Primate cities in the Third World are by now almost reflexively associated with the image of enormous environmental problems. Symptomatic of this association, the question of whether or not recent patterns of Third World urban growth are sustainable has attracted much scholarly attention in recent years (Nijkamp, 1990; Hahn and Simonis, 1991; Lowe, 1991; MacNeill et. al., 1991; Blowers, 1992; Olpadwala and Goldsmith, 1992; Yanarella and Levine, 1992; Söderström, 1993; Atkinson, 1994; Gilbert, 1994; Drakakis-Smith, 1995, 1996, 1997). While no definitive conclusions emerge from this literature, there is clearly a broad consensus that environmental problems are one of the most pressing issues connected with urban growth and will continue to be so into the future.

The question of whether a given pattern of urban growth is sustainable is, however, a vexed one--perhaps most fundamentally because there is no solid agreement on precisely how to define the notion of sustainability (Sneddon, 1997). In this paper, we will not rectify this problem, but we will of necessity work with a very basic conception of sustainability as a backdrop to our analysis. Following David Drakakis-Smith, we define sustainability in an urban context as a development pattern which (1) satisfies the requirement for equity, social justice, and human rights, (2) meets basic human needs, (3) allows social and ethnic self-determination, (4) promotes environmental awareness and integrity, and (5) promotes awareness of inter-linkages between various living beings across both space and time (Drakakis-Smith, 1995: 664). We identify sustainability in these terms in order to emphasize that sustainable outcomes necessarily require fulfilling social and biophysical demands simultaneously.

Our purpose, here, however, is not to defend a particular notion of sustainability but to analyze how one particular approach to the problem of sustainable Third World urbanization--namely, the push to decentralize industrial growth and to develop secondary urban centers in order to alleviate congestion and pollution in primate cities--has been deployed in the context of urban primacy and uneven development in Thailand. The Thai case is interesting to examine for a wide variety of reasons, including the fact that the country's economy has, until very recently, been one of the World Bank's star performers (World Bank, 1993). We emphasize that the Thai case is especially instructive because it illustrates the conundrums which attend attempts to promote greater urban sustainability through approaches which refuse to seriously discipline capital and instead rely narrowly on incentives for investors. Our suggestion is that such approaches would appear--insofar as the Thai case is representative--to fundamentally give away the game at the outset, by privileging growth and maintenance of class privilege over other socio-economic and ecological objectives.

Some authors have quipped that Thailand is less an industrializing country than Bangkok is an industrializing city (Hussey, 1993: 19). Such a view is backed to a certain extent by the statistics on Bangkok's primacy. (Tables 1 and 2.) In 1960, for example, Bangkok contained 8.1 percent of the country's inhabitants and 23.8 of its Gross Domestic Product (GDP); by 1996 it had increased its share to 11.8 percent of the population and 39 percent of GDP. Even more tellingly, Bangkok's share of national manufacturing product stood at 46 percent in 1981 and declined only slightly to 39 percent in 1996, while the share of the greater Bangkok Metropolitan Region (BMR) decreased from an overwhelming 72 percent to a still overwhelming 62 percent. Most of the limited decline in the BMR's manufacturing primacy is attributable to a decentralization of some manufacturing activity to the nearby provinces of the surrounding Central region--and if one counts the BMR as part of that region, its share of national manufacturing output remained steady at approximately 90 between 1981 and 1996. Indeed, the Central region including the BMR garnered over 70 percent of the GDP in 1996 while the Southern, Northern, and Northeastern regions, containing two-thirds of the population, garnered under 30 percent (NSO, various years).

Such a regionally skewed distribution of economic activities inevitably contributes to Thailand's high and rising national income disparities (Suganya and Somchai, 1988; Medhi, 1996; Kakwani and Medhi, 1997). It also creates enormous problems of urban pollution and congestion in Bangkok. For example, Bangkok's notoriously inadequate urban road network is hopelessly clogged by some 2 million cars (approximately 50 percent of the national total), which move at an average pace of 1 to 2 kilometers per hour during the peak of rush hour and contribute 70 percent of Bangkok's air pollution--including levels of suspended particulate matter which are frequently between 200 and 400 times the recommended levels of the World Health Organization (WHO) and carbon monoxide levels which are 50 percent higher than WHO standards (Pendakur, 1993: 56). In addition, inadequate transportation infrastructure has contributed to an enormous rise in motor vehicle fatalities, which have climbed from 3.0 per 100,000 nation-wide in 1957 to 22.8 in 1994, with Bangkok's rate rising from 14.5 to 37.2 between 1980 and 1994 alone (Ministry of Public Health, various years).

We will briefly discuss each of these policy approaches in turn, but we begin by noting the overarching role of the National Economic and Social Development Board (NESDB) in promoting provincial industry through its five-year development plans. The NESDB has given lip service for many years to the importance of improving rural incomes and decreasing the disparities between Bangkok and the rest of the country. Decentralization has been argued for on the grounds that it can relieve Bangkok's congestion and pollution while reducing the income gap between the capital city and the rest of the country (Medhi et. al. 1992: 216-17; Parnwell, 1992: 53; Parnwell and Suranart, 1996: 161).¹ Environmental benefits have been expected to be indirect, in that movement of manufacturing investment to the periphery will not by itself cure Bangkok's congestion and pollution problems but will, rather, slow the rates of growth in Bangkok and thus buy the state more time for implementing environmental clean-up policies. The Third National Economic and Social Development Plan (1972-1976) called for urban-industrial decentralization through encouragement of industrial relocation. The plan designated nine provincial growth poles, including Chiang Mai and Khon Kaen, which were to serve as the loci of the strategy. Implementation began under the Fourth Plan (1977-81) (Medhi et. al., 1992: 217; Parnwell, 1992: 55).

It can be argued that NESDB plans may have less effect than advertised--even on government policies, let alone on capitalist investment behavior--because the most the NESDB can do is to mandate certain priorities on the part of ministries. It is not clear that it can enforce performance standards.² Perhaps more significantly, however, the nature of the plans the NESDB develops is heavily influenced not only by the training of its officials but by the nature of the groups with whom it confers and collaborates in developing the plans. For the most part, this has meant that planning is driven by the needs of major investment groups who can offer the NESDB something in return for its own promises of infrastructure expenditure.³ As Michael Parnwell (1992: 56) puts it, planning has been growth-first in orientation and has relied on "the same kinds of urban-industrial centered, top-down strategies which were responsible for enhancing the uneven pattern of development in the first place..."

Whatever the designs, it is clear that throughout the years of BoI operation, promotional privileges have in fact done little to actually attract investment away from Bangkok. Indeed, projects started under BoI promotion have themselves been highly concentrated in the BMA, and only very recently have larger numbers begun to appear in outlying regions (Map 3.) It is clear that the promise of tax reductions, which is the main device available to BoI, is not adequate to compensate for Bangkok's advantages in infrastructure and other necessities. Indeed, as one of the authors found in interviews, even investors who have located up-country do not necessarily list BoI incentives as a primary factor in their location decisions, perhaps because the tax burdens on non-promoted projects are not so high as to make gaining promotional privileges a sufficient reason to forego Bangkok's agglomeration and urbanization economies.⁴

A further word about BoI's strategies is in order here. BoI is in many respects even more capital-friendly than the NESDB and, unlike the latter organization, has no formal contacts whatsoever with any groups besides investors and other state agencies.⁵ Thus, its policies are driven to a large extent by what investors want, even though it may not always be positioned to provide this. Thus, with the BoI as with the NESDB, it is not easy to discern how much the limited decentralization of its investment promotion has to do with forces outside its control, how much with its own limited mandate, and how much with its orientation towards promoting the biggest and the most powerful. Whatever the case, it is worth noting that BoI has promoted up-country investment more aggressively in the 1990s. However, this more active promotion coincides with a push by greater numbers of manufacturing industries to reduce rising production costs in Bangkok and thus can be seen in many ways as following more than leading investment.

Given the limitations on other tools for promoting decentralization, perhaps the most important state measures have had to do with the development of infrastructure. The overall effects of infrastructure development have tended to increase rather than decrease centralization, but the building of publicly-funded industrial estates and related facilities has been somewhat successful in attracting capital to regions removed from Bangkok. Industrial estates provide companies which locate there with roads, electricity, and amenities such as telephone service, garbage removal, and wastewater treatment, thus reducing production costs and the costs of compliance with environmental regulations.⁶ Some also have export processing zones where manufacturers may obtain more extensive tax exemptions and privileges. Finally, industrial estates allow foreign companies 100 percent ownership of land, something which is prohibited elsewhere in Thailand, while they also allow firms to bring in foreign technicians and experts without restriction and to take or remit foreign currency (Davidson and Ciambella, 1995: 262; IEAT, 1995/96: 39).

Up until the mid-1980s, all of the IEAT's industrial estates were located in the BMA or within the Central Region, fairly close to Bangkok.⁷ Thus, they had no effect on the development of manufacturing in the poorest and most peripheral regions. Furthermore, they have typically attracted highly capital-intensive industries with relatively low forward and backward economic linkages, so they may have had a relatively limited effect on overall economic growth even within the Central Region.⁸ Nonetheless, they have clearly contributed to rising manufacturing GDP growth rates in the extended BMA throughout the 1970s and 1980s (Maps 4 and 5). By the late 1980s, as will be discussed below, the first industrial estate established outside of Bangkok or the Central region--the Northern Region Industrial Estate (NRIE) in Lamphun--began to have similar effects in the North.

If the provision by the state of regional infrastructure is one of the major elements in promoting such manufacturing decentralization as has occurred, the role of the minimum wage policies cannot be ignored. The minimum wages of different areas for 1995 are shown in Map 6. While competitive wages are not by themselves decisive for attracting manufacturing investment to a particular place, they are nonetheless a necessary condition for attracting such investment.⁹ Thus, for example, a survey of Eastern Seaboard (ESB) manufacturers by Narathip Aramwattanano found low labor costs consistently listed by manufacturers as one of the 3 or 4 leading reasons for locating in the ESB (Narathip, 1991).¹⁰ In this regard, it is probably Thailand's general wage levels relative to other potential international production sites which is most significant, but regional differences of wages within Thailand may play some role in attracting investment to particular locations, as we will show. Of particular note, in this regard, is the fact that the minimum wage is poorly enforced--and even more poorly in up-country areas than in Bangkok. Thus, wages in up-country regions serve as some inducement to investors not only in the formal sense that minimum wages are lower but in the well-recognized informal sense that paying up-country workers less than the official minimum is easier. This reality is undergirded by the weakness of up-country unions and labor organizations, as will be discussed for the case of Chiang Mai, to which we now turn.

This section examines the process of state-promoted manufacturing growth in the Chiang Mai-Lamphun region as well as the environmental and human health consequence of such growth. We begin with a brief outline of the historical background to manufacturing development in the North. Against this background, we then present evidence that state policies promoting manufacturing growth in the region have been successful in fomenting manufacturing growth--including growth in manufacturing wages--but have done so without promoting sustainability, either in its ecological or social dimensions.

In spite of Chiang Mai's long history as an urban center and a regional trade hub--and in spite of various projects to modernize it and build it into a regional growth pole (Chakrit and Hagensick, 1973)--manufacturing in the Chiang Mai area before the 1980s remained rudimentary, overwhelmingly small-scale, and centered largely in agriculture. A 1982 survey of Chiang Mai province found that 81.1 percent of all factories, 74.3 percent of all employment, and 62.3 percent of all industrial investment was in agro-industries, which employed an average of 15 workers per factory. Agro-industries were much more likely to be spread throughout the province: 77 percent were located outside of the muang district (or municipal district), while 79 percent of non-agro-industries were located within the muang district (Wilawan, 1984: 26-27). The smaller agro-industries were not only more broadly distributed but were more likely (especially those with less than 50 workers) to procure raw materials, machinery, parts, and services from local suppliers. Smaller firms were also more likely than larger firms to sell products on the local market, while the largest firms were the most likely to sell products abroad (Wilawan, 1984: 41-59). In short, then, manufacturing development in Chiang Mai province into the early 1980s was centered around small-scale agro-industries with relatively strong forward and backward linkages to the local market, augmented by a much smaller number of non-agro-industry firms employing more capital and labor per unit but with much weaker local forward and backward linkages.

These points have important and widely-recognized implications for development planning which targets poverty and income disparity. Clearly, projects which enhance employment opportunities and growth potential in dispersed agro-industries and which build broader manufacturing projects out of this agro-industry base will have the most significant employment effects and will spread economic opportunities most equitably, both in a social and geographic sense. Larger-scale manufacturing projects which are not connected to agriculture are bound to have much less in the way of spread effects and may well contribute to greater social and geographic income disparities within the locale.¹¹ On the other hand, agro-industries are less likely to lead to rapid economic growth. Thus, continued reliance on them as leading industries within the provinces is likely to result in continued widening of the disparities in wealth between Bangkok and the rest of the country. This point has long been recognized by Thai academics and state agencies (Gray, 1990: 46). It can also be argued that smaller-scale and more agriculturally-based industries are likely to be less environmentally damaging than heavier industries, an argument which state planners have at least partially acknowledged. Yet, on the other hand, smaller and more dispersed industries are much more difficult for the state to regulate.

In the face of these conundrums, Thai state planners have attempted a multi-faceted strategy: promotion of local agro-industry, promotion of certain higher value-added light manufacturing industries, and promotion of tourism, which has significant potential in the Chiang Mai area. The importance of the tourism industry is particularly appreciated by government planners and this has shaped the kinds of manufacturing industries they choose to promote, with lighter industries favored because they are thought to be less polluting.¹² With regard to small-scale agro-industry or other small businesses, however, it appears that the state has done somewhat less in the way of promotion than what most local business leaders feel is necessary--a point also confirmed by the lending patterns mentioned earlier.¹³ In contrast, the state has devoted considerable effort to promoting larger-scale industry in the North--and particularly investment in non-agro-industries which can generate high rates of growth.

The center-piece of this strategy since the mid-1980s has been the Northern Region Industrial Estate (NRIE), located on 1,788 rai of land in Lamphun province, 23 kilometers south of Chiang Mai city and part of the broader Chiang Mai metropolitan area. The idea of the NRIE was first broached in the 1970s, construction was begun in April 1983, and the facility was completed in March 1985 (Tongnoi, 1983; NRIE, 1997). The estate's location, along the Chiang Mai-Lampang superhighway, is strategic in multiple ways (Manat et. al., 1992: 12; IEAT, 1995/96; NRIE, 1997). First, the general location of the NRIE, in a region cooler than Bangkok, makes it more attractive to electronics manufacturers, who can reduce necessary temperature control costs by locating in a less hot and humid area.¹⁴ Second the location in a specific area with a nearby urban center and a large labor force insures--or so the factories have hoped--an adequate labor supply.¹⁵ Third, the combination of being located near Chiang Mai city but outside of it allows companies to simultaneously take advantage of urban amenities, including skilled workers trained by local universities and technical schools, and the lower official minimum wages for non-urban areas. Fourth, the most significant specific advantage of being close to Chiang Mai city is the ease of access to Chiang Mai international airport, which can be reached by truck in about 30 minutes. For companies shipping goods out of the port in Bangkok, being located along the superhighway which connects with Bangkok (via an approximately 10-hour drive) is also advantageous.¹⁶ Fifth, the specific location within Lamphun province is close to required water supplies (ultimately connected to the Ping River) but is also an area of relatively sandy soils with low agricultural productivity. This led planners to believe that the losses to agricultural production from establishing the estate would be minimal and that local farmers would be willing to sell their land--the latter of which turned out to be true (Manat et. al., 1992: 10).

In spite of all these locational advantages, however, the NRIE did not originally attract strong investor interest. The estate planners had hoped that not only would electronics firms open operations in the estate but so would a number of more locally-oriented firms such as agribusinesses and ceramics manufacturers--the latter being courted because it was thought that the estate could help contain some of the pollution problems which they generate. However, in the early years of the estate there were few firms of any sort which opened projects. The smaller agribusiness firms could not afford land in the estate and in any event did not badly need the infrastructure and services which the estate provided. Likewise, the ceramics firms did not find the estate attractive, and electronics firms either did not know of it or had little incentive to invest in Lamphun instead of in Bangkok.¹⁷ Thus, by 1987, there had only been 10 units of land sold in the estate and as late as 1988 only 3 companies were actually operating (Gray, 1990: 47; Manat et. al., 1992: 12).

The economic boom of 1987-88 changed this. As land prices, wages, and other costs of production rose in Bangkok, investors began to look for alternative production sites. The issue was not so much whether or not to relocate--few firms with substantial sunk costs took this sort of option--but rather where to expand production or make new investments.¹⁸ In this context, the availability of sufficient infrastructure in Chiang Mai-Lamphun began to pay off, and by 1991 at least 86 units of land were sold (Manat et. al., 1992: 12). As of February 1997, 92 units, representing over 90 percent of all units available, had been sold, and 62 of the factories slated for these plots were in operation while another 19 were under construction (NRIE, 1997).

Of the investors in the estate, 28 had set up operations in the general industrial zone (GIZ) while 54 had set up in the export processing zone (EPZ), which has its own export clearing house to facilitate faster movement of goods through customs for firms engaging solely in export production.¹⁹ The EPZ also provides additional incentives such as exemption of import duty and value-added tax (VAT) on machinery and raw materials, exemption of export duty and VAT on all exports, and exemption of VAT on local goods used for production (IEAT, 1995/96: 38-39; NRIE, 1997). A total of 42 of the investors in both the GIZ and EPZ were receiving BoI privileges in addition to the IEAT privileges (NRIE, 1997).

The profile of the types of industries and the nationalities of investors in the NRIE reveals the ways in which Chiang Mai-Lamphun has been brought more fully into the circuits of both Bangkok-based and international capital. Of the 62 factories in operation, 16 were producing or assembling electronics components, 16 were in the agribusiness sector, 10 were involved in spare parts or machinery, 4 each were invested in jewelry and leatherware, while 3 produced clothing or wood products.²⁰ Of the 16 electronics firms, 12 were Japanese-owned or were joint ventures involving Japanese owners, while one was a joint-venture involving Hong Kong investors, one was a joint-venture involving Swiss investors, one was a joint-venture with Singaporean investors, and one was fully owned by Koreans. Of the 10 spare parts and machinery firms, 7 involved Japanese investors, and 1 each involved Swiss and Hong Kong investors. By contrast, of the 16 agribusiness firms, 8 were fully Thai-owned (though many of these were based in Bangkok), 3 were fully-owned by Taiwanese, and Japanese investors were involved in two joint-ventures with Thai investors while fully owning one operation (NRIE, 1997). In sum, foreign investors dominated the higher-technology production and assembly sectors, while Thai investors were dominant in agribusiness.

Overall, as of 1995, the NRIE was estimated to have received 19,925 million baht worth of investment and to be employing 22,000 persons (NRIE, 1997). The electronics and machinery firms were responsible for the greatest amount of invested capital and employment--probably over 90 percent of capital and 60 percent of employment.²¹ Thus, the NRIE had become a significant outpost of foreign investment in higher-technology export industries with some involvement from Thai agribusiness investors who employed less capital and labor but were still bigger and more capital-intensive than most local agribusiness firms.

The NRIE has been successful in helping spur manufacturing growth within Lamphun province and is part of a broader expansion of manufacturing in the Chiang Mai area which is illustrated in Table 3. In this sense, the state policies promoting manufacturing decentralization to Northern Thailand can be said to have been successful, even if they have not yet led to regions such as the North "catching up" with Bangkok economically. Yet if manufacturing decentralization policies are supposed to lead to more than just manufacturing growth outside of Bangkok--e.g., if they are supposed to lead to greater sustainability--then they have clearly not been successful in the Chiang Mai case. Indeed, rapid economic growth has increased economic inequality and environmental damage within the North, without substantially lessening national income disparities or environmental problems within Bangkok.

Meanwhile, increased numbers of factories and automobiles are producing rapidly rising burdens of air pollution--a problem exacerbated by Chiang Mai-Lamphun's location within a valley, surrounded in all directions by mountains. Growing automotive traffic and inadequate development of either transportation infrastructure or public transportation have rendered Chiang Mai's roadways badly congested and noxious. Measurements of air quality along major transportation routes have registered high levels of carbon monoxide, nitrous oxides, sulfur dioxide, suspended particulate matter, and lead (NESDB, 1996: sor47). In addition, the inadequacy of transportation facilities has taken an increasing toll on human life, with the incidence of traffic mortality shooting up from 10.3 per 100,000 in 1980 to 28.6 per 100,000 in 1995 (Ministry of Public Health, 1995).

We can see this by briefly examining the organization which is responsible for regulating the workplace environment, the Ministry of Labor and Social Welfare. The Ministry's Department of Labor has inadequate personnel to monitor up-country factories. For example, there are only 8 factory inspectors in the Chiang Mai Department of Labor and the head of the Department says that a minimum of 50 would be needed to adequately cover all the establishments in the province.²² The absence of labor unions in the North interacts with such weaknesses in the Thai state regulatory apparatus, since even under the best of conditions, factory inspectors cannot hope to adequately cover establishments if there are not workers reporting or protesting conditions--and in the absence of unions, such behavior by workers is unlikely.²³

Within this broad context of limited regulatory capacity, the NRIE factories have not necessarily been the worst offenders of labor rights, but their record has not been good. Beginning in 1993, a number of workers at NRIE electronics firms began to die from unexplained causes which involved headaches, inflamed stomachs, and vomiting. The exact number of workers who died is uncertain but by 1994 the total may have reached as high as 23 (Forsyth, 1994: 35; Theobald, 1995: 18).²⁴ The causes of the deaths have not been definitively determined but both local NGOs and health officials believe that they were probably the result of sustained exposure to heavy metals such as lead. This was a particularly salient problem for young women working long hours in the electronics factories, who were suffering sustained exposures to chemicals by working overtime in order to make bonus pay (Forsyth, 1994: 36).

Local NGOs reacted to the deaths with concern, and related them to environmental problems which had developed in Lamphun since the opening of the NRIE. For example, Lamphun activists had been complaining that waste and untreated water from the NRIE was being dumped into local fields and streams, causing fish kills and other problems.²⁵ Indeed, the Ministry of Industry's Pollution Control Department acknowledged that the NRIE's incinerator was too small for the volume of waste produced at the estate and had been designed for general waste, not industrial waste.²⁶ When the worker deaths occurred, the environmental and worker health concerns synergized, leading to an outpouring of public anger over the effects of the NRIE on Lamphun. Simultaneously, many frightened workers left their jobs to seek work elsewhere. A Chiang Mai doctor who had investigated some of the cases of worker illness estimated that after the deaths as many as 50 percent of the workforce at the NRIE may have quit because of health concerns and the fear that they would be next to die.²⁷

The reaction of leading institutions of the Thai state to this public outcry was defensive. For example, the head of the IEAT, Somchet Thinaphong, claimed that the deaths were caused by AIDS, a claim which could not be sustained by medical records (Forsyth, 1994: 36). Despite this defensiveness, however, the popular concern about worker deaths and waste dumping may not have gone entirely unrewarded. Both some of the factories and the Labor Department regulatory bodies claim to have taken more aggressive measures to ensure that workers wear protective gear and that tighter safety precautions are maintained. State representatives and business leaders lament that popular attention has focused on the NRIE firms when, they point out, conditions are worse and more difficult to regulate outside of the estate--a perception which is simultaneously self-serving and accurate.²⁸ Yet even so, the fear of having a bad public image--and perhaps of having greater difficulty maintaining workforce stability--seems to have had a few salutary, if limited, effects. Indeed, some of the NGO groups themselves felt by 1997-98 that the health situation at the NRIE had improved, though they continued to be concerned about longer-term health consequences of work-place chemical exposures as well as local environmental effects of the NRIE. Moreover, struggle over health at the NRIE continues, with the local NGOs and their supporters attempting to create a health center for workers at the estate, which would disseminate information and train workers in identifying health hazards.²⁹

Overall, however, the picture which emerges from the case of manufacturing development in Northern Thailand is of the expansion of advanced capitalist accumulation strategies to a region which is so far ill-prepared to regulate the activities or deal with the human and environmental consequences. One can summarize the problems by saying that while the state has shown the capacity to help bring industrial development and economic growth to Bangkok's periphery it has shown neither the capacity nor the will to regulate this development in ways that minimize its negative aspects.

One of the notable omissions in discussions of regional development and growth pole strategies applied to Khon Kaen and the Northeast region of Thailand is the lack of attention given to the ecological implications of different urban and urban-focused regional planning strategies. While previous work has addressed shortcomings in the government's agricultural production strategies aimed at the Northeast (Dixon, 1977), the politically-driven nature of regional development (Somrudee, 1991), and the flaws in regional industrialization policies and practice (Walton, 1996; Parnwell and Khanamarong, 1996), there has been little examination of the links between the Thai state's efforts to promote "regional cities" in the Northeast and the ecological impacts of these efforts on the targeted urban areas and their rural hinterlands.

In the remainder of this section, we want to draw attention to this lacuna by addressing the environmental consequences of Khon Kaen's recent growth from two angles. The first looks at a specific industrial operation, Phoenix Pulp and Paper Company, its role in industrial development in Khon Kaen, and the socio-ecological impacts accruing to the factory. The second angle focuses on environmental concerns within Khon Kaen municipality related to water. In both instances, we argue that the role of the state--and its contradictory functions as development promoter and environmental manager--has been a critical contributor to what are arguably unsustainable processes.

Until the early 1960s, the municipality of Khon Kaen was essentially a large village. Located in Northeast Thailand³⁰, which by income levels and other measures of social welfare is the country's poorest region, government planners identified Khon Kaen as a "regional city" as part of an overarching regional development initiative begun in the 1960s.³¹ Throughout this decade, the state centered development efforts (primarily road and irrigation system construction) on rural areas due to concern over subversion in the countryside, led by the Communist Party of Thailand. The emphasis was not on "long-term efforts to alleviate the poverty of the rural population," but on "short-term, piecemeal attempts to win the people's loyalty" (Somrudee , 1991: 194).³²

Throughout the 1970s and subsequent decades, Khon Kaen emerged as the centerpiece of the government's growth-pole strategy in the Northeast. As outlined in official state planning documents, the city was to be promoted as a locale suitable for industrial development which would, theoretically, deflect rural-urban migration from the Northeast to Bangkok by absorbing non-farm labor. As highlighted earlier (see pp. 4-5), the Fourth Plan (1977-1981) encouraged decentralization through regional development and laid out several financial and monetary incentives for provincial industrial development (Walton, 1996: 113-114).³³ The NESDB also incorporated the development of regional cities as an explicit goal in the Fourth Plan (Somrudee 1991: 95).

Yet regional development through the promotion of Khon Kaen as growth pole is also linked to state activities in the area of rural development. Despite the overwhelmingly agricultural character of the region, rural industrialization³⁴ has long been a primary aim of development planning in the Northeast, as revealed through a series of "regional" reports from the past three decades (CDN, 1961; Louis Berger, 1972; Biwater, 1987). These reports highlight the Northeast's potential in terms of agro-processing industries and promote an export-oriented strategy of industrial development as the most likely path toward economic growth. We turn now to a representative example of this strategy in the Northeast, the Phoenix Pulp and Paper factory, and its socio-ecological impacts.

While industrialization in Thailand during the period of the economy's rapid growth from the 1970s to the 1990s was generally founded on the manufacturing sector, industrialization in the Northeast has followed a somewhat different path. The Northeast is the least industrialized region in the country, leading one set of authors to conclude that one's "lasting impression is of a largely unindustrialised, and certainly underindustrialised, regional economy" (Parnwell and Suranart, 1996: 166). The industrial structure of the region clarifies the degree to which industry depends on the agricultural base. Over "four-fifths of industrial enterprise in the North-East consists of rice-milling, and much of the remainder is also based on the processing of agricultural and other primary produce" (Ibid.: 167). Non-resource-dependent industries in the region have been very few. Part of this can be explained by the labor market, which is tied to agriculture. There is a pronounced seasonal demand for labor due to the agricultural character of the region. In the off-season, silk-weaving, village handicrafts and similar activities pick up the slack for some laborers. However, many migrate to Bangkok and other regions in search of non-farm and farm labor opportunities (Walton 1996: 118).

John Walton (1996) cites the Phoenix Pulp and Paper Mill as a "notable exception" in the failure of the province to attract private investment despite heavy public sector investment in infrastructure and government administration and educational services (117). However, we argue that the limited "success" of the Phoenix plant as an exemplar of the type of industrial venture state regional development policy is designed to attract is drawn into question by the deleterious socio-ecological consequences of the mill.

In the late 1970s, Phoenix Pulp and Paper Company (hereafter "Phoenix") selected a site for their plant in the province of Khon Kaen on the right bank of the Nam Phong ("River Phong") approximately 40 kilometers northwest of Khon Kaen municipality. The company chose this location on the basis of the then plentiful supply of kenaf (a grass-like plant whose fiber is useful in gunny sack and pulp production) in the immediate vicinity. Other locational factors in the company's decision to construct a plant in Khon Kaen included easy access to sufficient water (a crucial input in the production process and a medium of byproduct disposal)³⁵, availability of cheap power, and availability of land for purchase (Phoenix, 1977: 277-285).

In 1992 at the height of the dry season (March-April), a series of devastating pollution incidents on the Nam Phong transfixed the nation. A molasses spill from a sugar factory left a nine kilometer slug of black ooze that moved slowly down the Phong to the Chi and Mun Rivers and eventually to the Mekong, leaving in its wake a trail of dead and dying fish eventually estimated at 500 metric tons.³⁶ In late April of the same year, Phoenix was charged with polluting the river and ordered to halt production by the Provincial Governor. Officials from the Department of Industrial Works allowed the plant to reopen in June 1992 following promises from plant management they would strive to improve the quality of the treated waste water they were releasing into Huay Chote, a small swamp adjacent to the Nam Phong that served as Phoenix's main repository of treated effluent (Kunsiri 1992).

After discovery of more dead fish (an estimated 4,000 kilograms) downstream of the Phoenix plant in May 1993, Khon Kaen officials issued another closure order. After negotiations with several state agencies including the Office of Environmental Policy and Planning (OEPP), the Department of Industrial Works (DIW), and the Pollution Control Department (PCD), Phoenix agreed to a government-conceived program to "recycle" its waste water--Project Green. The idea was rather simple: divert two-thirds to three-quarters of the plant's treated effluent away from Huay Chote swamp where it would eventually reach Nam Phong and deliver it instead via a network of pipes to the approximately 300 rai (48 has) of eucalyptus fields surrounding the factory where it could be reused for irrigating the trees.³⁷

Project Green went into effect in early 1994 and quickly engendered its own set of problems. Documents of the company show that water delivered to eucalyptus plots would be treated to meet the standards for eucalyptus plantations and would be delivered only to the fields of those farmers who signed contracts with the company.³⁸ Soon after implementation of the project, villagers living near the plant observed dead fish in their family ponds, dying trees in rice fields, and seemingly healthy rice plants with atrophied grains or no grains at all.³⁹ One villager who had signed on to the project observed: "At first we thought the company will allow us to control the water to be used in the plantation. But since the project began, the company has sent in more water than the eucalyptus needs" (quoted in Walakkamon, 1994).

In August and September of 1995, following several unfruitful efforts to negotiate directly with Phoenix, farmers from the villages near the Phoenix factory staged demonstrations at the Khon Kaen Provincial Hall. They demanded just compensation for the rice plants, trees and fish lost to Project Green (Walakkamon, 1995). Phoenix had agreed in principle to pay Bt2 million (US$80,000) to affected villagers, but seemed to be backpedaling, claiming there was no conclusive evidence that Project Green had caused the problems. Early in the project Phoenix's environmental manager admitted that "many problems had emerged" due to Project Green and that it was a "big headache" (Ploenpoch, 1994). However, Phoenix has stressed throughout the controversy that it was the government, specifically the Department of Industrial Works, who stipulated reuse of the treated waste water.⁴⁰

Throughout this period, Phoenix employed an effective strategy of deflecting the pollution charges, first, by vehemently denying responsibility for the fish kills (Pichaya, 1993) and, second, through apologists in the Thai government. The Minister of Industry queried: "If we punish them, who will want to invest here?" (Bangkok Post, 29 May 1993). Following this latest incident, several academics from Khon Kaen University (KKU), who were concerned with the pollution levels in the Nam Phong and publicly sympathetic to the concerns of residents living near Phoenix, began pushing the idea of declaring the Nam Phong a "pollution control zone," as stipulated under the recently enacted 1992 Environmental Promotion and Protection Act. Dr. Wanchai Wattanasap, KKU's president, noted the Nam Phong's "chronic pollution problems" and, in addition to declaring it a pollution control zone, argued for the creation of a river basin conservation plan (Bangkok Post, 4 June 1993).

The experiences of Khon Kaen as growth pole over the past three decades led Somrudee Nicrowattanayingyong to conclude that "no provincial town in Thailand has ever experienced the central government's urban development planning for as long a period of time as Khon Kaen" (1991: 18-19). Despite this attention, Khon Kaen has largely failed to achieve the development goals set by planners. Part of the logic of the growth pole strategy was to slow the flow of people to Bangkok and thus relieve some of the negative social and ecological impacts of large numbers of people. In fact the "town's [Khon Kaen's] experience has not mitigated, but accelerated, Northeastern rural migration to Bangkok (and to other big cities)" (Somrudee, 1991: 25).⁴¹ Furthermore, the greatest beneficiaries of Khon Kaen's growth have not been members of the rural populace who were the planners' ostensible targets, but rather urban middle class in-migrants (lecturers, bureaucrats, businesspeople) who have been able to take advantage of the service-oriented opportunities in the growing city, and, in a very real sense, Bangkok banks who directed savings from other parts of the region to Khon Kaen and were able to take advantage of the increased level of business transactions filtering through Khon Kaen (Somrudee, 1991: 249-254).

Despite the apparent failure of Khon Kaen to spur rural and industrial development in any meaningful way, its evolution into a regional city has been accompanied by most of the problems afflicting large cities throughout the Third World. The municipality itself covers 46 square kilometers with a population of 142,313 in 1996, although the actual figure may be double due to the presence of unregistered migrants. Despite being the first provincial center in Thailand to have a town plan (Sternstein, 1977: 106), the rapid growth of the city has generated several problems, particularly concerning public services such as electricity, piped water, garbage collection, and road construction and repair. However, the most serious environmental problems--both currently and in the future--revolve around water supply, waste water treatment and the community waste water drainage system. While a dilemma confronting all of Khon Kaen's residents, the environmental problems related to water affect different social groups within the city's boundary in unlike ways.

Given the enormous amount of literature on Bangkok's sustainability problems, we need do little here to substantiate the argument that the Thai state's decentralization policies have been unsuccessful as a means for reducing either social inequality or environmental problems in Bangkok. All of the problems connected with Bangkok's primacy, which we discussed earlier, have grown dramatically worse throughout the period in which the Thai state has been promoting (with at least limited success) industrial growth in up-country areas such as Chiang Mai and Khon Kaen. At the same time, as we have shown, the decentralization policies--insofar as they have had some success--have actually added new layers of concern over urban sustainability. Decentralization of economic and manufacturing growth has not satisfied the criteria of sustainability suggested by Drakakis-Smith either in Bangkok or in Thailand's major secondary cities.

We will conclude by briefly suggesting why such an outcome should by no means be surprising. Our argument is that the policies we have discussed do not in fact have sustainability goals as their major purpose--and this is less because of whatever technical or ideological orientations might be ascribed to them than because of their class basis. This matter is one which is neglected in most discussions of what manufacturing decentralization could or should do, but one which is crucial for both the structure and outcomes of actual policies.