Effects of Land Use, Market Integration, and Poverty on Tropical Deforestation: Evidence from Forest Margins Areas in Central Sulawesi, Indonesia

Abstract

This dissertation aims to accommodate the complex inter linkages between various factors affecting the tropical deforestation processes by examining the deforestation problem in the vicinity of Lore Lindu National Park in Central Sulawesi, Indonesia. This dissertation consists of three independent papers that analyse tropical deforestation from many perspectives. The first paper identifies driving forces and characteristic processes of land use changes, in this case forest and non-forest. An adapted Hausman-Taylor panel approach was used in a binary dependent variable, alongside with key socioeconomic, geographical, and geophysical factors to examine the determinants of forest cover dynamics. The results indicate that population growth is negatively correlated with forest cover. Furthermore, deforestation occurred even in remote locations, and in areas with steep slopes. The second paper links the drivers of deforestation which occur as a result of cash crop expansion and market integration, using price transmission analysis between domestic and international markets. In order to attain this objective, a wide range of econometric techniques including panel model, cross section and time series analysis was applied. Results indicate that a high degree of integration between local and international markets is one of determinants of the area of land under cacao cultivation. The results confirm that decisions to expand the cultivation of cacao will depend more on expectations of future prices than prevailing prices. The only stabilising effect against cacao expansion is the presence of paddy rice fields in the village. The third paper contributes to the debate on the link between poverty and deforestation, by providing an alternative approach using a village level perspective and broadening the range of poverty measures used to include subjective well-being (SWB) and poverty proxies. Since poverty is a complex phenomenon, both subjective and objective approaches have been utilised. For the subjective approach this in cludes using proxies of different aspects of poverty and SWB at the village level. The relationship between deforestation and a number of explanatory variables has been estimated in a Beta regression framework. Results suggest that there is a non-linear relationship between SWB as well as other proxies of poverty and deforestation. The initially poorest and wealthiest villages have high deforestation rates, whilst the moderately well-off villages have lower deforestation rates. In contrast, the relative poverty assessment of the objective view shows no empirical evidence that poverty increases the deforestation rate. Moreover, additional objective view proxies measuring particular aspects of poverty show unclear patterns; for each aspect used, variables might increase or decrease the deforestation rate. As an example, high illiteracy rates and less access to markets increase deforestation rates, but an increase in the percentage of irrigated land in a village reduces deforestation rates. Nevertheless, overall between 2001 and 2007, the improvement of village well-being helped to reduce the deforestation rate in this region. The results of the study are consistent and highly complementary, and finally provide the more effective policy suggestions within a local context, which will better harmonise the goals of economic growth and environmental sustainability.