JULIA HELIE

This paper examines how reelection incentives influence politicians' decisions on deforestation, using Indonesian district-level panel data from 2005 to 2019 along with pixel-level environmental data. By exploiting quasi-random election timing and term limits, the study isolates the impact of reelection incentives on deforestation. The results show a significant increase in deforestation in the year leading up to an election, only when the incumbent is eligible for reelection. This effect is most pronounced in highly competitive political environments. Furthermore, the paper investigates the underlying mechanisms driving these reelection incentives effects, including alignment with the electorate and securing campaign funding. Firstly, legal land use classification and post-deforestation land use are employed to identify beneficiaries and political motivations. Subsequently, the study refines its analysis by examining fire-induced deforestation to assess sensitivity to more observable and local externalities. Finally, by matching land permits and public procurement data with a unique campaign donations dataset, the study explores potential links between deforested areas and campaign donations.

This paper investigates whether electoral accountability disciplines politicians. This question has been thoroughly studied and the political agency model suggests that this is the case. However, showing it empirically faces a big identification challenge: disentangling accountability from the selection and competence effects, which are going in the same direction. This paper is the first to properly disentangle both by exploiting a unique identification relying on the unexpectedly retroactive introduction of term limits in Chile. With this method, we analyze the behavior of incumbents when they learn, while in office, that they are not eligible anymore for reelection. We use detailed public procurement data spanning from 2007 to 2022 to build proxies of corruption. Our results indicate that when politicians lose their electoral accountability they move towards more discretionary spending.

Numerous studies rely on remote sensing science to monitor land cover dynamics and their

implications in current socio-environmental issues. With the continuous and simultaneous

development of Earth Observation (EO) sensors and computing through the Big Data era, many maps are produced and open promising expectations for land cover monitoring. However, they may also confuse end-users in choosing suitable data for their work. This is verified for the oil palm (OP) monitoring, widely studied at different spatial and temporal scales with remote sensing images. OP is one of the most important crops in the world’s food supply. Indonesia, the leading global producer of crude palm oil, has seen its forest areas impacted in recent decades with an expansion of the OP surfaces from 4 million hectares in 2006 to 15 million hectares nowadays. This OP growth has huge socio-environmental impacts (e.g. deforestation or biodiversity loss), requiring accurate maps at a high spatial resolution, which still represents an essential challenge for the remote sensing community. In this work, we combined four existing OP maps with the Dempster-Shafer Theory (DST) to (1) produce an improved OP map at the Sumatra scale and (2) localize uncertainty or confidence areas for the OP class, which represents relevant information for further studies and the end-users. Results show that DST improves statistical validation for Sumatra island (i.e. Kappa = 0.780). Results also provide maps with high confidence or uncertainty areas for the OP class. Besides, the spatial variability related to specific OP practices through industrial and smallholder plantations shows higher confidence for the class "industrial OP" than the class "smallholder OP".

This paper explores the impacts of gold mining, specifically artisanal and small-scale mining (ASM) versus large-scale mining (LSM), on local communities' welfare in Tanzania. While LSM generates substantial government revenue, it also leads to conflicts and poor labor conditions, while ASM offers economic opportunities but also poses social and environmental challenges. Using spatial and temporal variations, the study infers causal effects on local households' welfare. The results reveal that an increase in gold prices positively affects welfare near LSM sites but has no significant impact near ASM sites. Moreover, a distributive analysis suggests that the positive effect of mining intensification on the welfare of households living close to LSM is driven by households at the top of the distribution whereas ASM seems to have a positive effect on local consumption for households at the bottom of the distribution.