IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group II: Impacts, Adaptation and Vulnerability Sectoral and regional estimates

The literature on costs and benefits of adaptation to sea-level rise is relatively extensive. Fankhauser (1995a) used comparative static optimisation to examine the trade-offs between investment in coastal protection and the value of land loss from sea-level rise. The resulting optimal levels of coastal protection were shown to significantly reduce the total costs of sea-level rise across OECD countries. The results also highlighted that the optimal level of coastal protection would vary considerably both within and across regions, based on the value of land at risk. Fankhauser (1995a) concluded that almost 100% of coastal cities and harbours in OECD countries should be protected, while the optimal protection for beaches and open coasts would vary between 50 and 80%. Results of Yohe and Schlesinger (1998) show that total (adjustment and residual land loss) costs of sea-level rise could be reduced by around 20 to 50% for the U.S. coastline if the real estate market prices adjusted efficiently as land is submerged. Nicholls and Tol (2006) estimate optimal levels of coastal protection under IPCC Special Report on Emissions Scenarios (SRES; Nakićenović and Swart, 2000) A1FI, A2, B1, and B2 scenarios. They conclude that, with the exception of certain Pacific Small Island States, coastal protection investments were a very small percentage of gross domestic product (GDP) for the 15 most-affected countries by 2080 (Table 17.2).

Table 17.2. Sea-level rise protection costs in 2080 as a percentage of GDP for most-affected countries under the four SRES world scenarios (A1FI, A2, B1, B2)

Protection costs (%GDP) for the 2080s  
SRES scenarios A1FI A2 B1 B2 
Micronesia 7.4 10.0 5.0 13.5 
Palau 6.1 7.0 3.9 9.1 
Tuvalu 1.4 1.7 0.9 2.2 
Marshall Islands 0.9 1.3 0.6 1.7 
Mozambique 0.2 0.5 0.1 0.8 
French Polynesia 0.6 0.8 0.4 1.0 
Guinea-Bissau 0.1 0.3 0.0 0.6 
Nauru 0.3 0.4 0.2 0.6 
Guyana 0.1 0.2 0.1 0.4 
New Caledonia  0.4 0.3 0.2 0.4 
Papua New Guinea 0.3 0.3 0.2 0.4 
Kiribati 1.2 0.0 0.3 0.0 
Maldives 0.0 0.2 0.0 0.2 
Vietnam 0.1 0.1 0.0 0.2 
Cambodia 0.0 0.1 0.0 0.1 

Source: Adapted from Nicholls and Tol (2006).

Ng and Mendelsohn (2005) use a dynamic framework to optimise for coastal protection, with a decadal reassessment of the protection required. It was estimated that, over the period 2000 to 2100, the present value of coastal protection costs for Singapore would be between US$1 and 3.08 million (a very small share of GDP), for a 0.49 and 0.86 m sea-level rise. A limitation of these studies is that they only look at gradual sea-level rise and do not generally consider issues such as the implications of storm surges on optimal coastal protection. In a study of the Boston metropolitan area Kirshen et al. (2004) include the implications of storm surges on sea-level rise damages and optimal levels of coastal protection under various development and sea-level rise scenarios. Kirshen et al. (2004) conclude that under 60 cm sea-level rise ‘floodproofing’ measures (such as elevation of living spaces) were superior to coastal protection measures (such as seawalls, bulkheads, and revetments). Meanwhile, coastal protection was found to be optimal under one-metre sea-level rise. Another limitation of sea-level rise costing studies is their sensitivity to (land and structural) endowment values which are highly uncertain at more aggregate levels. A global assessment by Darwin and Tol (2001) showed that uncertainties surrounding endowment values could lead to a 17% difference in coastal protection, a 36% difference in amount of land protected, and a 36% difference in direct cost globally. A further factor increasing uncertainty in costs is the social and political acceptability of adaptation options. Tol et al. (2003) show that the benefits of adaptation options for ameliorating increased river flood risk in the Netherlands could be up to US$20 million /yr in 2050. But they conclude that implementation of these options requires significant institutional and political reform, representing a significant barrier to implanting least-cost solutions.

Adaptation studies looking at the agricultural sector considered autonomous farm level adaptation and many also looked at adaptation effects through market and international trade (Darwin et al., 1995; Winters et al., 1998; Yates and Street, 1998; Adams et al., 2003; Butt et al., 2005). The literature mainly reports on adaptation benefits, usually expressed in terms of increases in yield or welfare, or decreases in the number of people at risk of hunger. Adaptation costs, meanwhile, were generally not considered in early studies (Rosenzweig and Parry, 1994; Yates and Street, 1998), but are usually included in recent studies (Mizina et al., 1999; Adams et al., 2003; Reilly et al., 2003; Njie et al., 2006). Rosenzweig and Parry (1994) and Darwin et al. (1995) estimated residual climate change impacts to be minimal at the global level, mainly due to the significant benefits from adaptation. However, large inter and intra-regional variations were reported. In particular, for many countries located in tropical regions, the potential benefits of low-cost adaptation measures such as changes in planting dates, crop mixes, and cultivars are not expected to be sufficient to offset the significant climate change damages (Rosenzweig and Parry, 1994; Butt et al., 2005).

More extensive adaptation measures have been evaluated in some developing countries (see, for example, Box 17.3). For the 2030 horizon in Mali, Butt et al., (2005) estimate that adaptation through trade, changes in crop mix, and the development and adoption of heat-resistant cultivars could offset 90 to 107% of welfare losses induced by climate change impacts on agriculture.

Box 17.3. Adaptation costs and benefits for agriculture in the Gambia

Njie et al. (2006) investigated climate change impacts and adaptation costs and benefits for cereal production in the Gambia. Under the SRES A2 scenario the study estimated that for the period 2010 to 2039, millet yield would increase by 2 to 13%. For the period 2070 to 2099 the outcome is highly dependent on projected changes in precipitation as it could range from a 43% increase to a 78% decrease in millet yield. Adaptation measures such as the adoption of improved cultivars, irrigation, and improved crop fertilisation were assessed in a framework accounting for projections of population growth, water demand and availability. These measures were estimated to increase millet yield by 13 to 43%, while reducing interannual variability by 84 to 200% in the near term (2010 to 2039). However, net adaptation benefits (value of higher production minus cost of implementation) were not necessarily positive for all adaptation strategies. In the near term, net adaptation benefits were estimated at US$22.3 to 31.5 million for crop fertilisation and US$81.1 to 88.0 million for irrigation. The authors conclude that irrigation is more effective to improve crop productivity under climate change conditions, but the adoption of improved crop fertilisation is more cost efficient. Meanwhile, much uncertainty remains regarding the cost of developing improved cultivars. In the distant future, potential precipitation decrease would make irrigation an imperative measure.

In addition to their effect on average yield, adaptation measures can also smooth out fluctuations in yields (and consequently social welfare) as a result of climate variability. Adams et al. (2003) found that adaptation welfare benefits for the American economy increased from US$3.29 billion (2000 values) to US$4.70 billion (2000 values) when their effect on yield variability is included. In the case of Mali, Butt et al. (2005) show that adaptation measures could reduce the variability in welfare by up to 84%.

A particular limitation of adaptation studies in the agricultural sector stems from the diversity of climate change impacts and adaptation options but also from the complexity of the adaptation process. Many studies make the unrealistic assumption of perfect adaptation by individual farmers. Even if agricultural regions can adapt fully through technologies and management practices, there are likely to be costs of adaptation in the process of adjusting to a new climate regime. Recent studies for U.S. agriculture found that frictions in the adaptation process could reduce the adaptation potential (Schneider et al., 2000a; Easterling et al., 2003; Kelly et al., 2005).

With regard to adaptation costs and benefits in the energy sector, there is some literature – almost entirely on the United States – on changes in energy expenditures for cooling and heating as a result of climate change. Most studies show that increased energy expenditure on cooling will more than offset any benefits from reduced heating (e.g., Smith and Tirpak, 1989; Nordhaus, 1991; Cline, 1992; Morrison and Mendelsohn, 1999; Mendelsohn, 2003; Sailor and Pavlova, 2003; Mansur et al., 2005). Morrison and Mendelsohn (1999), meanwhile, estimate net adaptation costs (as a result of increased cooling and reduced heating) for the U.S. economy ranging from US$1.93 billion to 12.79 billion by 2060. They also estimated that changes in building stocks (particularly increases in cooling capacity) contributed to the increase in energy expenditure by US$2.98 billion to US$11.5 billion. Mansur et al. (2005), meanwhile, estimate increased energy expenditures for the United States ranging from US$4 to 9 billion for 2050, and between US$16 and 39.8 billion for 2100.

Besides sea-level rise, agriculture, and energy demand, there are a few studies related to adaptation costs and benefits in water resource management (see Box 17.4) and transportation infrastructure. Kirshen et al. (2004) assessed the reliability of water supply in the Boston metropolitan region under climate change scenarios. Even under a stable climate, the authors project the reliability of water supply to be 93% by 2100 on account of the expected growth in water demand. Factoring in climate change reduces the reliability of water supply to 82%. Demand side management measures could increase the reliability slightly (to 83%), while connecting the local systems to the main state water system would increase reliability to 97%. The study, however, does not assess the costs of such adaptation measures.

Box 17.4. Adaptation costs and benefits in the water management sector of South Africa

Callaway et al. (2006) provide estimates of water management adaptation costs and benefits in a case study of the Berg River basin in South Africa. Adaptation measures investigated include the establishment of an efficient water market and an increase in water storage capacity through the construction of a dam. Using a programming model which linked modules of urban and farm water demand to a hydrology module, the welfare related to water use (value for urban and farm use minus storage and transport cost) were estimated for the SRES B2 climate change scenario and the assumption of a 3% increase in urban water demand. Under these conditions and the current water allocation system, the discounted impact of climate change over the next 30 years was estimated to vary between 13.5 and 27.7 billion Rand. The net welfare benefits of adapting water storage capacity under current allocation rights were estimated at about 0.2 billion Rand, while adding water storage capacity in the presence of efficient water markets would yield adaptation benefits between 5.8 and 7 billion Rand. The authors also show that, under efficient water markets, the costs of not adapting to climate change that does occur outweigh the costs of adapting to climate change that does not occur.

N.B.: All monetary estimates are expressed in present values for constant Rand for the year 2000, discounting over 30 years at a real discount rate of 6%.

Dore and Burton (2001) estimate the costs of adaptation to climate change for social infrastructure in Canada, more precisely for the roads network (roads, bridges and storm water management systems) as well as for water utilities (drinking and waste water treatment plants). In this case, the additional costs for maintaining the integrity of the portfolio of social assets under climate change are identified as the costs of adaptation. In the water sector, potential adaptation strategies such as building new treatment plants, improving efficiency of actual plants or increasing retention tanks were considered and results indicated that adaptation costs for Canadian cities could be as high as Canadian $9,400 million for a city like Toronto if extreme events are considered. For the transportation sector, Dore and Burton (2001) also estimate that replacing all ice roads in Canada would cost around Canadian $908 million. However, the study also points out that retreat of permafrost would reduce road building costs. Also, costs of winter control, such as snow clearance, sanding, and salting, are generally expected to decrease as temperature rises.