Delacote, Robinson and Roussel (2016) built a spatially explicit patch model comparing four avoided-deforestation instruments — hotspot payments for environmental services, full-coverage payments, conservation areas, and an agglomeration bonus — under the assumptions of perfect, costless monitoring and direct policy costs only. They explicitly left open how costly, imperfect, and spatially heterogeneous monitoring and enforcement of deforestation and payment conditionality, together with the transaction costs of program implementation, alter spatial spillovers, leakage magnitudes, and cost-effectiveness rankings. Later work has partially advanced the enforcement-allocation half of the question for protected areas (Albers et al. 2019), but no work integrates monitoring and transaction costs for conditional-payment instruments into a leakage-aware landscape model. The problem remains open.

Background

Reducing Emissions from Deforestation and forest Degradation (REDD+) is the family of policies under which developing countries and projects are rewarded, through payments or carbon credits, for lowering forest loss. A central concern is leakage: deforestation reduced in one location may simply be displaced to other locations, so that the policy's net effect on the landscape is smaller than its local effect. Early economy-wide studies estimated such displacement from forest carbon programs (Murray, McCarl and Lee, 2004), and empirical work found 'slippage' of cropland and forest loss around conserved areas and payment programs (Alix-Garcia, Shapiro and Sims, 2012). Pfaff and Robalino (2012) emphasized that deforestation decisions are spatially contagious — one landholder's clearing changes the returns to clearing by neighbours — so policy impact prediction must be spatially explicit. Delacote and Angelsen (2015) showed theoretically that conservation and degradation-reduction efforts can generate either leakage or synergies depending on market structure.

Delacote, Robinson and Roussel (2016) embed these spatial interactions in a patch model. A landscape is divided into forest patches owned by agents of two types (high-deforestation 'X-agents' and others); each agent's per-period payoff from deforestation depends on its own clearing DitD_{it}, patch characteristics, and its neighbours' lagged clearing on adjacent patches, weighted by an interaction intensity αij\alpha_{ij} between 0 and 1. Deforestation is permanent and monitoring is assumed perfect and costless. They compare four policies: (i) a hotspot payment for environmental services (PES) — direct payments conditional on forest cover, targeted at high-deforestation patches; (ii) a full PES available to all agents; (iii) a conservation area (CA), a command-and-control set-aside in which deforestation is simply prohibited (the instrument whose siting interacts with enforcement, as in Robinson, Albers and Williams, 2011); and (iv) an agglomeration bonus (AB), which pays a direct transfer plus a bonus when adjacent patches are both conserved (Parkhurst and Shogren, 2008). Their methodological contribution is a taxonomy that decomposes policy-induced displacement into 'TNT leakage' (spillover from treated to non-treated patches — the usual definition), 'TT leakage' (spillovers among treated patches), and 'NTNT leakage' (secondary spillovers among non-treated patches), together with a dynamic leakage metric, under clustered versus dispersed landscape configurations. Their findings include: PES schemes are the least-cost and most cost-effective instruments for avoided deforestation but generate the most leakage per unit spent; the CA performs best when patch types are dispersed and worst when clustered; the AB is effective in clustered landscapes but at higher cost; and stronger interactions raise the temporal variability of landscape-level avoided deforestation, so policies must be evaluated over long horizons.

The model's cost accounting, however, includes only the direct payment or compensation costs of each instrument (footnote 25 of the paper explicitly excludes transaction costs, 'which may be very high for PES implementation'), and it assumes perfect, costless monitoring of deforestation and of compliance with payment conditionality. The concluding section states the open issue directly: in reality monitoring is costly, often imperfect, and monitoring costs most likely vary with location, so the spatial distribution of enforcement activities will in turn affect spatial spillovers — a consideration also stressed for reserve sizing by Robinson, Albers and Williams (2011), Pfaff and Robalino (2012), and Sims (2014) in the context of forest fragmentation. How these costs and imperfections change the leakage and cost-effectiveness conclusions was left unresolved.

Problem Statement

In a spatially explicit landscape model of avoided-deforestation policy with neighbour interactions — the patch framework of Delacote, Robinson and Roussel (2016), or an equivalent model in which agents' clearing decisions respond to neighbours' lagged clearing through interaction intensities αij\alpha_{ij} and policy-induced displacement decomposes into TNT, TT, and NTNT leakage — how are spatial spillovers, the magnitude and spatial distribution of leakage, avoided deforestation, and the cost-effectiveness ranking of avoided-deforestation instruments (hotspot payments for environmental services, full-coverage payments, conservation areas, and agglomeration bonuses) altered when monitoring and enforcement are imperfect, costly, and spatially heterogeneous, and when the transaction and enforcement costs of program implementation are included alongside direct payment costs? The source analysis assumes perfect and costless monitoring of deforestation and of compliance with payment conditionality and counts only direct policy costs, whereas monitoring costs vary over space, and enforcement capacity and its spatial allocation are chosen by the policy maker.

References

  1. Delacote, P., Robinson, E.J.Z., Roussel, S. (2016). Deforestation, leakage and avoided deforestation policies: A spatial analysis. Resource and Energy Economics 45:192–210. DOI 10.1016/j.reseneeco.2016.06.006, https://doi.org/10.1016/j.reseneeco.2016.06.006
  2. Murray, B.C., McCarl, B.A., Lee, H. (2004). Estimating Leakage from Forest Carbon Sequestration Programs. Land Economics 80(1):109–124. DOI 10.2307/3147147, https://doi.org/10.2307/3147147
  3. Alix-Garcia, J.M., Shapiro, E.N., Sims, K.R.E. (2012). Forest Conservation and Slippage: Evidence from Mexico's National Payments for Ecosystem Services Program. Land Economics 88(4):613–638. DOI 10.3368/le.88.4.613, https://doi.org/10.3368/le.88.4.613
  4. Pfaff, A., Robalino, J. (2012). Protecting forests, biodiversity, and the climate: predicting policy impact to improve policy choice. Oxford Review of Economic Policy 28(1):164–179. DOI 10.1093/oxrep/grs012, https://doi.org/10.1093/oxrep/grs012
  5. Delacote, P., Angelsen, A. (2015). Reducing Deforestation and Forest Degradation: Leakage or Synergy? Land Economics 91(3):501–515. DOI 10.3368/le.91.3.501, https://doi.org/10.3368/le.91.3.501
  6. Parkhurst, G.M., Shogren, J.F. (2008). Smart Subsidies for Conservation. American Journal of Agricultural Economics 90(5):1192–1200. DOI 10.1111/j.1467-8276.2008.01203.x, https://doi.org/10.1111/j.1467-8276.2008.01203.x
  7. Robinson, E.J.Z., Albers, H.J., Williams, J.C. (2011). Sizing Reserves within a Landscape: The Roles of Villagers' Reactions and the Ecological-Socioeconomic Setting. Land Economics 87(2):233–249. DOI 10.3368/le.87.2.233, https://doi.org/10.3368/le.87.2.233
  8. Sims, K.R.E. (2014). Do Protected Areas Reduce Forest Fragmentation? A Microlandscapes Approach. Environmental and Resource Economics 58(2):303–333. DOI 10.1007/s10640-013-9707-2, https://doi.org/10.1007/s10640-013-9707-2
  9. Albers, H.J., White, B., Robinson, E.J.Z., Sterner, E. (2019). Spatial protected area decisions to reduce carbon emissions from forest extraction. Spatial Economic Analysis 15(3):280–298 (print issue 2020). DOI 10.1080/17421772.2019.1692143, https://doi.org/10.1080/17421772.2019.1692143
  10. Albers, H.J., Preonas, L., Capitán, T., Robinson, E.J.Z., Madrigal-Ballestero, R. (2020). Optimal Siting, Sizing, and Enforcement of Marine Protected Areas. Environmental and Resource Economics 77(1):229–269. DOI 10.1007/s10640-020-00472-7, https://doi.org/10.1007/s10640-020-00472-7
  11. García, J.H., Orlov, A., Aaheim, A. (2018). Negative leakage: The key role of forest management regimes. Journal of Forest Economics 33:8–13. DOI 10.1016/j.jfe.2018.09.001, https://doi.org/10.1016/j.jfe.2018.09.001
  12. Moffette, F., Gibbs, H.K. (2021). Agricultural Displacement and Deforestation Leakage in the Brazilian Legal Amazon. Land Economics 97(1):155–179. DOI 10.3368/wple.97.1.155, https://doi.org/10.3368/wple.97.1.155
  13. Warnes, X., de Zegher, J.F., Iancu, D., Plambeck, E. (2023). Area Conditions and Positive Incentives: Engaging Local Communities to Protect Forests. SSRN working paper. DOI 10.2139/ssrn.4609761, https://doi.org/10.2139/ssrn.4609761
  14. Streck, C. (2021). REDD+ and leakage: debunking myths and promoting integrated solutions. Climate Policy 21(6):843–852. DOI 10.1080/14693062.2021.1920363, https://doi.org/10.1080/14693062.2021.1920363
  15. Filewod, B., McCarney, G. (2023). Avoiding carbon leakage from nature-based offsets by design. One Earth 6(7):790–802. DOI 10.1016/j.oneear.2023.05.024, https://doi.org/10.1016/j.oneear.2023.05.024
  16. Andrews, J., Hillis, V., Clark, M., Borgerhoff Mulder, M. (2025). Adaptive responses to inter-group competition over natural resources: the case of leakage. Ecology and Society 30(2). DOI 10.5751/es-15776-300216, https://doi.org/10.5751/es-15776-300216
  17. Ungar, M., Corredor-Garcia, J. (2026). Enforcing environmental law in the Amazon. People and Nature 8(7):2281–2294. DOI 10.1002/pan3.70342, https://doi.org/10.1002/pan3.70342
  18. Atmadja, S., Verchot, L. (2011). A review of the state of research, policies and strategies in addressing leakage from reducing emissions from deforestation and forest degradation (REDD+). Mitigation and Adaptation Strategies for Global Change 17(3):311–336 (print issue 2012). DOI 10.1007/s11027-011-9328-4, https://doi.org/10.1007/s11027-011-9328-4