5.3.3.3. Options for Responding to Leakage

To date, two approaches have been used /proposed to address leakage; these two approaches may be employed independently or simultaneously. One approach involves addressing leakage at the project level through either project design or re-estimating of net GHG benefits. Some people question whether project-level approaches can adequately ensure that leakage will be addressed, however. As a response, macro-level approaches to address leakage have been proposed that would involve developing regional or national baselines or establishing risk coefficients by project type or characteristic.

Project-level approaches: Leakage potential may be identified at the front end of project design and additional activities incorporated if the project appears vulnerable to leakage. If evidence of leakage emerges after project implementation has begun, project implementers may undertake additional activities to mitigate leakage or to monitor it and subsequently revise net GHG estimates.

Project design elements incorporated in projects: Although experience to date is limited, several elements have emerged that may help avoid leakage, depending on the socioeconomic and physical context of the project. Project design strategies that have been used to avoid leakage include providing socioeconomic benefits to local people that create incentives to maintain the project and its GHG benefits because of these associated benefits and using replicable or transferable technologies that can help avoid leakage because they allow project benefits to be duplicated outside project boundaries, so that social benefits are not restricted to a limited area. Incorporating these elements can help avoid leakage.

Multi-component projects may also help avoid leakage because they can combine project activities to fully address demands that drive land-use change (Chomitz, 2000). For example, the Costa Rican PAP generated carbon offsets by avoiding carbon emissions and through carbon sequestration. The PAP is consolidating approximately 570,000 ha of primary, secondary, and pasture lands within the National Parks and Biological Reserves of Costa Rica (Tattenbach, 1996; Stuart and Moura-Costa, 1998). The PAP plans to reduce deforestation of primary forest, thereby reducing carbon emissions resulting from deforestation. The PAP also plans to allow secondary forest and pasture to regenerate, thereby sequestering carbon through tree growth and accumulation of woody biomass. Concurrently, Costa Rica has also developed a parallel program, the PFP, which provides financial incentives for land owners outside the PAP area to opt for forestry-related land uses as opposed to agriculture-thereby generating a series of environmental services, such as CO2 fixation, maintenance of water quality, biodiversity, and landscape beauty (Forestry Law N. 7575, April 1996) (Stuart and Moura-Costa, 1998). The PFP is also expected to offset the effects of decreasing timber harvest in the project area, reducing possible leakage effects.

Another example of a multi-component project is the CARE/ Guatemala project, which increased fuelwood availability and agricultural productivity by encouraging agroforestry. The project also protected some forest areas, allowing degraded areas to regenerate. The CARE/Guatemala project began in 1988, and persisted through years of political strife and high demand for agricultural land because the project combined elements of forest protection with agricultural extension that provided social benefits that gave local people a stake in the project's success (Brown et al., 1997).

Re-estimation of net GHG benefits: Leakage cannot always be avoided at the outset or mitigated with additional activities. In some cases, GHG estimates can be recalculated. If project implementers can quantify the shortfall in output from the project, they can quantify the amount of leakage (Brown et al., 1997). To recalculate the original net GHG benefits, the project evaluator must determine approximately how much area must be logged or converted to agriculture to compensate for the decrease in output.

For example, the RIL project in Malaysia (see Table 5-2) was originally estimated to avoid 38,700 t of carbon emissions. However, the project may have resulted in carbon leakage because on 450 ha of the 1,400 ha project, timber production was decreased by approximately 49 m³ ha^-1 relative to conventional logging. Total timber shortfall was 450 ha x 49 m³ ha^-1 = 22,050 m³ of reduced timber output. To quantify the amount of potential leakage, it is possible to estimate the additional area that must be logged to make up for the deficit. The leakage potential could be roughly determined by estimating the amount of emissions resulting from logging to compensate for the 22,050 m³ of reduced output. Assuming that RIL makes up for the shortfall, leakage could be estimated as follows: RIL emits 108 t C ha^-1 and yields 103 m³ of timber per ha (Pinard and Putz 1997); therefore, harvesting 214 ha using RIL methods would make up for the reduced output. Leakage then equals 23,112 t C emitted (214 ha x 108 t C ha^-1). Thus, the estimated net carbon benefit is 38,700 t C (the original amount) - 23,112 t C (the leakage) = 15,558 t C.

These estimates are approximate, and they represent only one harvest cycle. They illustrate one means of quantifying leakage. In this example, RIL still results in a net carbon gain-which might or might not be the case for all projects. In addition, RIL projects are designed to increase output over time because there is less damage to young trees. In the long run, RIL sites may produce greater output than conventionally logged sites.

Macro-level approaches: Alternatives to project-based approaches have been proposed, including estimation of empirically based sectoral, national, or regional baselines that can potentially capture leakage, and development of adjustment coefficients for leakage risk and adjustment of net GHG estimates accordingly:

• National and regional baselines: Adopting sectoral or regional baselines on LULUCF emissions and sequestration is one alternative to the project-based approach. If a project attempted to reduce the rate or area of deforestation, the project-level effect would need to be demonstrated in subsequent monitoring and baseline estimates improved. By encompassing a large geographic area, leakage could be internalized. One proposal involves developing national, regional, or sectoral baselines on land-use change and management; this approach is based on the concept of tradable development rights. As above, a regional baseline deforestation rate would be determined by using land-use trends. A certain percentage of the forest would be protected, and the remaining forest could be developed (Chomitz, 2000). Allowance or development rights to the forest would be distributed; the owners of these development allowances could then in turn sell the development rights. In areas where development was allowed, the carbon sequestration services of the forest could be sold instead of developing the forest (Chomitz, 2000).
• Risk premiums and adjustment coeffiecients: To overcome the complexity of quantifying leakage, another approach is assigning specific leakage coefficients (Trexler and Kosloff, 1998). Project estimates could then be adjusted by this coefficient (Gustavsson et al., 1999). These coefficients could be developed at a regional or national level for different project types. The effect of risk premiums or adjustment coefficients is that the projects can claim only a portion of the estimated GHG benefits. A percentage of the net GHG benefit is retained in a buffer to cover the risk of leakage-the goal being to protect the atmosphere from added carbon emissions. For example, the PAP in Costa Rica has a reserve or buffer of carbon sequestered to insure against various risks, including leakage. The PAP assumes that some of the subsistence-based farmers who move from the forest may squat on new land, resulting in leakage. The PAP estimates a low risk of leakage because only 22,223 ha of the total of 530,498 ha is currently in private hands. Therefore, if 25 percent of all private owners choose to buy or occupy new areas of land and deforest them, the carbon offsets arising from over than 1 percent of the total project area would be negated. As a response, in its first year the PAP will offer only half of the estimated emission reductions for sale, with the rest serving as an insurance buffer for this and other estimated risks (Chomitz et al., 1999).

IPCC Homepage