USAID Proland Payment for Ecosystem Services_Case Study.pdf

PROLAND: CAN PES BE AN EFFECTIVE WAY TO ADDRESS THE PROBLEM OF DEFORESTATION? i

PRODUCTIVE LANDSCAPES

(PROLAND)

WHAT DO EXPERIENCE AND RESEARCH TELL US ABOUT

USING PES TO LIMIT DEFORESTATION? A SYNTHESIS OF THE

EVIDENCE AND CASE STUDIES FROM UGANDA AND

COLOMBIA

AUGUST 2020

This publication was prepared for the United States Agency for International Development contract number AID-OAA-I-13-00058/AID- OAA-TO-14-00050, Productive Landscapes (ProLand), under the Restoring the Environment through Prosperity, Livelihoods, and Conserving Ecosystems Indefinite Delivery Indefinite Quantity Contract. The authors’ views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government.

PHOTO CREDIT: JACOB SLUSSER, SMITHSONIAN INSTITUTE

Prepared for the United States Agency for International Development (USAID) contract number AID-OAA-I-13-00058/AID-OAA-TO-14-00050, Productive Landscapes (ProLand), under the Restoring the Environment through Prosperity, Livelihoods, and Conserving Ecosystems (REPLACE) Indefinite Delivery Indefinite Quantity Contract.

Tetra Tech Contacts: Mark Donahue, Chief of Party Mark.Donahue@tetratech.com

Ed Harvey, Project Manager

Ed.Harvey@tetratech.com

Tetra Tech 159 Bank Street, Suite 300, Burlington, VT 05401 Tel: (802) 495-0282, Fax 802 658-4247 www.tetratechintdev.com mailto:Mark.Donahue@tetratech.com mailto:Korinne.Nevins@tetratech.com

WHAT DO EXPERIENCE AND RESEARCH TELL US ABOUT USING PES TO LIMIT DEFORESTATION? A SYNTHESIS OF THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA i

TABLE OF CONTENTS

TABLE OF CONTENTS ............................................................................................................... i

THE PAYMENT FOR ECOSYSTEM SERVICES APPROACH

USE OF THE PES APPROACH TO ADDRESS DEFORESTATION: TWO EXAMPLES

CASE 1: UGANDA FOREST CONSERVATION PES PROGRAM

PES core concepts in the Ugandan case

CASE 2: REGIONAL INTEGRATED SILVOPASTORAL APPROACHES TO ECOSYSTEM MANAGEMENT PROJECT

(RISEMP)

PES core concepts in RISEMP

DECIDING ON USING THE PES APPROACH AND DESIGNING PES PROGRAMS FOR

LONG-TERM SUCCESS

NECESSARY INITIAL CONDITIONS

CRITICAL DESIGN ELEMENTS

THE CAREFUL DESIGN DECISIONS THAT MAKE PES WORK

REFERENCES

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 1

Forest protection plays a critical role in sustaining the world’s biodiversity, mitigating climate change, and supporting global economic growth. Over the past several decades, international donors, private-sector actors, and governments have increasingly turned to a “payments for ecosystem services” (PES) approach to stop, limit, or reverse damage to forests. Knowledge gaps have so far constrained our ability to refine the PES approach so that programs consistently achieve their objectives. Research based on 25 years of experience has nevertheless enabled us to identify factors critical to successful PES programming. Experience and research have helped hone our understanding of the key concepts of additionality, leakage, and permanence, and helped us identify important initial conditions and program design concepts. This ProLand Case Study describes the PES approach and presents examples of its use to address deforestation in Colombia and Uganda. It then uses those examples to illustrate contexts in which PES works best and describe essential design elements in using PES to effectively mitigate deforestation. By drawing on these lessons, governments and their partners can strengthen the use of PES as an approach to reducing global forest loss.

Authors: Dr. David Miller and Dr. Jonathan Winsten

THE PAYMENT FOR ECOSYSTEM SERVICES APPROACH

In our pursuit of increased food, fiber, and energy production over the past 60 years, humans have made unprecedented changes to Earth’s ecosystems (Millennium Ecosystem Assessment, 2007). Forest loss, land degradation, water pollution, and climate change are critical environmental problems that acutely affect the livelihoods and well-being of billions of people, especially those in the world’s poorer countries. The environmental impacts of land management choices inextricably link forest and agricultural lands to each of these crucial problems.

Frequently, however, the land-users’ decision-making does not take these consequences into account because these impacts and their associated costs rarely affect the land-users directly and exclusively. Rather, the costs are borne by people nearby, downstream, or globally.

Managing environmental externalities is important to many development objectives. Protecting human health, conserving biodiversity, addressing climate change and supporting the economic growth of impoverished peoples and nations all depend on the condition of the world’s ecosystems. When they address environmental externalities related to land use, policymakers select among several options.

They often turn to regulatory approaches so that actors take greater responsibility for the impacts of their land management decisions. However, regulations can be politically challenging and complex to administer and enforce. An alternative to regulation, the “payments for ecosystem1 services” (PES) approach, shifts the burden of design, implementation, and enforcement to parties beyond the government by creating incentives to motivate behavior change.2

1 In discussions of PES, the literature and practitioners use the terms “ecosystem services” and “environmental services” interchangeably.

Here, we use “ecosystem services”.

2 Some advocates also argue that PES is more efficient and fairer than regulatory approaches because it is less constraining on economic activity (Jack, Kousky, & Sims, 2008). We note that regulations may stimulate economic activity, and efficiency and equity cannot be assumed in PES programs; the challenges attaining them in PES programming is one theme of this paper.

PES employs voluntary, monetary contributions made by external buyers to maintain or enhance specific ecosystem services that have value to them. The buyers may be private-sector entities, governments, or

Payments for Ecosystem Services (PES) Theory of Change

If the beneficiary of an ecosystem service values that service more than it costs a natural resource manager to maintain or increase that service, and

If the beneficiary provides payments, and the natural resource manager accepts payments to voluntarily improve practices, Then both parties will benefit, and ecosystem services will be maintained or increase.

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 2

other public institutions. Buyers provide payments to land-users (e.g., farmers, ranchers, loggers, or households) to modify their practices to produce fewer negative externalities, such as water pollution, erosion, or carbon emissions. Land-users may agree to enhance ecosystem services by adopting new practices (such as planting trees) or by discontinuing current practices (such as harvesting trees).

Providers are paid only if they implement agreed-upon changes in practices, and sometimes only if there is additional evidence that implementing the practice has produced the targeted ecosystem service (Wunder, 2015).

Buyers can be motivated by several factors including corporate social responsibility goals; international commitments, such as the commitments to reduce greenhouse gas (GHG) emissions codified in the Paris Agreement; or the opportunity to avoid future costs, such as dredging silt from dams or building water treatment facilities. When the buyers receive more value in improved ecosystem services than their total payments, and land-users receive more in payments than it costs them to change their practices, both parties benefit. Since its emergence in the early 1990s, use of PES has grown rapidly, with growth since 2005 especially strong. PES currently accounts for an estimated $36 to $42 billion globally in annual payment transactions (Salzman, Bennett, Carroll, Goldstein, & Jenkins, 2018).

PES programs may be categorized according to the types of ecosystem services they support: those that support watershed-related services; those that support forest and land-use carbon services; and those that support biodiversity and habitats.3 More than 50 percent of the value of global PES transactions supports watershed services, but programs focused on forest carbon are also widespread and growing (Salzman et al., 2018). Commitments in the Paris Agreement on climate change and industry commitments such as Science Based Targets (SBT) may accelerate growth of forest carbon PES programs.

PES programs may also be categorized according to the source of the payments. Programs may be financed by users, governments, or environmental markets (Salzman et al., 2018).

3 These categories are not definitive; in a review of 55 PES projects worldwide, Ezzine-de-Blas, Wunder, Ruiz-Pérez, and Moreno-Sanchez (2016) proposed additional categories: agricultural environmental services, watershed environmental services, climate change, and biodiversity.

4 We have broadened the categorization presented in Salzman, Bennett, Carroll, Goldstein, and Jenkins (2018) to include voluntary environmental markets.

4 In user-financed programs, the beneficiaries pay the land-users for supplying the ecosystem services. In Costa Rica, private hydroelectric plants contribute to the National Forest Finance Fund (Fondo Nacional de Financiamento Forestal, FONAFIFO). The Fund negotiates contracts with land-users to conserve and restore forest cover, which regulates in-stream flows, reduces sedimentation, and, in some cases, increases dry-season flows (Blackman & Woodward, 2010).

In government-financed programs, the government (or an institution working with the government) pays the land-users on the behalf of the beneficiaries. An example is China’s Sloping Land Conservation Program, in which the government paid an estimated 120 million Chinese households to convert steep cropland to forest and grassland (Ouyang et al., 2016).

In environmental market PES programs, entities (usually companies) compensate for their own environmental externalities (e.g., greenhouse gas emissions, water pollution), by “offsetting” them through payments to land-users who produce specific ecosystem services. Some environmental market PES programs, such as cap-and-trade programs for greenhouse gas emissions, are driven by regulation requiring entities to compensate for their environmental externalities. In this case, companies purchase offsets (e.g., carbon or water quality credits) to comply with the regulation; environmental PES programs driven by a regulation are called “compliance” programs. An example is FACE, a consortium of Dutch electricity companies that offsets member emissions by financing the PROFAFOR program in Ecuador.

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 3

PROFAFOR established contracts with private landowners for carbon sequestration through reforestation and afforestation activities (Wunder & Albán, 2008).

PES programs may also be grouped by whether they encourage or discourage behavior, whether they are “asset building” or “use restricting.” Use-restricting programs provide incentives to land-users to discontinue practices. Asset-building programs provide incentives to land-users to adopt more sustainable practices. Programs that promote the adoption of new practices must often provide not only payments but technical guidance or make other investments to facilitate adoption. Some observers nevertheless argue that, in the long term, asset-building may cost less than use-restriction. Use restriction does not address demand, so even if a PES program discourages damaging practices in one geography, sustained demand can motivate production (and accompanying practices that degrade ecosystem services) elsewhere (leakage), thus undercutting the PES program’s goal. The market for wood will remain, and harvesting may be displaced to another forest. In contrast, participants in asset-building programs may continue new practices if they perceive them as beneficial (Pirard, Billé, & Sembrés, 2010; Pagiola, Honey-Roses, & Freire-Gonzalez, 2016). Some observers propose using a combination of “use restriction” to protect natural lands, such as forests, while using “asset-building” to improve the sustainability of land-user practices to reduce pressure on those lands (Karsenty, 2015).

Key Concepts in PES Schemes: Additionality, Leakage, and Permanence

No matter the ecosystem services supported, the source of payments, or whether they encourage or discourage specific behaviors, three fundamental concepts—additionality, leakage, and permanence— strongly influence all PES program design. As we demonstrate in this paper, successfully addressing the challenges inherent in these concepts requires specific initial conditions and program designs that take equity into account, generate payments effectively, incorporate adequate monitoring, and identify sites and participants strategically.

• Additionality: Programs should cause change that would not have otherwise taken place. Effective PES programs, for example, do not provide payments to loggers for letting trees stand that they were not intending to harvest. In programs where additionality is clear, it is unlikely that participants would have made the desired change without the conditional PES payments.

• Leakage: PES programs should not displace the target externalities to another location within or outside the project zone. PES programs often seek to reduce externalities at a national or even global scale. Programs providing payments for loggers to discontinue logging have no impact at these larger scales if, in response, loggers move and continue their work elsewhere. Leakage may even take place in another country, particularly when PES programs target behaviors associated with the production of international commodities.

• Permanence: PES programs may be evaluated by whether the changes they promote later revert to their original state. As with investment in international development generally, positive change during implementation is important, but long-term sustainability trumps immediate change. Payments in most PES programs eventually end, but in some cases land-users find the new practices beneficial and continue them. Ecosystem services may also be sustained after payments end when programs promote practices that produce changes that continue naturally. For example, trees planted in afforestation or reforestation programs continue to offer ecosystem services long after a program ends.

Key PES Concepts

Additionality – Activity and resulting increase in environmental services would not have occurred without the PES program.

Leakage – Activity that reduces environmental services is relocated from inside to outside the PES program boundaries.

Permanence – Land-users or other factors do not later reverse the production of environmental services paid for in a PES program.

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 4

To achieve additionality and permanence and to avoid leakage, PES programs need to consider large spatial—sometimes global—extents and indefinite time frames. It is not often possible to design PES programs for such large areas and long time frames, yet these ambitious objectives strongly shape the design of PES programs. The nature of this challenge—to apply the three foundational principles in a meaningful way and reduce impacts to the environment in a cost-efficient manner—means that PES programs tend to be complicated, demanding, and highly context-specific (Borner et al., 2017).

The complex, locally specific nature of PES programs makes their evaluation challenging. They rarely receive enough funding to produce the data necessary for rigorous evaluation (Burivalova et al., 2019a;

Pynegar, Jones, Gibbons, & Asquith, 2018; Baylis et al., 2015; Naeem et al., 2015; Samii, Lisiecki, Kulkarni, Paler, & Chavis, 2014). An assessment of the available research found that PES programs produce positive environmental outcomes slightly more than half of the time, and they produce positive economic and social outcomes slightly less than half of the time (Burivalova, Miteva, Salafsky, Butler & Wilcove, 2019b). These findings, however, are based on a small proportion of all PES programs; few have been rigorously evaluated (Salzman et al., 2018). Among the approaches available to establish causation in PES programs, the evaluation literature prizes studies that include some type of control, and that account for factors other than the project that may have caused the change (Burivalova et al., 2019b; Borner et al., 2017). Randomized control trials (RCT) are the gold standard when understanding a program’s effectiveness in other disciplines, but RCTs are particularly difficult to design and costly to undertake in the conservation setting.5

5 Standard in disciplines such as medicine, RCTs have their own limitations, and their application to conservation interventions poses specific challenges. Findings from RCTs conducted for PES programs are not easily transferrable to other locations. Also, they must assume target practices do not “spillover” through adoption by neighbors. RCTs may not be appropriate for PES projects with a small number of participants, or low rates of adoption. In this context, other less technically challenging methods—case studies, studies with a control or studies designed to account for confounding variables, and some quasi-experimental approaches—may provide useful insights and pose fewer challenges than RCTs (Baylis et al., 2015; Burivalova et al., 2019a).

We found only two RCT evaluations of PES programs in the research literature. We feature one of these, focused on a program in Uganda, in the first case study (below). The second RCT in the PES literature, conducted by Pynegar et al. (2018), found the application of the PES approach did not produce the hoped-for results due to design faults and limited participation.6

6 Pynegar, Jones, Gibbons, and Asquith (2018) underscore the challenges of voluntary participation, some of which are discussed in this paper. In the Pynegar et al. case, enrolled participants held only 2.5 percent of eligible land in treatment communities.

We do not feature it in this paper. The second case study presented below describes a PES program implemented in three Latin American countries. Although it did not employ an RCT evaluation, it provides a useful complement to the Ugandan example.

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 5

USE OF THE PES APPROACH TO ADDRESS DEFORESTATION:

TWO EXAMPLES

The two programs described below illustrate the application of the PES approach to the challenge of deforestation. One was implemented in the tropical forests of Uganda. The other, in ranchlands of Colombia, Costa Rica, and Nicaragua, although this paper focuses on the project’s implementation in Colombia. Together, these two programs suggest the variety of contexts in which organizations have applied PES to deforestation. These examples also highlight some of the persistent challenges to achieving, documenting, and evaluating durable improvements in ecosystem services through the PES approach.7

7 Research has also produced substantial literature regarding the large and long-term national PES programs of Costa Rica and Mexico. For information on the Costa Rican experience, see Borner et al. (2017); Arriagada, Sills, Ferraro, and Pattanayak (2015); Robalino and Pfaff (2013); Arriagada, Ferraro, Sills, Pattanayak, and Cordero-Sancho (2012); and Pagiola (2008). For the Mexican experience, see Alix-Garcia, Shapiro, and Sims (2012); and Corbera, Gonzalez, and Brown (2009).

CASE 1: UGANDA FOREST CONSERVATION PES PROGRAM8

8 This case study is based on Jayachandran et al. (2017). All information comes from that source, unless otherwise noted.

In Uganda’s Northern Albertine Rift landscape, local households have been clearing forests at a high rate over the past several decades (Twongyirwe, Bithell, Richards, & Rees, 2015). The Uganda Forest Conservation PES program, implemented from 2011 to 2013, tested the use of PES to address this rapid deforestation caused by agricultural expansion and local timber harvesting. Funded by the United Nations Environment Programme (UNEP) and the International Initiative for Impact Evaluation, the project’s goal was to rigorously study the impact of a well-designed PES program and assess whether it could significantly reduce the rate of tree cover loss. The program used forest cover as a proxy for ecosystem services such as carbon sequestration and wildlife habitat.

People in the project area are principally farmers who clear forest as necessary to extend their cultivation. They also supplement their incomes through the sale of charcoal and timber (Jayachandran, personal communication). The Chimpanzee Sanctuary and Wildlife Conservation Trust (CSWCT), a Ugandan nongovernmental organization (NGO), designed and implemented the program in partnership with a group of scientists from Northwestern University. CSWCT offered households annual payments to refrain from cutting trees on their forestlands. Approximately one third of the selected village households participated.

CSWCT paid participants the equivalent of $28 per hectare per year. The program team selected this payment rate because it was slightly greater than the opportunity cost to households for not cutting trees. Households reported making an average of $22 per hectare per year from the sale of timber in the three years prior to the project.9

9 Determining opportunity costs is not straightforward. In this region of Uganda, some households cultivate tobacco, a relatively lucrative crop, and have a much greater opportunity cost for participation. The project could have differentiated payments across participants, paying amounts corresponding more closely to a household’s opportunity costs. This might have secured higher participation rates and been more cost-effective but would also have increased implementation complexity and potentially created fairness issues among participants (Jayachandran, personal communication).

CSWCT staff monitored the forest areas for cutting. They also used QuickBird, a commercial satellite, to measure tree cover over the 2,461-square-kilometer study region at both the baseline and the end of the program. QuickBird produced imagery of sufficiently high resolution to allow the team to measure both deforestation and selective cutting.

To evaluate the impact of the payments, the program scientists conducted a randomized control trial in 121 villages, 60 of them in which households received payments; the remainder served as a control in

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 6

the study. Tree cover loss was about half as great in the treatment villages as it was in the control villages. The reduced cutting in treatment villages equated to averting 0.326 hectares of deforestation per household. The scientists also found that the program averted carbon dioxide (CO2) emissions at a total cost of $0.46 per metric ton (MT).10

10 After converting the carbon to CO2, the program researchers estimate that the program averted 183.5 MT CO2 per household. Forest cover in the area averages 153.5 MT C per hectare. Using the average payment per household of $37.80 over the two years, they calculated the payment to be $0.20 per MT of averted CO2 emissions. They also estimated the costs to administer the program at $0.26 per MT of averted CO2 emissions, making the total program costs $0.46 per MT of averted CO2 emissions.

Was this PES program worth it? Did the “buyers” receive more in improved ecosystem services than they made in total payments? Due to rigor in the design and data collection associated with the program, the team was able to conduct a cost-benefit analysis to propose an answer to this question.

Even with this monitoring system in place, the team had to begin with several assumptions. First, they used an estimate of the cost of CO2 emissions to society as a whole (i.e., the “social cost” of carbon or SCC) as a proxy for the program benefit.11 The SCC was $39 per MT of averted CO2. They based the

11 The social cost of carbon (SCC) is a commonly employed metric of the expected economic damages from CO2 emissions (Ricke, Drouet, Caldeira, & Tavoni, 2018). The economic damages result from both environmental impacts such as extreme weather events, agricultural losses and higher global temperatures; and the human health impacts, such as cancer, heart disease, and stroke.

cost on a median estimate of the SCC used by the U.S. Environmental Protection Agency in 2012. The team also assumed a ten-year lag from the time when logging would have occurred to the release of CO2 into the atmosphere and used a discount rate of 3 percent to calculate the present value of the future benefit of avoided emissions. They assumed that the SCC would increase by 1.9 percent annually.

Since the program was temporary, their findings also assumed that the land-users participating in the program would increase the rate of cutting after the payments ended and catch up with the cutting of the control households within four years. Using these assumptions, the team calculated the present value of delaying CO2 emissions to be $1.11 per MT. Given the total cost of averted CO2 emissions at $0.46 per MT, the benefit-cost ratio for the program was positive, at 2.4 ($1.11/$0.46). The program “buyers”—along with everyone else dependent on global climate—received over twice the value of the project’s total cost in ecosystem services.12

12 Much debate surrounds methods of valuing improvements in ecosystem services. In making their program cost benefit analysis, the Uganda research team had alternatives to comparing the costs of the PES program to the estimated societal costs of the CO2 emissions avoided.

They could have weighed the costs of the program against the cost of replacing the trees through reforestation; or the cost of compensating landowners for the loss of their trees; or, as we noted above, the same measure they used to establish payments, the cost of opportunities for income lost to participants. These methods of estimating ecosystem value are widely used in part because they are relatively easy to calculate. Other observers argue that a more accurate environmental valuation incorporates the complex social, economic, and environmental benefits valued by the multiple sets of stakeholders (Beecher, 1996; Gregersen, El Lakany, Karsenty, & White, 2010; Wegner & Pascual, 2011; (Anna & Nogueira, 2012).

PES core concepts in the Ugandan case

Additionality. The designers of the Uganda program sought to make sure the payments produced changes in behaviors that would otherwise have not occurred. To this end, they selected a zone with extreme deforestation pressure where households had demonstrated a clear pattern of harvesting trees.

Leakage. The Uganda program required that participating farmers enroll all their primary forestland, submitting the land in which they had control over the trees. This precluded individual households from transferring forest clearing and logging activities within their own property. The baseline and end-of-program remote sensing included the area surrounding the program zone, which revealed no evidence of leakage.

Permanence. Conducted as a trial, the Uganda program intended to delay deforestation and emissions for two years. If the program resulted in only that two-year delay, then after two years PES participants would have been cutting more trees than non-participants. Preliminary results three years after program

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 7

initiation found no difference in tree-cutting rate between PES participants and nonparticipants. If this finding is confirmed, it will mean that the reduction in harvesting did not reverse immediately when the project ended. The sustained benefits, while perhaps not permanent, could plausibly extend many years beyond a two-year hiatus in harvesting.

CASE 2: REGIONAL INTEGRATED SILVOPASTORAL APPROACHES TO ECOSYSTEM

MANAGEMENT PROJECT (RISEMP)

Cattle grazing is one of the three most-cited causes of deforestation throughout Latin America, together with agricultural expansion and infrastructure development (Armenteras, Espelta, Rodríguez, & Retana, 2017). Overgrazing has depleted soil fertility, reduced vegetative cover, increased soil erosion, and degraded water quality. This reduced pasture productivity in turn often results in clearing additional land for grazing, which increases deforestation pressure (Serrao & Toledo, 1992).

The Regional Integrated Silvopastoral Approaches to Ecosystem Management Project (RISEMP) sought to address the impact of cattle production on deforestation. Funded by the Global Environment Facility and implemented by the World Bank and local partners, RISEMP used a PES approach to promote the adoption of silvopastoral production systems (SPS), which integrate trees and shrubs into grazing systems. There are several potential benefits of SPS, including increased nutrient availability per hectare for improved meat or milk production, improved soil fertility, increased carbon sequestration, and increased wildlife habitat and biodiversity. The project’s two primary objectives were (1) to demonstrate the effectiveness of the PES approach in promoting farmer adoption of SPS on degraded pasturelands;

and (2) to measure the impacts of the resulting ecosystem services. RISEMP conducted one project in each of three countries (Colombia, Costa Rica, and Nicaragua) from 2003 to 2007.

Farmers in these countries had been resistant to integrating trees into livestock systems because of the high initial investment costs, tenure insecurity, and lack of access to technical assistance. To improve the chances of adoption, participating ranchers could select from 28 different SPS practices that experts believe generate biodiversity and carbon benefits, as well as private benefits for ranchers, by improving pasture productivity and providing more nutritious forage for cattle (Pagiola et al., 2007; Pagiola et al., 2016). Among the 28 practices were the planting of trees and shrubs for fencing and windscreens, and to serve as “fodder banks” in degraded pastures. This approach probably increased participation in RISEMP because land-users are more likely to participate in PES programs if they have the option to select the land management practices appropriate to their conditions and resources (Winsten, 2009).

The project worked with local institutions to help ranchers adopt SPS practices. In Colombia, for example, RISEMP worked with Centro para la Investigacion en Sistemas Sostenibles de Produccion Agropecuaria (CIPAV) to help 80 ranchers determine the practices best suited to their ranches. Much like the Ugandan PES project, the RISEMP used imagery from the QuickBird high-resolution satellite to track land use changes during the project and in the post-project assessment.

In this project, ranchers received payments based on an ecosystem services index (ESI) that contains values for the biodiversity conservation and carbon sequestration services associated with the specific SPS practices offered. The project used an expert panel to assign a score to each of the 28 practices. As an end-marker, primary forest was given the maximum score of 1 for biodiversity and 1 for carbon, so a combined ESI score of 2.0. Intensive SPS received a score of 0.6 for biodiversity and 1 for carbon, for a combined ESI of 1.6. Natural (unimproved) pasture with a high tree density was assigned an ESI score of

1.0. At the lower end of the spectrum, annual cropping and degraded pasture each received 0.0 ESI points.

RISEMP provided payments based on the ESI values assigned to the practices that each farmer implemented. Farmers could choose among the practices, but received annual payments based on the

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 8

difference between their farm’s ESI score that year and their ESI score at the beginning of the project.

By employing practices assigned higher scores, farmers increased their payments. RISEMP also provided a small one-time payment (US$10 per point) to farmers with higher baseline scores. These payments were not intended to purchase additional ecosystem services; their purpose was to minimize the disadvantage to farmers already employing the practices RISEMP intended to promote, as a matter of equity.

RISEMP in Colombia tested several aspects of the PES approach. The project randomly assigned participants into groups based on whether they received only payments, or payments plus supplemental technical assistance (TA). RISEMP provided supplemental TA to randomly selected ranchers in an effort to understand the impact of the payments versus the impact of the TA. This group received in-person, on-farm TA on how to implement and manage the selected SPS practices. The project also followed a control group of ranchers who expressed interest but did not receive program payments or TA because the program had already reached capacity when they expressed interest in participating (Pagiola et al., 2016).

In a summative evaluation of RISEMP, the project team assessed the increase in ESI scores per hectare and the percentage increase in ESI scores as the primary outcome variables. Across the four years of the project, the ESI per hectare increased by more than 60 percent for the project participants and by only 7 percent for the control group. From this result, Pagiola et al. (2016) concluded that the improved ESI scores would not have occurred in the absence of the program.

Four years after RISEMP ended, the project team again assessed ESI scores per hectare. They found that scores had not changed; participating ranchers had maintained the practices they had adopted during the project (Pagiola et al., 2016). This result likely emerged for two reasons. First, the program promoted practices that benefit ranchers over time by increasing the productivity of their land. The increased productivity provided an incentive for ranchers to maintain the SPS practices after project payments ceased. In effect, the land-users themselves were beneficiaries of the ecosystem services the payments promoted. Second, because RISEMP allowed ranchers to choose the practices to adopt, they selected those that best matched their land resources, operations, and preferences. This less prescriptive approach likely resulted in changes more appropriate to each farm, which land-users are more likely to maintain long-term than they are less-appropriate practices.

PES core concepts in RISEMP

Additionality. The results indicate that the land-use changes would not have occurred without the program payments. In contrast to the Ugandan example, payments in Colombia served to offset some of the costs of adoption (Pagiola et al., 2016) rather than offset opportunity costs.

Leakage. Because RISEMP increased, rather than diminished, output from the ranches, it did not displace production to other locations. However, the improvements in productivity could attract greater investment to cattle grazing, which could put increased pressure on forestland.

Permanence. Because RISEMP provided ranchers incentives to make changes that increased the productivity and profitability of their ranches, they were likely to maintain these changes even after the payments ceased. ESI scores had not changed four years after the program ended, and it is likely program achievements were sustained well beyond that date.

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 9

DECIDING ON USING THE PES APPROACH AND DESIGNING PES

PROGRAMS FOR LONG-TERM SUCCESS

Decades of experience designing and implementing PES programs provide lessons for policymakers and program managers. Certain initial conditions facilitate PES implementation, while specific design elements influence its success. Implementing PES programs in contexts with the necessary initial conditions and including desirable elements in the design will strengthen the viability, effectiveness, efficiency, and equity of PES programs.

NECESSARY INITIAL CONDITIONS

The PES approach will not be successful everywhere. Certain conditions are necessary for a PES approach to be successful and cost-effective. Stakeholders must be inclined to participate; the program must be able to clearly define property rights; and conditions must allow for a positive assessment of the costs and benefits from the perspectives of both program sponsors and land-users.

Stakeholder commitment and cohesion. A solid foundation of trust and collaboration is particularly important in PES programs. The payments may represent a considerable change in income, and both participants and nonparticipants must understand the rationale for the payment system and see the conditions for payment and the amounts as fair. By basing payments on an index of expert decisions, RISEMP introduced opacity and judgement into this process that could have, but apparently did not, weakened commitment to the program. Buyers and land-users need to have confidence in each other.

The engagement of a known, honest, and trusted intermediary organization to administer the program, and the participation of local social scientists on the project team, may help create the stakeholder engagement and transparent, clear communication process necessary to ensure participation, build stakeholder cohesion, and avoid conflict (Borner et al., 2017; Naeem et al., 2015; Leimona, van Noordwyk, de Groot, & Leemans, 2015; Hejnowicz, Raffaelli, Rudd, & White, 2014). In the Ugandan example, the fact that many households in the project area did not know of CSWCT may have reduced the participation rate (Jayachandran, personal communication). In Colombia, the fact that livestock producers knew and trusted CIPAV may have helped to increase participation (Pagiola, personal communication).

To strengthen commitment and cohesion from the outset, practitioners developing PES programs must solicit stakeholder input to design payment structures. These structures should reflect both the values buyers assign to improved ecosystem services, and the payments land-users are willing to accept to change their practices. Discussions and negotiation can be necessary to design a program adapted to the local landscape, land-use practices, and economics. Reinforcing this foundation of cooperation with transparent communication throughout implementation is also critical to maintaining participation and reducing the chance of discord among participants and the larger community.

Clearly defined property rights. Ronald Coase won the 1991 Nobel Prize in Economics for describing the conditions that allow for markets to efficiently address environmental externalities. One of the crucial conditions in the Coase Theorem (1960) is the existence of clearly defined property rights. For example, a farmer needs to have the right to cut down their trees, or to allow nutrients to pollute nearby waterways, to be paid to change behavior causing the external costs. The PES approach necessitates that people clearly hold the rights to produce the externalities, justifying payments for changing them. In the Ugandan example, households had secure rights to harvest the trees on their land, “An honest and trusted intermediary is one of the key factors of a successful PES scheme.”

—Leimona, van Noordwijk, de Groot, and Leemans (2015)

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 10

allowing payments to be offered to forego the harvesting. In Colombia, the ranchers had the right to manage their lands as they saw fit, allowing RISEMP to offer payments to add SPS practices.

Costs and payments acceptable to participants. Participation in PES programs is voluntary. Buyers must be willing to offer payments sufficient to attract land-users to participate. To do this, the ecosystem service value they assign to the changes they “purchase” must surpass the costs land-users assign to the changes they agree to make in their land management practices. Not all contexts meet these conditions. In some locations, the costs to land-users of mitigating an externality—of enhancing or maintaining certain ecosystem services—surpasses what buyers are willing to pay. The Ugandan team had estimated that their project could preserve trees that had a carbon value much greater than the overall cost of the project.

CRITICAL DESIGN ELEMENTS

Even where the necessary initial conditions outlined above exist, certain design elements are also needed for a PES program to succeed. These include questions of equity, effective payment systems, adequate monitoring and verification, and appropriate targeting of sites and participants.

Take equity into account. Program designers must understand and clarify the relative importance of environmental goals and poverty alleviation (Adhikari & Agrawal, 2013). Targeting poverty alleviation as an objective may reduce the cost-effectiveness of a PES program by prioritizing the distribution of payments over the generation of ecosystem services. Yet programs that do not account for differences in wealth among participants may advantage some over others and weaken program cohesion. Payments may bias participants already engaging in target practices and wealthy participants predisposed to adopt them. Programs require careful design and transparent communication to equitably and cost-effectively generate ecosystem services.

Effective payment systems. The structure of payments strongly influences the success of PES programs. Designs differ in the level of payments; their duration; those to whom payments are made;

and the mechanism used to deliver payments. Finally, the basis for payments, whether they are practices adopted or improvements in ecosystem services, also affects program success.

Level of payments. For a program to be cost-effective, payment rates should be high enough to motivate effective participation without inflating program cost. In the Ugandan case, the program intentionally set the payment rate slightly higher than the opportunity costs (i.e., the average value of timber sales) to the households. Because this amount represented 16 percent of median annual household income in the region, the program designers assumed it sufficient to secure an acceptable level of participation within the planned program area (Jayachandran, personal communication). The program accepted that households with higher opportunity costs (e.g., those cultivating tobacco, a high-value crop) might be unlikely to participate. A higher payment rate may have increased program participation, but at a higher marginal cost. RISEMP allowed participants to influence the payments they received by selecting from a list of practices, but by using an expert panel to create the index that set ecosystem values and relative payment levels, it also introduced judgement and opacity into the process.

Payment duration. The duration of payments influences, but does not determine, how long improvements in ecosystem services continue. Some programs extend payments indefinitely. For example, payments may derive from a continuous fee imposed on the beneficiary, such as a surcharge on the bill of a water user. Other programs provide payments for a limited period of time. Yet even in the two cases described, improvements in ecosystem services do not always stop when payments stop. In RISEMP, an asset-building PES program, ranchers experienced improved productivity and maintained the changes (SPS practices) long after payments ended. In the Ugandan program, which followed a use-

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 11

restricting approach, the design team did not expect households to refrain from harvesting after payments stopped.

Payment recipients. PES programs may offer payments based on the performance of individuals or of groups. They may also combine the two options. A winning solution submitted to the U.S. Nutrient Challenge described a PES program design that offered a primary incentive payment to individual farmers in targeted watersheds. The incentive payment was based on the amount of nutrient loss reduction each farmer could achieve and included a bonus payment to the entire group when nutrient loads at the mouth of the watershed dropped below a specified target (Winsten, 2014).

Payment delivery mechanism and basis. The type of mechanism used to provide payments may have a large impact on program costs, particularly if there are many participants. Making payments in conjunction with preexisting financial transfers (e.g., a reduced water bill) will incur significantly less time and expense than mailing many small payments to participants. Where viable, electronic transfers also reduce program costs.

Program costs also vary depending upon the criteria used to make payments, particularly whether a program distributes payments equally across a landscape or targets parcels where management changes have an outsize influence on ecosystem services. In most PES programs, simply implementing the agreed-upon practices triggers payments. As a result, for a given practice, they provide payments at a constant rate across a landscape even though the quantity of ecosystem services provided will vary due to factors such as soil type and slope. Other programs make payments based on a quantification of the amount of ecosystem services provided. Some programs of this type measure ecosystem services directly, while others rely on proxy variables (Winsten et al., 2011). In either case, performance-based payments motivate land-users to target critical fields and areas to maximize their program payments (Winsten, 2009). In a study of the use of PES to address water quality in the Chesapeake Bay, Ribaudo, Savage, and Aillery (2014) estimate that performance-based PES programs are at least four times more cost-effective than programs providing payments based on the simple adoption of practices.

Most PES programs do not employ performance-based payments because this program design requires monitoring or quantification to differentiate performance across the landscape. Analyzing the necessary data also consumes resources, which raises program costs, and deters program designers from this approach (Salzman et al., 2018). The practice-based RISEMP and the Uganda program paid for the adoption of practices at a constant rate across the landscape. The basis, however, was structured differently. The RISEMP paid rates to land-users based on the practices they adopted, while the Ugandan program paid a constant rate to everyone who decided to participate.

Adequate monitoring and verification. Monitoring and verification play a critical role in the implementation of PES programs and in evaluating overall program impact on ecosystem services (Naeem et al., 2015). The cost and type of monitoring and verification systems will vary by project based on program goals, on whether the program is performance- or practice-based, on other program design characteristics, and on budget. The level of trust among stakeholders may also influence the implementation, if not the design, of monitoring and verification systems (Winsten et al., 2011).

We have limited ability to quantify the changes in ecosystem services that result from new land management practices PES

Ways to Quantify Performance

(1) Using simple estimates based on land characteristics;

(2) Using proxy variables that are more easily measured than the actual environmental services;

(3) Using simulation models to estimate the production of environmental services; and

(4) Performing actual measurements (e.g., carbon accumulation in trees measured with a specific protocol).

THE EVIDENCE AND CASE STUDIES FROM UGANDA AND COLOMBIA 12

programs promote (Jack et al., 2008). As a result, PES programs often focus on verifying changes in the land management practices stipulated by programs. Where the results of changed practices are observable, this type of verification is relatively straightforward. In some cases, change may be more difficult to verify in asset-building programs than in use-restricting programs. For example, it may be harder to monitor the adoption of sustainable agricultural methods than to determine that trees have not been harvested from a local forest.

Because PES programs create incentives through payments and response to these incentives can be unpredictable, effective and transparent monitoring and verification play a critical role in implementation.

Participants may be inclined to cheat or game the system (Hart & Latacz-Lohmann, 2005). They may succumb, for example, to the moral hazard of intentionally reducing their baseline condition to enhance their program payment. A landowner could clear forest and then reap payments for planting trees on the now-treeless land. Poorly designed monitoring and verification systems can result in programs incentivizing ineffective land management changes and even produce negative impacts on intended ecosystem services (Wunder, 2008).

Land-users concerned about equity or corruption in the program may welcome the sense of trust fostered by more robust monitoring efforts (Kaczan, Swallow, & Adamowicz, 2013). On the other hand, land-users who fear it may be used against them (e.g., information on their farming practices used to regulate them in the future) may resist comprehensive monitoring and verification efforts. The monitoring and verification system for any project needs to be designed based on the specific context of the PES project, and with adequate stakeholder input and engagement to foster trust, accuracy, and efficiency.

Practitioners face challenging decisions when designing PES monitoring and verification systems that align with program objectives and measure outcomes accurately.