“Spillovers from agricultural processing” [paper] [presentation] New draft May 2019. Submitted and under review.


Abstract: This paper uses the proliferation of palm oil factories across Indonesia's undeveloped hinterland as a natural experiment to study industrial onset and estimate spillovers from agricultural processing. The main finding is signs of urbanization and structural change around factories: more non-agricultural employment, higher incomes, and more people, firms, and other economic and social organizations. These patterns can at least partially be explained by economic linkages, infrastructure and other public goods, and economies of scale in production. By focusing on subsistence rural regions in a large developing economy, this paper adds a globally-significant new case to a growing literature emphasizing the importance of agglomeration externalities for understanding the birth of new towns, the spatial distribution of economic activity, and structural transformation.


“Export agriculture and rural poverty: evidence from Indonesian palm oil”  [paper] [presentation] New draft March 2019. Submitted and under review.

Abstract: This paper measures the impacts of the world's largest modern agricultural export expansion—that of Indonesian palm oil since 2000—on poverty and consumption in producing communities. Identification exploits external demand growth and geographic differences in cultivation suitability in a difference-in-difference instrumental variables design. The main finding is that growth in palm oil sector lifted up to 2.6 million rural Indonesians from poverty this century. The median expansion led to 2.7 percentage points faster poverty reduction and 4 percent faster consumption growth. Divergent regional development trajectories can be explained by rising returns to labor and land, and indirect effects through increased household investment, government revenue, and rural economic and social infrastructure in producing communities.

“Causes of Indonesia’s forest fires”, with Wally Falcon, Matt Higgins, and Roz Naylor (Stanford) [paperSubmitted and under review.


Abstract: The economic costs of Indonesia’s 2015 forest fires are estimated to exceed US $16 billion, with more than 100,000 premature deaths. On several days the fires emitted more carbon dioxide than the entire United States economy. Here, we combine detailed geospatial data on fire and local climatic conditions with rich administrative data to assess the underlying anthropogenic causes of Indonesia’s forest fires at district and village scales. We find that El Nino explains most of the year-on-year variation in fire. The creation of new districts increases fire and exacerbates the impacts El Nino on fire. We also find that regional economic growth has gone hand-in-hand with the use of fire in rural districts. We proceed with a 30,000-village case study of the catastrophic 2015 fire season on Sumatra and Kalimantan and ask which villages, for a given level of spatial fire risk, are more likely to have  fire.  Villages more likely to burn  tend to be more remote, considerably less developed, and have a history of using fire for agriculture. Although central and district level policies have contributed to voracious environmental degradation, the close link between poverty and fire at the village level suggests that the current policy push for village development—and the strengthening of the village as an administrative unit—could offer opportunities to reverse this trend.

“Local impacts of resource booms”, with Paul Burke (ANU) New draft coming soon.


Abstract: We study the local impacts of resource sector booms with different factor intensities. Applying the synthetic control method to Indonesian district-level data over the 2000s, we examine palm oil in Sumatra, coal mining in Kalimantan, and natural gas extraction in West Papua. Findings are consistent with an inverse relationship between concentration of natural resource rents and benefit diffusion, including spillovers to other sectors of the local economy. All three resource booms boosted total economic output in per capita terms. Oil palm expansion raised agricultural, industrial, and services output, while coal mining reduced agriculture and services output. Oil palm and coal mining booms both appear to have delivered strong local poverty reduction. The natural gas project in West Papua delivered a massive increase in industrial and aggregate output, but appears to have had insignificant impacts on household welfare or poverty. Similar patterns are observed in additional cases and using all cases in a panel setup.


Naylor, R.L., Higgins, M.M., Edwards, R.B., and Falcon, W.P. (2019) “Decentralization and the environment: assessing smallholder oil palm development in Indonesia ”, Ambio, Jan, pp. 1—14. [DOI]

Edwards, R.B. (2017) “The Oil Palm Complex” (invited book review), Journal of Southeast Asian Economies, 34: 2, pp. 430-431 [DOI

Edwards, R.B. (2016) “Mining away the Preston curve”, World Development, 78, February 2018, pp. 22—36. [DOI]


“Fight fire with finance: a randomized field experiment to curtail land-clearing fire in Indonesia” with Wally Falcon, Grace Hadiwidjaja, Matt Higgins, Roz Naylor (Stanford), and Sudarno Sumarto (Stanford, TNP2K, and SMERU). AEA Registry [DOI] Completed January 2019. Papers coming soon.

fire2 2.jpg
fire3 2.jpg

Project summary: Payments for ecosystems services (PES) and conditional cash transfers (CCTs) are popular and often effective policy approaches to spur behavior change: paying people to undertake behaviors they otherwise would not. Usually, the behavior being “incentivized” benefits society, for example reducing deforestation with PES and vaccination and school attendance in CCTs. However, there remains limited evidence on the effectiveness of PES-type interventions, particularly in weak institutional settings, and no studies have focused on fire, the main way to clear land in the developing tropics. Our study is based in West Kalimantan, a province where Indonesia’s catastrophic 2015—16 fires were concentrated. We estimate the effects of conditional cash transfers to reduce fire—offered as PES contracts to village governments—by randomly assigning 75 villages to the program and 200 to a comparison group. Villages must not set any fire and promptly extinguish any natural fires to receive their ex-post payment. Hence, payment is conditional on performance, which we monitor using satellite data and field verification. Our intention is to scale up with additional treatment arms—different payment levels and information-only treatments—if the initial results are promising.

“The environmental impacts of agricultural processors on the frontier: evidence on fire, deforestation, and land use change in Indonesia”

“Electoral conflict spillovers” with Yusaku Horiuchi (Dartmouth) & Daniel Suryadarma (ANU and RISE).

“Plantation agriculture and child labor” with Eric Edmonds (Dartmouth).

“The local amenity and crime impacts of new marijuana retailers” with Brady Horn (UNM) & Paul Zachary (UCSD).

“Trade adjustment in a commodities-for-manufactures boom”