CHAPTER THREE The Geography of Security

NATIONAL VERSUS “REGIONAL” DEVELOPMENT

In the late 1970s and early 1980s, Laos’s forestry professionals, along with their various international advisers and donors, faced numerous and diverse challenges. Working in the aftermath of two decades of war, and before that a half century of colonial rule that combined exploitation with underinvestment, would-be foresters found that they needed to address the basic question of locating the forest resource itself and derivative problems of how to bring it to market. They also faced the wider issue of a rural population that was broadly interspersed within, and often heavily dependent on, this same resource. Details varied by region, of course. But the overlap was especially acute in the landscape 100 to 200 kilometers northeast of Vientiane, in the so-called panhandle region where Laos’s width is a mere ninety miles. Here, the nascent state forestry industry had access to a relatively rich and accessible resource, but this access was encumbered by a population that contained not only traditional upland communities but also recently displaced refugees and, at various times and places, an ongoing insurgency against the new government that lasted well into the late 1980s.

In such a context forest inventory presented a special problem, although one with historical echoes. In his account of the colonial period, historian Martin Stuart-Fox notes that “every French account of Laos listed the colony’s natural resources, almost as a prospectus for potential investors.”⁸ But such lists were rarely sufficient for sustained economic activity. Although long famed for its abundant forest resources—a 1937 “economic map” of French Indochina, for instance, shows Laos widely covered by a mix of “dense” and “thin” forest and adorned regionally with teak, pine, and various nontimber forest products (benzoin, stick lac, resins, “forest oil”)⁹—the colonial underinvestment in infrastructure noted in earlier chapters also extended to the resource-inventory process. Listing resources in “prospectus” form was thus essentially a kind of boostering, advertising investment potential in a general way without doing the expensive work of precise quantification, classification, and mapping. A US economic inventory of French Indochina conducted at the end of World War II made this point by noting that “most forest statistics in Indochina” omitted Laos entirely because it did not have even a basic forest bureaucracy. While noting that Laos “certainly has the largest forest area” of the five Indochinese states, its authors could only speculate hopefully: Laos “probably has much the largest quantity of exploitable timber” in Indochina, and “it seems likely that Laos will produce more timber when its industry is fully developed than of any of the other [Indochinese] states.”¹⁰

Despite significant attention between World War II and 1975—with foreign assistance, the Royal Lao government established a forest service at the national and provincial levels, and timber exports increased significantly in the years leading up to 1970¹¹—after 1975 forest inventory remained a source of ongoing frustration. Reflecting in 1983 on six years as the head of Sweden’s forestry cooperation project with the Lao government, Reidar Persson lamented the failure to create a “reliable national picture” of Laos’s forest resource. “This is a pity,” he wrote in Forestry in Laos, a self-published report written toward the end of his tenure, “because the higher Laotian authorities have a tendency to over-estimate the forest resources.… As long as I have been working with Laos’s forestry [efforts,] the lack of knowledge about forest resources has been one of the most serious obstacles.”¹² Three years later internal evaluators of the Lao-Swedish Forestry Project (LSFP) were even harsher. Focusing on the lack of profitable forestry operations despite almost a decade of assistance, they complained about the Lao government’s bureaucratic rigidity (“It is difficult to change a decision once it is taken”), and they seemed to include the LSFP’s original Swedish designers (including Persson) in their criticism of the project’s failure to adhere to what they called “the Swedish view, namely that all background facts should be known before a decision was taken.”¹³ Such an expectation, of course, was highly unrealistic, as Persson and other development professionals (both Lao and Swedish) knew well when they launched the LSFP in 1977. Indeed, it is precisely the circumstances that threw up the opacities and operational challenges that irked the LSFP’s evaluators that make it interesting for my purposes. By offering a window into Laos’s postwar industrial forestry landscape, the LSFP affords a chance to examine the specific geographies of field operations—both in forestry and otherwise—where population management work was explored and developed at a crucial moment of Laos’s postwar transition.

Persson’s 1983 report contained a map, reconstructed below as the upper-right portion of map 3.1, that provides a good jumping-off point. It showed the locations of the nine state forest enterprises (SFEs) created in the late 1970s and early 1980s, along with a table showing the area allocated and development partner associated with each.¹⁴ In contrast with various spatially precise forest categories that would be gazetted later in the 1990s and 2000s (see ch. 5), the original SFEs had only the most approximate target locations. And of the nine, only four—the two supported by Sweden, plus two more supported, respectively, by the Soviet Union and the Asian Development Bank—had quantified allocations of area. Moreover, as Persson’s map shows, even these were defined in only the roughest of terms.

State Forest Enterprise no. 1 (hereafter SFE 1), located in the northern portion of the Lao panhandle in what is now Bolikhamxai province, was in many ways the LSFP’s flagship operation. The Swedish also worked with the Lao Department of Forestry and with SFE 3, but both of these were assistance to institutions that already existed; SFE 1, in contrast, was built largely from scratch.¹⁵ Central to the LSFP’s approach to cooperation was what is now often called learning-by-doing; at the time, this was described as a decision to prioritize collaboration over preproject planning and, once started, to undergo “continuous review and adjustment until [the project] found its final form.”¹⁶ As many aid projects still do, this quest for “final form” began with the question of location. From a few competing possibilities, Lao and Swedish officials settled, after “some months” and “certain irritations” resulting from differences of opinion,¹⁷ on the area shown in Persson’s map for SFE 1, a broad swath of forest upriver from the city of Paksan (map 3.1). Even with this location, Swedish officials were apparently reluctant to fund “such a grandiose beginning in a remote and unknown area,” referring to the Lao government’s initial request to build SFE 1’s “heavy investments (sawmill, plywood, pulpmill, etc.)” in a place called Muang Houng, located some distance into the forest interior. As a compromise, Lao authorities agreed to locate SFE 1’s main facilities in a place called Muang Mai (today the district capital Bolikhan), located still inland from Paksan but only about a third of the way to Muang Houng, and still in the lowlands of the Mekong Plain.¹⁸

In developing a new forest enterprise at Muang Mai, it became rapidly apparent that the LSFP could not be limited to forestry. In 1981 it thus began a subproject, financed under Swedish support for SFE 1 but targeting the greater landscape surrounding both SFEs 1 and 3, called the Paksan Regional Project.¹⁹ As the study that helped launch this subproject explained, “Widespread shifting cultivation made isolated forestry development planning irrelevant” and required that the LSFP take “a broader view on the development potential and the development problems of the area.”²⁰ This addition of a rural development objective to the LSFP’s plans exemplified the abovementioned process of “continuous review and adjustment” and took the LSFP squarely into the realm of upland population management.

As the authors of a baseline study for the new subproject explained, the “regional” in the new project’s title had a very specific meaning:

[An] important aspect of the Muang Mai Project from a regional point of view is [to clarify and address] to what extent it is designed to benefit the region. For reasons which largely are acceptable, the Project has progressed to a considerable degree as an enclave in the regional economy with limited linkages and benefits to the region.… This is said not as a critique of SFE 1. It is merely a fact that a limited share of SFE 1’s expenditure has gone to the region. Looking into the future it is important to recognize that the forest exploitation and processing activities will not benefit the region to any considerable extent even if they will benefit the country.²¹

Thongphachanh and Birgegard, the authors of this study, emphasized that this tension between “regional” and national development came down to competition for forestland due to the prevalence of shifting cultivation as the pillar of local food security. They noted, however, that contrary to many assumptions (both then and now), shifting cultivation in the landscapes in and around SFEs 1 and 3 was “not a mountain phenomenon and not a cultivation method practiced only by hill tribes.”²² To drive the point home, they quoted the results of a districtwide survey that had found a full third of the population fully dependent on shifting cultivation, and over two-thirds relying on a mix of shifting and fixed lowland production.²³ These numbers were conservative, moreover, since the sample had been forced to exclude, for reasons elaborated below, “the high mountain areas” where shifting cultivation was ubiquitous.

The dependence on shifting cultivation, even in the “nonmountain” areas close to Muang Mai, reflected the area’s wartime and postwar history. During the early 1970s the Mekong lowlands, including the area around Paksan, had been a major destination for refugees fleeing the fighting in northern Laos, especially in the northeast. A 1976 report by USAID, for instance, which was intimately involved in support for internally displaced people before its departure in 1975, noted that in the Paksan area alone “twenty-one separate villages were constructed” by and for displaced people “in two long areas, one running 50 kilometers west from Paksan and the other area extending 27 kilometers north. Over 1,800 hectares of land were put under rice cultivation.”²⁴ Almost a decade later, Thongphachanh and Birgegard made it clear that this significant investment in new land development had nonetheless been insufficient to support the area’s new (and still growing) population. Northward expansion from Paksan had initially followed the potential for rice-paddy development along the Nam San River, but it was expanding into the uplands.²⁵

This put the “regional” population on a collision course with SFE 1, whose foresters had identified “considerable areas with reasonably dense forest” to the east of the Nam San and to the northeast of the Sayphou Nyou, a narrow band of limestone mountains running northwest to southeast that separates the lowlands of Paksan and Muang Mai from the interior forests around Muang Houng (see map 3.1). “The exact extent of these resources remains to be determined,” Thongphachanh and Birgegard noted (echoing Persson’s point about forest inventory), “but they are large enough to permit a substantial logging operation for many years to come.”²⁶ At the same time, however, shifting cultivation was expanding in precisely the same direction, “not so much because of a preference for hai [upland rice] cultivation” by the local population, Thongphachanh and Birgegard emphasized, which was mostly ethnic (“lowland”) Lao, “but as a result of the scarcity of wet field paddy (na) land and the relative abundance of land that could be used” for upland (hai) production.

Noting with alarm that a majority of the farmers surveyed for the Paksan Regional Project reported upland fallow periods of less than five years, they predicted that shifting cultivation would continue to expand northward as “farmers will have no alternative but to look for new hai land,” first in the areas along the Nam San and then in the watersheds above.²⁷ Arguing that what was “a moderate conflict at present” was “likely to develop into a more serious one over time,”²⁸ Thongphachanh and Birgegard emphasized the key question confronting the project: how to reconcile “the forest interests represented by SFE 1 and the legitimate interests of the local population to satisfy their basic needs for food.”²⁹

AN “UNCOMMITTED” POPULATION

Efforts to reconcile or at least mitigate this conflict focused on a suite of practices that Laos’s Council of Ministers would later codify as “population management work.” Why were these developed? If enclosing and developing a strategically important resource was an urgent national priority—and by most accounts of the time, forestry was atop a very short list of development options in the Lao PDR’s first decade³⁰—why not resort to coercion and violence, the “blood and fire” that Marx famously associated with primitive accumulation?³¹ Why was it so important for foresters and associated development specialists to address the looming conflict between forestry and food security using “softer” methods rather than via what Thongphachanh and Birgegard called “law enforcement, fencing and guarding,” which they recommended strongly against?³² A partial answer no doubt comes from the ideological commitments of both of the governments involved. The Lao government, after all, represented a newly established “People’s Democratic Republic” that, while ideologically willing to deploy coercion in the name of enforcing “the people’s will,” also faced practical threats to its legitimacy if it was seen as dispossessing the common people on a large scale. Similarly, the Swedish government, whose support for the new Lao government had as much to do with its earlier outspoken criticism of US imperialism during the “Vietnam” War as it did with forestry competence per se, was already well-known for its commitment to rights-based approaches to development cooperation.³³ Helping a Marxist-Leninist government dispossess its own people would not have been high on its agenda.

But there is much more to the story that comes through if we look closely at the landscape around SFE 1. My reference point here is security, both in the narrow politico-military sense in which the term is often used (and in which LSFP staff like Thongphachanh and Birgegard used it) and in the wider sense, also including economic issues, that appears in the instruction on population management work. The latter meaning echoes as well Michel Foucault’s 1977–78 lectures on governmentality, which he titled Security, Territory, Population.³⁴ His focus was European history, but as geographers and other scholars have noted, his work highlighted a tension between two key forms of state power that have wider and contemporary relevance.³⁵ Inspired in part by Jeremy Bentham’s ideas about designing prisons, factories, and workhouses for the poor, Foucault called the first of these “disciplinary” power; this focused on spatial arrangements that optimized social control, and the principles that emerged from it have been applied to numerous cases, from buildings like the ones mentioned above, to town planning, to state regimes for keeping track of landownership.³⁶ As Foucault also noted, however, economic processes like town planning, trade, and farming also depended on maintaining flows or adequate circulation. These ideas derived not from Benthamite notions about restriction and control but from French “physiocrat” economists who sought to “govern with nature” (physio-cracy), and whose ideas about enhancing productivity led them to embrace natural models like blood circulation through the body in their quest for “healthy” and “efficient” economies.³⁷

This tension between control-based and productivity-based governing is a useful way to think about the often-conflicted spatialities of politico-military versus economic security. Laos’s postwar experiences of industrial forestry exemplify this tension. Population management work sought to achieve both control and productivity, sometimes of the same people, as in the case of managed upland settlement; and sometimes of very different things, as in the case of upland farmers being kept away from forestry operations. Its methods were inherently spatial and often came up against the tension between control and production in their specific deployments. They also made use of force when they needed to; as Foucault and various political theorists have long pointed out, carrots work better when there are also sticks at hand.³⁸ But a key reason that Lao authorities and development planners were so keen to minimize the use of force, and instead manage the social geographies of hinterland settlement and production carefully, had to do with what LSFP personnel like Thongphachanh and Birgegard referred to as “the security situation.” In the landscape of SFEs 1 and 3, there was an inherent tension between forestry and upland farming via the competition for land. But in addition, an antigovernment insurgency had emerged in the wake of the Lao PDR’s establishment in 1975, largely in the footprints of former upland maquis areas. (STOL sites, discussed in ch. 2, are one indicator of this geography, and are included on map 3.1 above.)³⁹ Much of the insurgency was located north of the SFEs in what came to be called the Saysomboun Special Zone, but enough of it spilled south into the areas of SFEs 1 and 3 that it affected their operations significantly. The security situation was already sufficiently serious to warrant Thongphachanh and Birgegard’s attention in 1981–82, and according to LSFP and other accounts it seems to have gotten worse in the years that immediately followed.⁴⁰

The insurgency tended to be strongest in remote and forested areas; this has often been the case with so-called peasant insurgencies around the world, and in this case it followed specifically from the geography of earlier US war-making. In a context where shifting cultivation was socially and geographically widespread, and where it was not limited to “upland” ethnic groups, the inherent economic conflict between forestry and upland farming risked spillover into political sympathy for the rebels. In such a context, a 1986 LSFP report’s comment that “it is often said, today, that the local population generally look upon the state forestry enterprises as their enemies” carries multiple potential meanings.⁴¹

Thongphachanh and Birgegard had interesting things to say about security and development, and their analysis and recommendations heavily inform the end of this section and the first part of the next. First, though, it is worth detouring to examine a key silence in their and other LSFP accounts: the heavy Vietnamese military presence in and around state forestry operations. This may have been deliberate, an example of the depoliticization that often accompanies development discourse.⁴² But it also may have been too obvious to need saying. The presence of the Vietnamese military during this period is widely noted by scholars, and would have been well-known to the consultants’ intended audience of government officials, technical staff, and LSFP advisers.⁴³ In contrast, it leapt off the page in the few popular accounts that were produced for a wider audience at the time.

In May 1987, for instance, a reporter for the Swedish newspaper Svenska Dagbladet visited Muang Mai to write a feature on Swedish forestry aid to Laos. The Vietnamese military figured centrally in the account:

Swedes and other foreigners who visit Muang Mai say they are escorted by Lao soldiers or militiamen. But during Svenska Dagbladet’s visit, the Lao escort was replaced by a Vietnamese escort in Muang Mai. About twenty well-armed Vietnamese soldiers kept watch over our trip to the forest and our stay there.

“Security is better now, but ever since 1984 we have had many clashes with Lao groups that came across the Mekong River from Thailand. A shipment was attacked recently,” said Vietnamese Lieutenant Pham Van Thu, who has spent nine years in Laos.

There are between 40,000 and 50,000 Vietnamese soldiers in Laos. A sizeable force is also stationed at the main camp in Muang Mai. They move about the area freely and sit in groups talking to the Lao workers. In the evenings, they go down to the little private market.⁴⁴

A 1978 account that appeared in the Far Eastern Economic Review is also relevant. Even though it focused on an area significantly to the south of SFE 1, in Savannakhet province, it offers a glimpse of the negative entanglements between local agrarian livelihoods, security measures in general, and the Vietnamese military in particular. Written by an anonymous “Western student” who snuck into Laos for eight days, the account—based on “conversations with villagers, guerillas, an army officer who recently defected with ten men and their weapons, and an escapee from ‘Seminar’ (a re-education labor camp)”—made explicit the neocolonial dimension of the Vietnamese presence from the perspective of Lao farmers:

The Laotians are fighting not only a rigidly authoritarian regime but, as they see it, one that is kept in being by the armed forces of an occupying power, the Vietnamese. The Vietnamese are said to be invariably in control of their Laotian counterparts.… Village leaders said that their villages were visited and searched by Vietnamese units up to five times a month.…. The villagers had an unalloyed hatred of the Vietnamese and disgust with Laotian officials and soldiers who were seen to be working for the Vietnamese and against their own people.⁴⁵

It is in this light that Thongphachanh and Birgegard’s comments on security deserve to be read. Outlining “three aspects of this problem,” they distinguished first between “reactionary groups and bandits” and the rest of the population.⁴⁶ While not unconcerned by the former, it was largely with the latter—the main portion of the rural population—that their focus lay. This “reserved and uncommitted” population’s allegiance needed to be won, they argued, counterinsurgency-style: “The situation has placed the rural population in a difficult position. The reactionaries are trying hard to alienate the population from the Government by threats and harassment. As an understandable reaction, the attitude of the population is reserved and uncommitted. Under these conditions mobilization for development becomes more difficult.”⁴⁷

“Mobilization for development” is often used today to mean uncompensated sacrifice to the national community.⁴⁸ There is certainly a hint of this in Thongphachanh and Birgegard’s use of the phrase here. But given their concern for what might happen if the government pushed its citizens too hard in the name of the “national” interest, it is better to think of “mobilization” in this context via the quest for “win-win” efforts described by the Council of Ministers above. Given the right socio-spatial relations, planner-experts like Thongphachanh and Birgegard sought to achieve the simultaneous benefit of both citizens and nation through the judicious application of population management techniques. But this required taking seriously the needs of the local population so that their “attitude” would shift from “reserved and uncommitted” to mobilizable for various forms of development, which inevitably called for sacrifice.

Thongphachanh and Birgegard’s final point on security, quoted above in the chapter’s epigraph, outlined a conflict between development and security-related resettlement. The paradox they described built on their second point and highlighted the need to address upland resettlement with the utmost care: “Thirdly, the security situation may demand measures, e.g., in terms of movement and relocation of population, which are not conducive to development. Resettlement of whole villages and even entire tasseng [subdistrict] populations for such reasons may upset both short-term and long-term development efforts. As an improvement of the living conditions of the population is probably the most effective way of overcoming the security problems, there is a paradox that development efforts are most needed where the security situation makes them most difficult to carry out.”⁴⁹

Reading these comments in light of the Vietnamese military presence, the conflict between local and “national” development is especially delicate. If it was difficult to ask villagers to sacrifice for Laos’s national development, the close working relationship between the Vietnamese military, the Lao military, and the Muang Mai forestry operations made this proposition even tougher. In this context, Thongphachanh and Birgegard’s sympathy for the local population’s “reserved and uncommitted” attitude toward SFE 1 and government intervention more generally, highlights the daunting nature of population management work in the postwar industrial forest.

THE GEOGRAPHY OF POPULATION MANAGEMENT

This characterization of the local population as understandably reserved and uncommitted vis-à-vis the Lao government was reflected in Thongphachanh and Birgegard’s proposals for population management. The approaches they proposed followed a three-way distinction between the lowlands, the inner uplands, and the outer uplands, and sought to balance the need for state forestry in the latter two landscapes with the need for livelihood development in all three. Their recommendations anticipated and exemplified the rhetoric of win-win development and security that the Council of Ministers would articulate later in the decade, and the techniques that they proposed would reappear in the 1990s and 2000s as staples of state development practice. Given the lingering debates about the line between voluntary and coerced participation that these schemes trod, both in the rubber sector and elsewhere, their genealogies are worth examining.

Although Thongphachanh and Birgegard were skeptical of the potential for lowland rice production to meet the needs of the whole population of Muang Mai, they made paddy improvement and expansion one component of their population management strategy. Focusing on the Mekong lowlands southwest of the Sayphou Nyou, they recommended distributing improved plows to increase yields, redistributing fields that had been abandoned, and building new paddy land in areas where forest quality was already poor. This would help keep the “source” population that threatened the SFEs anchored as much as possible, while also following official policy preferences for lowland rice as much as the landscape would allow. The first of these also had an overtly strategic reasoning, targeting lowland farmers who already had paddy land of their own. Plow distribution, they argued, exemplified the “clear political dimension” of development work since it would “reach a large segment of the population, can be implemented without delay and without too much difficulty, [and will] have a visible impact.”⁵⁰ In this, Thongphachanh and Birgegard echoed the kinds of high-visibility development that counterinsurgency strategists in the region had long advocated.⁵¹

They then turned to the area beyond the Sayphou Nyou, where the real challenges lay. They differentiated between the river and stream valleys of the inner frontier zone and the “high mountains” populated by the “Lao Soung”—a term that translates literally as “Highland Lao” and, in this context, would have referred to the Hmong. For the inner zone, Thongphachanh and Birgegard pushed for a generalized effort to “improve and transform” shifting cultivation through a mix of yield improvements and conversion to “permanent up-land cultivation.” Proposing a long-term approach that would draw heavily on “experiences in other countries,” they hoped that Laos would employ a version of the taungya cultivation system developed by British colonial foresters in Burma to enclose forests for state-managed extraction without provoking too strong a backlash from locals: “Protection of forest resources should be achieved by integration of shifting cultivators into forestry activities rather than by law enforcement, fencing and guarding. The integration should primarily take the form of development of agro-forestry (taungya) systems for reforestation. Areas for reforestation should be selected at the fringes of forest reserves where the encroachment by shifting cultivators is a threat.”⁵²

As this passage makes clear, the “agro-forestry” they had in mind was not the sort practiced in, for example, smallholder rubber systems in Indonesia.⁵³ Rather, “agro-forestry” was more of a euphemism for state-managed cultivation in which the agriculture was temporary and the forestry was long-term and state-owned; it was, in short, a form of managed enclosure. But it was intended to be subtle rather than overt. For Thongphachanh and Birgegard, physical alienation of land bred political alienation of the population; the line was almost one-to-one. Their recommendation against the use of “law enforcement, fencing and guarding” was based on the rationale that such an approach risked pushing locals toward the “reactionaries” discussed above.⁵⁴ “Integration”-based approaches that would keep the population in place while developing and managing the forest resource in their midst were, from this perspective, far more preferable because of their softer touch.

This approach seems to have taken root, although unevenly. In 1991 the anthropologist Ing-Britt Trankell spent seven weeks in the same part of Laos on a self-described “hit-and-run” anthropology project studying the effects of a road upgrade.⁵⁵ In some of the villages her team visited, she described a reforestation effort that had the essential features of the taungya method that Thongphachanh and Birgegard had advocated a decade earlier: “Villagers now receive plots from the forest company on which they are allowed to work for 2–3 years, on the condition that they plant tree saplings for the reforestation of the area. The work is performed simultaneously with the planting of rice. After the harvest, the area is marked with a fence in order to prevent cattle from damaging new plants. Villagers approve of reforestation programs and plea[d] for more such work to be done, but at the same time they resent the idea of being excluded from an area which they have themselves cleared and worked.”⁵⁶This ambivalence is a hallmark of many land allocation schemes that aim to rationalize upland farming through the offering of work (in this case) or formal tenure security (in others) in exchange for using less land. In Trankell’s case, however, equally telling were her experiences in other villages, where residents were unwilling to even speak to her research team after having had their fishponds destroyed and “nutritionally valuable secondary growth” (i.e., swidden fallow forest) cut down following an earlier socioeconomic survey. As Trankell reported, “The issues of land use and land rights with regard to forestry and agricultural land are presently the most difficult and crucial problems.… Clashes with forestry company staff and forestry programs due to mutually conflicting views regarding the right to and the use of forestry products are reported.”⁵⁷ Such heavy-handed approaches testify that the taungya-style approach of managed enclosure was not the only form of population management work local authorities were conducting to control village-scale land use; tactics varied significantly by time and place.

A second key strand of population management, echoing Thongphachanh and Birgegard’s final point on security, was the managed resettlement of upland communities into concentrated areas. Spanning the entire decade of the 1980s—Thongphachanh and Birgegard did their fieldwork in 1981, Trankell did hers in 1991—in the areas around SFE 1, this focused on Hmong communities who populated what Thongphachanh and Birgegard described as the remote areas comprising SFE 1’s mandate, and who Trankell (a decade later) noted were still “regarded as responsible for the occurrences of insurgent activities that regularly haunt the area.”⁵⁸ (Trankell had been prevented from visiting what she called “villages in the hinterland,” and had lamented the “certain ‘shortage’ of villages suitable for our project” as a result.) Concentrated resettlement was far less subtle and ambivalent than the taungya-style managed enclosures described above; provincial authorities in Bolikhamxai, for instance, described the establishment of the focal site at Muang Houng, discussed below, as requiring “tremendous effort to open the road in the middle of the forest” in part because of the “security problems [that] were an issue a few years ago.”⁵⁹ They wrote this in 2000, implying that the insurgency persisted well into the 1990s. Trankell’s description from a decade earlier, when the insurgency was still in full swing, described the “resettlement programs” as having been seen by local authorities “as a way to pacify the Hmong by bringing them down from the hills.” Provincial authorities, she noted, “reported that at the time they had three different programs for ‘the settlement and education’ of Hmong groups in lowland areas.”⁶⁰

Thongphachanh and Birgegard’s remarks about concentrated resettlement are especially interesting and relevant because they capture a key strand of official thinking at a key formative moment in Lao state forestry. In their report, concentrated resettlement offered a way to resolve the “paradox” mentioned above by bringing the population out of both the areas haunted by insurgency and the forests targeted for development by the SFEs. By stabilizing and fixing land use, they argued, it had the benefit of remedying the problems of what they took to be the Hmong’s especially destructive style of shifting cultivation, which they described as moving from place to place “with no intention to return to the abandoned land after a fallow period” and leaving land with “little value either as agricultural land or as forest land.”⁶¹ Thongphachanh and Birgegard termed their resettlement-based approach “intensive ‘sub-area’ development,” and they initially described it via “some general principles” due to their inability to access many of the areas they were seeking to manage:⁶²

Firstly it is suggested that the development of the mountain areas in the country as a whole as well as in Muang Paksan can not imply a development of the entire mountain areas (in terms of road networks, electrification, social services, etc.). Rather, development has to be concentrated to certain sub-areas and pockets with a relatively good economic potential.… A strategy of selective (and intensive) development in the mountains most likely presupposes a gradual out-migration from these areas. The reason is that it may be difficult (and economically unacceptable) to develop a sufficient economic base in selected areas for the entire mountain population.⁶³

This vision of upland relocation implied a squeezing of the upland population’s economic base that would leave room for some, but not for all. The remainder of the population, they suggested, would be induced to relocate to the lowlands. Using a crude but clear graphic (reconstructed as fig. 3.1), Thongphachanh and Birgegard elaborated their “general principles” via a spatio-temporal proposal: “People from the mountains are to be encouraged to settle in the sub-areas selected for intensive development. This movement is illustrated by the arrows marked (1) in the [figure above]. The second movement involves an out-migration from the mountain sub-areas to the plains (arrow marked 2). After an initial period of settlement in the sub-areas (arrows marked 1), the two movements are expected to go on simultaneously. The time perspective involved is very long (25–50 years).”⁶⁴

“INTENSIVE ‘SUB-AREA’ DEVELOPMENT”: THE CASE OF MUANG HOUNG

The distinction between general principles and actual action plans dissolved in a place called Muang Houng. Located at the end of the road that offered access to the forest resource beyond the Sayphou Nyou, Muang Houng had been the Lao government’s original proposal for where to build SFE 1, as noted above. After SFE 1 was moved to Muang Mai because of pushback from Swedish advisers, Muang Houng retained a focus within SFE 1’s “regional” subproject, not for forestry but as a destination for upland resettlement. Muang Houng thus exemplified the Paksan Regional Project’s cross-sectoral mandate of doing nonforestry development work that would nonetheless enable forestry to proceed.

Thongphachanh and Birgegard devoted almost four pages in their study to the “construction of a lower secondary school for the mountain tassengs [subdistricts]” that, they suggested, be built in Muang Houng. In contrast to the “general principles” above, their proposal for this school was quite detailed, covering “the proposed activity,” “the buildings,” “benefits and justifications,” “implementation responsibility,” “inputs required,” a “cost estimate,” and “agricultural activities to reduce boarding costs.”⁶⁵ Education of upland communities was central to bringing state forestry and regional development into alignment, they argued, because it would facilitate the out-migration that was needed to alleviate the demographic pressure on forestry operations: “The provision of social infrastructure and services should purposively be used to influence future population settlement patterns. This means that priority should be given to (i) the plains along [the] Mekong; (ii) the pockets in the mountain areas identified for intensive development; and (iii) locations where agro-forestry systems are introduced (as an incentive to involve the farmers). To speed up out-migration of the high mountains it is proposed that even stronger efforts than hitherto are made to favor education of the children among the hill tribes.”⁶⁶

In the years that followed their proposal, Muang Houng became the home of Bolikhamxai’s first “focal development” site. A Socio-Economic Profile with Emphasis on District Development, published in 2000 by provincial authorities and the United Nations Development Program (UNDP), describes in detail both focal-site development in general and Muang Houng in particular. Throughout the 1990s, focal-site development had been increasingly embraced as official rural development policy, and in 1998 Laos’s central government had announced plans to scale up the “approach” nationwide.⁶⁷ Despite having arguably helped pioneer the approach via two decades of work in Muang Houng and its surrounding areas, Bolikhamxai authorities took care to frame their work in the language of official policy: “The essence of our Focal Site-based rural development is ‘an area approach targeting rural poverty.’ Our ‘Focal Site’ strategy is hence the ‘bringing together of development efforts in an integrated and focused manner within a clearly defined geographical area, aiming at the eradication of poverty and at promoting sustainable development.’ ”⁶⁸

Quoting official policy language like this downplayed the more heavy-handed dimensions of focal-site development, namely the massive relocation it often entailed and the fact that it was frequently a key part of active counterinsurgency operations. In their description of focal-site development in the district that now contains Muang Houng and Muang Mai, provincial authorities referenced Muang Houng as the first of the province’s original two focal sites and emphasized its “accelerated development” compared to a second site “without any Lao Soung” where the administrative “structure was established but no development activities yet” undertaken. In contrast, the Muang Houng site was “half Lao Loum and half Lao Soung, with a minority of Lao Theung,” and contained ten thousand people: ten times the population of the other focal site and a full half of the district’s population.⁶⁹

This concentration of population in Muang Houng and its immediate environs suggests a great deal of state intervention in the areas that Thongphachanh and Birgegard had called the inner and outer forest frontiers. In contrast to the situation in the early 1980s, when the greater landscape of SFE 1 was extensively populated but only partly inventoried owing to the security situation, by the late 1990s “accelerated development” had rearranged both of these. As the provincial profile makes clear, these efforts spanned the spectrum between coercion and consent: “For us, focal site development must be ambitious and, at first, necessarily provincial government driven because of the basic need to first provide access, land clearing, etc.… We are aware that once the initial effort has been provided, a ‘softer’ approach has to take over.”⁷⁰

This “accelerated” process (“access, land clearing, etc.”) was a sanitized way of referring to what the profile described elsewhere as a mix of logging and military operations that, in the case of Muang Houng, proved especially daunting. This began with “open[ing a] new road to the remote areas (with logging activities),” and as noted above, required “tremendous effort to open the road in the middle of the forest.”⁷¹ Part of this effort was certainly physical and logistical: Muang Houng was located well into the panhandle’s forested interior, in the rolling uplands that ascend from the Mekong to the crest of the Annamite Mountains along the border with Vietnam. But the need to begin with “ambitious” rather than “softer” approaches also reflected the political difficulties that confronted the state push into the upland interior. The profile acknowledged as much, although it did so from the comfort of hindsight: “While security problems were an issue a few years ago, no troubles have occurred since accessibility has been improved, because improved accessibility meant improved access to socio-economic development.”⁷²

Although the success of these “socio-economic development” activities is debatable,⁷³ it is clear that the development of the Muang Houng focal site figured centrally in the efforts—first of SFE 1, and then of provincial authorities, who took over the Paksan Regional Project in 1986 and SFE 1 in the years immediately after⁷⁴—to separate the upland population from the forest resource. While the details of this process are beyond my present scope, the “two provincial saw mills” referenced in the Bolikhamxai provincial profile testify to its general direction, as does the almost complete surrounding of Muang Houng by provincial logging areas in the 1990s and 2000s.⁷⁵

Finally, the Bolikhamxai provincial profile described the responsibilities of district-level Agriculture and Forestry officials. These were a far cry from the “integration of shifting cultivators into forestry activities” envisioned by Thongphachanh and Birgegard and observed fleetingly by Trankell in 1991. Instead, they focused on conducting land-use zoning—the LFA process, which was then in its heyday of nationwide implementation—and actively “prohibiting logging and wildlife hunting, as well as wildlife trade, by enforcing the law and punishing offenders according to regulations and rules.”⁷⁶ This was precisely the “law enforcement, fencing and guarding” approach that Thongphachanh and Birgegard had written off in the early 1980s as both politically infeasible and unadvisable because of its security implications. By the late 1990s, in contrast, it was a different era entirely, and population management techniques had changed to fit the times.

In his lectures on governmentality, Foucault offered a conceptual distinction that is relevant to population management work. The population, he explained, was the collective who not only followed their own interests and desires but also acted in a way that was rational and could thus be governed, in pursuit of the greater good, using various policy instruments. This was the collective who would “do as they ought,” as Bentham put it, with the right mix of incentives, rules, and, if necessary, force. As an essentially governable group, the population stood in stark contrast to what Foucault called “the people”: the unruly mob who demanded that their needs be met even if doing so proved to be socially inconvenient. The people, unlike the population, placed themselves outside the collective, and thus by “refusing to be the population, disrupt the system” as a whole.⁷⁷ In the context of Laos’s postwar uplands, this distinction between “population” and “people” was not inherent to the groups involved, at least from the perspective of the advisers, managers, and state authorities involved in and around industrial forestry operations and associated upland projects. Rather, the distinction was contingent on socio-spatial relations. Working in the name of development, state authorities and their advisers sought to manipulate these relations in ways that would render uplanders into the desirable category of population, “allowing” them, as the Council of Ministers put it, to direct their collective mastery to the “two strategic tasks” of development and defense.

The reality was considerably messier. But this messiness sheds important light on Laos’s upland landscape at a key historical moment, showing how population management was not something that could be theorized from afar and then applied uniformly. Instead, it required a local praxis that experimentally combined analysis of the immediate terrain, prediction about future scenarios and competing needs, and deployment of available resources. The particular applications—taungya-style managed enclosures in some areas, focal sites in others; the deliberate avoidance of “fencing and guarding” in some circumstances and the embrace of laws, rules, and regulations in others—highlight the variation across space and time. But more importantly, they point to the family resemblance of different types of population management work that were being developed as a repertoire, a field of practical experience that could be drawn on in later situations. Scholars of development in Laos have tended to view the relationship between politico-military and economic issues through a lens of transition, emphasizing the shift from security-oriented interventions in rural development in the 1980s to development-oriented interventions in more recent decades.⁷⁸ This relationship, however, is better conceptualized as one of interaction. This theme of security as always both political and economic sits at the heart of population management work.

The Council of Ministers’ 1988 instruction came in the early phase of Laos’s official turn toward market-based development. This “Renovation Policy” or “New Economic Mechanism” was promulgated in 1986 and led, through the late 1980s and throughout the 1990s, to the devolution and in some cases privatization of various state enterprises, the writing and adoption of a constitution and beginnings of a legal code, and a range of state- and donor-led efforts to recruit foreign direct investment. This period gave birth to a number of widely known policies and interventions, including the formalization of focal-site development and the creation of the village-scale LFA zoning program.⁷⁹ It also launched a new discourse of postwar security talk that helped animate the application of these efforts by foregrounding the need for what state and party leaders called “heightened vigilance” as the country proceeded onto “the new battlefield where no gunfire can be heard.”⁸⁰ An excerpt from a Lao radio broadcast from seven months after the council issued its instruction on population management work is exemplary, highlighting political leaders’ anxiety about increased contact with the outside world and, tellingly, suggesting a link between social conflict and external interference that we will see again:

The enemies have taken advantage of the conveniences of traveling in our country to send their spies in the guise of businessmen, traders, tourists, or workers to gather information about us and launch propaganda to create rifts between Laos and its friendly countries, between soldiers and police and civilians, between state officials and cadres and people, and among the people of different ethnic groups[. They] hope to cause mutual suspicion, antagonism and distrust between the lower and higher echelons, triggering internal conflicts so as to start riots and uprisings as they did in other countries. These are the most dangerous, subtle, and cruel of the enemies’ tricks.⁸¹

As a way to combat these schemes, state authorities emphasized the need for diligence in population management work. Echoing the 1988 instruction, another radio broadcast from early 1989 exhorted cadres to “vigorously turn to the grassroots and build them into all-around strong localities,” calling this “a foremost strategy in our party’s and state’s national defense and public security maintenance work.”⁸² As the examples above illustrate—from state forests to managed enclosure to focal sites—this idea of “all-around strong localities” captures the essence of multidimensional security intended by population management work.

Today, and even at the time it was issued, this security discourse carried the hollow ring of propaganda. Yet we should not confuse form with content. Events in the Lao uplands were existentially serious, both for the Lao state and for the many members of the population who found themselves living amid struggles both political and economic over which they had limited control. And population management work cannot be dismissed today as mere propaganda. As the 1980s turned into the 1990s and especially the 2000s, the government’s quest for “all-around strong localities” drew increasingly on the outside world for economic resources and expertise. But to organize these in place, on the ground, Lao authorities drew crucially on the population management methods developed and honed during the earlier postwar period. It is with this history in mind that we can now return to the northwest.