Untangling the proximate causes and underlying drivers of deforestation and forest degradation in Myanmar
Abstract
enPolitical transitions often trigger substantial environmental changes. In particular, deforestation can result from the complex interplay among the components of a system—actors, institutions, and existing policies—adapting to new opportunities. A dynamic conceptual map of system components is particularly useful for systems in which multiple actors, each with different worldviews and motivations, may be simultaneously trying to alter different facets of the system, unaware of the impacts on other components. In Myanmar, a global biodiversity hotspot with the largest forest area in mainland Southeast Asia, ongoing political and economic reforms are likely to change the dynamics of deforestation drivers. A fundamental conceptual map of these dynamics is therefore a prerequisite for interventions to reduce deforestation. We used a system-dynamics approach and causal-network analysis to determine the proximate causes and underlying drivers of forest loss and degradation in Myanmar from 1995 to 2016 and to articulate the linkages among them. Proximate causes included infrastructure development, timber extraction, and agricultural expansion. These were stimulated primarily by formal agricultural, logging, mining, and hydropower concessions and economic investment and social issues relating to civil war and land tenure. Reform of land laws, the link between natural resource extraction and civil war, and the allocation of agricultural concessions will influence the extent of future forest loss and degradation in Myanmar. The causal-network analysis identified priority areas for policy interventions, for example, creating a public registry of land-concession holders to deter corruption in concession allocation. We recommend application of this analytical approach to other countries, particularly those undergoing political transition, to inform policy interventions to reduce forest loss and degradation.
Abstract
esDilucidando las Causas Proximales y los Conductores Subyacentes de la Deforestación y la Degradación en Myanmar
Resumen
Las transiciones políticas comúnmente desencadenan cambios ambientales sustanciales. Particularmente, la deforestación puede resultar de una interacción compleja entre los componentes de un sistema – actores, instituciones y políticas existentes – que se están adaptando a oportunidades nuevas. Un mapa conceptual dinámico de los componentes de un sistema es particularmente útil para los sistemas en los que múltiples actores, cada uno con formas de ver la vida y motivaciones diferentes, pueden estar intentando simultáneamente de alterar las facetas diferentes del sistema, sin conocer los impactos sobre los otros componentes. En Myanmar, un punto caliente de biodiversidad global con una de las áreas forestales más extensas del sureste asiático continental, las continuas reformas políticas y económicas tienen la probabilidad de cambiar la dinámica de los conductores de la deforestación. Por lo tanto, un mapa conceptual fundamental de esta dinámica es un prerrequisito para las intervenciones que buscarán reducir la deforestación. Usamos una estrategia de dinámica de sistemas y un análisis de red causal para determinar las causas proximales y los conductores subyacentes de la pérdida y degradación del bosque en Myanmar de 1995 a 2016 y para articular las conexiones entre ellos. Las causas proximales incluyeron al desarrollo de infraestructura, la extracción de madera y la expansión agrícola. Estas causas estuvieron estimuladas principalmente por las concesiones formales de agricultura, tala, minería y de energía hidráulica y la inversión económica y los temas sociales relacionados con la guerra civil y la tenencia. La reforma de las leyes del suelo, la conexión entre la extracción de recursos naturales y la guerra civil, y la asignación de concesiones agrícolas influirán sobre la extensión de la futura pérdida y degradación de los bosques en Myanmar. El análisis de red causal identificó áreas prioritarias para las intervenciones políticas, por ejemplo, creando así un registro público de concesionarios de tierras para disuadir la corrupción en la asignación de concesiones. Recomendamos la aplicación de esta estrategia analítica en otros países, particularmente aquellos pasando por una transición política, para informar a las intervenciones políticas para reducir la pérdida y degradación del bosque.
Introduction
Forest degradation and conversion are among the most significant ongoing transformations of the Earth's surface (Smil 2013). Forest transformation is driven by a complex interplay of multifarious drivers related to human activity, including governance, policy, infrastructure, economics, institutions, markets, and technology (Geist & Lambin 2002). The scale and urgency of the problem has spurred a range of policy measures, such as protected areas, payments for ecosystem services, and community forestry schemes (Angelsen 2010).
It is critical, however, that forest conservation policy be developed with an appreciation and understanding of both the proximate and the underlying drivers of forest change, as well as their potential interactions. For example, Thailand's logging ban in 1989 addressed the proximate cause of forest degradation (logging) but not the underlying cause (demand for timber) and therefore merely shifted forest degradation to Cambodia and Myanmar (Le Billon 2000; Asia-Pacific Forestry Commission 2001). Interventions to increase yields to reduce the area needed for food production—and thereby spare land for forests—will also increase the profitability of agriculture and may therefore incentivize agricultural expansion (Phelps et al. 2013). The nature of dynamic and complex systems makes intervention challenging because actions taken to address one process could potentially affect the behavior of various agents in the system in ways that are often delayed over time, distributed over space, and unforeseen (Sterman 2000; Liu et al. 2013).
Periods of political transitions present challenges for forest conservation entities and environmental managers, government and nongovernment alike. Political transitions, such as democratization, can greatly affect deforestation rates (Buitenzorgy & Mol 2010). The uncertainty of power relations during reforms may drive short-term rent seeking (Fredriksson & Mani 2004) and thus incentivize increased forest resource extraction, as occurred in Indonesia following the overthrow of Suharto (McCarthy 2000). The weakening of institutions and monitoring during Nepal's ongoing political transition has led to increased illegal logging (Paudel et al. 2013). Political competition can increase pressures to accumulate or extract forest resources for networks of support; this occurred in the aftermath of political transitions in Cambodia (Le Billon 2000) and Kenya (Klopp 2012).
Myanmar has been transitioning from an economically isolated military dictatorship to a more democratic state that is increasingly open to foreign investment. Given the emerging threats to Myanmar's forests (Prescott et al. 2017), there is an urgent need to clearly articulate the drivers of deforestation and forest degradation (Webb et al. 2012, 2014).
The challenges of understanding complex systems in transition require the use of systematic conceptual tools. Systems thinking (Sterman 2000) allows for the holistic study of complex systems by considering the interconnections among the underlying drivers explicitly. Constructing causal networks—visual representations of the interconnections between underlying processes in a system—is a key component of the systems-thinking approach and is useful for identifying the key linkages that underpin a complex process. Systems-thinking and causal-network analysis are tools that have a wide range of applications, including business (Sterman 2000), public health (Ziegler et al. 2016), and land-use change (Geist & Lambin 2002).
We used systems-thinking and causal-network analysis to conceptualize and articulate the relationships among proximate and underlying drivers of forest loss and degradation in Myanmar in the ∼20 years prior to the transition to a democratically elected government. We analyzed relevant literature through the lens of systems thinking to develop a visual representation of the drivers of forest loss and degradation and their interrelationships in Myanmar. In doing so, we established a baseline to consider how forest conversion and degradation may be affected by the ongoing political transition. Such an analysis is critical to inform the ongoing discussion shaping new forest policies as the Myanmar Government continues its governance transition.
Methods
Literature Search
We conducted a systematic search to retrieve articles published from 1995 to 2016 with a title, abstract, or lists of keywords containing “Myanmar” OR “Burma” AND “deforestation,” OR “forest loss,” OR “forest decline.” We searched BioOne, JSTOR, Google Scholar, ScienceDirect, and Wiley Online Library (Fig. 1). We cross-referenced citations, including gray literature, until we could not locate new articles related to deforestation in Myanmar. We obtained 126 documents that met these criteria (Supporting Information).
While reviewing these 126 documents, we included those that had qualitative or quantitative evidence of deforestation trends and an original evaluation of drivers for the case study analysis. We did not include papers that did not substantiate claims of deforestation or simply stated deforestation trends without elaboration as case studies, but we kept them for reference. This elimination process left us with 38 documents, which all together documented 46 individual case studies (Fig. 1 & Supporting Information). The individual case study—which we defined as an original investigation occurring at a specific study site—was the unit of analysis.
We identified and coded all data regarding drivers of forest loss or degradation from the 46 case studies (Strauss & Corbin 1990; Supporting Information) and grouped proximate and underlying drivers into separate categories. We determined links between the codes based on our interpretation of the case studies. We documented links with NVivo version 10 (QSR International 2015), a qualitative data-analysis software that facilitates data identification, indexing, and retrieval.
Systems-Thinking Analysis and Causal-Network Diagrams
Causal-network diagrams are flexible tools for conceptualizing the structure of a system because they show the links between variables (Sterman 2000). Arrows indicate the direction of influence, from cause to effect. The polarity of arrows (plus or minus) indicates whether the effect is increasing or decreasing. Furthermore, causal-network diagrams help identify feedback structures within a system.
We mapped the linkages among deforestation drivers in a causal-network diagram (Anderson & Johnson 1997; Sterman 2000) constructed in VenSim version 3.6G (Ventana Systems 2015) and yEd version 3.15.0.2 (yWorks 2015). We drafted causal-network diagrams for each case study and contextual article and then meshed the diagrams to create a holistic representation of deforestation in Myanmar (Supporting Information). To facilitate comprehension, we created 2 separate but linked causal-network diagrams. The first represented the nexus between forest loss and degradation and their proximate causes; the second represented the relationships among proximate causes and underlying drivers.
Caveats and Limitations
Despite the recognized rigor of causal-network analysis (Sterman 2000), there are 3 primary caveats or limitations to our study. First, as with all reviews of multiple studies that vary in methods and analysis, case-study data were inadequate for quantitative evaluation of contributing factors to forest loss or degradation—a common limitation of analyses of complex and large-scale socioecological systems (Rissman & Gillon 2017). Specifically, several studies lacked information on the spatial extent of forest loss or degradation.
Second, we restricted the analysis to published case studies, which in aggregate did not constitute a spatially random sample. This could exclude important region-specific dynamics. For example, although some states subjected to civil war were very well represented in our study (e.g., Kachin and Shan states, which contained 33% and 15% of case-study locations respectively), other states such as Rakhine were not (Fig. 2).
Finally, there is significant cultural, historical, and floristic variation both between and within different states and regions. The individual dynamics for any given state, region, or site within the country may therefore vary. However, we believe the repetition of common themes throughout the geographic range of the case studies supports the creation of a generalized model for Myanmar.
Results
We identified a range of proximate causes of forest loss and degradation in Myanmar from 1995 to 2016 (Fig. 3): agricultural expansion, wood extraction, and infrastructure, mining, and hydropower development. The most common causes were agricultural expansion (78% of cases), wood extraction (85%), and infrastructure development (50%) (Supporting Information). Commercial mining and hydropower development were identified less frequently (41.3% and 39.1%, respectively [Supporting Information]). For brevity, we abbreviated the frequently recurring deforestation-causing nexus of mining, logging (timber extraction), hydropower (development), and agriculture as MiLoHyAg. Underlying drivers of forest loss and degradation included weak land tenure, economic pressures, civil war, environmental policies and concerns, and corruption (Supporting Information).
Agricultural Expansion
Agricultural expansion was identified in the majority of case studies we reviewed as a proximate cause of forest loss or degradation. Although we were unable to quantify the relative contributions of specific agricultural practices to forest area lost, more studies identified commercial agriculture as a cause of deforestation than subsistence agriculture (36 vs. 27, respectively [Fig. 3 & Supporting Information]).
The majority of land granted for agricultural concessions was located in heavily forested and politically contested regions (e.g., 68% of the agribusiness concessions awarded from 2010 to 2012 were in Kachin and Tanintharyi [Donald et al. 2015]). Agricultural concessions were frequently logged but not subsequently converted to agriculture. By one estimate, <25% of the 2 million ha of agricultural concessions awarded from 2012 to 2013 was converted to agriculture because private investors obtained agricultural concessions primarily to gain access to timber (Woods 2015) (Fig. 4).
Following decades of overharvesting, environmental concerns and a 2014 ban on the export of raw logs led to a decrease in the number of logging concessions awarded. At the same time, the agricultural master plan and land policies (e.g., Foreign Investment Law; Vacant, Fallow and Virgin Lands Management Law; and Farmland Law) aimed to increase the number of agricultural concessions awarded to develop Myanmar's agricultural sector. However, in the face of continued demand for Myanmar's timber, agricultural concessions were often logged but not subsequently converted to agriculture, because agricultural concessions were acquired as a front for logging (Woods 2015) (Fig. 4).
Wood Extraction
Wood extraction was a documented cause of forest loss or degradation in 39 case studies (Supporting Information). Commercial logging was more frequently attributed as a cause than wood extraction for fuelwood or charcoal (39 vs. 26, respectively [Fig. 3 & Supporting Information]), which are common energy sources for many households (Sovacool 2012).
Although Myanmar was arguably the paragon of scientific forest management during the British colonial era, political and financial pressures of an isolated military dictatorship led to a pattern of unsustainable timber extraction for export (Global Witness 2003). Ceasefire agreements between the Burmese regime and armed insurgent groups led to extensive deforestation in ethnic uplands. For example, along the Kachin–Yunnan border, ceasefire deals enabled the Burmese Government to gain greater territorial control over ethnic territories through the allocation of resource concessions (Woods 2011). These concessions facilitated timber trade and its legalization solely through military or state channels that were backed by foreign governments and investors. As a result, both timber profits for the Union Military and allied companies were legitimized, and access to resource rents by ethnic insurgent groups was removed (Woods 2011). Illegal extraction of valuable timber (such as rosewood) for export to China was reported in militarily and politically contested areas in Kachin (EIA 2015).
Infrastructure Development
Infrastructure development (mainly roads, bridges, communication networks, and educational and health facilities) was a key proximate cause (28 case studies [Fig. 3 & Supporting Information]) that not only contributed directly to forest loss or degradation, but also catalyzed the development of several other proximate causes (Fig. 3 & Supporting Information). For example, the Myawaddy–Thingannyinaung Road in Kayin bisected several protected forest areas and facilitated logging (Burma Environmental Working Group 2011).
Hydroelectric Dams and Mining
Hydroelectric dams were identified in 18 case studies (Supporting Information) as a cause of deforestation. Logging, for example, was reported in anticipation of the proposed Tasang Dam in Shan State (Burma Environmental Working Group 2011). As an illustration of potential land-use change, International Rivers (2009) estimated that completion of the (currently suspended) Myitsone Dam would inundate 76,600 ha of forest and displace over 15,000 people (Burma Environmental Working Group 2011).
Mining was reported as a cause of deforestation in 19 cases (Fig. 3 & Supporting Information). Mining affected a relatively small aggregate area (83,647 ha in 2015 [LaJeunesse Connette et al. 2016]) but was reported to be causing forest clearing for small-scale artisanal mines, particularly along the Chindwin and Irrawaddy Rivers in Kachin (Burma Environmental Working Group 2011).
Weak Land Tenure
Tenure security is weak in Myanmar. The state retains ultimate ownership of all land and the right to withdraw land-use rights if use conditions are unmet (Scurrah et al. 2015). The Ministry of Agriculture and Irrigation's 30-year Master Plan for the Agricultural Sector (2000–2030) aims to convert over 4 million ha of “wasteland” to commercial agriculture (Scurrah et al. 2015)—two laws enabled this at the expense of smallholder land rights (Fig. 5). The 2012 Farmland Law (Fig. 5) allowed farmland cultivation rights to be attained and traded through land-use certificates (LUCs). However, the bureaucratic and financial costs of obtaining LUCs effectively limited the acquisition of LUCs to approximately 15% of farmers (Displacement Solutions 2015; Scurrah et al. 2015). Farmers without LUCs were (and remain) effectively without statutory land-use rights and vulnerable to land confiscations.
The 2012 Vacant, Fallow, and Virgin (VFV) Land Management Law (Fig. 5) allowed VFV lands of up to 20,234 ha to be leased to public citizens, private-sector investors, government entities, and nongovernmental organizations for MiLoHyAg (and other) purposes for up to 30 years. The vague definitions of land use in this law have enabled forested land and land occupied by farmers lacking LUCs to be legally designated as VFV and therefore eligible for allocation as an agricultural concession. Consequently, some lands have been confiscated from smallholder farmers, allocated as land concessions for MiLoHyAg activities by investors, and subsequently deforested (Oberndorf 2012; Scurrah et al. 2015).
Economic Pressures
Although liberalization of Myanmar's previously socialist economy began in 1988, economic growth and foreign investment in the 1990s and 2000s were limited by economic mismanagement, the Asian financial crisis of 1997, and U.S. and EU sanctions (Findlay et al. 2016). Political reforms in 2011 led to the partial lifting of the U.S. and EU financial and trade sanctions. With improved international trade conditions, a revised Foreign Investment Law 2012, and ongoing international demand for natural resources, Myanmar became more attractive to investors (Fig. 5): foreign investment into Myanmar increased from US$0.901 billion in 2010 to US$4.08 billion in 2015 (World Bank 2017). However, because official figures do not include a large informal trade across Myanmar's land borders in conflict areas, increased investment may be conflated somewhat with increasing regulation of the informal economy. Because foreign investment was generally channeled into Myanmar through joint ventures with domestic (often military linked) companies as a safeguard against political instability, bureaucratic burdens, taxes, and the perceived risk of asset confiscation (Woods 2012), it was difficult to differentiate the respective effects of increasing foreign and domestic investment on forest loss and degradation.
Civil War
Contested territorial control is at the root of many social and political processes that affect forest loss and degradation. Civil war, which began soon after independence in 1948, has been characterized principally as the struggle between the Union Military and nonstate armed groups fighting for greater autonomy for non-Bamar ethnic groups and their homelands (Callahan 2007).
The extraction of natural resources and development of commercial agriculture have been entwined with the dynamics of civil war and cease-fire deals (Fig. 5). The latter were used as opportunities by the Union Military and state to confiscate land, grant concessions for logging and agriculture, extract timber, and develop large-scale plantations of crops such as rubber (Woods 2011). Lands confiscated directly in war, or in the aftermath of cease-fire deals, were often allocated for MiLoHyAg concessions. For example, land previously left fallow by upland swidden agriculturalists was in some cases reclassified as “wasteland” and converted to rubber plantations (Woods 2011). The grievances resulting from these land seizures in the wake of cease-fire ultimately undermined these truces and incentivized further conflict (Brenner 2015) (Fig. 5).
Profits from MiLoHyAg activities on confiscated lands in conflict zones—such as the minerals (particularly jade) and logs illegally exported across Myanmar's land borders—were key sources of revenue for both the Union Military and nonstate armed groups such as the Kachin Independence Army (Woods 2011; EIA 2015) (Fig. 5). This in turn generated further revenue for the Union Government and Military and consolidated Union political and territorial control (Woods 2011).
War resulted in tragic costs to people in Myanmar's border regions, where many suffered human-rights abuses at the hands of the Union Military, including forced evictions (South 2007). By 2014, an estimated 642,000 people had been displaced within Myanmar (IDMC 2014). Internal displacement in some cases has led to localized deforestation around relocation sites. In turn, the land they were displaced from was in some cases degraded and deforested by government and private companies for MiLoHyAg activities (South 2007; MacDonald 2013).
Environmental Policies and Concerns
Forest-management concerns led to policies (the 2003 Rules on Protection of Wildlife and Protected Area Conservation Law in the 1994 Protected Areas Law and the 2014 log-export ban [Aung 2007; Springate-Baginski et al. 2016]) that helped reduce the overall forest loss and degradation by limiting the areas in which MiLoHyAg concessions could be granted (Fig. 5). Many protected areas, however, were poorly managed and limited in resources and capacity (Rao et al. 2002; Aung 2007).
The 2014 ban on exporting raw logs was enacted to reduce the pressure on unsustainably exploited timber resources (Springate-Baginski et al. 2016). However, timber exports persisted despite the ban (Woods 2013). There was a reported increase in logging prior to enactment of the ban as harvesters and traders were awarded a “final round” of harvesting opportunities in anticipation of the ban (Springate-Baginski et al. 2016). Civil war also undermined enforcement, and the illegal export of logs persisted after the ban, particularly across the Kachin–Yunnan border (EIA 2015).
Corruption
Corruption was prevalent in, for example, the application processes for permits or documents, in which bribes were reported to be unavoidable (Springate-Baginski et al. 2016). Corruption in the case studies we analyzed affected both the channeling of investment into the granting of MiLoHyAg concessions (e.g., the bribing of forestry officials to obtain agricultural concessions in forestry reserves [Woods 2015]) and in the development of these concessions, such as the payment of bribes by domestic companies to subcontract and overharvest logging concessions of the Myanmar Timber Enterprise (Springate-Baginski et al. 2016) (Fig. 5).
Discussion
The key proximate causes of forest loss and degradation in Myanmar from 1995 to 2016 have been documented in many other parts of the world (Angelsen & Kaimowitz 1999; Geist & Lambin 2002; Gibbs et al. 2010; Laurance et al. 2015). However, the underlying drivers (i.e., the key points for policy intervention) were deeply embedded within Myanmar's socioeconomic and political context (Fig. 5). Key issues specific to Myanmar included the influx of foreign investments via domestic companies in the wake of sanctions being lifted, natural resource and land appropriation during and after civil war, and laws that favored the Union Government and foreign investors over smallholder rights.
Increased demand for agricultural expansion is a fundamental driver of deforestation. A range of case studies showed that both commercial and subsistence agriculture drive forest loss and degradation, although more studies implicated the former (Fig. 3 & Supporting Information). There is an ongoing debate over the relative contributions of commercial and subsistence agriculture (including swidden agriculture) to forest loss and degradation. Although we have insufficient data to comment on relative areas affected by different types of agriculture, previous analyses elsewhere suggest the role of swidden agriculture in causing permanent forest loss may be overemphasized (van Vliet et al. 2012).
The allocation of commercial agriculture concessions, a central point in the causal network, is likely to be the critical point in the system to target (Fig. 5). Particular vigilance may be warranted to preempt the surreptitious use of agricultural concessions to circumvent the log-export ban (Woods 2015). However, because the underlying demand is likely to increase, it will be important to ensure that future expansion or intensification in agricultural production does so in ways that protect smallholder tenure security and are located away from priority areas for forest conservation. A new Environmental Impact Assessment Law (MOECAF 2015) should help avoid development in areas of high conservation value, provided that relevant capacity for monitoring and enforcement is developed.
The narratives in many of the case studies suggested that tenure insecurity facilitated land seizures that led to forest conversion and incentivized deforestation to claim land-tenure rights and preclude seizure (Woods 2015). The narratives also suggested that improvements to land-tenure security would lower the possibility of land confiscation for MiLoHyAg activities and concomitantly lower the possibility of forest loss and degradation (Fig. 5). However, the association between greater tenure security and reduced deforestation remains uncertain because global analyses suggest that greater tenure security could incentivize long-term forest conservation in some cases but incentivize long-term investment in land conversion in others (Busch & Ferretti-Gallon 2014; Robinson et al. 2014). This ambiguity is reflected in our causal-network diagram; increasing land-tenure security would increase smallholder commercial agriculture but also reduce land confiscation for MiLoHyAg activities (Fig. 5). Understanding the tenure-deforestation linkages within Myanmar is a priority area for future research to highlight areas where improving tenure security may lead to unintended deforestation outcomes.
Our results suggest, in alignment with previous studies (Woods 2011; EIA 2015), that civil war plays a role in the cycle of forest loss and degradation. Some areas of Myanmar outside direct Union Government and Military control have retained high forest cover under the watch of nonstate armed groups—perhaps due to conservation action by nonstate political groups such as the Karen National Union, which has set up its own protected-area network (KNU 2014), and the deterrence of large-scale commercial activities. A key dynamic described in the case studies was the confiscation of forested land by Union Military in conflict zones for development into MiLoHyAg concessions, often by private interests with connections to the Union Military (Woods 2011). These land seizures, particularly for agricultural concessions, served the dual purpose of profit generation and the establishment of effective Union control over contested landscapes (Woods 2011). Although the majority of deforestation occurred in Union-controlled territory, both Union Military and nonstate armed groups such as the Kachin Independence Army were reported to have profited from illegal logging (EIA 2015). The dynamics of establishing agribusiness (and other MiLoHyAg) concessions in disputed territories as a way of bringing the land under government control had precedent in the responses of the Thai, Malaysian, and Indonesian governments from the 1950s through the 1970s to insurgencies based in highly forested areas (Peluso & Vandergeest 2011). Our synthesis of the case-study evidence suggests that land conflict is closely linked to territorial control and revenue generation from the rich resources that can be found in rebel territories. An important caveat is that because our case studies did not cover all studies equally (e.g., no case studies from Rakhine [Fig. 2], which had high levels of internal conflict and displacement [Callahan 2007]), it is possible that we have missed some regional variation in the ways conflict affected forest conservation.
Explicit discussion of the responsibility and capacity for managing forest resources sustainably may need to be an important component of a lasting peace settlement. However, these dynamics are likely to be region specific within Myanmar, indicating a need to carefully consider the regional historical context. Conservation planning needs to be mindful both of ongoing conflict dynamics (Hammill et al. 2016) and the need to enact inclusive proconservation policies for the postconflict landscape (Baptiste et al. 2017).
Our conceptual model allows us to look at reforms begun or enacted since 2015 and to consider their potential effects. For example, the Myanmar Investment Law under discussion at time of writing in 2017 is intended to streamline the investment process by merging the different processes for foreign and domestic investors (Table 1). Based on our causal network, we believe this new law has the potential to accelerate the allocation and conversion of forested land into MiLoHyAg concessions. The National Land Use Policy (NLUP) of 2016, however, is intended to enhance tenure security; our analysis suggests that this will both reduce deforestation and improve the livelihoods of smallholders by reducing the potential for land confiscation (Fig. 5 & Table 1). However, the reforms in the NLUP will affect only privately held land. Measures to reduce the conversion of forest in land managed by the Forest Department—the source of many recent agricultural concessions—will require active policy interventions, including measures to reduce corruption.
| Policy reform | Predicted effects on causal network | Potential effect on forest loss or degradation |
|---|---|---|
| Existing policies | ||
| National Land Use Policy (2016) | Intends to improve land tenure security for smallholders. If successful, likely to decrease land confiscations and increase smallholder commercial agriculture. | unclear—causal network suggests increase in forest loss and degradation, but author narratives suggest decrease |
| EIA law (2015) | Intended to strengthen existing EIA procedures, thereby strengthening the inhibitory effect of environmental concerns on allocation of MiLoHyAg* concessions. This may ultimately decrease forest loss/degradation. | decrease |
| Myanmar Investment Law (2017) | Merges regulations for foreign and domestic investment. A streamlined process may lead to increased investments in MiLoHyAg. | increase |
| Suggested policies | ||
| Create a public registry of concessionaire owners and responsibilities. | Combat corruption in the allocation and development of MiLoHyAg concessions. | decrease |
| Improved funding for protected areas. | Strengthen the inhibitory effect of protected areas on the allocation of MiLoHyAg concessions. | decrease |
| Generate revenue from forests through payments for ecosystem services (including, but not limited to, carbon storage). | By providing alternative sources of corporate, government, and smallholder revenue, the economic pressure to develop MiLoHyAg activities would be reduced. | decrease |
| Improved planning for infrastructure—explicit spatial planning that links population centers without bisecting forested areas. | This would reduce the impacts of forest loss and degradation through infrastructure construction. | decrease |
- *Mining, logging, hydropower, and agricultural activities.
Because opportunities for investment are likely to increase, we expect that more revenue will be generated, boosting the attractiveness of Myanmar's economy to outside investors and strengthening a reinforcing loop that is likely to accelerate the proliferation of environmentally destructive industries. A long-term strategy to break this reinforcing loop may be needed to decouple government revenue from environmentally destructive activities (Table 1). A recent ecosystem services valuation estimated that the economic net gains of forest conservation would be US$20 billion over the next 20 years (Emerton & Aung 2013). The key challenge, as in other developing countries with extensive remaining forests, will be to establish effective financial mechanisms to capture the value of standing forests, including payments for ecosystem services (e.g., carbon sequestration) by industries (in natural resource extraction and tourism) or foreign donors.
Our systems-thinking approach and literature review allowed us to situate forest loss and degradation in Myanmar within a dynamic, complex system with nonlinear and history-dependent characteristics. The importance of approaching environmental problems as complex and adaptive social-ecological systems has been emphasized before (Levin et al. 2013). The approach we took showed how case studies can be used to systematically identify and map components of the system. Using this approach, we found that the system in which the drivers and causes of forest loss and degradation are situated is governed by reinforcing feedbacks that accelerate forest loss and degradation because the revenue generated is reinvested to drive more economic activity in commercial agriculture and logging. Understanding these dynamics offers insight into the key areas where policy solutions are required: ensuring that MiLoHyAg concessions are not allocated in forested areas, decoupling government revenue from environmentally destructive activities, improving land-tenure security, tackling corruption, and ensuring that negotiations over natural resources are included in peace settlements (Table 1). Understanding these threats and possible solutions improves the likelihood that Myanmar's transition, unlike those in other countries, does not result in widespread forest loss or degradation.
Acknowledgments
This research was supported by an Academic Research Grant (Tier 1) from the Ministry of Education and MOE grant number MOE2015-T2-1-131. We thank R. Oberndorf, O. Springate-Baginski, G. Connette, K. LaJeunesse Connette, and J. Richardson for technical advice. We also thank J. Phelps and 2 anonymous referees for comments that improved the manuscript.
