Development, Environment and Political Economy in Nepal
GoLFor-DEEPN aims at evaluating the effects of forest management decentralization in Nepal on (i) forest cover, (ii) collective action at local level and (iii) political economy at the municipality level. After decades of deforestation and forest degradation, tree cover slightly improved over the last 20 years. There is very little work done to understand these changes at the country level taking into account competing explanations such as economic growth, rural exodus, civil war, the development of alternative energy sources, tree plantations on private land.<br /><br />GoLFOR-DEEPN aims at quantifying the effects of Community Forest Users Group on the stabilization of tree cover in Nepal, while shedding lights on the underlying mechanisms and discussing the distributional impacts on welfare at the local level and on power relation. We structure the project around 5 testable hypothesis:<br />a) Has decentralization of forest management increased forest cover ?<br />b) Has the creation of user groups negatively affected nearby areas? Are negative spillover driving the creation of new groups?<br />c) Has restricted access to forest resources fostered the adoption of new technologies such as biogas installations or improved cooking devices? <br />d) Is public good provision at the local level by udser groups driven by inequalities?<br />e) Can local management of a natural resource be considered as a proto-democratic institution in a context where they were no local elections during 20 years? Do (good?) managers of user groups have more chance to be elected at higher constituency level?
The projects combine 3 disciplinary approaches, namely economics, remote sensing and human geography, with their own methods. On top of that, GoLFor-DEEPN sparks numerous discussions and interactions between researchers by organizing regular meetings and joint field missions. This allow the partners to develop, by couple of discipline and with the 3 disciplines altogether to produce a joint knowledge.
In economics, our approach is firmly quantitative, using large samples representative at the country level and relying on statistical inference to characterize relations between variables and estimating causal relationships. Testes hypothesis are fed by field work of human geographers and by measures of forest canopy cover developed by the remote sensing part of the project.
In remote sensing, our approach mostly intersects intra-annual variations in greenness due to the phenological cycle to inter-annual variations in tree cover. On one side, it is the best way to separate trees from grass because they do not have the same seasonality, on the other, interannual variations is what matters while assessing the impact of community forestry, both on areas that they managed and on adjacent areas.
When it comes to human geography, our approach relies on semi-structured interviews with users, managers and people indirectly affected. These interviews require a long time frame to cross-validate answers and go deeper than the “official stories”.
If disciplinary approaches are very different, we could rely on regular meetings and one joint mission in the first 18 months to foster the cooperation between the 3 axis, to confront hypothesis, and to define complementary approaches and a work method that is going to yield a specific scientific production based on a common methodology.
As of April 2020, the main output of the project relates to the data collection, as planned in the first part of the project. We could test our hypothesis during a first joint mission in November 2018 and set the basement of upcoming data collection. The first part of this data collection process happened in the first half of 2019. We scanned constitutions and operational plans of more than 3000 community forest user groups. This is an exhaustive work in 14 randomly selected districts of Nepal (out of 75). Digitized copies are currently analysed using optical character recognition software specially adapted to Nepali and to our specific context. Processed data currently available are incremental to answer hypotheses (a) and (e).
In parallel, we carry on the analysis of the effect of decentralized management on forest cover, both on the economics side than on the remote sensing part of the project. Community forestry appears as a vector of forest regrowth, especially in remote areas in high altitude regions where the value of wood is rather small and collective action easier to organize. In the low lands, where timber is more valuable and where markets are more developed, the story points towards a potential negative effect on tree cover. In remote sensing, we do rely on intra-annual variations in greenness to produce a measure of forest density, on top of discussing forest area, a choice better suited to the Nepali context.
With respect to human geography, we did a first specific field work in Sunsari district. We administered a first set of semi-structured interviews in peri-urban forests and compared community forest management to community water management in these areas. This work sets milestone for further field work in other places over longer time spans.
The prospects for the project are rather optimistic, as long as the ongoing sanitary crisis does not limit too much our ability to do some field work. This is critical, especially for the qualitative part of the project.
In the coming months, the priority is to process the scanned documents of the 3000+ community forest user groups that we collected last year. This is a crucial step to get some comprehensive information about group members, forest resources, and forest polygons over a wide range of user groups (we have such documents for more than 1/8 of all the groups in the country). It is incremental in estimating the link between access to forest resources and inequality, linking group members and electoral outcomes, etc. All project hypothesis can be addressed using these data, except hypothesis d). Bad quality and low coverage of annual accounts prevent us from using them in a consistent way. This hypothesis will be addressed through the use of secondary information and qualitative work.
In remote sensing, the short term goal is the validation of the first measures showing an increase in leaf area over the last 20 years and disentangling competing explanation such as a longer period vegetation, an increase of herb layer within forests or an increase in tree cover. We plan to find the optimal months maximizing the spread in greenness between the tree and the herb layer. We rely on middle range resolution sensor (Sentinel 2) to cover the whole country and then scale up the method to use sensor with a longer historical depth to match the tracking of forests with the time span on the management reform.
In human geography, extensive field work should have started in April 2020 but we have to postpone the start of this work in Autumn 2020 or maybe even Spring 2021.
Database on more than 3000 community forest user groups in Nepal. even if the department of forest in Nepal initiated the program, this is a peice of information that does not exist in a centralized way.
Scientific communications during workshops (Germany and India)
GoLFor-DEEPN aims at evaluating, at the national level, the effect of handing over the management of forest from the Department of Forest to Community Forest User Groups on (i) forest cover, (ii) public good provision at the local level, and (iii) local political governance. After decades of continuous deforestation, forest cover slightly improved over the last 20 years. No existing large scale study analyses precisely the cause of this improvement. Economic growth, emigration from rural area, conflict, alternative energy development, or plantation of trees on private plots are other confounding factors that could explain a reduction of forest degradation.
GoLFor-DEEPN wants to quantify the contribution of this decentralization programme to the stabilization of forest cover. It aims at carving out the underlying mechanisms and discuss the distributional consequences of community forestry both with respect to local livelihoods and to local politics. It is structured around five testable hypothesis:
(i) Does handing-over the management of forests to their local users increases forest cover?
(ii) Does the creation of CFUGs cause negative externalities on adjacent areas? Do these potential spillovers increase the probability of the emergence of local collective action in nearby villages?
(iii) Do restrictions to the harvesting of forest products imposed by CFUGs increase the adoption of substitute energy (gas, kerozen) and alternative technologies (such as biogas or improved cookstoves) among villagers living close to forest plots newly managed at the local level?
(iv) Does the provision of local public goods depend on initial inequalities?
(v) Can we see local management of natural resources as form of proto-democracy which helps to select leaders for elections at a higher level?
In terms of methodology, it will rely on state-of-the-art statistical analysis, remote sensing and qualitative work. The econometric analysis will rely on a combination of first hand collected data, survey data, administrative records, and of course remote sensing images. The qualitative approach will be carried on in parallel to inform and to be questioned by the econometric approach, to get deeper insights into mechanisms, and to assess the potential discrepancies between administrative data and the reality on the field.
Related to the seminal work about collective action (Olson, 1965), the tragedy of the commons (Hardin, 1968), and the governance of commons by communities (Ostrom, 1990) or their failure (Diamond, 2005), GoLFor-DEEPN builds on the results of Kosfeld and Rustagi (2015), Mansuri and Rao (2013), Baland et al. (2010a) and Somanathan et al. (2009). It innovates in several dimensions. First, its methodology does take into account the environmental and institutional spillovers on areas not managed by communities combining advanced methods in remote sensing with econometrics techniques. Second, it combines qualitative work in villages to country-wide scale statistical analysis at the frontier of the most recent standards in economics research. This interaction will be very useful to understand the external validity of findings, for other resources, in Nepal or elsewhere. Third, the timing of the project is perfect to analysis how local governance can shape an emergent democracy at the local level since the first local elections after 20 years took place in 2017. This looks very much as a bottom-up approach where leaders with restricted responsibilities in terms of population and set of actions run for the elections and will occupy positions with broader scopes. This transition from a proto-democracy to a formal democracy is seldom discussed in the literature.
Monsieur François Libois (ECOLE D´ ECONOMIE DE PARIS)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
EEP ECOLE D´ ECONOMIE DE PARIS
Help of the ANR 189,000 euros
Beginning and duration of the scientific project: September 2018 - 42 Months