Are Environmental Social Movements Socially …personal.lse.ac.uk/FORSYTHT/Forsyth_Thailand_social...Are Environmental Social Movements Socially Exclusive? An Historical Study from

World Development Vol. 35, No. 12, pp. 2110–2130, 2007� 2007 Elsevier Ltd. All rights reserved

0305-750X/$ - see front matter

www.elsevier.com/locate/worlddevdoi:10.1016/j.worlddev.2007.01.005

Are Environmental Social Movements

Socially Exclusive? An Historical Study from Thailand

TIM FORSYTH *

London School of Economics and Political Science, London WC2A 2AE, UK

Summary. — Environmental social movements in developing countries are often portrayed asdemocratizing but may contain important social divisions. This paper presents a new methodologyto analyze the social composition and underlying political messages of movements. Nearly 5,000newspaper reports during 1968–2000 in Thailand are analyzed to indicate the participation of mid-dle and lower classes, and their association with ‘‘green’’ (conservationist) and ‘‘red-green’’ (liveli-hoods-oriented) environmental values. Results show middle-class ‘‘green’’ activism has dominatedforests activism, but lower-class ‘‘red-green’’ activism has grown for forests and pollution. News-papers, however, portray all environmentalism as ‘‘democratization,’’ suggesting that the possibleexclusiveness of some environmental norms is unacknowledged.� 2007 Elsevier Ltd. All rights reserved.

Key words — environmentalism, social movements, governance, democratization, Asia, Thailand

* This research was conducted with assistance from the

UK Economic and Social Research Council. The author

would like to thank his colleagues Tim Dyson, Monica

Di-Gregorio, Diana Weinhold, and two anonymous re-

ferees for valuable comments. Final revision accepted:January 22, 2007.

1. INTRODUCTION

Do environmental social movements enhancedemocratization in developing countries? Inrecent years, various analysts, journalists, andcampaigners have claimed they do. This asser-tion is based on the belief that resistance againstcompanies and states engaged in resourceexploitation might empower less powerfulgroups and increase environmental protection(Peet & Watts, 1996; Peritore, 1999). Moreover,as Cohen and Arato (1992, p. 492) have noted,‘‘social movements constitute the dynamic ele-ment in processes that might realize the positivepotentials of modern civil societies.’’

Against this, however, a growing number ofsocial scientists have cited important reasonsto question this optimism. First, individualsand social groups may be unable to participateequally in social movements, and consequentlymovements may not represent nor benefit allpeople (Covey, 1995). Second, ‘‘environmental-ism’’ itself is highly varied, and dominantthemes may arise at the expense of alternativeperceptions, including those of poorer people(Guha & Martinez-Alier, 1997). And third,activists might not shape political progress asmuch as they think, as the reporting and inter-pretation of social movements may be con-

211

trolled by other people in accordance withtheir own values and relations with the state(Tilly, 1994). Consequently, environmental so-cial movements may not always empower lesspowerful people, or redefine environmentalagendas in their favor (Offe, 1985). Indeed, un-der these conditions, environmentalism mightnot be socially inclusive, but contain or prolongpatterns of social exclusion.

This paper contributes to research about so-cial movements and democratization by assess-ing the relationship between the socialcomposition and political values of environ-mentalism in Thailand. In particular, the paperassesses the relative importance of lower andmiddle classes within social movements, andthe co-existence of these with the so-called‘‘green’’ environmental agenda (which focuseson wilderness conservation and the protectionof nature against people), and the ‘‘red-green’’agenda (which is more livelihoods oriented

0

ARE ENVIRONMENTAL SOCIAL MOVEMENTS SOCIALLY EXCLUSIVE? 2111

and seeks to protect people and environmentsimultaneously). Some analysts have arguedthat the ‘‘green’’ agenda is associated with mid-dle classes, and the ‘‘red-green’’ agenda is an‘‘environmentalism of the poor’’ (Martinez-Alier, 2002; Nash, 1982). But, to date, littleempirical work has tested this association. Thispaper seeks to do so by asking three key ques-tions: How have different social classes partici-pated in environmentalism? What have beenthe main political objectives and varieties ofenvironmentalism? And are there associationsbetween the social composition and varietiesof environmentalism? These questions areasked for all environmental protest in Thai-land, and then specifically for activism concern-ing forests and industrial pollution.

To achieve these objectives, the paper pre-sents a new methodology for analyzing socialmovements based on historic newspaper re-ports. Newspaper reports are useful because—where they exist in good detail—they indicateboth the historic events and social participationwithin social movements, as well as how eachwas reported. ‘‘Journalism,’’ so the saying goes,‘‘is history’s first draft,’’ 1 and accordinglynewspapers give a record of social movementsthat builds a common reference point of howincidents of activism became historically signif-icant (Wakefield & Elliott, 2003). The databaseused in this paper of nearly 5,000 news reportsduring 1968–2000 is the largest empirical surveyof environmentalism of its type, and provides apowerful supplement to analyses based on casestudies (Buergin & Kessler, 2000; Forsyth,2004). The method described in this papermay provide both quantitative and qualitativeanalyses, although this paper chiefly presentsa quantitative overview of its findings.

Thailand is an appropriate location for thisresearch because it has experienced many inci-dents of activism concerning forest protection,opposition to dams, or concerns about pollu-tion that have been linked to processes ofdemocratization (Hirsch, 1997a, 1997b; Somc-hai, 2006). The paper starts, however, with asummary of debates about environmental so-cial movements, and the benefits of using news-papers to analyze them.

2. ENVIRONMENTALISM AND SOCIALMOVEMENT THEORY

In recent years, various activists and analystshave claimed that environmental social move-

ments may enhance democratization and theempowerment of poorer people in developingcountries. For example, writing about environ-mentalism, the development theorist, ArturoEscobar (1996, p. 65) wrote: ‘‘We need newnarratives of life and culture . . . this is a collec-tive task that perhaps only social movementsare in a position to advance.’’ But can socialmovements achieve this?

Social movements have been defined as col-lective challenges by people with common pur-poses, usually in interaction with elites andauthorities (Eyerman & Jamison, 1991; Ibarra,2002; Tarrow, 1994). The nature of socialmovements and their method of analysis, how-ever, vary widely. According to Della Porta andDiani (1999, p. 3), social movements analysismay be divided into four main approaches ofcollective behavior, resource mobilization,political process, and new social movements.These approaches focus, respectively, on thestructural rebalancing of political systems; thecognitive tactics activists use to influence pol-icy; the political opportunities for movementswithin institutional contexts; and new socialidentities in advanced societies. All of these ap-proaches also offer insights into the relationshipof movements on democratization. Social activ-ism may lead to the greater representation ofpoorer people in politics, enhance politicalrights, and create new participatory public are-nas of deliberation (McCoy, 2001).

According to much writing on social move-ments, environmentalism may be considered a‘‘new’’ social movement because it emerged asan important political force and source of iden-tity along with women’s and peace movementsin Europe and North America in the 1960s(Maheu, 1995). Moreover, unlike ‘‘old’’ socialmovements, which were based on industrialclass divisions and their material interests, envi-ronmentalism and other ‘‘new’’ social move-ments have claimed to transcend old classdivisions, even if activists themselves tendedto be middle class or highly educated (Eder,1993; Touraine, 1981). Environmentalism maytherefore have some tensions for social repre-sentation because it claims to be ‘‘typically apolitics of a class but not on behalf of a class’’(Offe, 1985, p. 833; emphasis in original).

Many writers about environmentalism indeveloping countries, however, have sometimesclaimed that lower social classes have moreagency (Taylor, 1995; Wignaraja, 1993). Writ-ing about Thailand, for example, James Fahn(2003, p. 324), a US-born journalist, considered

2112 WORLD DEVELOPMENT

environmentalism to be a universal social normby writing: ‘‘the green urge is universal andtranscends national and cultural boundaries.’’Yet, he also claimed (2003, p. 7):

Whereas the green movement in the North tends tofocus on the middle class, in the South not only isit centered more on the farmers and fishermen whorely on natural resources, but it’s also concernedmore with who gets to use resources, not just withhow they are used. . . In Asia, such activists are themainstream because the majority of the populationthere is still underprivileged, and the middle class re-mains a minority (emphasis in original).

And Phil Hirsch (1997a, p. 179), an academic,has commented:

The environmental movement in Thailand has be-come a significant force in recent years. . . The move-ment has drawn in a wide range of social, economicand political actors in Thai society. . . In this respect,environmentalism represents an opposition force,but one that has, ironically, been increasingly inclu-sive (emphasis in original).

But against this optimism, other authorshave suggested that social movements maynot be so socially inclusive. First, can differentsocial classes participate equally in social move-ments? Clearly, many movements in developingcountries have included richer and poorer peo-ple. But these also have unequal resources andinfluence. Covey (1995), for example, found inthe Philippines that political alliances betweennational NGOs and local grassroots organiza-tions generally represented the national NGOinterests more effectively. Others have arguedthat grassroots organizations willingly suppresssome objectives in order to gain visibility andwin the support of the state or other actors(Lohmann, 1995).

Second, do class differences still matter? Var-ious analysts have argued that environmentalvalues may reflect social class or socio-eco-nomic contexts. Nash (1982), for example, ar-gued that the perception of wilderness asbeautiful or threatened in the United Stateswas correlated with the growth of urbanizationand industrialization, and was therefore mostfelt by urban, rather than rural, dwellers, oftenrelated to social class. Others have argued thatthe so-called ‘‘green’’ and ‘‘red-green’’ environ-mental agendas may also be linked to middleand lower social classes, and that a ‘‘green’’agenda (based on conserving wilderness) mayplace tensions between urban middle-class per-ceptions of environment and people who live inthe so-called wilderness areas, such as pastoral-

ists or shifting cultivators who are inherentlymore concerned about the ‘‘red-green’’ (liveli-hoods-oriented) agenda (Guha & Martinez-Alier, 1997; Neumann, 1998). 2 Indeed, recentdebates about community forestry in Thailandhave frequently counter-posed views of whetherpeople and forests can co-exist (Johnson & For-syth, 2002). Environmentalism should thereforenot be considered as one single norm, andmovements may result in different varieties ofenvironmentalism being communicated accord-ing to the needs and relative strength of differ-ent participants.

And third, are both environmental valuesand structures of social movements shaped his-torically, and by wider political norms? Rangan(2000, p. 181), for example, claims that the fa-mous anti-logging Chipko movement in north-ern India reflected various environmentalconcerns from local people, including the‘‘red-green’’ concern to protect forest suppliesto local sawmills. 3 But the campaign becameportrayed as a ‘‘green’’ example of women en-gaged in ‘‘tree hugging’’ because this imagegained the sympathy of urban-based environ-mentalists, academic and international environ-mental lobbies. Hajer (1995, pp. 64–65) hasargued that this phenomenon can be calledthe evolution of ‘‘storylines,’’ in which conflictsbecome associated with predefined notions ofsocial blame and responsibility in hidden ways.Consequently, some themes within environ-mentalism—such as opposition to logging ordams—may carry historic associations, whichmay not be shared or relevant to all activistsor contemporary disputes, or apparent in howsocial movements are reported (Tilly, 1994).In Thailand, environmental policies have alsobeen linked to nationalism and citizenship,indicating a strong involvement of the state inhow environment is defined (Vandergeest,2003).

As a result of these concerns, many analystshave questioned whether environmentalism isindeed a universal norm or inclusive force. Inpart, this is because of the difficulty of applyingthe concept of class—as discussed under newsocial movements—to societies that compriserural peasantry and urban workers, and whichare further divided by ethnicity, gender, orcaste. At the same time, it is important not todismiss or stereotype middle-class environmen-talism as unhelpful to poorer people, or as inef-fective (Kim, 2000; Mawdsley, 2004). (Theidentification of class in this paper is discussedfurther in the methodology section.)


Similarly, it is sometimes difficult to agreeupon a definition of ‘‘environmental’’ concernsthat can be shared by different groups. Forexample, much contemporary environmental-ism is associated with the ‘‘green’’ agenda, yetsome aspects of environmental justice or the‘‘red-green’’ agenda may not immediately beconsidered environmental by some activists. Inthe United States, for example, discussions ofenvironmental justice and environmental racism(concerning topics such as the location of wastetreatment facilities near disadvantaged groups)may also address general questions of publicpolicy and social attitudes as well as ecologicalrisks (Pellow, 2002). In such cases, social activ-ism may help broaden the definition of environ-mental concerns, or find it use existingdefinitions in order to gain visibility and supportfrom other activists and the state. Under theseconditions, existing forms of environmentalismmay exclude alternative conceptions of environ-mental priorities from less powerful actors.

Consequently, rather than adopting strictdefinitions of social class or environmentalism,it is more useful to look at how social activismand environmental norms co-evolve, andwhether either implies social exclusion. Look-ing at historic newspaper reports is one wayto do this.

3. ANALYZING SOCIAL MOVEMENTSWITH NEWSPAPERS

Social movement research has adopted vari-ous techniques including interviewing activists,or analyzing campaigning literature (Diani &Eyerman, 1992). Analyzing historic newspapersmay complement these approaches in the fol-lowing ways:

— Newspapers, or any regular media report-ing, provide a basic historical record ofevents as they happened. This record, how-ever, depends on the detail of reporting, edi-torial policy, and freedom from statecensorship. The style of news reporting alsovaries between news sources and the type ofnews report, such as front-page news or fea-tures 4 (Wakefield & Elliott, 2003).— Newspapers allow quantitative measure-ments of social actors described in eachreport as activists or sources of information(Jensen & Jankowski, 1991).— Textual or discourse analysis may be per-formed on newspaper reports to indicatesocial and political values at the time of writ-

ing. These values may be about the topicunder consideration, or about the actorsdescribed in the report, and the rolesascribed to them. This analysis may be per-formed on straight news reports, and/or fea-tures and opinion pieces (Bauer, 2000;Donati, 1992; Silverman, 1993; van Dijk,1991).— Newspaper reports for a specific topic orconflict may be analyzed collectively to indi-cate how political values or social actorswere represented over time. This analysismay indicate the symbolism or ‘‘storyline’’attributed to the activism (D’Anieri, Ernst,& Kier, 1990; Hajer, 1995).

Of course, newspapers may also be selectivehistorical sources, and avoid some cases ofpolitical activism. Newspaper censorship andeditorial styles may change over time. Individ-ual journalists may demonstrate personalbiases. In response to these concerns, it shouldbe remembered that analyzing newspapers doesnot aim to provide an objective history of socialmovements, but to indicate how activism wasgiven meaning and hence became historicallysymbolic. The limitations of newspapers canbe reduced by seeking alternative histories ofevents, or by interviewing journalists aboutreporting practice and censorship (Porter,1992). Comparing similar events from differentnews sources may also indicate biases.

4. THE STUDY

Historic newspapers were used to measurethe social composition and political content ofenvironmentalism in Thailand during 1968–2000. The study asked three key questions:

— How do different social classes participatein different episodes of environmentalism?— What are the main political objectives(including varieties of environmentalism) ofdifferent episodes of activism?— Are there associations between the partic-ipation of different social classes and varie-ties of environmentalism?

To achieve these objectives, the study ana-lyzed different individual episodes of environ-mentalism, and then compared episodeslinked to forest protection with those concern-ing industrial pollution. In this paper, the anal-ysis focused on middle- and lower-class actors,and ‘‘green’’ and ‘‘red-green’’ environmental-ism, although (as discussed below) further ac-tors and political themes were researched. The


study was restricted to reporting of events inThailand, and only included internationalevents if they were connected to Thai disputes.Some brief comparison with women’s move-ments was conducted to describe another newsocial movement. 5

5. RESEARCH DESIGN

To conduct the study in Thailand, variousdecisions were necessary. First, which newssource? The Bangkok Post was chosen for var-ious reasons. It is considered by many to beThailand’s leading authoritative broadsheet,and has reported environmental stories sincethe 1960s. The Bangkok Post is published inEnglish, although its target audience are edu-cated Thai people, and domestic news is writtenexclusively by Thai journalists. 6 The use ofEnglish language assisted the speed of analysis(although the researcher can speak and readThai). Like all Thai newspapers, various statepressures have influenced the Bangkok Postover the decades, and especially under militarygovernments during the 1970s when the Viet-nam War and the fear of insurgency encour-aged the closure of some forest areas topublic access (Vandergeest, 1996). More re-cently, the Bangkok Post has adopted criticalpositions to governments, especially since theera of formal democratic elections from1988. 7 Indeed, during the military governmentof 1991–93, the paper famously printed anempty front page to protest against governmentsuppression of news about the May 1992 Bang-kok pro-democracy demonstrations (McCargo,2000). Pragmatically, the Bangkok Post also of-fered a comprehensive library of news clippingssince the 1960s, where newspaper staff had al-ready organized into different themes or stories.

A brief comparative study of Thai-languagenewspapers 8 was also undertaken for selectedthemes. This comparison indicated the differ-ence between the day-to-day reporting stylesof the Bangkok Post with more sporadic andpopulist reporting of other newspapers. Bang-kok Post journalists and editors were also inter-viewed for their personal histories andviewpoints. (These results are not discussedfor reasons of space.)

Second, how to identify environmental con-flicts? Environmental conflicts were definedgenerally as cases where various citizens,NGOs, or other activists undertook acts ofpublic demonstration or lobbying of govern-

ment in order to change policy or preventunpopular developments. Specific conflictswere identified using the Bangkok Post’s ownfiling classification in order to avoid imposingoutside definitions of grouping news reports.In effect, each clippings file presented the com-plete ‘‘storyline’’ of different episodes of con-flict. Some specific episodes (such as the‘‘Nam Choan dam’’ dispute) were convenientlyidentifiable in time. Others, such as general fileson ‘‘logging’’ or ‘‘air pollution,’’ were ongoing.Table 1 lists each episode of environmentalism.Specialists on Thailand will be familiar withthese episodes, although there is insufficientspace to allow a full description of each event.(Perhaps, surprisingly, there has been relativelylittle public activism on matters of climatechange or genetically modified crops in Thai-land.)

Third, what period to study? The total periodof 1968–2000 was selected largely pragmaticallyas no files for environmentalism existed beforethis date, and the files of paper clippings werereplaced by an electronic database in 2000.However, these dates also conveniently coveredthe period from the rise of environmentalism asa global political force in the late 1960s. Asnoted below, some years in the 1970s werenot analyzed in order to save time, and to avoidperiods of obvious censorship.

Fourth, which episodes to include? All spe-cific incidents of environmental social move-ments and conflict were included in the study,starting with the Chiang Mai cable car dispute(dating from the early 1970s), but more fullyfrom 1982 with the start of the Nam Choandam dispute, and in 1986 with the conflict con-cerning tantalum mining in Phuket (see Table1). Files on logging and pollution started from1968. These files sometimes contained episodesof major conflict (such as the political cam-paign for the 1989 logging ban), but usuallyspecific episodes on these themes were filed sep-arately, and hence the subject matters over-lapped. To avoid periods of well-known statecensorship, the decision was taken to analyzeonly the periods of 1968–71, 1978–81, 1988–91, and 1996–2000 for logging, air, and waterpollution. (When the cumulative research forforest- and pollution-based activism was ana-lyzed, these files were collated with specific inci-dences of activism.)

Some stories were excluded. Reports aboutfishing, mangrove forests, and aquaculture werenot included in this survey because of shortageof time, and because these activities—mainly

Table 1. Environmental conflicts in Thailand, 1978–2000

Conflict Dates of first and lastnews reportsa

Brief summary

Mainly ruralLogginga,b January 68 (periodically

to) December 2000Ongoing, long-term coverage of public debatesabout logging, including the 1989 logging ban

Chiang Mai cable carc March 69–April 94 Unsuccessful proposal to build cable car to NationalPark and Buddhist temple in Doi Suthep

Nam Choan dam April 82–February 98 Unsuccessful proposal to build a dam in rainforest inwestern Thailand (Kanjanaburi province)

Pak Mul dama April 89–December 2000 Dam on river in northeast Thailand, and resettlementand compensation of fishing communities

Pa Kham, Buri Ramforest evictions

June 89–February 95 Government plans to resettle farmers in eucalyptusand pine plantations, northeast Thailand

Other forest evictionsd February 94–April 94 Government plans to resettle farmers in eucalyptusand pine plantations, northeast Thailand

Suan Kitti scandal January 90–December 90 Senator accused of flouting logging/plantation laws(Chachoengsao province)

Encroachments intoNational Parkse,a

May 93–December 99 Ongoing opposition to logging/agriculture/tourism/mining in national parks

Yadana pipeline January 95–December 99 Construction of gas pipeline from Burma throughrainforests in western Thailand

Community Forestryf,a January 96–December 2000 Ongoing debate about laws governing forest accessTha Chana scandal January 96–September 98 MP accused of flouting logging laws

(Surat Thani province)Chom Thong watersheda December 96–December 99 Highland–lowland dispute about alleged damage of

upland agriculture in Chiang Mai provinceSalween National

Park scandalJanuary 98–May 2000 Illegal logging discovered in national park in Mae

Hong Son provinceOpposition to The Beach October 98–March 99 Opposition to government decision to allow filming

of a Hollywood movie in two national parks

Mainly industrialPhuket tantalum mining May 86–June 92 Environmental and economic effects of tantalum

mining in Phuket, a tourist and fishing provinceGeneral pollutiona June 88–December 99 Ongoing concern about non-specific forms of pollutionWater pollutionb,a January 68 (periodically to)

December 2000Ongoing concern about water pollution, includingwaste water

Mae Moh power planta October 92–December 99 Concern about a power plant that uses lignite fuelin northern Thailand (Lampang province)

Air pollutionb,a January 68 (periodically to)December 2000

Ongoing concern about air pollution, especially in cities

Nam Pong riverpollution

April 93–August 98 Alleged pollution by Phoenix pulp and paper plantinto river in northeast Thailand(Khon Kaen province)

Lamphun industrialpoisoning

February 94–October 96 Public debate about industrial poisoning in an industrialestate in northern Thailand (Lamphun province)

Rayong, Map Ta Phutindustrial estatea

March 94–December 99 Concern about pollution and waste treatment at anindustrial estate on the eastern seaboard of Thailand

Prachuab Khiri Khanpower planta

November 98–July 2000 Proposed construction of coal-fueled power plantin fishing area with coral reef (eventually rejected)

a Ongoing stories where the end date of assessment is not the end date of activism.b Logging, water pollution and air pollution comprise the total of four sub-periods: 1968–71, 1978–81, 1988–91, and1996–2000. (This paper does not compare these sub-periods because of shortage of space.) The most important eventfor logging was activism before and after the logging ban, eventually passed in 1989. There was no file for ‘‘generalpollution’’ before June 1988.c The Chiang Mai cable car dispute had two distinct periods of reporting: March 1969–March 1971 and April 1985–May 1989 (the last report was in April 1994).d Other evictions include Dong Yai (in Buri Ram province, March 1994); Thap Lan (in Nakhon Ratchasimaprovince, February, 1994); and Tha Takiab (in Chachoengsao province, March 94).e Community Forestry also includes the localized dispute about forest encroachment and community forestry inDong Larn, northeast Thailand.f This category of Encroachment into National Parks does not include the specific case of opposition to the filming ofThe Beach.



in southern Thailand—also refer to complexquestions of Muslim identity that may compli-cate the analyses of social class. Furthermore,the influence of mangroves alone on forestsactivism is low compared to other forests (seeFlaherty, Vandergeest, & Miller, 1999). Numer-ous other subjects such as ‘‘housing’’ or ‘‘urbanplanning’’ were also not included as specific top-ics, as they were not specifically labeled as envi-ronmental, but could be considered by futureresearch.

Fifth, how to analyze social composition ofactivism? This was achieved both quantitativelyand qualitatively. The number of different ac-tors cited (in active or passive roles) in eachnews report was identified and counted. Theoriginal research counted various categoriesincluding state actors (e.g., politicians, forestryagencies), international organizations, scientificexperts, businesses, as well as members of mid-dle and lower classes. Most of these actors wereself-evident (it is acknowledged that non-stateactors may sometimes overlap with state). Sci-entific experts were identified when an individ-ual was cited as giving formal expertise.Where the individual came from a state agencyor company, etc., both ‘‘expert’’ and the othersector were marked.

This paper, however, focuses only on middleand lower-class actors. The identification ofthese was more complex, and inevitably in-volved simplifications (see Table 2). Identifica-tion was only done after guidance from Thaijournalists, NGO workers, and academics. 9

In practice, ‘‘middle class’’ implied relativelywealthy individuals and/or those educated totertiary level. It also included people of obvious

Table 2. Summary of guidelines used for id

‘‘State’’ actors ‘‘Middle class

—Politicians (e.g., ministers)—Government ministries—Government agencies (e.g.,Royal Forest Department,and their spokespeople)

—Intellectuals speaking oenvironment (e.g., Boons—Educated, or elite Thai(including university studspeaking as students)—Members of Thai roya(except the highest echelo—Some NGOs with highinvolvement (e.g., Seub FDhamanaat Foundation)—Monks who take a stanabout environment

NB: these actors are all within Thailand, and exclude inter

‘‘upper’’ class such as aristocrats (although thehighest members of Thai royalty were treatedseparately). ‘‘Lower class’’ frequently referredto factory or agricultural workers in menialtasks, or various villagers or slum dwellers.Politically active monks were generally consid-ered ‘‘middle class’’ when they fitted thedescription of intellectuals voicing concern forall society, even though many monks are frompoor backgrounds and some work specificallywith the poor. (Despite their high visibility,monks were reported prominently in only fewdisputes, notably the Chom Thong watershedstruggle, and the evictions of villagers from for-est plantations in northeastern Thailand.)

The classification of middle- or lower-classorganizations needed much care, and raisedthe possibility of autocorrelation when compar-ing actors with varieties of environmentalism.Organizations were classified primarily accord-ing to their membership rather than viewsabout environment. Nonetheless, this approachdid result in classifying some organizations as‘‘middle class’’ that were overtly campaigningto protect wilderness or architectural sites,including some who sought to exclude farminggroups from these areas, particularly in ChomThong (see Table 2). ‘‘Lower-class’’ organiza-tions included trade unions and NGOs cam-paigning for workers’ or poor farmers’ rights.It was acknowledged that this classificationwould occasionally simplify complex socialrelations, but it assumed the approach adoptedwould provide a useful platform for findings.

And sixth, how to analyze political content ofreports? News reports were analyzed for under-lying political messages and assumptions about

entifying political actors in news reports

’’ ‘‘Lower class’’

ut aboutong Legakul)citizens

ents

ltyns)middle-classoundation,

ce

—Factory workers, agriculturalworkers, residents of poor housing—Trade unions— NGOs or alliances associated withpoor people (e.g., Assembly of the Poor)—Farmers’ organizations or movementsassociated with poor (e.g., Chomthonglowland farmers’ group; NorthernFarmers’ Network)

national actors.


environment. This analysis took two stages. Inthe first stage, an open-ended discursive analy-sis of text was applied to identify the politicalviewpoint underpinning the subject matter ofreports. For example, some themes identifiedat this stage included concerns about ecology;concerns about state failure (including corrup-tion); the threat of unregulated business; socialinjustice; community strength; or various opti-mistic views such as the benefits of wealth ortechnology. These themes existed alongsidethe subject matters of reports such as forestsor dams. The incidence of different themes ineach report was recorded both quantitatively(the number of times these themes occurred)and qualitatively (using textual analysis).

The second stage of analysis then accumu-lated these diverse frames into four overviewframes that simplified the initial classificationof political content. These broader framesaimed to describe the initial classification, butin ways relevant to wider debates about envi-ronment and development. These overviewframes were ‘‘green’’ and ‘‘red-green’’ formsof environmentalism (defined above); ‘‘democ-ratization’’ (referring to any sub-frame thatemphasized state failure or the benefits of com-munity action); and ‘‘modernization’’ (referringto sub-frames stressing the benefits of economicgrowth, technology and an unreformed state).Table 3 gives an expanded definition of howeach overview frame was defined.

The advantage of identifying frames in twostages is that it allows further studies to usethe sub-frames to create different overviewframes if needed. It should be noted that it waspossible for each report to have scores for morethan one of these overview frames, and that fre-quently some frames would often co-exist.

Finally, it was important to consider changesin newspaper reporting itself. Journalists andeditors at the Bangkok Post were interviewed

Table 3. Summary of guidelines used for ide

‘‘Green’’environmentalism

‘‘Red-green’’environmentalism

—‘‘Natural environmentis threatened’’

—‘‘Nature is beautiful’’—‘‘People are a threat

to natural environment’’

—‘‘Environmental changeis a threat to people’shealth and livelihoods’’—‘‘People need resourcesto live’’—‘‘Poor people knowhow to look afterenvironment’’

to assess their own viewpoints, and changes ineditorial policies. The Bangkok Post now hasincreased its page numbers over the years,and, in keeping with international changes injournalism, has increasingly adopted an anec-dotal style of reporting, which emphasizesdescriptions of individuals rather than a distantsummary of events. The methodology used inthis paper cannot remove these biases, butcan acknowledge these changes.

6. ANALYSIS

In total, 4,672 new reports were analyzed (atotal of 88,315 ‘‘column inches’’ of text). Themethods above produced the following infor-mation:

— Quantitative measurement of differentactors cited per report— Quantitative measurement of differentpolitical themes cited per report— Qualitative examples of text demonstrat-ing environmental storylines

(For reasons of space, however, this currentpaper focuses chiefly on quantitative findings).

Information was initially presented on a dailybasis, but was totaled for each specific episodeof activism and for each month and year. Thisstructuring allowed the comparison of differentepisodes of activism; average total of activismon a year-by-year basis; or between differentthemes of activism, specifically concerning for-ests and industrial pollution.

Tests of association were also applied toquantitative information about the different ac-tors and political themes of activism. Simplelinear regression was used to test associationsbetween class composition and different politi-cal themes, using Pearson’s ‘‘r’’ coefficient asthe indicator of association. These tests, how-ever, risked certain statistical weaknesses. Tests

ntifying overview framings of news reports

‘‘Democratization’’ ‘‘Modernization’’

—‘‘The state is corrupt orfailing and needs reforming’’—‘‘Communities can governeffectively’’—‘‘Social movements arepositive forces of politicalreform’’

—‘‘Economic growthis good’’—‘‘Technology is a usefulsolution to environmentalproblems’’—‘‘The state is strong anddoes not need reforming’’


could be performed on data collected on adaily, monthly, or yearly basis. In all cases,the clearest relationships were found whenusing monthly or yearly data, but these datasetsinevitably had fewer observations. It wasdecided to accept this limitation in order to ex-plore which trends became apparent. In orderto make charts clearer, ‘‘r’’ coefficients werepresented in an accumulated fashion (i.e., oneyear’s total was added to the previous year),meaning that the strongest long-term associa-tions are shown by rapidly ascending lines.

(a) Social composition of activism

Figure 1 shows the growth of new reports onenvironmental subjects in Thailand since 1968.Table 4 indicates statistical summaries concern-ing the social composition and overall politicalmessages contained in each episode of conflict.Figure 2 charts the long-term trend in thesenumbers.

Figure 1 is consistent with most histories ofenvironmentalism in Thailand since the 1960s(Buergin & Kessler, 2000; Hirsch, 1997a,1997b; Ubonrath, 1991; Vandergeest, 1996).News reports about pollution and logging be-gan in 1968, but the first public campaign on‘‘environment’’ was to stop a cable car on theBuddhist shrine of Doi Suthep in Chiang Maiin the early 1970s. 10 This ‘‘green’’ focus onenvironmentalism grew in the 1980s with a suc-cessful campaign (involving internationalNGOs such as the World Wildlife Fund) to pre-

0

10

20

30

40

50

60

70

80

90

1968

1970

1972

1974

1976

1978

1980

1982

Activism before and after 1989 logging ban

Phuket tantalum

Nam Choan dam

Figure 1. Number of total environmental reports, by month, w

text for explanation of sampling and th

vent the construction of a large dam in theNam Choan forest in the western rainforestsof Kanjanaburi province, and concerns aboutthe impacts of tantalum mining on tourismand fishing in Phuket in 1986. By the late1980s, various organizations were campaigningto achieve a national logging ban (which waspassed in 1988, the same year as the cancella-tion of the Nam Choan dam). Since 1988, therehave been several test cases of the logging ban(often including alleged rule bending by politi-cians such as in the Suan Kitti and Tha Chanascandals), and various worries about encroach-ment into national parks (see Table 1).

After the logging ban, however, environmen-talism arguably changed toward a more ‘‘red-green’’ agenda by considering the social justiceof relocating villages in northeastern Thailandby a military government (1991–92) and laterby the Royal Forest Department, and thelong-standing debate about the ecological andsocial impacts of the Pak Mul dam on poor fish-ing communities. Concerns about industrialpollution also highlighted social justice by con-sidering the impacts of coal-fired power stations(Mae Moh and Prachuab Khiri Khan). Per-haps, most importantly, the debate about thenew community forestry bill (still ongoing) dis-cussed rules for greater public participation inforest governance, and posed conservationist(‘‘green’’) activists against livelihoods-oriented(‘‘red-green’’) campaigners. In particular, dis-putes about community forestry have ques-tioned how far upland minorities in northern

1984

1986

1988

1990

1992

1994

1996

1998

2000

Yadana pipeline, Salween scandal, community forestry

ith key peaks indicated. Source: The Bangkok Post. (See

e gap in data between 1972–1977).

Table 4. Statistical summaries for environmental conflicts in Thailand, 1968–2000

Basic descriptions,conflict (see Table 1)

Total newsreports

Averagecolumn inches

Actors pernews report

Political frames ofstories per news report

‘‘State’’ ‘‘Middle class’’ ‘‘Lower class’’ ‘‘Green’’ ‘‘Red-Green’’ ‘‘Democratization’’ ‘‘Modernization’’

Women’s rights 180 55.93 0.72 1.14 0.57 n/a n/a n/a n/aLogging (total of

four periods)1,342 16.54 1.23 0.17 0.26 0.20 0.09 0.21 0.67

Chiang Mai cable car 36 11.11 1.28 0.36 0 0.25 0 0.11 0.75Nam Choan dam 147 19.23 0.91 1.20 0.07 0.34 0.02 0.40 0.17Pak Mul dam 365 22.61 0.64 0.38 0.69 0.18 0.19 0.45 0.30Pa Kham, Buri Ram

evictions48 21.67 1.15 1.02 0.90 0.08 0.65 0.31 0.40

Other forest evictions 26 17.77 1.65 0.46 0.81 0.23 0.19 0.15 0.33Suan Kitti scandal 63 24.22 1.70 0.06 0.10 0.06 0.03 0.43 0.19National Park

Encroachments78 28.83 1.44 0.81 0.17 0.29 0.04 0.46 0.28

Yadana pipeline 209 22.53 0.72 1.05 0.13 0.42 0.19 0.35 0.28Community Forestry 242 24.79 1.36 0.33 0.55 0.28 0.12 0.35 0.40Tha Chana scandal 79 21.41 1.80 0.10 0.13 0 0 0.28 0.70Chom Thong watershed 29 34.66 1.07 0.76 1.38 0.17 0.21 0.59 0.14Salween scandal 104 22.22 2.22 0.05 0.21 0.13 0.02 0.47 0.56Opposition to The Beach 42 20.36 0.90 1.07 0.19 0.71 0 0.40 0.19Phuket tantalum mining 100 23.95 1.27 0.16 0.13 0.06 0.01 0.13 0.88General pollution 276 24.45 0.96 0.28 0.26 0.20 0.22 0.31 0.24Water pollution 549 17.04 1.08 0.10 0.32 0.28 0.48 0.22 0.36Mae Moh power plant 103 14.49 1.14 0.01 0.90 0.01 0.39 0.36 0.34Air pollution 412 17.79 0.99 0.05 0.33 0.13 0.50 0.15 0.53Nam Pong river pollution 48 16.90 0.98 0.29 0.25 0.35 0.13 0.17 0.67Lamphun poisoning 33 32.06 1.18 0.03 0.85 0.03 0.52 0.55 0.12Rayong industrial estate 95 16.73 1.47 0.13 1.07 0.19 0.55 0.49 0.34Prachuab Khiri Khan

power station66 52.52 2.45 0.64 1.67 0.67 0.06 1.56 0.70

Source: Analysis performed on reports in The Bangkok Post.

AR

EE

NV

IRO

NM

EN

TA

LS

OC

IAL

MO

VE

ME

NT

SS

OC

IAL

LY

EX

CL

US

IVE

?2119

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1968

1971

1974

1977

1980

1983

1986

1989

1992

1995

1998

STATE

MIDDLE CLASS

LOWER CLASS

Figure 2. Representation of state, middle-class, and lower-class actors in all environmental stories 1968–2000 (average

actors per news report). Source: Bangkok Post.


Thailand should be considered as destroyers offorest, or whether the state and some NGOshave overstated environmental concerns to en-hance visions of national security (Forsyth &Walker, 2008; Vandergeest, 2003). The ChomThong dispute (concerning the campaign toresettle Hmong farmers in Chiang Mai prov-ince) is one example of this (see Table 1),although previous attempts to resettle ethnicminorities from national parks (such as the Ka-ren from the Thung Yai Naresuan national parkin the western forest complex, or other resettle-ment of Hmong) before the 1990s were gener-ally not reported. A new Assembly of the Pooremerged in 1997 to campaign about poverty,livelihoods, and environment (Somchai, 2006),and the 1997 Constitution asserted the rightsof communities to participate in decisionsaffecting resources.

Figure 2 shows the statistical trends in the so-cial composition of environmentalism duringthis period. As shown in Figure 2, the represen-tation of different classes has changed markedlyin environmental news reporting. State actorshave been most prominent, with a relatively sta-ble average of approximately 1.2 states actors

mentioned per report. However, lower-class ac-tors have increased markedly, from approxi-mately 0.1 actors per report in the early 1970sto high points of 0.8 in the late 1990s. Middle-class actors, however, have declined from a highpoint (during the Nam Choan rainforest dis-pute) in 1982, of more than one actor per report,to approximately 0.4 in the late 1990s. At surfacevalue, this suggests that environmental newsreporting has become more inclusive of lowerclasses, and therefore may achieve a form ofdemocratization. However, these results do notindicate whether the higher citation of lower-class actors reflects a growing presence of poorpeople in social movements; changes in journal-istic practices; or—more controversially—thedesire of journalists or NGOs to report lower-class voices to gain political legitimacy.

The results for individual episodes of activ-ism, however, are more complex. The study ofwomen’s movements showed that middle-classactors were cited at a high level of 1.14 perstory, compared with 0.57 lower-class actors.This imbalance is consistent with theoriesabout new social movements that suggest thiskind of identity-based politics is (as stated by


Offe above) ‘‘of [the middle] class, but not nec-essarily on behalf of a class.’’ But this pattern isnot necessarily matched by all environmentalconflicts. The five conflicts with most middle-class actors were the Nam Choan dam (1980s;1.20 actors per report); opposition to TheBeach (1998–99; 1.07); the Yadana gas pipelinefrom Burma (1990s; 1.05); protests against vil-lage evictions in northeastern Thailand (1990s,1.02 actors); and concern at encroachment ofnational parks (1990s; 0.81). (It is worth notingthat the encroachment in the Salween nationalpark reported hardly any middle-class actors,and was largely reported as an affair involvingstate actors.) The average of middle-class actorsfor all reports, for the entire database, was 0.3,compared with 0.37 for lower-class actors and1.17 for state actors.

Lower-class actors were generally less re-ported, but featured more frequently in specificepisodes. The five conflicts with most lower-class actors were the Prachuab Khiri Khan pro-tests against a power plant (1998–2000; 1.67 ac-tors per report); the Chom Thong actionagainst upland agriculture in watersheds(1996–99; 1.38); pollution in the Rayong indus-trial estate (1990s; 1.07); pollution at the MaeMoh power plant (1990s; 0.90); and the evic-tions in northeastern Thailand (1990s, 0.90).

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1968

1970

1972

1974

1976

1978

1980

1982

Figure 3. Themes of ‘‘green’’ and ‘‘red-green’’ environmental

of frames per report). S

These results also indicated that some epi-sodes had higher incidences of lower and mid-dle classes either being present or absent atthe same time. The evictions in northeasternThailand had the highest mutual presence ofboth middle and lower classes. Other eventswith both included the Chom Thong dispute(where lowland poor farmers allied with mid-dle-class environmentalists against uplandfarmers), and to a lesser extent, the PrachuabKhiri Khan power station (where fishermen re-sisted the damage to livelihoods, but environ-mentalists expressed concern about theimpacts on a coral reef).

Examples of conflicts where both lower andmiddle classes were absent were the Suan Kittiscandal, the Thailand Chana scandal, and ille-gal logging in the Salween national park. Allof these episodes involved alleged corruptionof politicians or government agencies, and newsfocused on state actors.

(b) Political content of activism

Table 4 also summarizes the incidence ofpolitical messages in reports about differentevents. Figures 3 and 4 also chart the changesof these messages over time.

1984

1986

1988

1990

1992

1994

1996

1998

2000

GREENRED-GRN

ism in all environmental news, 1968–2000 (average number

ource: Bangkok Post.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1968

1971

1974

1977

1980

1983

1986

1989

1992

1995

1998

DEMOCMOD

Figure 4. Themes of ‘‘democratization’’ and ‘‘modernization’’ in all environmental news, 1968–2000 (average number of

frames per report). Source: Bangkok Post.


Figure 3 demonstrates that ‘‘green’’ environ-mentalism reached its peak during the start ofthe Nam Choan dam campaign in the early1980s. This is not surprising as the dam was athreat to an internationally regarded rainforestarea, and most overtly environmental groups atthat time were concerned with threats to wilder-ness and national heritage. Yet, perhaps sur-prisingly, ‘‘green’’ and ‘‘red-green’’environmentalism have been reported relativelyequally at all other times, and indeed ‘‘red-green’’ environmentalism even outreached‘‘green’’ views in 1992 because of concernsabout pollution, especially at the Mae Mohpower plant, and in 1994 because of the north-eastern evictions. These results suggest that—with some explicit exceptions—environmental-ism in Thailand has reflected both concernsabout ecological damage alone, as well as theimpacts of degradation on people.

It is not clear, however, how far these reportsof ‘‘green’’ and ‘‘red-green’’ environmentalismindicate changes in the underlying definitionsof ‘‘environment.’’ For example, the mostovertly ‘‘green’’ story reported was the opposi-tion to the filming of The Beach in two nationalparks (1998–99), which was claimed to be envi-ronmentally damaging. The evictions of villag-ers from plantation areas in northeast Thailand(mainly 1994–95) were highly ‘‘red-green’’ be-

cause they portrayed social injustice. But didthese incidents accurately reflect the levels ofenvironmental threats, or use existing conceptsof environmentalism to gain political attention?In the case of opposition to The Beach, activistswere worried about the apparent corruption ofthe state in apparently privileging a foreign filmcompany to enter and change the landscape inparks. Activists deliberately used language ofecological risk to gain legitimacy. Newspaperreports (in the Bangkok Post at least) reflectedthis language.

Similarly, the environmental concern shownfor industrial pollution in cases such as MaeMoh power plant or poisoning of workers inthe northern region industrial estate (in themid-1990s) showed a new attention onto thehealth risks faced by factory workers and citi-zens, and the inadequacy of the state inaddressing them. But the opposition to thecoal-fired power station in the coastal site ofPrachuab Khiri Khan (1998–2000) mainly de-fined environmental damage in terms of im-pacts on a local coral reef (i.e., a ‘‘green’’environmental concern), rather than on healthrisks to people. Similarly, in disputes concern-ing forest evictions, some reports were clearly‘‘red-green’’ because they focused on the injus-tices of eucalyptus and pine plantations for thelivelihoods and rights of poor villagers. But


reports still maintained concern for ecologicalfragility, and newspapers would carry simulta-neous ‘‘green’’ reports about deforestation else-where in Thailand. Consequently, many reportswould use pre-existing environmental defini-tions or norms to discuss social justice, ratherthan engage in redefining pre-existing defini-tions of degradation.

More generally, however, these disputes havenonetheless been portrayed as cases of ‘‘democ-ratization.’’ Figure 4 demonstrates that news re-ports since the 1960s have increasingly adoptedthe theme of ‘‘democratization’’ while focusingincreasingly less on ‘‘modernization.’’ Thesetrends reflect the decline in state censorshipand increasing level of open political debate inThailand, and the tendency for early environ-mental reporting to emphasize technologicaland economic benefits rather than social con-cerns. For example, many activists claim thatthe Chiang Mai cable car dispute and Phukettantalum mining controversy were instrumentalin solidifying environmentalism and civil societyin Thailand, yet, the Bangkok Post largely re-ported them both as ‘‘modernization.’’ Thissuggests that newspaper reporting and popularhistories held by NGO workers are different inthese cases. Reports concerning the Nam Choandam in the early 1980s, however, did notemphasize modernization as highly.

Consequently, ‘‘green’’ and ‘‘red-green’’ envi-ronmentalism have remained largely equalthemes in environmental reporting, but thetheme of environmentalism-as-democratizationhas increased over time. These trends are consis-tent with political histories of Thailand, whichlink environmentalism with resistance againstmilitary or unaccountable governments (Hirsch,1997a, 1997b). Yet, it is not clear if this focus ondemocratization has allowed greater diversity inhow environmental risks are portrayed, orwhether journalists use the theme of environ-mental risks as a means to demonstrate the lackof state accountability. Much reporting on pol-lution has brought attention to the risks facedby poorer people and their angry responses.But, as shown by opposition to The Beach andthe Prachuab Khiri Khan coral reef, the themeof ecological fragility may still take precedencein newspaper reports over more livelihoods-ori-ented approaches to environmental risk.

(c) Tests of association

Pearson’s ‘‘r’’ coefficients were calculated tomeasure the association of middle and lower

classes with the overview themes of ‘‘green,’’‘‘red-green,’’ environmentalism and ‘‘democra-tization.’’ (‘‘Modernization’’ is not assessed inthis paper for reasons of space.) To make thesetests clearer, episodes of environmentalismwere collated into groups concerning forestsor pollution. 11 The aim of this analysis wasto contribute to general debates about the roleof social class in determining different kinds ofenvironmentalism, and whether there was anyassociation between different social composi-tions of environmentalism and popular beliefsabout ‘‘democratization.’’

As discussed above, charts use accumulated‘‘r’’ coefficients (i.e., values were added year-on-year). The strongest long-term associationsare indicated by rapidly ascending lines. Figure5 shows there is a high, on-going, association ofmiddle classes with ‘‘green’’ environmentalismsince the 1980s. The next highest influence isin lower-class ‘‘red-green’’ environmentalism,which has risen sharply since the 1990s, fol-lowed by middle-class ‘‘red-green’’ environ-mentalism. The weakest overall association isbetween lower classes and ‘‘green’’ environmen-talism, although this rises slightly in the late1990s.

These results reinforce the expected trendthat middle-class activism has been strongestin ‘‘green’’ issues. This was especially apparentduring the Nam Choan dispute of the early1980s, but has always been strong. However,two other periods can be identified: during1988–94, there is a strong correlation betweenmiddle- and lower-class ‘‘red-green’’ environ-mentalism; and during 1994–2000, there is aweakening in ‘‘red-green’’ environmentalism,and a relative strengthening in ‘‘green’’ envi-ronmentalism from both lower and middle clas-ses. These later two periods seem to representpublic concern about strong-state policies dur-ing the return of the military government1991–92, and its legacy in resettling villages inthe northeast; and the later debate about com-munity forestry from the mid-1990s. The infor-mation in Figure 5 suggests that resettlementpolices in the early 1990s were considered asdamaging to people, but that discussions ofcommunity forestry have generally been linkedto a more ‘‘green’’ agenda. This latter stage iscontroversial because activists have arguedwhether community forestry is a class issue,and whether poorer villagers agree or disagreewith proposals to limit devolved governanceor certain land uses (Jintana & Routray,1998). The trends in Figure 5 cannot confirm

0

1

2

3

4

5

6

7

8

9

10

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

Middle Class-GreenMiddle Class-Red GreenLower Class-GreenLower Class-Red Green

‘R’ for total (yearly) data:Middle class: Green = 0.90 Lower class: Green = 0.78 Middle class: Red-Green = 0.78 Lower class: Red-Green = 0.83

Figure 5. Forests activism: association of social class and varieties of environmentalism (accumulated Pearson’s ‘‘r’’).


whether such views were indeed voiced by poorvillagers, or spoken on their behalf by NGOs orjournalists. But this research can indicate thatthe Bangkok Post did show an increased associ-ation of lower-class actors and ‘‘green’’ think-ing in the late 1990s.

Pearson’s ‘‘r’’ coefficients were also calcu-lated for yearly data of activism for forests,and revealed very strong associations of 0.9for middle-class actors and 0.78 for lower classand ‘‘green’’ thinking; and 0.78 for middle-classand 0.83 for lower class and ‘‘red-green’’ think-ing. Such statistics should obviously be treatedwith caution for inferring about general rela-tionships concerning social classes and environ-mentalism. But they do support previousassumptions that the ‘‘green’’ environmentalagenda is more likely to be associated with mid-dle classes, and ‘‘red-green’’ environmentalismwith lower classes.

Figure 6 charts the relationships between so-cial class and different types of environmental-ism in stories concerning pollution. Here, theresults are different. Firstly, there is no one,clearly influential, social class or type of envi-

ronmentalism over the entire period. Yet, it isclear that the perception of pollution has chan-ged from largely ‘‘green’’ by both middle andlower classes during 1984–96, toward a moredominant ‘‘red-green’’ perception by lowerclasses since the early 1990s. The weakest ofall environmental agendas for pollution hasbeen middle-class ‘‘red-green’’ thinking. Thesetrends again suggest that middle classes weremore concerned at the impacts of pollutionon ecology generally, rather than on poorerpeople. Yet, the results also suggest a move-ment toward democratization within environ-mentalism, because the lower classes arebecoming more prominent in defining debatesabout pollution, and using ‘‘red-green’’ termsto do so.

‘‘R’’ coefficients were calculated for yearlydata for pollution activism, and revealed muchweaker associations than with forests. Associa-tions with ‘‘green’’ thinking were 0.46 for mid-dle classes and 0.44 for lower classes.Associations with ‘‘red-green’’ thinking were0.47 for middle classes and 0.61 for lower clas-ses. These results suggest that—despite the

0

1

2

3

4

5

6

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

Middle Class-Green

Middle Class-Red Green

Lower Class-Green

Lower Class-Red Green

‘R’ for total (yearly) data:Middle class: Green = 0.46 Lower class: Green = 0.44 Middle class: Red-Green = 0.47 Lower class: Red-Green = 0.61

Figure 6. Pollution activism: association of social class and varieties of environmentalism (accumulated Pearson’s ‘‘r’’).


trends observed above—the overall associa-tions of environmentalism and pollution arestill relatively weak, and that most overtly‘‘environmental’’ discussions in Thailand areassociated with forests rather than pollution.

Finally, Figure 7 charts the relationships be-tween the reporting of different social classes,and the presence of ‘‘democratization’’ as apolitical theme of news reports in storiesinvolving forests and pollution. Here, the mostobvious trend is that ‘‘democratization’’ ishighly associated with middle-class activismconcerning forest issues since the late 1980s.This is consistent with other discussions ofenvironmentalism in Thailand that suggest thecampaigns against the Nam Choan dam andlogging ban in the 1980s became symbolic ofwider campaigns for democracy. The associa-tion between middle-class activism and democ-ratization, however, became weaker in the early1990s. By the late 1990s, lower-class activismwas increasingly associated with democratiza-tion concerning both forests and pollution.This last trend is all the more noteworthy,as—before the 1990s—democratization and

lower-class forests activism were very weaklyassociated.

Why have lower-class forest activists becomeincreasingly linked with democratization? Inthe past, newspapers often portrayed forests-based activism by poor people as law breakingor ecologically destructive (for example, byburning pine and eucalyptus plantations innortheastern Thailand, or by performing shift-ing cultivation in watershed zones). Evidence,however, does not suggest that these activitiesare now viewed in different ways. One explana-tion might be the growth in public activism fol-lowing the 1997 Constitution, which invitedpublic participation in debates about naturalresource management. More controversially,however, this finding might also indicate the de-sire of journalists or activists to influence thedebate about community forestry legislationby adopting the language of ‘‘green’’ environ-mentalism and ‘‘democratization.’’ Figure 5shows that ‘‘green’’ environmental activismfrom lower classes has increased during the1990s as well as the presentation of this activityas ‘‘democratization.’’ In principle, these trends

0

2

4

6

8

10

12

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

Forests: Middle ClassForests: Lower ClassPollution: Middle ClassPollution: Lower Class

‘R’ for total (yearly) data:Middle class: Forests = 0.71 Lower class: Forests = 0.87 Middle class: Pollution = 0.91 Lower class: Pollution = 0.96

Figure 7. Association of forests and pollution activism on perceptions of democratization (accumulated Pearson’s ‘‘r’’).


could indicate a growing conversion of agricul-tural smallholders to the ‘‘green’’ environmen-tal agenda; the desire of these people toappear ‘‘green’’ in order to win support fromother NGOs and the state; or the wish of otheractivists and journalists to portray poor forestusers in these terms. (Or, indeed, a mixture ofthese factors.) Some analysts have argued thatthe debate about community forestry has delib-erately portrayed some poorer minorities asenvironmentally friendly in order to win politi-cal support from both ‘‘green’’ and ‘‘red-green’’lobbyists (Walker, 2001). The observed rise in‘‘green’’ environmentalism and ‘‘democratiza-tion’’ among lower classes might provide fur-ther evidence for this claim.

The association of ‘‘democratization’’ andenvironmentalism was also calculated usingyearly data. The ‘‘r’’ coefficients of ‘‘democrati-zation’’ with forests activism were relativelystrong, at 0.71 for middle-class activism and0.87 for lower classes. But these associationswere extremely high for stories concerning pol-

lution, with 0.91 for middle classes and 0.96 forlower classes. These results indicate that allforms of environmentalism are highly associ-ated with ‘‘democratization,’’ and that theinvolvement of lower classes (even in the‘‘green’’ agenda) is especially considereddemocratizing. It is also worth noting that theassociations between social classes and‘‘democratization’’ are higher than those be-tween social classes and ‘‘green’’ or ‘‘red-green’’environmentalism. These findings indicate thatenvironmentalism may be considered democra-tizing in its own right, even if poor people donot necessarily participate in redefining envi-ronmental problems.

7. CONCLUSIONS

This paper has sought to investigate thedemocratizing impact of environmental socialmovements in developing countries by present-ing a new methodology based on analyzing the


historic representation of political activism innewspapers. Nearly 5,000 newspaper reportsfrom Thailand during 1968–2000 were studiedto identify the social composition and politicalcontent of environmentalism, and whether thepresence of social class was linked to ‘‘green’’(conservationist) or ‘‘red-green’’ (livelihoods-oriented) varieties of environmentalism, andwhether such acts were portrayed as ‘‘democra-tization.’’ This methodology presents a poten-tially useful way to analyze both the short-term membership and subject matter of socialmovements, and the long-term symbolismattributed to them.

Many discussions of environmentalism anddemocratization in developing countries havesuggested that environmentalism tends to bedominated by the middle classes, and that envi-ronmental norms of social movements tend toreflect the ‘‘green’’ rather than ‘‘red-green’’agenda. This paper has generally confirmedthat environmentalism in Thailand has beendominated by middle-class ‘‘green’’ activism,especially concerning forests. Yet, the presenceof lower-class activists has increased signifi-cantly since the 1990s, and environmentalismhas increasingly considered ‘‘red-green’’ topicstoo. Moreover, at the surface level, newspapershave linked environmentalism with ‘‘democra-tization,’’ or the criticism of the state and beliefin community action. Environmentalism is veryhighly associated with popular concepts of‘‘democratization,’’ especially involving lower-class activism, and particularly concerning pol-lution.

It would therefore appear that environmen-talism in Thailand has been a democratizingforce, both by forming an arena to criticizethe state, and by becoming more diverse andsocially inclusive itself (Hirsch, 1997a, 1997b).But this paper disagrees that environmentalnorms associated with environmentalism areuniversal, or shared by different classes (Fahn,2003). Some approaches to community for-estry, for example, include methods such aslarge-scale forest plantations or the resettle-ment of villages that impact negatively on rurallivelihoods, especially among upland minoritiesin northern Thailand. Various ‘‘red-green’’

activists and campaigners have opposed theseapproaches to environmental management,but not all newspaper reports have consideredthis kind of activism to be democratizing orenvironmental. It is also not clear if theobserved rise in lower-class ‘‘green’’ environ-mentalism concerning all forests since the late1990s reflects an innate ‘‘green’’ agenda frompoorer people; the imposition of this viewpointon them by journalists as a means to gainlegitimacy for conservationist approaches tocommunity forestry; or the tactical adoptionof a ‘‘green’’ agenda (or ‘‘storyline’’) byvillagers as a way to gain political visibility.Further methods are necessary to clarify thistrend.

Consequently, this paper raises some impor-tant questions for the relationship of environ-mental social movements and democratizationin developing countries. At one level, it is clearthat environmentalism has attracted variouspolitical criticisms of the state, and that news-papers have considered lower classes to play alegitimate role in democratization. But at an-other level, the environmental norms associatedwith environmentalism may actually be morediverse and socially divisive than commonlyportrayed, and yet newspapers have frequentlydisplayed any form of environmentalism asdemocratizing. In other words, environmental-ism may sometimes be portrayed as an ‘‘envi-ronmentalism of the poor,’’ but it is notalways clear if the poor have defined what itmeans to be environmental. For these reasons,environmentalism—in Thailand at least—can-not be seen as a classic new social movementin which activists claim new social identitiesthat represent all social classes. Rather, a moreeffective assessment of environmentalism anddemocratization needs to look at the evolutionof environmental norms as well as the apparentparticipation in any form of activism. There is aneed to look more closely at the social divisionsimplied by ‘‘green’’ versus ‘‘red-green’’ environ-mentalism; how, why, and by whom differentactivists or reporters link political activism toenvironmental norms; and how far differentenvironmental norms empower or disempowervarious political objectives.

NOTES

1. This statement is commonly attributed to PhilipGraham, the publisher of the Washington Post 1946–63.

2. These varieties of environmentalism, of course,reflect other theoretical approaches to environment of


ecocentrism, or deep green environmentalism (thatemphasizes ecological fragility and the existence of a‘‘nature’’ external to human influence); and technocen-tric, or anthropocentric environmentalism (that accen-tuates the human experience of environmental change,and adaptation to ecological limits). Debates are dividedon how far these matters are class related. Feministscholars have also argued that gender is an importantinfluence on the types of environmentalism.

3. The Chipko Andolan (Embrace the Tree) movementin the present Himalayan state of Uttaranchal in Indiastarted in 1973 when villagers spontaneously demon-strated against the logging of Ash trees by a sports goodscompany. Over time, the activity grew into a resistancemovement against the contract system of logging wherevillage people would embrace trees (inspiring the term,‘‘tree hugging’’). Many key activists wanted to protectlocal sawmill cooperatives. But in 1977, a section of theChipko activists decided to oppose all tree-fellingactivities in the region, and attracted attention frominternational environmental NGOs. Others have alsoalleged that Chipko became absorbed into widerregional politics of defining Uttaranchal province (Ran-gan, 2000).

4. Newspapers, and newspaper articles, vary greatly.Serious broadsheets may provide the most detailed newsreports. International wire-news agencies (such as Reu-ters or Associated Press) may also serve the same service,but be written in a style that is communicable tointernational readers. Both broadsheets and wire agen-cies also produce news ‘‘features,’’ which provide morein-depth discussions or illustrations than straight newsreports, and occasionally opinion pieces (‘‘op-eds’’ oreditorials), which provide one writer’s personal view.

5. ‘‘Women’s movements’’ was a file kept at theBangkok Post library. This was later merged with the

campaign for a 90-day maternity leave. The total yearsfor women’s movements recorded was 1988–96.

6. Some features or opinion pieces are written byfarang, or western journalists. Most international news iswritten by international wire news agencies.

7. Thailand had short experiments with universal adultsuffrage during the 1970s, but this principle is generallyconsidered to have been adopted on a permanent basisin 1988. However, further military governments oc-curred during 1991–92, and in September 2006.

8. Thai Rath and Matichon.

9. Various people were interviewed, but I wouldespecially like to thank Chayan Vadhanaputhi, JiUngkaporn. Kamol Sukin, Noi Thammasatien, PasukPongpaichit, Prawase Wasi, Sanitsuda Ekichai, SuparaJanfitcha, Surapon Duangkhae, Ubonrath Siriyuvasak,and Wasant Techawongtham.

10. At this time, there were generally few environmen-tal NGOs, and these tended to focus on the concerns ofelites about the protection of national architectural ornatural heritage (such as the Society for the Conserva-tion of National Treasures and Environment establishedby Boonsong Legakul).

11. The only major environmental conflict excludedfrom this classification was the Pak Mul dam, whichfocused on the ethics of building a dam in northeasternThailand, with associated displacement of fishing com-munities. Mangrove forests were also not specificallyincluded in ‘‘forests’’ because of the decision to avoiddebates about aquaculture—although many generaldiscussions of forests included mangroves, and the totallevel of reporting about mangroves was small.

REFERENCES

Bauer, M. (2000). Classical content analysis: a review. InM. Bauer, & G. Gaskell (Eds.). Qualitative research-ing with text, image and sound (pp. 131–151). Lon-don: Sage.

Buergin, R., & Kessler, C. (2000). Intrusions andexclusions: democratization in Thailand in the con-text of environmental discourses and resource con-flicts. GeoJournal, 52, 71–80.

Cohen, J., & Arato, A. (1992). Civil society and politicaltheory. Cambridge, MA: MIT Press.

Covey, J. (1995). Accountability and effectiveness inNGO policy alliances. In M. Edwards, & D. Hulme(Eds.), Non-governmental organizations—perfor-mance and accountability: Beyond the magic bullet.London: Earthscan.

D’Anieri, P., Ernst, C., & Kier, E. (1990). New socialmovements in historical perspective. ComparativePolitics, 22, 445–458.

Della Porta, D., & Diani, M. (1999). Social movements:An introduction. Oxford: Blackwell.

Diani, M., & Eyerman, R. (Eds.) (1992). Studyingcollective action. Sage modern politics series (Vol.30). London: Sage.

Donati, P. (1992). Political discourse analysis. In M.Diani, & R. Eyerman (Eds.), Studying collectiveaction. Sage modern politics series (Vol. 30,pp. 136–167). London: Sage.

Eder, K. (1993). The new politics of class: Socialmovements and cultural dynamics in advancedsocieties. London: Sage.


Escobar, A. (1996). Constructing nature: elements for apoststructural political ecology. In R. Peet, & M. C.Watts (Eds.). Liberation ecologies: Environment,development, and social movements (pp. 46–68). Lon-don: Routledge.

Eyerman, R., & Jamison, A. (1991). Social movements:A cognitive approach. Cambridge: Polity.

Fahn, J. (2003). A land on fire: The environmentalconsequences of the Southeast Asian boom. Boulder:Westview.

Flaherty, M., Vandergeest, P., & Miller, P. (1999). Ricepaddy or shrimp pond: Tough decisions in ruralThailand. World Development, 27(12), 2045–2060.

Forsyth, T. (2004). Industrial pollution and socialmovements in Thailand. In R. Peet, & M. Watts(Eds.), Liberation ecologies: Environment, develop-ment, and social movements (2nd ed., pp. 422–438).London: Routledge.

Forsyth, T., & Walker, A. (2008). Forest guardians,forest destroyers: The politics of environmental knowl-edge in northern Thailand. Seattle: University ofWashington Press.

Guha, R., & Martinez-Alier, J. (1997). Varieties ofenvironmentalism: Essays north and south. London:Earthscan.

Hajer, M. (1995). The politics of environmental discourse.Oxford: Oxford University Press.

Hirsch, P. (1997a). The politics of environment: oppo-sition and legitimacy. In K. Hewison (Ed.), Politicalchange in Thailand: Democracy and participation(pp. 179–195). London: Routledge.

Hirsch, P. (Ed.) (1997b). Seeing forests for trees:Environment and environmentalism in Thailand.Chiang Mai: University of Washington Press.

Ibarra, P. (Ed.) (2002). Social movements and democracy.Basingstoke: Palgrave Macmillan.

Jensen, K., & Jankowski, N. (Eds.) (1991). A handbookof qualitative methodologies for mass communicationresearch. London: Routledge.

Jintana, Amornsanguansin, & Routray, J. (1998). Plan-ning and development strategy for effective manage-ment of community forestry: lessons from theThai experience. Natural Resources Forum, 22(4),279–292.

Johnson, C., & Forsyth, T. (2002). In the eyes of thestate: negotiating a ‘‘rights-based approach’’ to forestconservation in Thailand. World Development, 30(9),1591–1605.

Kim, S. (2000). Democratization and environmentalism:South Korea and Taiwan in comparative perspective.Journal of Asian and African Studies, 35(3), 287–302.

Lohmann, L. (1995). No rules of engagement: interestgroups, centralisation and the creative politics of‘‘environment’’ in Thailand. In J. Rigg (Ed.),Counting the costs: Economic growth and environ-mental change in Thailand (pp. 211–234). Singapore:ISEAS.

Maheu, L. (1995). Social movements and social class: Thefuture of collective action. London: Sage.

Martinez-Alier, J. (2002). The environmentalism of thepoor: A study of ecological conflicts and valuation.London: Elgar.

Mawdsley, E. (2004). India’s middle classes and theenvironment. Development and Change, 35(1),79–103.

McCargo, D. (2000). Politics and the press in Thailand:Media machinations. London: Routledge.

McCoy, P. (Ed.) (2001). Political opportunities, socialmovements and democratization. Oxford: Pergamon.

Nash, F. (1982). Wilderness and the American mind (3rdrevised ed.). New Haven: Yale University Press.

Neumann, R. (1998). Imposing wilderness: Struggles overlivelihood and nature preservation in Africa. Berkeley:University of California Press.

Offe, C. (1985). New social movements: challenging theboundaries of institutional politics. Social Research,52(4), 817–868.

Peet, R., & Watts, M. (Eds.) (1996). Liberation ecologies:Environment, development, social movements. Lon-don: Routledge.

Pellow, D. (2002). Garbage wars: The struggle forenvironmental justice in Chicago. Cambridge, MA:MIT Press.

Peritore, N. (1999). Third World environmentalism.Miami: University of Florida Press.

Somchai, Phatharathananunth (2006). Civil society anddemocratization: Social movements in NortheastThailand. Singapore: NIAS.

Porter, D. (1992). Life histories in the analysis of socialmovement activists. In M. Diani, & R. Eyerman(Eds.), Studying collective action. Sage modern poli-tics series (Vol. 30, pp. 168–193). London: Sage.

Rangan, H. (2000). Of myths and movements: RewritingChipko into Himalayan history. London: Verso.

Silverman, D. (1993). Interpreting qualitative data:Methods for analysing talk, text and interaction.London: Sage.

Tarrow, S. (1994). Power in movement: Social move-ments, collective action and politics. New York andCambridge: Cambridge University press.

Taylor, B. (Ed.) (1995). Ecological resistance movements:The global emergence of radical and popular environ-mentalism. New York: State University Press.

Tilly, C. (1994). Social movements as historically specificclusters of political performances. Berkeley Journalof Sociology, 38, 1–30.

Touraine, A. (1981). The voice and the eye. Cambridge:Cambridge University Press.

Ubonrath S. (1991). The environment and popular culturein Thailand. Southeast Asian Affairs (pp. 298–308).

Vandergeest, P. (1996). Mapping nature: Territorializa-tion of forest rights in Thailand. Society and NaturalResources, 9, 159–175.

Vandergeest, P. (2003). Racialization and citizenship inThai forest politics. Society and Natural Resources,16(1), 19–37.

van Dijk, T. (1991). The interdisciplinary study of newsas discourse. In K. Jensen, & N. Jankowski (Eds.). Ahandbook of qualitative methodologies for mass com-munication research (pp. 108–120). London: Routl-edge.

Wakefield, S., & Elliott, S. (2003). Constructing thenews: The role of local newspapers in environmentalrisk communication. The Professional Geographer,55(2), 216–222.


Walker, A. (2001). The ‘‘Karen Consensus,’’ ethnicpolitics and resource-use legitimacy in northernThailand. Asian Ethnicity, 2(2), 145–162.

Wignaraja, P. (Ed.), (1993). New social movements in theSouth: empowering the people. Zed: London.

Available online at www.sciencedirect.com

Documents

Are Environmental Social Movements Socially …personal.lse.ac.uk/FORSYTHT/Forsyth_Thailand_social...Are Environmental Social Movements Socially Exclusive? An Historical Study from