Lessons

• REDD can deliver additional near-term emissions reductions that are critical for maintaining global options for stabilising GHG concentrations. • Estimated costs of REDD vary with the modelling approach used. However, all models indicate that cost savings from REDD could buy deeper and faster emissions cuts than would be achieved with the same global expenditure but without REDD. • Cost-e˜ective and large-scale REDD is available for a limited time only, thus adding value to protecting tropical forests now.• The risk of REDD supply ‘˚ooding’ the carbon market can be contained by policy designs ranging from strict and long-term targets with ‘banking’ to modest limits on the use of REDD and other types of credits.


Introduction
Scientific evidence indicates that avoiding dangerous interference with the climate system-for example, warming greater than 2 °C by the end of the centuryrequires rapid and large-scale reductions in greenhouse gas (GHG) emissions from both developed and major emitting developing countries.Reducing emissions from tropical forests offers an immediate opportunity to mitigate a major emissions source at relatively low estimated costs.Reducing emissions from deforestation and forest degradation (REDD) efforts could also offer a 'bridge strategy' of reducing near-term emissions while buying time to adapt to a low carbon future.
This infobrief looks at some important questions for decisions over the policy and architecture of REDD: What will REDD cost? 1.
How will REDD affect the overall strategy for 2.
reducing emissions?How will REDD affect the carbon price and thus 3.
efforts to reduce emissions in other sectors?
This paper focuses on ways in which different economic models provide different answers to these questions.
CIFOR infobriefs provide concise, accurate, peer-reviewed information on current topics in forest research

No. 18
November 2008 What will REDD cost?

Types of REDD costs
Estimated costs of REDD vary with the data and modelling approach used and the types of cost considered.Studies report costs in terms of supplying or buying REDD, or both.Most estimates focus on the 'opportunity costs' of avoiding deforestation from a landowner's perspective, without the costs of developing institutional capacities and actually implementing and transacting a REDD programme.These 'opportunity costs' are the foregone economic benefits from alternative land uses.
Some economic models have estimated 'supply curves' ('marginal cost curves') that indicate a cost spectrum for incremental reductions in forest emissions (Figure 1).The cost curves slope upwards, showing that for small emissions reductions, costs can be kept low by protecting just the lowest-cost lands; with greater reductions, the added incremental or 'marginal' costs rise as protection must extend to higher-cost lands.
For example, estimates of total opportunity costs more than double in moving from 94 per cent to 100 per cent protection of the Brazilian Amazon forest, because of the high agriculture potential of just 6 per cent of the lands (Nepstad et al. 2007).
The costs of implementing REDD policies comprise upfront costs of 'capacity building'; ongoing 'administrative costs' of monitoring, enforcement and other activities needed to run a REDD programme; and 'transaction costs' involved in successfully connecting buyers and sellers.Implementation costs will vary with national capacities and strategies.One-time needs for capacity building and policy reform for REDD in 40 countries were recently totalled at USD 4 billion (Eliasch 2008).In addition, the costs of generating valid REDD credits will crucially depend on the baseline-setting rules for how REDD efforts shall be compensated.

Modelling approaches
Most estimates of REDD costs come from 'bottom-up' or 'engineering' studies based on detailed information on particular activities in particular locations, at fixed prices.In contrast, 'top-down' models are more aggregate and take into account commodity market interactions-both demand and supply.Top-down models have generally yielded higher estimates for the costs of large-scale REDD, partly because they account for market feedbacks (see Table 1).Feedbacks occur as reductions in deforestation lower timber harvests and land conversion to agriculture.Consequent lower growth in the supply of soyabeans, cattle and timber will raise their prices, thereby raising the incentives to deforest, as long as the unsatisfied demand does not abate completely.Such feedbacks will raise the costs of REDD and increase the risk of 'leakage' or 'displacement', by providing incentives to shift deforestation elsewhere.
Differences in the modelled 'baseline' scenario of what deforestation would be without REDD policies also affect the estimated costs of REDD.Greater  Opportunity cost curves are estimated.Total costs above include opportunity costs of supplying emissions reductions plus the 'rents' (pro ts) earned by REDD providers in selling reductions at a single market price (Figure 1).This is the expenditure for a buyer in a competitive market; the seller's 'rents' are a redistribution of resources, not a cost to society as a whole.However, the rents a ect the cost e ectiveness or ability of a REDD programme to maximise reductions for a limited budget.

Comments
Commodity prices xed.Market e ects incorporated (e.g.price rises as supply falls), which tends to raise costs.-Scope of incentives: more complete coverage lowers leakage and thus costs.-Targeting of incentives: targeting payments at emissions reductions lowers transfers to non-emitters and thus costs (to buyers), but avoiding 'leakage' and ensuring equity must also be considered.
forecasted deforestation under 'business-as-usual' would potentially reduce emissions, but may also mean greater modelled pressures on forests and thus higher costs of forest protection.Other differences in data and assumptions contribute to variation in estimates of REDD costs (Table 2).
How will REDD a ect the overall strategy for reducing emissions?
Consideration of deforestation and other land-based options for reducing emissions within climate models is a relatively new field.However, results from the Energy Modeling Forum 21 (Fisher et al. 2007) and related efforts suggest that reducing deforestation, in addition to planting trees, changes in forest management, and other land-based options to mitigate GHGs, may provide important cost savings to reach climate stabilisation goals over the next century (Table 3).
These cost savings may enable greater global emissions reductions than could be achieved without REDD for the same overall cost.Estimated savings of USD 2 trillion through global forestry mitigation could finance a 10 per cent stricter target or 0.25°C less of warming over the century depending on the modelled scenario (see Table 3).The potential gains from REDD depend on the target GHG concentrations in the atmosphere and the menu of available options for reducing emissions.More alternatives bring more potential sources of cheap reductions and reduce the reliance on any single option in reducing emissions at least cost.Another critical assumption affecting the estimated role of REDD across models is the expected development of future biofuel technologies (Table 3).In particular, biomass production for electricity generation combined with carbon capture and sequestration could, in theory, be a powerful competitor for land if it became a feasible means to generate energy with negative carbon emissions (e.g.Obersteiner et al. 2001).Most studies of REDD focus on the economic potential assuming that institutional frameworks and capacities are readily available to immediately implement REDD worldwide.However, not all countries will choose to join an international climate agreement or be able to effectively reduce deforestation emissions in the near term.These institutional and political barriers lower the realistic scale of reductions and their effective global impact.Inconsistent incentives for REDD and other GHG reductions across countries would create the potential for international emissions 'leakage', with reductions in one country potentially being offset by increases elsewhere.For example, Gan and McCarl (2007) estimate international leakage as high as 42-95 per cent in the forestry products industry.
How will REDD a ect the carbon price and e orts to reduce emissions in other sectors?
The potential cost advantages of REDD may detract from abatement in other sectors, if REDD credits were made fully interchangeable with other GHG credits.
A perceived risk is that REDD may 'flood' the carbon market, dampening the price signal to develop and deploy clean energy technologies.The effect of REDD on carbon prices and technology incentives depends on several factors: Alternative land uses represent the 'opportunity costs' of REDD: clearance for rice cultivation in Jambi, Indonesia.According to Eliasch (2008), introducing REDD credits along with modest quantitative limitations on REDD has a negligible estimated effect on the European Union's carbon price, even if countries can satisfy 50-85 per cent shares of their abatement through international credits, depending on the stringency of the European Union target.The precise proportional impact of REDD on the price depends on the assumptions determining the shape of the cost curves, including the costs of the potential alternatives.
Sufficiently ambitious and credible long-term targets anticipated by market participants also provide incentives for saving up credits for use under tighter future targets.Taking 'banking' into account, Piris-Cabezas and Keohane ( 2008) estimate a global REDD programme would lower the global carbon price by 14 per cent, while using all forestry mitigation options would reduce the price by 31 per cent, for a fixed emissions reductions target.Doubling the estimated supply of REDD credits has a relatively small effect on the modelled price, as additional credits are 'banked' and used gradually over time.If REDD helps build a store of relatively low-cost emissions reductions, this 'bank' can also dampen price volatility by providing a buffer against unexpected price spikes in the future.

Conclusions
The economic impact of REDD will depend on the overall climate targets and policy architecture, the design and implementation of REDD, and its fungibility with the rest of the GHG market.The latest science suggests that only a global programme that begins almost immediately and achieves large reductions in global GHGs by mid-century can preserve options to avoid dangerous interference with the climate system.Despite different assumptions, a range of economic models indicates that REDD can make a significant contribution to cost-effectively stabilising GHG concentrations at this scale and speed.
Both the cost and timing of REDD are critically important.Estimated cost savings from REDD could buy greater and faster reductions than could be achieved with the same global expenditure without REDD.Early emissions reductions have particular value as a global insurance policy for maintaining climatic options in light of scientific uncertainty (Fisher et al. 2007).As tropical forests are disappearing, REDD is also a cost-effective opportunity for reducing emissions that is available for a limited time only.The time-limited and irreversible nature of REDD-once deforestation occurs, it cannot be avoided in the future-adds further value to protecting tropical forests now rather than foreclosing future options for lowering global emissions.
Ongoing deforestation means that cost-e ective and large-scale REDD is available for a limited time only, thus adding value to protecting tropical forests now: timber harvesting in Ghana.

Photo by Cecilia Luttrell
For further information, please contact: Ruben N. Lubowski RLubowski@edf.org A full length version of the paper can be found at: http://www.cifor.cgiar.org/globalredd/For general inquiries contact cifor@cgiar.org Center for International Forestry Research (CIFOR) advances human well-being, environmental conservation, and equity by conducting research to inform policies and practices that a ect forests in developing countries.CIFOR is one of 15 centres within the Consultative Group on International Agricultural Research (CGIAR).CIFOR's headquarters are in Bogor, Indonesia.It also has o ces in Asia, Africa and South America.CIFOR works in over 30 countries worldwide and has links with researchers in 50 international, regional and national organisations.www.cifor.cgiar.org

Figure 1 :
Figure1: Supply and demand for REDD 'credits' Tavoni et al. (2007) estimate that global implementation of REDD plus A/R and changes in forest management would delay deployment of some technologies and reduce investment in energy research and development by about 10 per cent, for a fixed emissions reduction target.Anger and Sathaye (2008) find a 40 per cent carbon price reduction from introducing REDD into a market that also allows unlimited credits for developing country mitigation through the clean development mechanism.Other studies find more muted impacts, depending on the policy scenario.

Table 1 .
Halving global deforestation: comparison between bottom-up and top-

Table 2 .
E ects of including di erent features on the estimated costs of REDD Select features included in the model E ect on costsPrice feedbacks: lower supplies of timber, crops, etc., raise prices and thus opportunity costs of forest protection.+Number of deforestation drivers modelled: accounting for more drivers, such as timber and agriculture, will raise opportunity costs of forest protection.Accounting for new future drivers, such as biofuels, rather than extrapolating past drivers can also increase estimated costs.+Implementation and transaction costs, investment risks.+Landconversionbenets as opposed to costs: one-time bene ts from timber harvests upon forest clearance raise costs of forest protection.+Greaterassumedparameterfor the 'elasticity of transformation' (the convertibility of forest land to other uses) raises costs in some models.+Carbondensity/releases: greater emissions avoided per hectare protected lower cost per tonne.-Timber bene ts from protected forests (e.g.sustainable forest management).-Scope of the REDD model (forestry activities, sectors, countries, gases): greater scope implies less leakage and more opportunity for low-cost global reductions.

Table 3 :
Estimated potential of REDD to lower costs and buy additional emissions reductions: comparison of models