documentation. Thefocus on well-resourced environmental assessment was alsointended to facilitate the identification of cases with thepotential to illustrate a variety of causal mechanisms.Practical considerations—particularly the willingness ofstakeholders to participate, which was partly related towhether a consent decision had been taken and thepossibility of legal challenges—were also addressed in thesampling protocol. The cases selected for investigation areoutlined in Table 1; they are referred to simply as minesstabilization, offshore wind farm, and land remediation inorder to protect the identity of research participants. For thesake of brevity, further contextual information on individualcases is provided only as and when required in analysing theresearch results.

X

S

p1, p2, p3

l1, l2, l3

Object X,havingstructure S

Necessarilypossessing causalpowers (p) andliabilities (l)

= Necessary relation

Fig. 1. The critical realist conception

A comprehensive contextual understanding of causationwas developed for each case using two interlinked methods.Firstly, content analysis (see Silverman, 2001) of environ-mental assessment related documentation was undertaken.This provided a detailed insight into the formal portrayalof the role and functioning of the environmental assess-ments. Secondly, in-depth qualitative interviews (seeChirban, 1996; McCracken, 1988) were undertaken with29 stakeholders. During the interviews, the issue ofcausation was addressed largely indirectly, through discus-sions on stakeholders’ involvement with the projects andenvironmental assessments, their expectations of environ-mental assessment, and their perspectives on its effective-ness. The use of a semi-structured interview protocolmeant, however, that the precise range and nature ofinterview questions was interview-specific; the questionsasked depended on such factors as the findings of thecontent analysis, comments made by other stakeholders,and comments made during the interview (Arksey andKnight, 1999; Robson, 2002).The interviewees were selected to encompass representa-

tives from the principal categories of environmentalassessment stakeholders (Glasson et al., 2005), as follows:developers and their facilitators (n ¼ 7); decision-makers(n ¼ 7); statutory consultees (n ¼ 9); and, non-statutoryconsultees (n ¼ 6). Background data on the worldviews,scientific philosophies and interpretations of environmentalassessment of each interviewee were collected prior to theinterviews using a self-completion questionnaire. Whereinterview or questionnaire data are used in the article, theabbreviations MS, OW and LR, followed by a uniquereference number, are used to refer to research participantsinvolved with the mines stabilization, offshore wind farmand land remediation case, respectively.The data collected were used to develop causal models,

based on critical realist principles. Quasi-quantitative datafrom the questionnaires and content analyses wereanalysed statistically and/or graphically. The generalanalytical framework of Miles and Huberman (1994) wasemployed to structure the analysis of qualitative dataobtained from the interviews. This analytical frameworkconsists of three cyclical elements:

c1

c2

c3

e1

e2

e3

Underspecificconditions(c)

Willproduceoutcomeex

= Contingent relation

of causation. Source: Sayer, 1984.

ARTICLE IN PRESS

Table 1

Overview of the case studies

Mines stabilization Offshore wind farm Land remediation

Developer Public sector Private sector Public sector

Development

characteristics

Stabilization of between 80% and 90% of an

estimated 400,000m2 mine void, primarily

using foamed-concrete

Thirty wind turbines located approximately

7 km from land. Maximum generation

capacity of 108MW

Excavation of waste from the site of a former

chemical weapons establishment

Infrastructure for connection of the wind

farm to the electricity distribution system

Construction of a landfill site (maximum

capacity of 50,000m3)

Construction of a temporary vapour

containment structure and waste transfer

station

Notable

environmental

constraints

Development falls within a UNESCOa World

Heritage Site and affects a candidate Special

Area of Conservation

In close proximity to a candidate Special

Area of Conservation, a Special Protection

Area, and a Special Marine Area

Existing waste disposal sites notified as

‘special sites’ under Part IIA of the

Environmental Protection Act 1990

Socio-political

context

Affected community (over 1000 people) live

above unstable shallow mine workings. The

problem has been known for some years, and

the community feels let down by the lack of

action. A previous proposal for stabilization,

however, was abandoned due to public

opposition. The environmental assessment

conducted for this former proposal lacked

meaningful stakeholder participation

Essentially a novel form of development in

the UK at the time it was proposed and much

uncertainty existed about potential

environmental impacts and construction

risks. Main consent decisions were made by

national government departments through a

relatively opaque process. Development of

offshore wind energy has high-level political

support

Deep-seated community resentment

concerning the land-use, and anger about the

activities carried out at the site and the

secrecy surrounding these activities. Local

community believes human health and

fisheries have been adversely affected. High

profile issue, with site visits by high-ranking

politicians

Approximate

development

budget

£155 millionb More than £100 millionc £15–£20 milliond

Other

comments

The environmental assessment was

commended for featuring a high level of

public involvement using a wide variety of

methodse

Essentially a novel form of development in

the UK context. Deemed to be the highest

quality documentation produced under the

first round of offshore wind farm

developmentsf

Environmental assessment methodology

represents a good quality example of a

traditional technical approach,f combined

with high levels of stakeholder involvement

Highly unusual and high profile development

aUnited Nations Educational, Scientific and Cultural Organisation.bMS#8.cProject website, December 2006.dLR#2 and a confidential source.eConfidential report produced by the Institute of Environmental Management and Assessment.fBased on a review using the Institute of Environmental Management and Assessment review criteria (see Fuller, 1999).

M. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481236

data reduction: transformation of the data through theapplication of descriptive or inferential coding; � data structuring: through the use of display techniques,

such as matrices, flow diagrams and networks;

� data synthesis: a continuous process of extracting

meaning from the data.

Where possible, triangulation of the research results wasundertaken to ensure the analyses of causation wererobust. This was not always feasible since some outcomeswere inherently specific to individuals or particular interestgroups. There is, in consequence, a level of indeterminacyin the data. Nevertheless, this does not detract from theresearch findings; it is the search for the meaning behindwhat is reported in documents and interviews (i.e.necessary and contingent conditions) that is the primaryanalytical concern (Silverman, 2001).

3. Observations on causation in the empirical realm

3.1. The contribution of environmental assessment to design

and consent decisions

The characteristics of the documentation on environ-mental concerns produced for each of the case studydevelopments, despite evident limitations in the scientificrigour of the analyses, superficially indicates that environ-mental assessment practices were modelled upon conven-tional theoretical premises. Large numbers of impactpredictions were made; a variety of alternatives considered;and, numerous mitigation measures were proposed. How-ever, no incidences were identified wherein analyses ofpotential environmental effects played an instrumental rolein design and consent decisions.In relation to consent decisions, case officers and

decision-makers acknowledged that they did not have time

ARTICLE IN PRESS

1The land remediation project was exempt from planning regulations,

but it was nevertheless decided that planning consent would be applied

for, and that this would be the ultimate determinant of whether the

development was commissioned.2The consenting process for offshore wind farms is complex, and can

involve a variety of permissions for particular aspects of the development

from different decision-makers, at both the national and local levels.

M. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–1248 1237

to read, or, in some instances, the expertise to understand,all of the environmental assessment documentation, in thecases under investigation, and more generally (MS#2,OW#5, LR#1, #8). Instead, they relied extensively uponcomments made by statutory consultees and advisorybodies, a finding also reported by Wood and Jones (1997).Yet passive information provision does not appear to haveconsistently or markedly influenced the comments made bystatutory consultees or advisory bodies; it was repeatedlystated that these organizations would have raised the samegeneral concerns irrespective of whether an environmentalassessment had been undertaken (e.g. MS#2, OW#12).Local politicians, where involved as decision-makers,emphasized the importance of public opinion, rather thantechnical information. As one local councillor stated: ‘‘Ican read it [environmental assessment documentation] andI can make my own assessment. That’s fine. But if it comesout that this is absolutely fine for the community, but theydon’t want it or don’t like it, that’s what I’m there for’’(MS#6). Furthermore, comparative assessments invariablywere not instrumental in design decisions because, asreported elsewhere (e.g. Shepherd and Bowler, 1997;Taylor, 1984), choices were made before the environmentalassessment had been undertaken (e.g. mines stabilizationand land remediation cases) or after it was completed (e.g.land remediation and offshore wind farm cases) (MS#1,OW#1, #12, LS#1, #10).

Whilst there was no evidence of an instrumental role forpassive information provision, environmental assessmentwas evidently material to decision-making, for it wasactively recruited in all of the case studies, in essentially anadvocacy manner, to achieve the proponents’ immediategoal of gaining development consent. This was freelyacknowledged by one environmental assessment projectmanager, who when asked about their interpretation of thepurpose of environmental assessment replied, ‘‘[t]he mainpurpose is to get [development] consent’’ (OW#1). Never-theless, the predominant form of advocacy was not flagrantanalytical bias, although there was evidence in each of thecases of post hoc rationalization of decisions. Advocacyconsisted, instead, primarily of the employment of con-ciliatory mechanisms intended to engender support for thedevelopment—or, at least, minimize opposition—amongstinfluential stakeholders.

The selection of what conciliatory mechanisms toemploy was influenced by the effectiveness of differentmethods at engendering support amongst stakeholders.This, in turn, reflected the worldviews and priorities(institutional or otherwise) underpinning stakeholders’agendas. Dialogical mechanisms (e.g. working groupdeliberations, an information centre, and public meetings)played a more prominent role than analytical mechanismsin initiatives to gain the trust and support of localcommunities. This may reflect their familiarity with theuse of dialogical methods to participate in informationdissemination and evaluation. Statutory consultees’ expec-tations of environmental assessment, in contrast, invariably

reflected Enlightenment ideals. Consequently, reasonablyrigorous scientific analyses were required to engendersupport amongst these stakeholders, seemingly irrespectiveof the use made of these data in consent decisions.Provision for impact mitigation was also instrumental inengendering the support of statutory consultees in certaininstances (MS#1, #2, #8, LR#1).The agenda of the statutory consultees usually reflected

practical interpretations of their institutional mandate toprovide advice and conserve or protect certain resources,such as landscapes, species or human health. Yet theiragenda was also influenced by power. For example, EnglishNature were said to have limited interest in ecologicalimpacts arising from wind farms sited outside of protectedareas due to their legal remit and powers (OW#4). Incontrast, impacts upon a site and species protected underEuropean legislation in the mines stabilization casecontributed to English Nature adopting an approach onestakeholder described as, ‘‘so intransigent’’ (MS#8). Theconciliatory mechanism employed to overcome EnglishNature’s reservations in this instance involved a commit-ment by the developer to implement an £11 million habitatrecreation scheme (MS#8). The power of the public andnon-statutory groups, on the other hand, was more contextspecific, depending to a large extent on mobilisation ofpolitical interest and/or influence.The ultimate determinant of the characteristics of the

environmental assessment, within broad legislative andgood practice constraints, was the agency of stake-holders—that is, the ability of different stakeholders toinfluence consent decisions—for this defined conciliatoryrequirements. Stakeholder agency was mainly defined bythe rules and norms of the decision-making process underwhich individual cases were consented. In the minesstabilization and land remediation cases, which wereprimarily consented under the land-use planning system1,influential stakeholders (i.e. those with power and agency)included statutory consultees, case officers, the public, localpoliticians, and decision-makers. The conciliatory mechan-isms employed in these cases, therefore, had to address thevarying demands and expectations of a broad range ofstakeholders (see Fig. 2). Nonetheless, the influence ofother variables was apparent. In the land remediation case,for example, the attitude of a local politician wasinstrumental in bestowing agency upon comparativelypowerless stakeholders.Influential stakeholders were restricted in the offshore

wind farm case, which was consented primarily outside theremit of the land-use planning system,2 to statutoryconsultees, advisory bodies, and decision-makers. Unlike

ARTICLE IN PRESS

Acceptable

assessment

Acceptable design

and mitigation

Acceptable design

and mitigation

Acceptable

assessment

Conformance with

policy framework

Reassurance

Public support

Political supportSupport of the

planning committee

Development

consent

Support of case

officer

Formulation of

planning conditions

Support of statutory

consulteesAcceptable design

Mitigation proposed; commitment

to mitigation; involvement in the

environmental assessment;

analytical rigour; statutory

consultees' analyses of impacts;

necessity of projects.

Participation; openness; trust;

preferred alternative;

comprehensiveness; willingness

to adjust design; perception of

influence.

Example

contingent

conditionsKey stage

Key:

Fig. 2. Conciliation of stakeholders in the mines stabilization and land remediation cases.

M. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481238

the other two developments, there was limited representa-tion of the affected communities (directly or by electedofficials), for the most important consent decisions weremade by national Government departments (i.e. theDepartment of Trade and Industry and the Departmentfor the Environment, Food and Rural Affairs). The publicthus could not exert a political influence on decision-making as readily as in the case of developments consented‘locally’ under the planning system. The institutionaldecision-making context for offshore wind farms further

constrained the agency of the public due to its opacityand remoteness from the potentially affected communities.In consequence, the accountability of decision makerswas low.Whilst power, agency and agenda primarily defined the

characteristics of environmental assessment processes(within legislative and good practice constraints), addi-tional contextual variables exerted a significant influenceon the effectiveness of different conciliatory mechanismsthrough their operation as contingent conditions. For

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–1248 1239

example, the choice of engineering method for stabilisingthe mines was highly influential in overcoming localcommunity opposition to the development, despite theassessment of alternatives constituting post hoc decisionrationalization.3 The effectiveness of the environmentalassessment at engendering support for the developmentwas not a product of the technical quality of comparativeassessments, but arose because the local communityperceived the selected option to be preferential to the‘‘nightmare scenario’’ (MS#5) of a previous planningapplication (MS#3, #6). Linked to this interpretation ofeffectiveness was a belief that the purpose of environmentalassessment was, ‘‘not to make things worse’’ (MS#3).Similarly, public distrust of the developer was a significantissue in the land remediation case due to deep-seatedresentment of the military presence and the history ofsecrecy surrounding their operations in the region. Estab-lishing trust—through the employment of inclusive, trans-parent and responsive stakeholder involvement methods—was thus contingent to engendering support for the projectamongst local and regional communities.

Although there was no substantive evidence that thepassive provision of scientific analyses instrumentallyinformed consent and design decisions, it did operatecausally. The main contribution to engendering supportmade by this causal mechanism resulted from thereassurance it provided to stakeholders by creating aperception of due diligence; it illustrated to the public thattheir concerns had been considered in detail and demon-strated to statutory consultees that the proponent under-stood the potential impacts and the need for impactmitigation (MS#3, #6; OW#12). It thereby helped tolegitimize decisions to grant development consent (Holder,2004; Novek, 1995).

3.2. Additional transformative potentialities

The research results provide empirical evidence that notonly can conventional theory on causation be criticized forneglecting the institutional and political character ofdecision-making, but in focusing almost exclusively onconsent and design decisions, it has misrepresented thebreadth of environmental assessment’s transformativepotentialities. The empirical investigation of causationidentified a series of outcomes that environmental assess-ment contributed to in addition to its influence uponconsent and design decisions. This does not mean that aseries of novel occurrences were recorded in this research.Rather, certain phenomena conventionally consideredprocedural means-to-an-end (e.g. stakeholder participationin environmental decision-making and the public avail-ability of information) were deemed in this research to

3The environmental assessment and design project manager acknowl-

edged that the preferred option was selected based on expert judgement at

the stage of tendering for the contract and that the subsequent analyses

were necessary in order to justify the choice to stakeholders.

constitute outcomes that had the potential to make apositive contribution to sustainable development andmodern principles of environmental governance. Otheroutcomes pertained to phenomena that are unvalidated orunder-validated within the context of environmentalassessment (e.g. institutional reform and forms of learn-ing). The observed outcomes in their totality weredelineated into a model of the transformative potentialitiesof environmental assessment consisting of four maincategories: developmental outcomes; learning outcomes;governance outcomes; and, attitudinal and value changes.Examples of the findings on which this model is based areincluded in Table 2.No evidence was observed to substantiate the long-

standing belief that environmental assessment has thepotential to transform governmental bureaucracies byaltering the rules and norms of decision-making (Bartlett,1986; Caldwell, 1991). It should be noted that theidentification of institutional reform can be problematicas it may be a long-term and subtle phenomenon,particularly within an institutional context such as that ofEngland, where experience is divested in a large number ofgeographically dispersed individuals. Yet the observationof change in an non-governmental organisation (NGO)(see Table 2) indicates that institutional reform is aplausible theory, and one which has application beyondthe conventional concern of policy analysis, the govern-mental bureaucracy. The small size of the NGO meant thatchanges were transmitted through the organization com-paratively rapidly and that the causes were readilyidentifiable.The development of critical realist models of causation

for each of the observed outcomes, through the retro-duction of causal mechanisms and contingent conditionsfrom the interview and content analysis data, illustrated thecomplexity of causal relations in the empirical realm. Anindication of this complexity is provided in Fig. 3, whichshows a simplified4 model of causation for one particularoutcome: the occurrence of institutional reform in anNGO. This individual model is used as an example tohighlight features of complexity in causation common to avariety of transformative potentialities.Firstly, outcomes were invariably the product of a chain

of linked, and often interacting, causal mechanisms.Institutional reform in an NGO occurred as an immediateconsequence of attitudinal and value changes in its staff,and the consequential effects this had upon organizationalculture and decision-making. Ultimately, however, institu-tional reform was a product of a multiplicity of mechan-isms that influenced staff attitudes and values. In someinstances, causal mechanisms operated singularly: aspreviously stated, ultimately it was the causal liabilities ofattitudinal and value changes that produced institutional

4Despite outlining causation in more detail than the data in Table 2,

Fig. 3 is a substantially simplified model of causation, particularly in

relation to the operation of contingent conditions.

ARTICLE IN PRESS

Table 2

The transformative potentialities of environmental assessment

Category of

transformative

potentialities

Sub-category Empirical example(s) (case(s) drawn from) Outline of causation

Learning

outcomes

Social learning Enhanced community awareness about the existence of

internationally significant bat populations and the

cultural value of the mines (MS)

Knowledge about the existing environment and its values

acquired through information in documentation, formal

participatory processes, and informal dialogue

Improved understanding of other stakeholders’ beliefs

and values (LR)

Stakeholders’ understanding of each others’ attitudes

develops based on information in documentation, formal

participatory processes, and informal dialogue

Development of communication skills (MS) Community involvement in, and reflection upon, formal

participatory processes and informal dialogue advances

understanding of the effectiveness of different

communication methods and strategies

Technical learning Better understanding of the ecology of the area resulted

in upgrading the conservation status of a bird

population (OW)

Knowledge acquired through the collection and

interpretation of baseline data

Enhanced public understanding of EU governance

procedures (MS)

Detailed knowledge of EU provisions for environmental

governance gained through independent research into

legislation and communication with EU politicians and

bureaucrats. This was undertaken in order to check the

validity of the developer’s actions

Governance

outcomes

Availability of

information

Publication of extensive information on baseline

environment, design options, and mitigation (all)

Legislative provisions concerning the environmental

assessment report combined with the need to appease

influential stakeholders and principles of good practice

results in the production of extensive documentation

Stakeholder

participation

(amount, form and

timing)

Extensive involvement of statutory consultees (LR) Early, sustained and in-depth involvement of statutory

consultees in order to ensure their concerns are addressed

and thus engender their support

Transparency and

accountability in

decision-making

Documented justification of design choices publicly

available (MS & LR)

Good practice expectations concerning the assessment of

alternatives and the political sensitivity of design

decisions necessitate a thorough justification of choices

Local autonomy Elements of delegated power in decision-making (MS &

LR)

The need to appease influential stakeholders and an

influential individual’s belief in accountability result in

the employment of involvement mechanisms that devolve

a modest amount of responsibility for decision-making

Local democracy More inclusive decision-making (LR) Broad representation of stakeholders (particularly

bureaucrats) in the involvement methods in order to

ensure their concerns are addressed and thus engender

their support

Honing of debating skills (MS & LR) Stakeholder participation (formal and informal) and

subsequent reflection upon experiences provides

opportunities to practice and refine debating skills

Network

development

Linking a fragmented government bureaucracy (LR) The need to appease influential stakeholders results in

early, sustained and in-depth involvement of statutory

consultee. A network develops as inter-personal

relationship are built and information exchanged

Development

outcomes

Design choices Documented justification of design choices (MS & LR) Good practice expectations concerning the assessment of

alternatives and the political sensitivity of design

decisions necessitate a thorough justification of choices

Consent decisions Overcoming local community opposition (MS) Comprehensive analysis, participatory processes, choice

of infill material and provision for impact mitigation

cumulatively provide local community with reassurance,

directly and indirectly

Attitudinal and

value changes

Changing attitudes Greater community cohesion (MS) Local community dynamics are affected by social

learning and the spaces for interaction environmental

assessment generates. The community is also united in

their support for the preferred in-fill material

Political support for environmental assessment (MS) Environmental assessment is viewed by the electorate as

successful and community opposition to the development

is overcome

M. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481240

ARTICLE IN PRESS

Table 2 (continued )

Category of

transformative

potentialities

Sub-category Empirical example(s) (case(s) drawn from) Outline of causation

Changing values Cultural resource valued more highly (MS) Public values change as they learn about the cultural

resource and in light of its imminent destruction

Institutional reform Changes in methods of operation, recruitment and

organizational culture in an non-governmental

organization (OW)

Changes in the norms of governance are perceived to

influence the effectiveness of different strategies for

influencing decisions

M. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–1248 1241

reform. More frequently, however, causal mechanismsoperated synergistically: the existence of environmentalassessment legislation, requirements for stakeholder in-volvement, and the publication of an environmentalassessment report combined to enhance the profile ofenvironmental concerns in decision-making, symbolicallyor otherwise.

Secondly, outcomes influenced the operation of upstream, aswell as downstream, causal mechanisms through feedbackloops. This was particularly apparent in relation to attitudinaland value changes. Institutional reform, for example, wasobserved to be a self-reinforcing outcome, for the attitudinaland value changes to which environmental assessmentcontributed resulted in reforms that served to buttress thenew status quo (e.g. the recruitment of like-minded individuals).

Thirdly, in addition to the direct influence of the causalmechanisms illustrated in Fig. 3, others exerted an indirecteffect through their influence upon contingent conditions.For example, one way in which environmental assessmentcontributed to changes in governance norms was byenhancing opportunities for stakeholders to critique pro-posals by providing them with information on the devel-opment. Whether or not such opportunities were exploitedwas contingent, amongst other things, upon the interest ofstakeholders in the development. This, in turn, was affectedby causal mechanisms that influenced the profile of thedevelopment and of environmental issues in decision-making. The model in Fig. 3 thus omits, for the purposeof effective communication, further networks of causalmechanisms that influenced contingent conditions. Thedynamism of contingent conditions was an important causeof complexity in environmental assessment causation.

Finally, although conventional theory often implies(through omission) that environmental assessment operatesautonomously of extraneous contextual variables (Cash-more et al., 2004), this is a position which is not supportedby the empirical data. The influence upon causation offactors external to environmental assessment was perva-sive, for in practice no outcomes were truly independent ofextraneous contextual variables.5 In the case of institu-

5The pervasive influence of extraneous contextual variables is the reason

why environmental assessment is described as a contributory factor to the

observed outcomes. Reliable measurement of its contribution is a

methodologically challenging objective that this research was not designed

to address.

tional reform in the NGO, it was recognized thatenvironmental assessment was only one of multiple causalmechanisms promoting change (OW#2). Reform alsoreflected the changing worldviews of the organization’sstaff and their perception that the norms and rules ofdecision-making were altering. Indeed, it was probable thatinstitutional reform would have occurred in the absence ofenvironmental assessment (OW#2); it was merely one ofmultiple contributory factors.The influence of extraneous variables was also evident in

the two more participatory cases (mines stabilization andland remediation), for it was contextual variables, ratherthan environmental assessment ideology per se, that madethe greatest contribution to the decisions to employparticipatory methods. The high level of stakeholderinvolvement in these cases resulted from such contextualfactors as: historical failures to involve affected commu-nities and the dissatisfaction this generated (MS#6, #9;LR#2, #7), and imperatives to gain development consent incomparatively short time-frames (MS#8, LR#1). Environ-mental assessment essentially served as a convenientframework (conceptually, methodologically, and possiblyfinancially) around which to structure stakeholder involve-ment in these cases.The effectiveness of the participatory practices at

reassuring the local community were also pervasivelyinfluenced by contextual contingencies. In the case of themines stabilization case, for example, the need for thedevelopment was undisputed and it was the local commu-nity who would (eventually) benefit from the development;the preferred design option was overwhelmingly viewed aspreferential to a previous proposal; and, a communityassociation was instrumental in ameliorating local com-munity concerns. The community association’s favourabledisposition to the development, however, arguably re-flected the political interests of a majority of its electedBoard. It also reflected the fact that the communityassociation was funded by the developer and, latterly, usedthe developer’s newsletter to disseminate information tothe local community. This reputedly led to a perceptionthat the community association could not afford to beoverly critical of the developer (MS#9).The influence of contextual contingencies was also

apparent in relation to the potential for institutionalreform to occur in the governmental bureaucracy. It wasevident that contextual contingencies—including the loca-

ARTICLE IN PRESS

Fig.3.Causalmodel

forinstitutionalreform

inanon-governmentalorganization.

M. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481242

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–1248 1243

lisation of experiences in individuals, infrequent involve-ment with environmental assessment, high staff turnover,and low staff morale—represented powerful barriers toinstitutional reform.

4. Discussion

The research has evident methodological limitations thatconstrain the conclusions which can be drawn and theconfidence placed in them. The analysis was designed toproduce a richer contextual understanding of causation,rather than generalisable results; a limited number of cases,all from one jurisdiction and at the project level ofdecision-making, were studied; and, the results are based,in part, upon the research team’s (subjective) interpreta-tions of stakeholders’ (subjective) beliefs and (partial)experiences. Nevertheless, in combination with otherresearch, the observations on causation provide valuableempirical insights that have the potential to advance debateon key theoretical premises concerning the role andfunctioning of environmental assessment. The followingdiscussion focuses on the implications of three issueshighlighted by the empirical analysis: the failure to achieverationalist goals; the complexity of causation; and, theexistence and implications of multiple expectationsamongst stakeholders.

4.1. Advocacy not rationality

Environmental assessment, like various other appraisaltools (notably cost-benefit and life-cycle analysis), is basedon a normative theory of causation grounded in philoso-phies of rationality in science and decision-making. Thisresearch, like a burgeoning number of empirical studies(e.g. Flyvbjerg et al., 2003; Hills, 2005; Phillips, 2002),highlighted the practical inadequacies of such theoreticalpremises. No instances were observed wherein analyses ofenvironmental consequences were instrumental to decision-making in a rationalistic sense, nor were scientific practicestruly rational.

Failures to achieve rationalist goals are frequentlyattributed to implementation deficits (Owens et al., 2004),examples of which are pervasive in environmental assess-ment practices (Byron et al., 2000; Lee et al., 1999). Theresults of this research, however, provide evidence tosupport the belief that initiatives to improve scientificrationality—or at least technical rigour—will not necessa-rily result in rational decisions being taken (Amy, 1990;Boyle, 1998; Dalal-Clayton and Sadler, 2005; Deelstra etal., 2003); other contextual barriers appear more signifi-cant. There is also evidence in the analysis (e.g. in relationto the consideration of alternatives), and in the history ofscientific practices (Jasanoff and Wynne, 1998; Latour,1988; Wynne, 1984), that an emphasis on technical qualitymight not promote true scientific rationality, merely betterconcealment of its socio-political construction. This is notto suggest that the case study environmental assessment

documentation was faultless from a technical perspective,despite purposeful selection of cases to limit the influenceof technical contingencies. Nonetheless, in the casesexamined, environmental assessment was not intended tobe used in a neutral manner, but became co-opted in apolitical process of conciliation designed to engendersupport for the development amongst influential stake-holders.The use of environmental assessment in an advocacy

manner or the influence of power, agency and agenda arenot remarkable observations given empirical researchundertaken in the policy sciences (Flyvbjerg et al., 2003;Owens and Cowell, 2002; Sabatier, 1988) and anecdotalevidence concerning environmental assessment practices(Beattie, 1995; Caldwell, 1991; Malik and Bartlett, 1993).What is more noteworthy is the influence of these variableson the methodological conduct of environmental assess-ment. Power and agency determined which stakeholderswere influential (and hence needed to be involved in asubstantive manner), whilst the agenda of influentialstakeholders (incorporating legislative remit and preor-dained worldviews) affected the selection and design ofconciliatory mechanisms. Although undertaken withinbroad legislative limits and in accordance with notions ofgood practice, there was extensive perversion of methodo-logical practices by behavioural considerations. Thepoliticisation of science (de Bruijn and ten Heuvelhof,1999; Jasanoff, 2004b) was subtle, but pervasive.To what extent does this finding support Flyvbjerg’s

(1998) central contention that power defines rationality,and thereby what counts as reality? In the case of thisresearch, power, agency and agenda operated to dictatewhat knowledge was influential, and hence important,rather than what counted as knowledge per se; that is,they determined which rationalities and realities environ-mental assessment procedures needed to engage with and,in part, how. Power and agency were divested in a varietyof stakeholders pursuing different agendas in themines stabilization and land remediation cases. Conse-quently, it was necessary for a variety of rationalities to beaddressed through the environmental assessment process,and a range of conciliatory measures were employed.In the offshore wind farm case, in contrast, the compara-tive powerlessness of the public, in combination witha political drive to develop offshore wind farms (Depart-ment of Trade and Industry, 2003), resulted in engagementwith a more constrained range of rationalities; agencywas primarily limited to government institutions, whopursued modernist ideals. This finding supportsFlyvbjerg’s contention, but emphasizes that the distribu-tion of power within society can result in the co-existenceof multiple influential realities that it is necessary tosimultaneously confront. The results also illustrate theimportance of considering agency and agenda in parallelto power.Empirical limits to environmental policy appraisal tools

fulfilling rationalist goals, on their own, are not grounds

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481244

for abandoning the rationalist project. As has beenemphasized elsewhere (e.g. Forester, 1989; Simon, 1969),an empirical understanding of constraints can aid theformulation of measures to promote greater rationality. Itis also arguably the case that, despite considerable effort(e.g. Habermas, 1984, 1987), there is no single, unproble-matic alternative (Richardson, 2005). Nonetheless, thecontinued employment of policy appraisal tools based onabstract theories that are empirically naıve is a folly thatcannot be justified in a world of limited resources andpressing environmental problems, even in the absence of astraightforward alternative. As Wynne (1984, p. 316)forcefully argued more than 20 years previously: ‘‘Objec-tive examination of technical options in a supposedinstitutional vacuum is illusory and worse because it onlyconceals the particular institutional frameworks implicit inthe analysis, and thereby consigns policy (and analysis) tokindergarten immaturity’’. Given the apparent importanceof environmental assessment as a site of political engage-ment in the English context, it is also perhaps undesirable,as well as illusory, to promote ideals of rationality indecision-making.

4.2. The complexity of causation

Power, agency and agenda exerted potent influences oncausation, but a large number of other contextual variableswere observed to operate contingently in this research. Thecomplexity of causation was further accentuated by thevalidation of certain theories that environmental assess-ment has transformative potentialities over and above itscontribution to design and consent decisions. Numerousforms of learning were observed, and not merely in relationto dialogical interactions, upon which empirical research(Fitzpatrick, 2006; Fitzpatrick and Sinclair, 2003; Webleret al., 1995) has thus far focused; governance relatedprocedural by-products were deemed to constitute valuabletransformative potentialities in their own right; and,various attitudinal and limited value changes wererecorded. There were also multiple interlinkages betweencausal mechanisms; these often acted additively andsometimes synergistically, as the example of institutionalreform in an NGO illustrated. Additionally, causalmechanisms contributed indirectly to outcomes throughtheir influence on dynamic contingent conditions. Thesefactors combined to create a model of causation character-ized by complexity.

Observations on the complexity of causation have anumber of important implications for theory development,subsequent empirical research, and the theory-practicenexus. Firstly, although provisional, the results provideempirical support for reforming theory on the transforma-tive potentialities of environmental assessment, for theconventional causal model misrepresents the breadth ofways in which it can contribute to sustainable develop-ment. Although practice, to some extent, is already leadingtheory in this respect (e.g. European Commission, 2005),

reformation is important for, amongst other things, itwould legitimize and emphasize the importance of theadditional transformative potentialities. This would requiremethodological modifications to be made in order todesign approaches for the purposeful targeting of theadditional transformative potentialities.The methodological limitations outlined in the introduc-

tion to the discussion and the contextuality of causationhighlight the need for additional empirical work onenvironmental assessment’s transformative potentialities.However, the existence of direct and indirect interconnec-tions between causal mechanisms has important implica-tions for the design of such work. Foremost amongst theseis that, whilst reductionist research will be important indeveloping a detailed understanding of particular trans-formative potentialities, the individuation of theories—apernicious tendency within the literature (e.g. Bartlett andKurian, 1999)—will misrepresent causation in the empiri-cal realm. It is important, therefore, that future research isframed within the context of a strategy designed to producean holistic and integrated understanding of causation,despite the evident methodological and intellectual chal-lenges this poses. Furthermore, given the pervasiveinfluence of contextual contingencies on causal mechan-isms, greater attention must be given to the situated natureof environmental assessment in future research.Finally, the breadth of transformative potentialities and

the influence of contextual contingencies raises questionsabout expectations of theory in a number of respects.Environmental assessment, particularly at the project level,has typically been presented as an all-embracing approach,which can be applied according to a set heuristicirrespective of the nature of a development or of socio-political context (Petts, 1999a). This arguably is a reflectionof modernist beliefs in the universality of the scientificmethod and in rationality. The contextuality of causationbrings into doubt the adequacy of a universal heuristic andof unproblematic learning from the particular, for con-textual variables (including notions of science and govern-ance) are invariably historically situated and dynamic(Jasanoff, 2004a; Pope et al., 2005). Variations ingeopolitical realities also make it appropriate to placedifferentiated emphasis on individual transformative po-tentialities spatially and temporally.Linked to the issues of universality, is that of expecta-

tions of the theory-practice nexus. It is evident that theconventional form of theoretical proposition (e.g. all othervariables being constant, more x results in more y) is ofquestionable relevance in dynamic policy environmentswhere things are virtually never equal (Hills, 2005;O’Toole, 2004). Can environmental assessment theory thusreally be expected to explain and predict? It is anunavoidable (if somewhat uncomfortable) conclusion ofthis research that whilst theoretical horizons must bebroadened beyond the traditional concern with technicalissues, straightforward recommendations for proceduralpractices may not necessarily emerge. This is not to suggest

ARTICLE IN PRESS

6This is a simplification because analytical models involved the

participation of certain stakeholders (mainly statutory consultees) in

defining an acceptable scope and approach. The interactions, however,

were primarily limited to technical issues, whereas deliberation with local

communities covered a considerably broader range of issues.

M. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–1248 1245

that there are no transferable lessons in environmentalassessment; rather, it is argued that considerably morecaution must be taken in formulating procedural andmethodological generalisations than has hitherto been thecase in much research. The more modest goal of the theory-practice nexus, it is suggested, should be for theory toprovide a rich knowledge base that facilitates reflexivepractices in a given context (Lawrence, 2003; Richardson,2005).

4.3. Multiple expectations amongst stakeholders

In addition to highlighting the influence of behaviouralvariables, the use of environmental assessment in anadvocacy manner is interesting from a theoretical perspec-tive for it illuminated the varying expectations of stake-holders. Expectations differed in relation to the linkedissues of the purposes of environmental assessment, thelegitimate means by which these could be achieved, and thesubjects that needed to be addressed within assessments.This reflects the normative effect of epistemological beliefsbeyond methodological prescriptions (Kolakowski, 1972).It also highlights the exclusivity of the dominant environ-mental assessment paradigm (Emmelin, 1998; Petts, 2003),for rarely has effort been made to engage stakeholdersother than academics and practitioners in debates(explicitly or implicitly) concerning it’s societal purposes.In combination with limited attention to capacity devel-opment in the cases, through which a common under-standing of purposes might be fostered, it is not surprisingthat expectations about environmental assessment wereobserved to vary significantly in the empirical realm. Thus,Wynne (2002) bemoans what he terms the ‘‘hermeneuticsanitization’’ of appraisal through the ‘‘cultural reificationof risk’’ (pp. 470 and 468, respectively [original emphasis]);that is, the subversion of societal expectations based on animposed presumption that assessment and management ofrisks are the salient issues.

A prominent example of multiple expectations in thecase studies was the division between those stakeholderswho interpreted environmental assessment as a processprimarily for generating scientific-based understanding ofenvironmental consequences (primarily the statutory con-sultees) and those who saw it more as a democratic processfor addressing the interlinked development-related con-cerns of a diverse polity (primarily the public and theirelected representatives). The need to appease both ration-alities in cases where each group was influential (i.e. minesstabilization and land remediation) produced a palpablepolarization of environmental assessment procedures intoanalytic and deliberative models.

The juxtaposition of multiple expectations is notnecessarily problematic in itself; after all, the integrationof deliberation into analytical procedures is what is said todifferentiate environmental assessment from technocraticpolicy appraisal tools (Petts, 2003; Wood, 2003). Based onthe observations of causation, however, it was evident that

analytical and deliberative procedures operated largelyautonomously, essentially as discrete environmental assess-ments running concurrently.6 The imperative to appeasedivergent rationalities resulted in environmental assess-ment becoming, ‘‘all things to all people’’ (OW#12).Epistemological, political and practical factors may have

contributed to the largely autonomous operation ofanalytical and deliberative procedures (see also Levidov,1999; Owens et al., 2004; Petts, 2003), not least of which isthat the precise role of public involvement in environ-mental assessment remains ambiguous (Petts, 1999b).However, the failure to achieve in practice an integratedbroadening of the epistemological spectrum from science toknowledge also reflects inherent tensions created by theexistence of multiple expectations. These are likely to beparticularly significant in jurisdictions like England, whereenvironmental assessment has been firmly embedded with-in an established policy framework. Local communities didnot understand, or view as legitimate, boundaries betweendeliberations related to environmental and broader socio-political concerns. Furthermore, in the absence of otheropportunities for early and repeated public involvement, itwas inevitable that deliberations purportedly linked to theanalysis of environmental consequences became loci fordebate on a broader range of issues. Attempts to narrowlyfocus the agenda upon environmental concerns are unlikelyto prove successful in such circumstances. What might beviewed by some as dilution or even hijacking of theenvironmental agenda, could also be interpreted as acomponent of policy learning, through which the publicinfluence political agendas from the bottom-up (Owens andCowell, 2002).The conceptual challenges generated by these observa-

tions are considerable. In consequence, can environmentalassessment’s role be articulated independently of context?If so, which expectations are ‘correct’ and who chooses?Are multiple expectations epistemologically reconcilable?And can the assessment of environmental effects mean-ingfully be separated from broader policy issues?One solution proposed in the literature is to reassert the

environmental focus of this policy appraisal tool, perhapsthrough the introduction of a substantive mandate for it(Jay et al., 2007; Morrison-Saunders and Fischer, 2006).Given the research findings outlined in this article, Hajerand Wagenaar’s (2003) recommended approach of thepursuit of a better ‘fit’ between knowledge and politicalimplementation contexts appears a more appropriate wayforward. Societal expectations (essentially a contextualconsideration) are clearly a fundamental component of this‘fit’,7 so too is institutional context; hence, (re)focusing onenvironmental issues will only be appropriate where

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481246

institutional circumstances ensure legitimate societal ex-pectations concerning development and governance can bemet. The issue of environmental assessment’s role andfunctioning can thus be re-interpreted as a context-specificquestion of the degree to which existing arrangementshamper finding effective solutions to the developmentissues stakeholders want resolved (Hajer and Wagenaar,2003). A broadening of environmental assessment tosustainability appraisal is arguably inherent in such avision in the English context, unless major institutionalreforms were to occur.

5. Conclusions

Environmental assessment in the contexts investigateddid not operate in accordance with the conventional modelof causation. This was to be expected given the welter ofevidence that science and decision-making rarely conformto rationalist ideals. The empirical critique is important,nonetheless, for far from being a ‘straw man’, rationalistideals have far-reaching normative implications for theoryon environmental assessment’s role and functioning. Thenotion of scientific analyses objectively informing designand approval decisions reflects interpretations of the policyproblem and methodological solution that are blinkered byenlightenment ideals. This research has provided empiricalevidence that environmental assessment has the potentialto contribute to sustainable development through a multi-plicity of additional, and often interlinked, transformativepotentialities. Theory should be reformed to reflect this.

The research also highlights the contextuality ofenvironmental assessment’s role and functioning. In termsof the advancement of theory, this is important for at leastthree reasons. Firstly, the range of transformative potenti-alities observed indicates that environmental assessmenthas the potential to mobilize various interpretations ofsustainable development (see also Owens and Cowell,2002). This means environmental assessment can purpose-fully adapt to geopolitical contextualities, such as govern-ance and scientific traditions, and socio-economicnecessities. Secondly, maintaining civil legitimacy inenvironmental assessment practices requires that consid-eration is given to institutional context and stakeholderexpectations. There is, in consequence, an imperative topursue a better ‘fit’ between purposes, practices andcontext (Hajer and Wagenaar, 2003). Finally, the apparentcase-specific contextuality of causation raises significantchallenges in relation to our understanding of environ-mental assessment’s functioning and to traditional expec-tations of theory. Can the historical predilection fordecontextualized ‘best practice’ be maintained when out-comes are so pervasively influenced by dynamic contextualcontingencies? Collectively, these observations imply that

7This is not to suggest that management of expectations will not be

required in some instances, for unrealistic expectations are arguably a

cause of disillusionment with appraisal tools (Owens and Cowell, 2002).

the concept of a universal heuristic is no longer tenable or,in fact, desirable. The more modest expectation of theoryshould be for it to provide a knowledge base that can aidreflexive contextual design of environmental assessmentsystems and practices (Jasanoff, 2004a).This research has provided valuable insights into the

development of an alternative causal theory, but consider-able further research and reflection are required. Validationof outcomes in additional contexts, investigation of theinfluence and operation of behavioural contingent condi-tions, and understanding the relationship between trans-formative potentialities and sustainable development arebut a few of the topics requiring additional study. Further,the interfaces between causal and other aspects ofenvironmental assessment theory, combined with the factthat rationalist ideals are deeply embedded in manybureaucracies, is likely to make progress slow andcumbersome. As Owens and Cowell (2001) pragmaticallyconclude in policy appraisal, and arguably science moregenerally, ‘‘we struggle, experiment, learn and (for the mostpart) edge forward’’. Environmental assessment practicesprovide valuable opportunities to ‘struggle’ and hopefullylearn not only about human actions and their impacts, butalso in relation to sustainable development, governance,and the policy-science interface. Such learning mayconceivably prove to be of equal significance as environ-mental assessment’s more direct contributions to trans-forming development.

References

Amy, D.J., 1990. Decision techniques for environmental policy: a critique.

In: Paehlke, R., Torgerson, D. (Eds.), Managing Leviathan: Environ-

mental Politics and the Administrative State. Belhaven, London, pp.

59–79.

Arksey, H., Knight, P., 1999. Interviewing for Social Scientists: An

Introductory Resource with Examples. Sage, London.

Bartlett, R.V., 1986. Rationality and the logic of the National Environ-

mental Policy Act. The Environmental Professional 8, 105–111.

Bartlett, R.V., Kurian, P.A., 1999. The theory of environmental impact

assessment: implicit models of policy making. Policy and Politics 27,

415–434.

Beattie, R.B., 1995. Everything you already know about EIA (but

don’t often admit). Environmental Impact Assessment Review 15,

109–114.

Beck, U., 1998. Democracy Without Enemies. Polity Press, Cambridge.

Bhaskar, R., 1978. A Realist Theory of Science. Harvester, Hassocks.

Boyle, J., 1998. Cultural influences on implementing environmental impact

assessment: insights from Thailand, Indonesia and Malaysia. Environ-

mental Impact Assessment Review 18, 95–116.

Byron, H.J., Treweek, J.R., Sheate, W.R., Thompson, S., 2000. Road

developments in the UK: an analysis of ecological assessment in

environmental impact statements produced between 1993 and 1997.

Journal of Environmental Planning and Management 43, 71–97.

Caldwell, L., 1991. Analysis-assessment-decision: the anatomy of rational

policymaking. Impact Assessment Bulletin 9, 81–92.

Cashmore, M., 2004. The role of science in environmental impact

assessment: process and procedures versus purpose in the development

of theory. Environmental Impact Assessment Review 24, 403–426.

Cashmore, M., Christophilopoulos, E., Cobb, D., 2002. The quality

of environmental impact statements in Thessaloniki, Greece.

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–1248 1247

Journal of Environmental Assessment Policy and Management 4,

371–395.

Cashmore, M., Gwilliam, R., Morgan, R., Cobb, D., Bond, A., 2004. The

interminable issue of effectiveness: substantive purposes, outcomes and

research challenges in the advancement of EIA theory. Impact

Assessment and Project Appraisal 22, 295–310.

Chirban, J.T., 1996. Interviewing in Depth: The Interactive-Relational

Approach. Sage, London.

Dalal-Clayton, B., Sadler, B., 2005. Strategic Environmental Assessment:

A Sourcebook and Reference Guide to International Experience.

Earthscan, London.

Dalkmann, H., Jiliberto, R., Bongardt, D., 2004. Analytical strategic

environmental assessment (ANSEA) developing a new approach to

SEA. Environmental Impact Assessment Review 24, 385–402.

de Bruijn, J.A., ten Heuvelhof, E.F., 1999. Scientific expertise in complex

decision-making processes. Science and Public Policy 26, 179–184.

Deelstra, Y., Nooteboom, S.G., Kohlmann, H.R., van den Berg, J.,

Innanen, S., 2003. Using knowledge for decision-making purposes in

the context of large projects in the Netherlands. Environmental Impact

Assessment Review 23, 517–541.

Department of Environment Transport and the Regions, 1999. Environ-

mental impact assessment. Circular 02/99. Department of the

Environment, Transport and the Regions, London.

Department of Trade and Industry, 2003. Energy White Paper. Our

Energy Future—Creating a Low Carbon Economy. Stationery Office,

London.

Emmelin, L., 1998. Evaluating environmental impact assessment sys-

tems—Part 1: theoretical and methodological considerations. Scandi-

navian Housing and Planning Research 15, 129–148.

European Commission, 2005. Impact Assessment Guidelines. SEC(2005)

791. European Union, Brussels.

Fitzpatrick, P., 2006. In it together: organizational learning through

participation in environmental assessment. Journal of Environmental

Assessment Policy and Management 8, 157–182.

Fitzpatrick, P., Sinclair, A.J., 2003. Learning through public involvement

in environmental assessment hearings. Journal of Environmental

Management 67, 161–174.

Flyvbjerg, B., 1998. Rationality and Power: Democracy in Practice.

University of Chicago Press, London.

Flyvbjerg, B., Bruzelius, N., Rothengatter, W., 2003. Megaprojects and

Risk: An Anatomy of Ambition. Cambridge University Press, Cam-

bridge.

Forester, P., 1989. Planning in the Face of Power. University of California

Press, Berkeley, CA.

Fuller, K., 1999. Quality and quality control in environmental impact

assessment. In: Petts, J. (Ed.), Handbook of Environmental Impact

Assessment. Volume 2: Environmental Impact Assessment in Practice:

Impact and Limitations. Blackwell Science, Oxford, pp. 55–82.

Glasson, J., Therivel, R., Chadwick, A., 2005. Introduction to Environ-

mental Impact Assessment: Principles and Procedures, Process,

Practice and Prospects. Spon Press, London.

Habermas, J., 1984. The Theory of Communication Action. Volume 1.

Reason and the Rationalization of Society. Beacon, Boston.

Habermas, J., 1987. The Theory of Communication Action. Volume 2.

Lifeworld and System. Beacon, Boston.

Hajer, M., Wagenaar, H., 2003. Introduction. In: Hajer, M., Wagenaar, H.

(Eds.), Deliberative Policy Analysis. Understanding Governance in

the Networked Society. Cambridge University Press, Cambridge,

pp. 1–30.

Hills, S., 2005. Falling through the cracks: limits to an instrumental

rational role for environmental information in planning. Environment

and Planning A 37, 1263–1276.

Holder, J., 2004. Environmental Assessment: The Regulation of Decision

Making. Oxford University Press, Oxford.

Jasanoff, S., 2004a. Design on Nature: Science and Democracy in Europe

and the United States. Princeton University Press, Princeton.

Jasanoff, S., 2004b. Science and citizenship: a new synergy. Science and

Public Policy 21, 90–94.

Jasanoff, S., Wynne, B., 1998. Science and decisionmaking. In: Raynor, S.,

Malone, E. (Eds.), Human Choice and Climate Change: The Societal

Framework. Battelle Press, Columbus, OH, pp. 1–87.

Jay, S., Jones, C.E., Slinn, P., Wood, C., 2007. Environmental assessment:

retrospect and prospect. Environmental Impact Assessment Review

27, 287–300.

Keat, R., Urry, J., 1982. Social Theory on Science. Routledge and Kegan

Paul, London.

Kolakowski, L., 1972. Positivist Philosophy. Penguin, Harmondsworth.

Latour, B., 1988. Science in Action: How to Follow Scientists and

Engineers Through Society. Harvard University Press, Cambridge,

MA.

Lawrence, D.P., 1997. The need for EIA theory-building. Environmental

Impact Assessment Review 17, 79–107.

Lawrence, D.P., 2003. Environmental Impact Assessment: Practical

Solutions to Recurrent Problems. Wiley, Hoboken, NJ.

Lee, N., Colley, R., Bonde, J., Simpson, J., 1999. Reviewing the quality of

environmental statements and environmental appraisals. Occasional

Paper No. 55, University of Manchester, Manchester.

Levidov, L., 1999. Democratising technology or technologising democ-

racy. Technology in Society 20, 211–226.

Malik, M., Bartlett, R.V., 1993. Formal guidance for the use of science in

EIA: analysis of agency procedures for implementing NEPA. The

Environmental Professional 15, 34–45.

McCracken, G., 1988. The Long Interview. Sage, London.

McDonald, G.T., Brown, L.A., 1995. Going beyond environmental

impact assessment: environmental input to planning and design.

Environmental Impact Assessment Review 15, 34–45.

Miles, M.B., Huberman, A.M., 1994. Qualitative Data Analysis. Sage,

London.

Morgan, R., 1998. Environmental Impact Assessment: A Methodological

Perspective. Kluwer Academic Publisher, London.

Morrison-Saunders, A., Fischer, T.B., 2006. What is wrong with EIA and

SEA anyway? A sceptic’s perspective on sustainability assessment.

Journal of Environmental Assessment Policy and Management 8,

19–39.

Nilsson, M., Dalkmann, H., 2001. Decision making and strategic

environmental assessment. Journal of Environmental Assessment

Policy and Management 3, 305–327.

Novek, J., 1995. Environmental impact assessment and sustainable

development: case studies of environmental conflict. Society and

Natural Resources 8, 145–159.

O’Toole, L.J., 2004. The theory-practice issue in policy implementation

research. Public Administration 82, 309–329.

Owens, S., Cowell, R., 2001. Planning for sustainability—new orthodoxy

or radical challenge? Town and Country Planning, pp. 170–172.

Owens, S., Cowell, R., 2002. Land and Limits: Interpreting Sustainability

in the Planning Process. Routledge, London.

Owens, S., Rayner, T., Bina, O., 2004. New agendas for appraisal:

reflections on theory, practice, and research. Environment and

Planning A 36, 1943–1959.

Petts, J., 1999a. Introduction to environmental impact assessment in

practice: fulfilled potential or wasted opportunity? In: Petts, J. (Ed.),

Handbook of Environmental Impact Assessment. Volume 2. Environ-

mental Impact Assessment in Practice: Impact and Limitations.

Blackwell Science, Oxford, pp. 3–9.

Petts, J., 1999b. Public participation and environmental impact assess-

ment. In: Petts, J. (Ed.), Handbook of Environmental Impact

Assessment. Volume 1. Environmental Impact Assessment: Process,

Methods and Potential. Blackwell Science, Oxford, pp. 145–177.

Petts, J., 2003. Barriers to deliberative participation in EIA: learning from

waste policies, plans and projects. Journal of Environmental Assess-

ment Policy and Management 5, 269–293.

Phillips, F., 2002. The distortion of criteria after decision-making.

Organisational Behavior and Human Decision Processes 88,

769–784.

ARTICLE IN PRESSM. Cashmore et al. / Journal of Environmental Management 88 (2008) 1233–12481248

Pope, J., Morrison-Saunders, A., Annandale, D., 2005. Applying

sustainability assessment models. Impact Assessment and Project

Appraisal 23, 293–302.

Pratchett, L., 2004. Local autonomy, local democracy and the ‘new

localism’. Political Studies 52, 358–375.

Richardson, T., 2005. Environmental assessment and planning theory:

four short stories about power, multiple rationality and ethics.

Environmental Impact Assessment Review 25, 341–365.

Robson, C., 2002. Real World Research: A Resource for Social Scientists

and Practitioner-Researchers. Blackwell Publishing, Oxford.

Russel, D., Jordan, A., 2006. Gearing-up governance for sustainable

development: patterns of policy appraisal in UK central government.

CSERGE Working Paper EDM-2006-02. University of East Anglia,

Norwich.

Sabatier, P., 1988. An advocacy coalition framework of policy change and

the role of policy-oriented learning therein. Policy Sciences 21,

129–168.

Sadler, B., 1996. Environmental assessment in a changing world:

evaluating practice to improve performance. Final Report of the

International Study of the Effectiveness of Environmental Assessment.

Canadian Environmental Assessment Agency and International

Association for Impact Assessment, Ottawa.

Sayer, A., 1984. Method in Social Science: A Realist Approach.

Hutchinson, London.

Sayer, A., 2000. Realism and Social Science. Sage, London.

Schout, A., Jordan, A., 2005. Coordinated European governance: self-

organizing or centrally steered? Public Administration 83, 201–220.

Shepherd, A., Bowler, C., 1997. Beyond the requirements: improving

public participation in EIA. Journal of Environmental Planning and

Management 40, 725–738.

Silverman, D., 2001. Interpreting Qualitative Data. Methods for Analys-

ing Talk, Text and Interaction. Sage, London.

Simon, H.A., 1969. The Sciences of the Artificial. MIT Press, Cambridge,

MA.

Steinemann, A., 2001. Improving alternatives for environmental impact

assessment. Environmental Impact Assessment Review 21, 3–21.

Taylor, S., 1984. Making Bureaucracies Think: The Environmental

Impact Statement Strategy of Administrative Reform. Stanford

University Press, Stanford.

Webler, T., Kastenholz, H., Renn, O., 1995. Public participation in impact

assessment: a social learning perspective. Environmental Impact

Assessment Review 15, 443–463.

Weston, J., 2000. EIA, decision-making and screening and scoping in UK

practice. Journal of Environmental Planning and Management 43,

185–200.

Wood, C., 2003. Environmental Impact Assessment: A Comparative

Review. Pearson Education, Harlow.

Wood, C., Jones, C.E., 1997. The effect of environmental assessment on

UK local planning authority decisions. Urban Studies 34, 1237–1257.

Wynne, B., 1984. The institutional context of science, models and policy:

the IIASA energy study. Policy Sciences 17, 277–320.

Wynne, B., 2002. Risk and environment as legitimatory discourses of

technology: reflexivity inside out? Current Sociology 50, 459–477.

Yin, R.K., 1994. Case Study Research: Design and Methods. Sage, London.