Show Summary Details

Page of

date: 27 April 2017

# Global and Regional Economic Damages from Climate Change

## Summary and Keywords

Economic damage from climate change includes several aspects that need to be considered at the global and regional levels to achieve an equitable common solution to global warming. The economic literature reviewed here analyzes this issue under three general perspectives.

First, the analytical estimation of the linkages between damages in monetary terms and climate variables, as projections of temperature, precipitation, and frequency of extreme events, is rapidly evolving. Damage functions are included in complex economic models in order to calculate the economic impact of the climate change on economic output and growth, thus informing the debate on the amount of resources that should be devoted to reducing greenhouse gas (GHG) emissions and limiting climate damages. The choice of the geographical aggregation in this respect is a crucial aspect to be considered if policy advice is to be formulated on the basis of model results. The higher the level of regional detail, the more reliable the results are in terms of geographical distribution of economic damages.

Second, the precise estimation of the costs associated with different damages caused by climate change is attracting growing interest. Climate costs present a wide range of heterogeneity for several reasons, such as the different formulation of the damage function adopted, the modeling design of the economic impact, the temporal horizon considered, and the differentiation across sectors. Two broad categories of analysis are relevant. The first refers to the choice of the sectoral dimension under investigation, where some studies cover multiple sectors and their interactions, while others analyze specific sectors in depth. The second classification criterion refers to the choice of the economic aspects estimated, where a strand of literature analyzes only market-based costs, while other analyses also include non-market (or intangible) damages. The most common sectors investigated are agriculture, forestry, health, energy, coastal zones and sea level rise, extreme events, tourism, ecosystem, industry, air quality, and catastrophic damages. Most studies consider market-based costs, while non-market impacts need to be better detailed in economic models.

Third, the computation of a single number through the analytical framework of the social costs of carbon (SCC) represents a key aspect of the process of adapting complex results in order to properly inform the political debate. SCC represents the marginal global damage cost of carbon emissions and can also be interpreted as the economic value of damages avoided for unitary GHG emission reduction. Several uncertainties still influence the robustness of the SCC analytical framework, such as the choice of the discount rate, which strongly influences the role of SCC in supporting or not mitigation action in the short term.

Although the debate on the economic damages arising from climate change is flourishing, several aspects still need to be investigated in order to build a common consensus within the scientific community as a necessary condition to properly inform the political debate and to facilitate the achievement of a long-term equitable global climate agreement.

# Importance of Economic Assessment of Climate Change Damage

A central piece of information for climate policy is the damage (or benefit) that will occur as a result of climate change—when it will occur and where. While the impacts of climate change have first to be put in physical terms that information has to be taken a stage further wherever possible so as to give decision-makers an estimate of the impacts in monetary terms. Monetary measures, in spite of all their uncertainty, play an important role in deciding how much effort to put into reducing climatic impacts and how much to devote to adapting to the impacts that are going to take place. These assessments will vary from one country to another and will influence countries’ negotiating positions and efforts in fighting climate change. Where damages are found to be higher than perceived, countries will increase their efforts in developing adaptation programs and looking to finance them. The balance between mitigation and adaptation will also be affected by updated knowledge on the time profile of damages from climate change.

Given the pivotal importance of information on climate change damage, it is surprising how many gaps there are in our knowledge about the expected levels as a function of climate trends across regions and economic sectors. Efforts to assemble the data fall into three broad categories (Tol, 2009): (1) the enumerative method, which adds up the values of different impacts of climate change obtained from natural science papers, which in turn may be based on a combination of laboratory experiments and climate and impact models (DARA, 2012; Fankhauser, 1994); (2) the statistical approach, in which regressions are used to estimate the welfare impact as a function of projected temperature and precipitation, given the impacts experienced so far; and (3) an economic modeling approach in which an economy-wide computable general equilibrium (CGE) model or an integrated assessment model (IAM) is used to assess the consequences of a given set of physical impacts as reported in results from enumerative studies (Tol, 2015). CGE models have a more detailed economic structure (but with little physical modeling) linking different sectors of the economy in one country and in linking countries through trade, while IAM models combine physical climate models with aggregate economic models to project changes in the economy as a result of climate change.

Whatever the methodology used, a large number of uncertainties (both scientific and economic) must be taken into account (Markandya, 2014). There are several steps in the assessment of climate impacts, each of which generates uncertainties, with a cumulative effect on the final valuation. Figure 1 shows the elements in the assessment. First is the path of future emissions, which translates into changes in the atmospheric concentrations of CO2; second, the radiative forcing influences the transformation from concentrations to temperature increase; the next stage is the downscaling of the climate models to local conditions and to the system impacts for the sectors under consideration. In parallel, estimates of impacts will depend on future states of the world, which requires drawing up socioeconomic scenarios of population, economic activity, etc. These interact with the physical system level climate impacts to generate socioeconomic impacts, which in turn form the basis for adaptation measures.

Click to view larger

Figure 1: Structural elements in the assessment of climate change impacts illustrating the uncertainty cascade.

(Source: Markandya, 2014)

Given uncertainties of future climate change impacts, it is not surprising that results from numerous studies are very heterogeneous and that the quantitative assessment is still uncertain and incomplete (Tol, 2015). However, while climate cost estimates vary greatly, all these studies agree on one fact: climate change is affecting the world, and developing countries are suffering and will suffer the highest costs.

The literature reviewed in this article falls into three categories. First is the estimation of the damage functions linking damages mainly to projections of temperature, precipitation, and frequency of extreme events. Second are the actual estimates of damages emerging from the use of these functions. In both cases the data can be quite detailed and disaggregated by region and sector. The third component of the literature seeks to provide a single number that applies to all sectors and region. Given that emissions of GHGs have the same effect on concentrations irrespective of wherever they are emitted, it is possible to calculate the global impact of a ton of GHGs emitted at a given point in time on global damages for as long as that ton adds to concentrations in the atmosphere. Estimates of these damages over time and across space are referred to as the social cost of carbon. The rest of this article summarizes and reviews the literature on these three sets of information on climate change damages, with a final section offering conclusions on the current state of knowledge in this field.

# Damage Functions

Damage functions link damages in money terms to climate variables such as temperature or precipitation. Typically temperature is the variable most commonly used. These functions then feed into IAMs, which calculate the impact of the damages on economic output and growth globally or in a given region. Several IAM models have been used, with differences that are in part a matter of subjectivity in the modeling design.1

One of the best known IAMs is DICE (Dynamic Integrated model of Climate and the Economy) (Nordhaus & Sztorc, 2013). It is a policy optimization model based on neoclassical economic growth theory, extended to include the natural capital of the climate system. It includes different modules according to which the CO2 emissions due to fossil fuel consumption affect the carbon cycle and, therefore, the climate system, resulting in economic impacts on different sectors of the ecosystem. More precisely, the model follows a production function approach with a constant-returns-to-scale Cobb-Douglas of capital, labor, and technological change that also considers climate damages and abatement costs. The impacts sectors considered are agriculture and forestry, water resources, coastal zones, energy consumption, air quality, and human health. However, in order to include the effects of non-monetized impacts (as losses from biodiversity, extreme and catastrophic events, very long-term warming and uncertainty), an increase of 25% has been introduced to reflect these impacts. The updated version assumes that climate damages are a quadratic function of (global mean) temperature change, thus a fraction of output in each period is expressed as:

$Display mathematics$

where Q(t) is output net of damages and abatement, A(t) is total factor productivity, K(t) is the capital stock, while Ω‎(t) and Λ‎(t) represent, respectively, climate damages and abatement costs. The damage function is given by the following expression:

$Display mathematics$

where TAT is the atmospheric temperature and ψ1 is a parameter of the equation. The function is an aggregate of damages to agriculture and forestry, water resources, coastal zones, energy consumption, air quality, and human health and depends solely on temperature. It does not include thresholds or tipping points or potential catastrophic impacts and has been calibrated to account for temperature increases from 0 to 3°C, owing to the difficulty of finding reliable estimates associated with larger warming levels. The functional form adopted ensures that damages do not exceed 100% of output.

The RICE (Regional Integrated model of Climate and the Economy) is the regionalized version of the DICE and shares the same modeling structure (Nordhaus & Sztorc, 2013), with the following 12 regions: the United States, the European Union (EU), Japan, Russia, Eurasia, China, India, Middle East, sub-Saharan Africa, Latin America, other high income countries, and other developing countries. Accordingly, each region has a social welfare function and optimizes its consumption, GHG policies, and investment over time.

The FEEM-RICE model is an extended version of the RICE 99 model by Nordhaus and Boyer (2000) with the same representation of the damage function but with a detailed focus on technological aspects (Bosetti, Carraro, & Galeotti, 2006). Climate change impacts are expressed as a fraction of world output through a polynomial function of global mean temperature change accounting for agriculture, health, rise in sea level, and non-market and catastrophic damages similar to RICE.

MERGE is a Model for Estimating the Regional and Global Effects of GHG reductions, developed and updated by, respectively, Manne, Mendelsohn and Richels(1995) and Manne and Richels (2005). It is an intertemporal model of optimal long-term economic growth that includes a climate change module that distinguishes between market (or economic) and non-market (or intangible) damages. With respect to the former, considering an equilibrium climate sensitivity of 2.5°C,2 the model assumes that such a temperature rise would lead to GDP losses of 0.25% in the high-income nations and to losses of 0.50% in the low-income nations. For different temperature changes, the market losses are assumed proportional to the change in mean global temperature from its level in 2000. For non-market damages, on the other hand, which are the focus of the model, the expected losses are assumed to increase with the temperature rise squared in order to represent the risk that losses from possible catastrophes could increase radically. The damage function in this case is based on the willingness to pay (WTP) to avoid a temperature rise. For high-income countries, the authors assume a WTP equal to 2% of GDP, to avoid a 2.5°C global temperature rise w.r.t. its level in 2000. This percentage is chosen because it is the share devoted by the United States to all forms of environmental controls (Manne & Richels, 2005), while lower-income regions are assumed to have lower WTP (1% GDP). Accordingly, the damage costs are introduced in the model through the following economic loss factor (ELF):

$Display mathematics$

where x is the temperature rise. The parameters hsx, the “hockey-stick” parameter representing the (quadratic) loss due to temperature rise, and catt, the catastrophic temperature parameter, define the WTP. For high-income countries, hsx takes the value of 1 and catt the value 17.7°. Accordingly, the ELF represents the fraction of consumption that remains available for conventional uses by households and by government. The authors acknowledge that the two parameters representing the WTP are still highly speculative (Manne & Richels, 2005).

The WITCH (World Induced Technical Change Hybrid) model is another top-down neoclassical optimal growth model based on a development of the FEEM-RICE model (Bosetti et al., 2006b; De Cian, Bosetti, & Tavoni, 2012). In the last model update (Bosello & De Cian, 2014), a new set of damage functions and adaption cost curves have been introduced, separating benefits from negative impacts due to climate change. They distinguish between market damage component from non-market damages, based on data from the EU FP7 ClimateCost project (Bosello, Eboli, & Pierfederici, 2012) and from Nordhaus (2007). The WITCH updated damage function, accounts for impacts on the sea level rise, and changes in crop productivity and in energy demand for a 1.9°C temperature increase (excluding tourism, net primary productivity of forests and floods). The market impacts are calculated as a loss (% change) in the regional GDP and estimates of coastal land loss due to the sea level rise are based on the DIVA (Dynamic Integrated Vulnerability Assessment) model (Vafeidis et al., 2008). The impacts on the agriculture sector are calculated as changes in the average productivity of crops from the ClimateCrop model (Iglesias, Garrote, Quiroga, & Moneo, 2009; Iglesias Quiroga, & Garrote, 2010). The change in residential energy demand due to increasing temperatures are derived from the POLES (Prospective Outlook on Long-term Energy Systems) model (Criqui, 2001; Criqui, Mima, & Menanteau, 2009). The non-market damage component, on the other hand, includes damage estimates for non-market health impacts and catastrophic damages (Nordhaus, 2007) and ecosystem losses. In this case, the WITCH model follows the approach used in the MERGE model, but uses an updated proxy for the WTP, considering the EU average per capita environmental expenditure (0.6% of EU25 GDP in 2001 due to a 2.5°C increase) linked to the per capita income, as follows:

$Display mathematics$

where γ‎ and ε‎ are parameters, ∆T is the temperature rise based on a climate sensitivity of 2.5°C, and GDP and POP are the GDP and population projections in the given year based on a model that assumes a climate sensitivity of 2.5°C. The parameters γ‎ and ε‎ are been calibrated to give exactly 0.6% of GDP loss when per capita income is 28,780 USD and Δ‎T = 2.5°C. The S-shaped relationship between per-capita income and WTP has then been used to compute the WTP in the different model regions.

Another IAM, the ICAM (Integrated Climate Assessment Model), represents the interaction between socioeconomic and natural system, as follows:

$Display mathematics$

where the income Y for region j in time t under policy k is a function of income, cost of mitigation (CM), climate change impacts (CC), dead weight losses due to energy consumption (DE), and the welfare impact of tax collection (Dτ‎) in t−1 as well as the exogenous rates of population growth (ρ‎) and welfare enhancing technical growth (γ‎) at time t. The impact from climate change (CCj,k,t−1) is calculated as a function of the temperature change, its rate of increase, an estimate of the agricultural sector as a fraction of the economy, and coastal zone damages due to sea level rise (Dowlatabadi, 1998).

The FUND (Climate Framework for Uncertainty, Negotiation and Distribution) model is a non-CGE policy optimization model that includes a complex damage module (Waldhoff, Anthoff, Rose, & Tol, 2014). The impacts due to climate change involve agriculture, forestry, sea level rise, health (in term of cardiovascular and respiratory disorders related to cold and heat stress, malaria, dengue fever, schistosomiasis), energy consumption, water resources, unmanaged ecosystems (Tol, 2002a, 2002b), diarrhea (Link & Tol, 2004), and tropical and extra-tropical storms (Narita, Anthoff, & Tol, 2009, 2010). The impact of sea level rise is calculated first in terms of premature deaths, migration, and losses of dryland and wetlands and then monetized, while for the other impacts there is no representation in geophysical terms but only in monetary values. For energy consumption, agriculture, cardiovascular and respiratory diseases, the sign and magnitude of the impacts depend on the distance from the assumed climatic optimum (determined by a variety of factors, including plant physiology and the behavior of farmers), while the remaining impacts are modeled as simple power functions, either negative or positive of temperature, depending on whether the actual climate conditions are moving closer to or further away from that optimum climate. Furthermore, vulnerability to climate change depends on population growth, economic growth, and technological progress, and while market damages affect both economic growth (through investment) and welfare (through consumption), the impact of non-market factors is only on welfare.

PAGE2009 (Hope, 2010, 2011) follows the FUND model approach and, accordingly, the damage function is a polynomial function that depends on the rate and magnitude of temperature increase. Relative to the economic and non-economic sectors, it adopts the enumerative approach, which adds the non-economic to economic impacts. Among the main extensions included in PAGE2009, nitrous oxide is included as a polluting gas and the climate sensitivity depends on both the transient climate response and the feedback response time. In addition, a carbon cycle feedback is introduced (as a linear feedback from global mean temperature to a percentage gain in the excess concentration of CO2); regional temperature is adjusted for the effective latitude and the land-based nature of the region; sea level rise is explicitly modeled (as a lagged linear function of global mean temperature). Furthermore, PAGE2009 links directly the impact to the per capita GDP, “letting the impacts drop below their polynomial on a logistic path once they exceed a certain proportion of remaining GDP to reflect a saturation in the vulnerability of economic and non‐economic activities to climate change, and ensure they do not exceed 100% of GDP” (Hope, 2010, p. 3). Discontinuity impacts are modeled in a way that they do not occur immediately but within a delayed lifetime, and the equity weighting scheme proposed by Anthoff et al. (2009) is also used.3

The above models are relatively aggregated economic models and do not take account of intersectoral linkages in the way that a computable general equilibrium (CGE) model does. CGE models are very demanding in terms of details at the sector level so as to construct the social accounting matrix that is the basis for the modeling of changes or shocks to the economic system. Over time, changes in these linkages will occur, and it is very difficult to account for all of them in a credible way for a long period of time. Since the IAMs are looking to 2100 and beyond, the CGE modeling framework is not suitable. An attempt has been made to look at the medium term using the CGE model and then link it for the longer term to an IAM. The OECD’s CIRCLE project (Costs of Inaction and Resource Scarcity: Consequences for Long-term Economic Growth Project) is such a hybrid (OECD, 2015). It provides an assessment of the economic implications of climate change damages. The damage function to 2060 follows a production function approach to link climate change impacts to specific drivers of growth in the dynamic general equilibrium model ENV-linkages and results in climate change damage expressed as percentage of GDP. It explicitly includes in quantitative terms several types of climate change impacts within the modeling framework: loss of land and capital due to sea level rise, capital damages from hurricanes, changes in crop yields, fisheries catches, labor productivity, tourism flows, health care expenditures due to diseases and heat stress, and energy demand for cooling and heating. The difference with IAMs is that to 2060 the climate impacts affect the productivity of individual sectors and thereby alter the economic structure in the economy.4 Beyond 2060 the analysis resorts to the standard IAM approach of looking at a more aggregate economy.

Table 1 summarizes the findings from the survey of damage functions. A number of points stand out. First is the heavy reliance on temperature alone on defining damages as a function of climate change. Other factors such as precipitation and wind are not directly used in the damage functions, which is clearly a weakness. Second, the functions are estimated from current temperature variations, which in the time series for given locations is relatively small. Drawing inferences from cross-section differences and applying them to project forward in time is problematic. Third, a lot of ad hoc assumptions have been made, such as a percentage addition for non-market influences, proportionality of market damages to temperature changes, and so on. Fourth, many impacts are not directly accounted for. In particular, tipping points and discontinuities are missing or are included in a very ad hoc way, as are effects of social disorder, conflicts, and so on. Finally, most of the damage functions are applied to an aggregated model. Although they have been put together from separate impacts in different areas (sea level rise, agriculture, health, etc.), the overall model applies—in most cases, a single production function for a given region. The exception is the OECD model for the medium term to 2060, which disaggregates the economy into 40 sectors.

For all these reasons estimates of climate change damages as estimated by these functions can only be regarded as orders of magnitude, with a risk that the numbers will be on the low side. In general the estimates from these models have been used to inform the debate on the amount of resources that should be devoted to reducing GHG emissions and thereby limiting climate damages and less to set adaptation policies at the sectoral level (i.e., to decide on how to cope with the climate impacts on agriculture, health, etc.). Furthermore, while differences between models on GHG mitigation due to climate damage estimates can be considerable, these are less than the differences due to the choice of the discount rate. In addition, it is important to remember that the damage estimates are subject to wide ranges of uncertainty within each model, and different ways of dealing with this uncertainty can have different implications for emissions reductions.

The next section looks at the actual numbers for the damages by sector and which lessons can be drawn from them.

Table 1. Damage Functions for Climate Change

Study

Region

Sector

Damage function form

Fully Integrated IAM

DICE-2013R

Global; single product (sector)

Based on Tol (2009). Agriculture and forestry, water, coastal zones, energy, air quality, and human health

Affects a fraction of world output: quadratic function of global mean temperature change

Adjustment of 25% of the monetized damages to reflect non-monetized impacts

RICE-2010

12 regions; single product (sector)

Follows DICE

Follows DICE

FEEM-RICE

8 regions; 1 single product (sector)

Follows DICE-2007: agriculture, health, rise in sea level, non-market and catastrophic damages

Affects a fraction of world output, polynomial function of global mean temperature change

WITCH

12 regions; 1 single product (sector)

Based on EU FP7 ClimateCost project (Bosello et al., 2012) and Nordhaus (2007):

• Market: sea level rise, energy demand, agricultural productivity, tourism flows, net primary productivity of forests, floods, and reduced work capacity because of thermal discomfort)

• Non-Market: ecosystem losses, non-market health impacts, catastrophic damages

Willingness-to-pay (WTP) approach

The updated damage function only uses the economic market impacts on

• rise in sea level

• changes in crop productivity

• and in energy demand

MERGE

5 regions; energy products (electric and non-electric)

Follows DICE-2007

Market as in DICE-2007; non-market willingness-to-pay (WTP)

Market losses proportional to change in mean global temp. w.r.t. 2000; non-market losses increase quadratically with the temperature rise

ICAM

17 regions; energy types: oil, gas coal and non-fossil

Agricultural sector as a fraction of the economy, and coastal zone damages due to sea level rise

Function of temperature change, its rate of increase

FUND

16 regions

Agriculture, forestry, sea level rise, cardiovascular and respiratory cold and heat stress, malaria, dengue fever, schistosomiasis, energy consumption, water resources, unmanaged ecosystems, diarrhea, tropical and extra tropical storms9

Specific climate impact module; includes market and intangible impacts

Attempt to include non-benchmark climate change and socioeconomic vulnerability

PAGE 2009

8 regions

Economic and non-economic sectors; follows the FUND model

Depends on the rate and magnitude of temperature increase; polynomial function of temp

Enumerative approach (sum of economic and non-economic impacts)

Agriculture, coastal zones, extreme events, health, energy and tourism demand; CGE model used to 2060, IAM model after; CGE model has 40 sectors

Replicates the net damages of the DICE model

Not covered: impacts on ecosystem, water stress, human security and tipping points (large-scale disruptive events)

# Estimates of Damage Caused by Climate Change

Estimates of damages from climate change derived from the damage functions can be directly used for policies relating to adaptation at the sectoral level as well as contributing to the debate on mitigation of GHGs. Several studies have made projections of damages globally and regionally across all sectors, and some have looked at specific sectors in greater detail.

## Studies Covering Multiple Sectors

Tables 2 and 3 summarize the global and regional studies covering several sectors. They report damages as a percentage of GDP in the relevant region at different points in time, as a function of projected changes in temperature or future concentrations of GHGs. A major difference in damages arises from the time period in the future that is being assessed. Table 2 looks at the medium-term estimates (to 2060), while Table 3 considers estimates for 2070 and beyond.

Table 2 shows damage estimates to be relatively small for the period to 2060: from a low of 0.3% of world GDP to a high of 2%. One of the first attempts is represented by the Mendelsohn model that estimates the global impact to be very small (0.3% of global GDP). In general the estimates do not vary much with projected emissions because such emissions do not have a major effect on climate change by 2060. Estimates can vary with the degree of coverage; for example, the study by Dellink, Lanzi, Château, Bosello, Parrado, and de Bruin (2014) has wider coverage of non-market impacts than the Mendelsohn study, which considered only five market sectors, namely agriculture, water, forestry, energy, and coastal zones.5 Among the studies included in Table 2, DARA (2012) is one of those that takes into account the wider range of impacts, both market-based and non-market based (e.g., losses from biodiversity, desertification), generally omitted by other studies. In particular, DARA follows the enumerative method using 22 indicators representative of four areas: environmental disasters, habitat change, health impact, and industry stress. Costs associated with each indicator come from the results of scientific research and from the application of specific models. According to this study the global cost due to climate change in 2010 was 1% of world GDP, and in 2030 it is expected to be about 2% of world GDP. By analyzing the impact for 184 countries, it also concludes that developing countries suffer the highest costs.

This finding of higher costs in developing countries is also present in other studies examined. The OECD study finds that by 2060 some regions such as OECD Europe only lose about 0.2% of their GDP, while sub-Saharan Africa and South and South-East Asia lose around 3.8 and 3.7%, respectively.

Table 3 is interesting in comparison with Table 2 in that it shows considerably higher damages in the longer term and, furthermore, a wider range of estimates. Global damages in the latter part of this century range from 1.7% of GDP (Nordhaus, 2011b) to 9% (OECD, 2015). Second, damages are sensitive to projected changes in temperature or concentrations of GHGs by the end selected date. With an increase in temperature of around 2.5°C, the loss is around 1.5 to 1.9%; with a 4°C increase, the loss is estimated at 6%, and with a 6°C increase it goes up to 9%.

Table 2. Damage Estimates in the Short to Medium Term (as % of GDP)

Study

Region

Year

Scenario: Temperature increase/ppm

Method

Estimates (GDP change)

Factors included

Mendelsohn et al. (1998)

Global

2060

2°C

Global Impact Model (GIM)

−0.3% of world GDP

Market impacts: agriculture, water energy, sea level rise, forests

Bosello et al. (2009)

Global

2050

1.2°C

ICES Model

−0.3%

Market and non-market impacts (extreme events excluded)

Global

2050

3.1°C

ICES Model

−1%

Bosello et al. (2012)

Global

2050

1.92°C

ICES Model

−0.5%

Sea level rise, energy demand, agriculture, tourism, forestry, health, floods.

DARA (2012)

Global10

2010

Enumerative method

−1%

Habitat change, Health impact, industry stress, environmental disasters (floods and landslides, storms, wildfires, drought)

Global

2030

IPCC SRES A1B

Enumerative method

−2%

Global

2060

>2.5°C

−1.5%

Sea level rise, health (changes in morbidity and demand for healthcare), ecosystems, agriculture, tourism, energy demand, and fisheries

OECD (2015)

Global

2060

3°C

−2%

Agriculture, coastal zones, extreme events (mortality, land and capital damages from hurricanes), health, energy demand, tourism demand

Middle East & North Africa

2060

3°C

−3.3%

South- and South-East Asia

2060

3°C

−3.7%

Sub-Saharan Africa

2060

3°C

−3.8%

Latina America

2060

3°C

−1.5%

Rest of Europe and Asia

2060

3°C

−2.1%

Third, the review shows that damages vary considerably by income quartile. The Mendelsohn, Dinar, and Williams (2006) study is critical in this regard by showing losses to the poorest quartile to be over 100 times those of the richest quartile. Moreover, this difference gets bigger as the extent of climate change increases. In 2100, if concentrations of GHGs rise from 685 ppm to 808 ppm, losses to the bottom quartile go up from 11.8% of income to 23.8%, while losses to the top quartile are unchanged.

Lastly, the finding of higher impacts for the poorest countries is further strengthened. The Moore and Diaz (2015) study shows that even with a 2°C increase, per capita income could be 40% lower than it would be without climate change (although the baseline level of income would have increased over time on account of underlying growth). Several factors can explain the higher economic costs of climate change for developing countries: first, their geographic position—most low-income countries are located in low-mid latitudes, characterized by higher temperature increases; second, developing countries are highly reliant on climate-dependent sectors, such as agriculture. Finally, they are less able to adapt because of the lack of institutions and financial resources (Tol, 2009).

Table 3. Damage Estimates in the Longer Term (as % of GDP)

Study

Region

Year

Scenario/Temperature increase/ppm

Method

Estimates (GDP change)

Factors included

Nordhaus and Boyer (2000)

Global11

2100

2.5°C

DICE99/RICE99

• −1.5%

• −1.88%

Agriculture; sea-level rise; market sectors; health; non-market impacts; human settlements/ecosystems; catastrophes

Mendelsohn et al. (2006)

Poorest Quartile

2100

685 ppm

Experimental studies

−11.8%

Market impacts

Poorest Quartile

2100

808 ppm

Experimental studies

−23.8%

Second Quartile

2100

685 ppm

Experimental studies

−2.4%

Second Quartile

2100

808 ppm

Experimental studies

−7.4%

Third Quartile

2100

685 ppm

Experimental studies

−0.3%

Third Quartile

2100

808 ppm

Experimental studies

−2.4%

Richest Quartile

2100

685 ppm

Experimental studies

0.1%

Richest Quartile

2100

808 pmm

Experimental studies

0.1%

Stern Review (2007)

Global

2100

2.4°C to 5.8°C

PAGE2002

−5%

Water, food, health, land, environment, abrupt, large-scale impacts

Nordhaus (2011b)

Global

2100

2.5°C

RICE2010

−1.7%

EC (2011)

EU

2100

Non-mitigation

PAGE09

−4%

Agriculture, health, energy, sea level rise, river floods, ecosystem

EU

2100

2°C

PAGE09

−0.5% to −1% of world GDP

Ciscar et al. (2014)

EU

2080s (2071–2100)

IPCC SRES A1B

GCM—RCM12—GEM-E3—DIVA

• −1.1% of EU GDP

• −1.8% of global welfare

Agriculture, energy, river floods, forest fires, transport infrastructure, coastal areas, tourism, human health

Global

2°C

−0.7% of EU GDP

Global

2100

>4 (between 2.4°C and 5.8°C

−3.5%

Sea level rise, health (changes in morbidity and demand for healthcare), ecosystems, agriculture, tourism, energy demand and fisheries.

OECD (2015)

Global

2100

4.5°C

−6%

Agriculture, coastal zones, extreme events (mortality, land and capital damages from hurricanes), health, energy demand, tourism demand

Global

2100

6°C

−9%

Moore and Diaz (2015)

Poorest countries

2100

2°C

DICE

−40% (gdp pc)

Note: Concentrations in ppm are related with some uncertainty to increases in temperature. A level of 685 ppm is likely to result in a temperature increase of 2.5–3.0°C, while a concentration of 808ppm is likely to be associated with a temperature increase of 3.2 to 3.5°C (IPCC, 2014).

## Studies Covering Specific Sectors

A number of studies provide sectoral estimates, including some that estimate the impact of climate change on a global scale. Tables A1A9 in the Appendix summarize the main findings for agriculture, forestry, health, energy, coastal zones, extreme events, tourism, ecosystem, and industry. The main findings are as follows:

• Agriculture: Although there is much discussion of agricultural impacts of climate change, the global level studies do not indicate significant damages either in the short term or even looking as far ahead as 2100. To be sure there are losses in specific regions, but these are compensated in part by gains in others and by trade in agricultural products. There is also a major element of autonomous adaptation as agricultural producers respond to climate change by modifying crops, using more resilient seed varieties, etc. Estimates for 2050/2060 are in the range of 0.3% to 0.8% of global GDP, while those for 2085/2100 are wider, at around 0.2% to 1.7%. It is important to note, however, that some countries and regions will suffer considerably greater losses. Sectoral studies such as Zhai and Zhuang (2009) find damages for India to be as high as 6.2% and for sub-Saharan regions to be in the range of 2.2% of GDP.

• Forestry: Damages to forestry from climate change are estimated as low relative to GDP—they range from negligible (Bosello et al., 2012) for 2050 to around $45 billion (Dara, 2012) for 2030, which is less than 0.1% of current global GDP and even less than that of projected GDP in 2030. It is important to note, however, that not all impacts are fully understood, and the possible impacts on biodiversity in forests are not valued in this assessment. • Health: Health effects of climate change are a major source of public concern, but the estimates in the short term are not significant in comparison to GDP. In fact, some studies indicate a small gain up to 2050 (around$11Bn; Bosello, Rosen, & Tol, 2006), while others come up with damages in the region of 0.6% of GDP. Gains arise partly as a result of higher winter temperatures reducing winter mortality. Other sources of health impacts include an increase in vector-borne diseases, an increase in food-borne diseases, mortality and morbidity from extreme events, loss of labor productivity, and illness from undernutrition in places where agricultural production declines. In the longer term, global estimates are not available, but some indications are for higher costs depending on which scenario is realized. A major issue with health effects is valuing loss of life. Methods based on the statistical value of life place a higher value on mortality in richer countries, which raises ethical issues. As a result, some IAM models place an equity weighting on loss of life in poorer countries, so all loss of life has a similar value irrespective of where it occurs. If one does that, the health costs are higher, although again not significantly so in the medium term.

• Energy: Energy costs relate to changes in the demand for heating (which goes down) and cooling (which goes up). Also, there are changes in hydropower capacity as a result of changes in water flows. Global damage estimates tend to be small (or even negative, indicating some gains). Regional variations are also not large and have a high level of uncertainty with respect to hydropower impacts.

• Coastal Zones: Sea level rise and storm surges are a major source of impacts of climate change. Estimates of damages are based on loss of land and costs of dealing with flood events. Also included are costs of some adaptation actions, such as sea dykes, where these are economically justified. The figures indicate a global damage of 0.1 to 0.2% of GDP in 2050, with a possible higher estimate for 2030 projected to be around $550 billion, or around 0.4% of projected global GDP in that year. By 2100, damages range from 0.12% (Nordhaus & Boyer, 2000, DICE model) to 0.32% (Nordhaus & Boyer, 2000, DICE model). Regional estimates vary substantially, and some coastal cities will need to spend a significant amount to deal with the impacts of climate change. A study by the Centre for Risk Studies at Cambridge University (Cambridge Centre for Risk Studies, 2015) estimates that from 2015 to 2030, climate change risks to the 300 most affected cities amount to around$156 billion, equal to about 0.04% of the value of their economic output over the period. Cities most likely to be affected are Taipei, Tokyo, Seoul, Manila, Istanbul, and New York, which have a significant share of this cost.

• Extreme Events: The table on extreme events shows perhaps the largest but also the most uncertain estimates of damages. Damages are from cyclones, floods, droughts, and similar events. In the short term the global figures range are around $307 billion (for 2030), equal to about 0.2% of global GDP. In 2100 figures for a temperature increase of 2.5°C range from as low as 0.007% of GDP (Narita et al., 2009) to 1% (Nordhaus & Boyer, 2000). With a 6°C increase by that year, losses are around 7% of GDP. Models of changes in the frequency and intensity of extreme events as a result of climate change are still not well developed, and so the estimates have a higher than average degree of uncertainty. • Tourism: Tourism is likely to be affected in places where temperatures rise and extreme events make them unattractive as destinations, while other areas may benefit from the change in holiday destination due to temperature increase. Globally damages are small and not particularly greater over time as are some other categories. In 2050/60 they range from 0.1% to 0.5%, and in 2100 one global study (Bigano, Hamilton & Tol, 2007) puts them at virtually 0% of GDP. • Ecosystems: Losses to ecosystems from climate change are among the most difficult to track and value. Most studies provide a qualitative assessment of the impacts, and only one study has made an attempt to value the damages. DARA (2012) estimates damages in 2010 and in 2030 from climate change to ecosystems through degradation of terrestrial and marine areas as well as losses of biodiversity. The figures are$115 billion in 2010 and $570 billion in 2030, which made the latter figure around 0.4 to 0.5% of projected global GDP in that year. Although the total may not be large, the damages are particularly important as they affect the livelihoods of some of the most vulnerable groups of people on the planet. • Industry: Damages to industry occur partly through the impacts already considered (effects on agriculture, forestry, energy, human heath, etc.). But there are also some direct effects as climate change damages infrastructure in transport and building. DARA (2012) estimates damages in these sectors as well as costs of fisheries under the industry category and comes up with figures of$150 billion to fisheries and $5 billion to transport in 2030. The World Bank (2010) estimates an average figure for the period 2010 to 2050 to cover costs of measures to address climate change impacts on buildings and transport infrastructure and comes up with an estimate in the range of$13 to $27 billion, depending on whether a wet or a dry climate scenario emerges over this period. Combining both the DARA estimates for fisheries and the World Bank estimates for infrastructure gives a total in the range$163 to $177 billion around 2030, which would make up around 0.1% of global GDP. # The Social Cost of Carbon (SCC) This section looks at damages from climate change captured in a single parameter: the social cost of carbon (SCC), which is an estimate of the discounted present value of economic damages associated with a small increase in carbon dioxide (CO2) emissions, conventionally one metric ton, that occur in a given year. It is thus the marginal global damage cost of carbon emissions, and it can also represent the value of damages avoided for a small emission reduction (i.e., the benefit of a CO2 reduction). The SCC costs have been reviewed in some depth in the literature (Anthoff & Tol, 2013; DEFRA, 2007; U.S. Government, 2013). Since the value is based on the discounted damages arising from a ton of CO2 over the long term, they are sensitive to the discount rate adopted. The higher the discount rate, the lower will be the value attached to future damages and hence the lower will be the discounted present value of the damages. This discounted value also increases over time as damages rise with time, so the SCC from one ton of CO2 released in 2050 will be higher than that of one ton released in 2015. The U.S. government review of 2013 is probably the most comprehensive recent assessment of SCC. Box 1 describes the elements in the calculation of the social costs of carbon from the different models that have been used and that are covered in this review. Box 1: Elements in the Social Costs of Carbon (SCC)  The SCC is calculated by running an Integrated Assessment Model (IAMs) in which future economic output is estimated under different scenarios for emissions of GHGs. By running the model with given emissions scenario, calculating the discounted present value of output and then running the model again with a small increase in emissions in the current period a second discounted present value is obtained. Subtracting the discounted value in the second run from the first gives an estimate of the damage caused by that small increase. Dividing the damage by the change in emissions gives the SCC today. The same calculation can be made starting the model in 2020, 2030 etc., to get the SCC for that year.The impacts of climate change taken into account vary from one model to another. Three major models are DICE, FUND and PAGE. All include the damage caused by seal level rise (SLR), agriculture and energy (higher demand for energy for cooling but less for heating). These also include additional costs of health treatment resulting from higher temperatures and extreme events. Models vary in the damage function they use (i.e. the link between emissions and climate change and between climate change and damages) and there is an element of arbitrariness about the functions. Elements not included in the models are:1) Incomplete treatment of non-catastrophic damages: current IAM’s do not assign value to all important physical, ecological and economic impacts of climate change, and it is recognised that even in future applications a number of potentially significant damage categories will remain non-monetised i.e. ocean acidification (not quantified be any of the 3 models), species and wildlife loss.2) Incomplete treatment of potential catastrophic damages: damage functions may not capture the economic effects of all possible adverse consequences of climate change: i.e., (i) potentially is continuous ‘tipping point’ behaviour in Earth systems; (ii) inter-sectoral and inter-regional interactions, including global security impacts of high-end warming; and (iii) imperfect substitutability between damage to natural systems and increased consumption.3) Uncertainty in extrapolation of damages to high temperatures: estimated damages are far more uncertain under more extreme climate scenarios.4) Incomplete treatment of adaptation and technological change: models do not adequately account for potential adaptation or technological change that may alter the emissions and resulting damages. Source: Markandya et al. (2016). The sensitivity of SCC to the discount rates can be seen from Table 4, taken from the U.S. government report. Three points should be noted in this table. First, the discount rate increases the SCC by more than a factor of 4 in 2010 and by just under a factor of 4 in 2050. Second, for any given discount rate, the figures given in the second to fourth columns are averages. Because of the uncertainty in damages, there is a distribution around the mean, and the last column gives the 95th percentile values for a discount rate of 3%. These are about three times greater than the mean. Lastly, there is a substantial increase in SCC over time: between 2010 and 2050 the value per ton of carbon rises by a factor of about 2.5 for a high discount rate and a factor of about 2 for a lower discount rate. Table 4. SCC Values in Dollars (US 2007) per Metric Tons of CO2. Discount Rate Year 5% Average 3% Average 2.5% Average 3% 95th Percentile 2010 11 33 52 90 2020 12 43 65 129 2030 16 52 76 159 2040 21 62 87 192 2050 27 71 98 221 Source: U.S. Government (2013). # Conclusions This article lays out the state of the art on the estimation of damages caused by climate change. A number of points stand out. To someone coming to the literature from the outside, perhaps one outstanding feature is the wide range of possible estimates of the social costs of carbon. If policies are to be based on these, a prior decision will have to be taken on which discount rate to take and how to treat the range of cost estimates once the discount rate has been pinned down. These choices influence the outcome significantly: a high discount rate supports little action on mitigation in the short term, while a low discount rate does the opposite. At the same time a wide range of values suggests that a precautionary approach would encourage policymakers to take mitigation action early. The second outstanding feature of the damage literature is the low level of aggregate damages up to middle of the 21st century and the possibly higher damages as we get closer to the end of the century. The former is a relatively robust result, although it has been questioned on the grounds that a number of impacts are not adequately valued in money terms. While work is ongoing in this area, it is unlikely that damages as a percentage of GDP will rise very much in the short to medium term. What these low values do not pick up is the significant effects the damages can have at specific locations and for specific people. Indeed, the literature already shows some whole regions and several developing countries suffering much higher damages as a percent of GDP than the average for the world as a whole. The last point to note is the relatively weak database for damages at the sectoral level. While the damage data may be good enough for policy at the aggregate level—how much to reduce GHGs and when, and how much to devote to adaptation in the aggregate—much more detail is needed for adaptation policies at the sectoral level. Much work is going on in this area, and it has to be local and site specific. As the data from these exercises are assembled, we will be able to use them to strengthen the damage database as a whole and improve policymaking in the climate change arena. ## Suggested Readings Markandya, A., Galarraga, I., & Sainz de Murieta, E. (2014). Routledge handbook of the economics of climate change adaptation. London: Routledge.Find this resource: Tol, R. S. J. (2009). The economic effects of climate change. The Journal of Economic Perspectives, 23(2), 29–51.Find this resource: ## References Anthoff, D., Hepburn, C., & Tol, R. S. J. (2009). Equity weighting and the marginal damage costs of climate change. Ecological Economics, 68(3), 836–849.Find this resource: Anthoff, D., & Tol, R. S. J. (2013). The uncertainty about the social cost of carbon: A decomposition analysis using fund. Climatic Change, 117(3), 515–530.Find this resource: Bigano, A., Hamilton, J. M., & Tol, R. S. (2007). The impact of climate change on domestic and international tourism: A simulation study. The Integrated Assessment Journal, 7(1), 25–49.Find this resource: Bosello, F., & De Cian, E. (2014). Documentation on the development of damage functions and adaptation in the WITCH model. Centro Euro-Mediterraneo sui Cambiamenti Climatici (CMCC) Research Paper, RP0228 (October 2014). Climate Impacts and Policy Division (CMCC): Lecce, Italy.Find this resource: Bosello, F., De Cian, E., Eboli, F., & Parrado, R. (2009). Macroeconomic assessment of climate change impacts: a regional and sectoral perspective. Impacts of Climate Change and Biodiversity Effects. Final report of the CLIBIO project, European Investment Bank, University Research Sponsorship Programme, Luxembourg.Find this resource: Bosello F., Eboli F., & Pierfederici, R. (2012). Assessing the economic impacts of climate change. An updated CGE point of view, FEEM, Working Paper No 2.12.Find this resource: Bosello, F., Roson, R., & Tol, R. S. J. (2006). Economy-wide estimates of the implications of climate change: Human Health. Ecological Economics, 58, 579–591.Find this resource: Bosetti, V., Carraro, C., & Galeotti, M. (2006). The dynamics of carbon and energy intensity in a model of endogenous technical change. Energy Journal [Special Issue: Endogenous Technological Changes and the Economics of Atmospheric Stabilization], 191–205.Find this resource: Bosetti, V., Carraro, C., Galeotti, M., Massetti, E., & Tavoni, M. (2006b). WITCH: A world induced technical change hybrid model. Energy Journal [Special Issue, Hybrid Modeling of Energy-Environment Policies: Reconciling Bottom-up and Top-down], 27, 13–38.Find this resource: Calzadilla, A., Rehdanz, K., Betts, R., Falloon, P., Wiltshire, A., & Tol, R. S. (2013). Climate change impacts on global agriculture. Climatic Change, 120(1–2), 357–374.Find this resource: Cambridge Centre for Risk Studies. (2015). World city risk 2025: Part I Overview and results. Cambridge, U.K.: Centre for Risk Studies, University of Cambridge.Find this resource: Ciscar, J. C., Feyen, L., Soria, A., Lavalle, C., Raes, F., Perry, M., et al. (2014). Climate impacts in Europe. The JRC PESETA II Project. JRC Scientific and Policy Reports, EUR 26586EN. Publications Office of the European Union, Luxembourg.Find this resource: Criqui, P. (2001). POLES: Prospective outlook on long-term energy systems. Institut d’Economie et de Politique de l’Energie. Grenoble, France.Find this resource: Criqui, P., Menanteau, P., & Mima, S. (2009). The trajectories of new energy technologies in carbon constraint cases with the POLES world energy model. IOP Conference Series: Earth and Environmental Science (Vol. 6, 21, p. 212004). IOP Publishing.Find this resource: DARA. (2012). Climate vulnerability monitor (2d ed.). A guide to the cold calculus of a hot planet. Fundacion DARA Internacional: Madrid.Find this resource: Darwin, R. F., & Tol, R. S. (2001). Estimates of the economic effects of sea level rise. Environmental and Resource Economics, 19(2), 113–129.Find this resource: DEFRA. (2007). The social cost of carbon and the shadow price of carbon: What they are, and how to use them in economic appraisal in the UK. Department for Environment, Food and Rural Affairs (DEFRA). London.Find this resource: Dellink, R., Lanzi E., Château J., Bosello F., Parrado R., de Bruin K., et al. (2014). Consequences of climate change damages for economic growth: A dynamic quantitative assessment. OECD Economics Department Working Papers, No. 1135, OECD Publishing, Paris.Find this resource: De Cian, E., Bosetti, V., & Tavoni, M. (2012). Technology innovation and diffusion in “less than ideal” climate policies: An assessment with the WITCH model. Climatic Change, 114, 121–143.Find this resource: Diaz, D. B. (2016). Estimating global damages from sea level rise with the Coastal Impact and Adaptation Model (CIAM). Climatic Change, 137, 143–156.Find this resource: Dowlatabadi, H. (1998). Sensitivity of climate change mitigation estimates to assumptions about technical change. Energy Economics, 20(5), 473–493.Find this resource: European Commission. (2011). Climate cost: The full costs of climate change. Summary of Results from the Climate Cost project, funded by the European Community’s Seventh Framework Programme. Published by the Stockholm Environment Institute, Sweden.Find this resource: Eyre, N., Downing, T., Hoekstra, R., Rennings, K., & Tol, R. S. J. (1999). Externalities of Energy, Vol 8: Global Warming. European Commission Community Research Studies. Office for Official Publications of the European Communities, Luxembourg.Find this resource: Fankhauser, S. (1994). Evaluating the social costs of greenhouse gas emissions. Centre for Social and Economic Research on the Global Environment, University College London and University of East Anglia, CSERGE Working Paper GEC 94-01. Norwich, UK.Find this resource: Galarraga I., & Markandya, A. (2009). Climate change and its socioeconomic importance. Bilbao, Spain: Basque Centre for Climate Change (BC3).Find this resource: Hope, C. (2010). The PAGE09 model: Estimating climate impacts and the social cost of CO2. Environmental Protection Agency, Washington, DC.Find this resource: Hope, C. (2011). The social cost of CO2 from the PAGE09 model. Economics Discussion Paper, 39. Kiel Institute for the World Economy, Germany. Available at http://www.economics-ejournal.org/economics/discussionpapers/2011-39. Iglesias, A., Garrote, L., Quiroga, S., & Moneo, M. (2009). Impacts of climate change in agriculture in Europe. JRC Scientific and Technical Reports. EUR 24107 EN. Joint Research Centre, Institute for Prospective Technological Studies. Luxembourg: Office for Official Publications of the European Communities.Find this resource: Iglesias, A., Quiroga, S., & Garrote, L. (2010). Report analysis for Europe. Deliverable D2B.2, Work package 2B Agriculture and water, ClimateCost project (FP7).Find this resource: IPCC. (2007). Summary for policymakers. In S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, et al. (Eds.), Climate change. The physical science basis. (pp. 1–18). Cambridge, U.K.: Cambridge University Press.Find this resource: IPCC. (2012). Managing the risks of extreme events and disasters to advance climate change adaptation. A special report of Working Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press.Find this resource: IPCC. (2014). Climate change 2014. Synthesis report. Summary for policy makers. Cambridge, U.K.: Cambridge University Press.Find this resource: Link, P. M., & Tol, R. S. J. (2004). Possible economic impacts of a shutdown of the thermohaline circulation: An application of FUND. Portuguese Economic Journal, 3(2), 99–114.Find this resource: Manne, A., Mendelsohn, R., & Richels, R. (1995). MERGE: A model for evaluating regional and global effects of GHG reduction policies. Energy Policy, 23(1), 17–34.Find this resource: Manne, A. S., & Richels, R. G. (2005). MERGE: An integrated assessment model for global climate change. In R. Loulou, J. Waaub, & G. Zaccour (Eds.), Energy and environment (pp. 175–189). Springer, US.Find this resource: Markandya, A. (2014). Incorporating climate change into adaptation programmes and project appraisal: Strategies for uncertainties. In A. Markandya, I. Galarraga, & E. Sainz de Murieta (Eds.), Routledge handbook of the economics of climate change adaptation (pp. 97–119). London: Routledge.Find this resource: Markandya, A., Saygin, D., Miketa, A., Gielen, D., & Wagner, N. (2016). The true cost of fossil fuels: Saving on the externalities of air pollution and climate change. Abu Dhabi: International Renewable Energy Agency.Find this resource: Mendelsohn, R., Dinar, A., & Williams, L. (2006). The distributional impact of climate change on rich and poor countries. Environment and Development Economics, 11(2), 159–178.Find this resource: Mendelsohn, R. O., Morrison, W. N., Schlesinger, M. E., & Andronova, N. G. (1998). Country-specific market impacts of climate change, Climatic Change, 45(3–4), 553–569.Find this resource: Moore, F., & Diaz, D. B. (2015). Temperature impacts on economic growth warrant stringent mitigation policy. Nature Climate Change, 5, 127–131.Find this resource: Moss, R., Babiker, M., Brinkman, S., Calvo, E., Carter, T., Edmonds, J., et al. (2008). Towards new scenarios for analysis of emissions, climate change, impacts, and response strategies. Geneva: Intergovernmental Panel on Climate Change.Find this resource: Narita, D., Anthoff, D., & Tol, R. S. J. (2009). Damage costs of climate change through intensification of tropical cyclone activities: An application of FUND. Climate Research, 39, 87–97.Find this resource: Narita, D., Anthoff, D., & Tol, R. S. J. (2010). Economic costs of extratropical storms under climate change: An application of FUND. Journal of Environmental Planning and Management, 53(3), 371–384.Find this resource: Nordhaus, W. D. (2007). Accompanying notes and documentation on development of DICE-2007 model: Notes on DICE-2007.delta.v8. Available at: http://aida.wss.yale.edu/~nordhaus/homepage/Accom_Notes_100507.pdf. Nordhaus, W. D. (2011a). Estimates of the social cost of carbon: Background and results from the Rice-2011 model. NBER Working Paper No. 17540. National Bureau of Economic Research, Cambridge, MA.Find this resource: Nordhaus, W. D. (2011b). Integrated economic and climate modeling. Cowles Foundation, Discussion Paper No. 1839. Yale University, New Haven, Connecticut.Find this resource: Nordhaus, W. D., & Boyer, J. G. (2000). Warming the world: The economics of the greenhouse effect. Cambridge, MA: MIT Press.Find this resource: Nordhaus, W. D., & Sztorc, P. (2013). DICE 2013R: Introduction and user’s manual (2d ed.). Available at http://www.econ.yale.edu/~nordhaus/homepage/documents/DICE_Manual_103113r2.pdf. OECD. (2015). The economic consequences of climate change. Paris: OECD. Available at http://dx.doi.org/10.1787/9789264235410-en.Find this resource: Ortiz, R. A., & Markandya, A. (2009). Integrated impact assessment models of climate change with an emphasis on damage functions: A literature review. BC3 Working Paper Series 2009–06. Basque Centre for Climate Change (BC3). Bilbao, Spain.Find this resource: Philibert, C. (2006). Discounting the future. In D. J. Pannel & S. G. M. Schilizzi (Eds.), Economics and the Future. (pp. 137–148). Edward Elgar Publishing.Find this resource: Stern, N. H. (2007). The economics of climate change: The Stern Review. Cambridge, U.K.: Cambridge University Press.Find this resource: Tol, R. S. J. (2002a). Estimates of the damage costs of climate change—Part 1: Benchmark estimates. Environmental and Resource Economics, 21(1), 47–73.Find this resource: Tol, R. S. J. (2002b). Estimates of the damage costs of climate change—Part II: Dynamic estimates. Environmental and Resource Economics, 21(2), 135–160.Find this resource: Tol, R. S. J. (2009). The economic effects of climate change. The Journal of Economic Perspectives, 23(2), 29–51.Find this resource: Tol, R. S. J. (2015). Economic impacts of climate change. Working Paper Series No. 75-2015. University of Sussex. Falmer, UK.Find this resource: U.S. Government: Interagency Working Group on Social Cost of Carbon. (2013). Technical support document: Technical update of the social cost of carbon for regulatory impact analysis under executive order 12866, United States Government. Washington, DC.Find this resource: Vafeidis, A. T., Nicholls, R. J., McFadden, L., Tol, R. S., Hinkel, J., Spencer, T., et al. (2008). A new global coastal database for impact and vulnerability analysis to sea-level rise. Journal of Coastal Research, 24(4), 917–924.Find this resource: Waldhoff, S. T., Anthoff, D., Rose, S., & Tol, R. S. (2014). The marginal damage costs of different greenhouse gases: An application of FUND. Economics: The Open-Access, Open-Assessment E-Journal, 8, 1–33.Find this resource: World Bank. (2010). The economics of adaptation to climate change. A synthesis report. The World Bank Group. Washington, DC.Find this resource: Zhai, F., & Zhuang, J. (2009). Agricultural impact of climate change: A general equilibrium analysis with special reference to Southeast Asia. ADBI Working Paper 131. Asian Development Bank Institute. Tokyo, Japan.Find this resource: # Appendix: Estimates of Climate Damages at the Sectoral Level Table A1. Damage Estimates—Agriculture Study Region Year Temperature increase/ppm Method Estimates Factors included Nordhaus and Boyer (2000) Global 2100 2.5°C DICE99/RICE99 • −0.13% of world GDP • −0.17% of world GDP Bosello et al. (2009) Global 2050 3.1°C ICES Model −0.6% (ca.) Land productivity Zhai and Zhuang, (2009) Global 2080 735 ppm by 2085 CGE/Linkage Model −1.4% globally, 6.2% in India, 2.2% in sub-Saharan Africa Productivity, trade, autonomous adaptation Bosello et al. (2012) Global 2050 1.92°C ICES Model −0.3% (ca.) of world GDP Crops productivity DARA (2012) Global 2010 Enumerative 50 billion USD loss (2010 USD) Change of agricultural output due to climate change Global 2030 IPCC SRES A1B Enumerative 350 billion USD loss Calzadilla et al. (2013) Global 2050 IPCC SRES A1B Scenario GTAP-W −0.29% of world GDP Water as a factor of production, rainfed, irrigated agriculture 2050 IPCC SRES A2 Scenario GTAP-W −0.28% of world GDP Ciscar et al. (2014) EU 2080s (2071–2100) IPCC SRES A1B GCM—RCM—GEM-E3—DIVA • −0.2% of EU GDP • −0.17% of global welfare Crop productivity change OECD (2015) Global 2060 3°C ENV-LINKAGES (IMPACT model) −0.8% of world GDP Changes in crop yields (including cropland productivity and water stress), changes in fisheries catches Note: A level of 785 ppm implies a temperature increase of 3.2 to 3.8°C. Table A2. Damage Estimates—Forestry Study Region Year Temperature increase/ppm Method Estimates Factors included Bosello et al. (2012) Global 2050 1.92°C ICES Model negligible Net primary productivity of forests DARA (2012) Global 2010 Enumerative$5 billion USD loss (2010 USD)

Change in forestry under climate change

Global

2030

Enumerative

$45 billion USD loss Table A3. Damage Estimates—Health Study Region Year Temperature increase/ppm Method Estimates Factors included Nordhaus and Boyer (2000) Global 2100 2.5 °C DICE99/RICE99 • −0.10% world GDP • −0.56% world GDP Climate-related diseases Bosello et al. (2006) Global 2050 1.03 °C GTAP-EF 11Bn. USD (1997 USD) Additional cost of illness Bosello et al. (2009) Global 2050 3.1 °C ICES Model Negligible Labour productivity; public and private expenditure EC (2011) EU 2050 Non-mitigation scenario 2°C Value of loss of life; DIVA for sea level rise, GTAP for economic effect −223Bn. USD Heat-related mortality; Foodborne disease; coastal flooding; labor productivity EU 2080 −333 Bn. USD Bosello et al. (2012) Europe 2050 1.92°C ICES Model Negligible Job performance (reduced work capacity due to thermal discomfort) DARA (2012) Global 2010 Enumerative$300 billion USD gain (2010 USD)

Labor productivity changes due to heat stress

Global

2030

IPCC SRES A1B

Enumerative

\$2.5 trillion USD gain

Ciscar et al. (2014)

EU

2080s (2071–2100)

IPCC SRES A1B

GCM—RCM—GEM-E3—DIVA

• −0.1% of EU GDP

• −1.23% of global welfare

Mortality, health expenditures

OECD (2015)

Global

2060

3°C

−0.9% of world GDP

Morbidity from heat and cold exposure (incl. heatwaves); mortality and morbidity from infectious diseases, cardiovascular

Note: See comments on SRES scenarios. EC (2011) values have been converted from euros to USD.

Table A4. Damage Estimates—Energy

Study

Region

Year

Temperature increase/ppm

Method

Estimates

Factors included

Bosello et al. (2009, 2012)

Global

2050

1.92°C—3.1°C

ICES Model

Negligible to +0.2% (ca.)

Households’ energy demand

EC (2011)

EU

2050

Non-mitigation scenario

POLES Model

• −43Bn. USD

• +49Bn. USD

• −2Bn. USD

• Cooling

• Heating

• Energy for water supply

EU

2100

• −157Bn. USD

• +174Bn. USD

• −7Bn. USD

• Cooling

• Heating

• Energy for water supply

DARA (2012)

Global

2010

Enumerative

+5 billion USD

Hydropower potential of climate change on river discharge

Global

2030

IPCC SRES A1B

Enumerative

+25 billion

Ciscar et al. (2014)

EU

2080s (2071–2100)

IPCC SRES A1B

GCM—RCM—GEM-E3—DIVA

Heating and cooling demand (residential and services)

OECD (2015)

Global

2060

3°C

Low impact

Changes in energy demand for cooling and heating

Note: See comments on SRES scenarios. EC (2011) values have been converted from euros to USD.

Table A5. Damage Estimates—Coastal Zones

Study

Region

Year

Temperature increase/ppm

Method

Estimates

Factors included

Nordhaus and Boyer (2000)

Global

2100

2.5°C

DICE99/RICE99

• −0.32% of world GDP

• −0.12% of world GDP

Darwin and Tol (2001)

Global

0.5 m SLR

FUND Model

−42.9 Bn. USD

Dryland lost due to SLR

Global

0.5 m SLR

FARM Model

−25.0 Bn. USD

Dryland lost due to SLR

Bosello et al. (2009, 2012)

Global

2050

1.92 to 3.1°C

ICES Model

Negligible to −0.1% of world GDP

Land loss due to SLR

EC (2011)

EU

2050

• Non-mitigation scenario

• 2°C

DIVA Model

• −15.8 Bn. USD

• −16.6 Bn. USD

Flooding and other impacts due to SLR (e.g., erosion)

EU

2080

• Non-mitigation scenario

• 2°C

DIVA Model

• −36.0 Bn. USD

• −23.1 Bn. USD

Flooding and other impacts due to SLR (e.g., erosion)

DARA (2012)

Global

2010

Enumerative (DIVA Model)

−85 Bn. USD (2010 USD)

Tidal basin, beach and wetland nourishment cost; land loss costs; migration costs; river dike costs; river floods costs; salinity intrusion costs, sea dike costs, sea flood costs

Global

2030

IPCC SRES A1B

Enumerative (DIVA Model)

−550 Bn. USD

Ciscar et al. (2014)

EU

2080s (2071–2100)

IPCC SRES A1B

GCM—RCM—GEM-E3—DIVA

• −0.7% of EU GDP (24.5 Bn. USD)

• −0.4% of global welfare

Migration coast, sea floods coast

OECD (2015)

Global

2060

3°C

−0.2% of world GDP

Loss of land and capital from SLR

Diaz (2016)

Global

2050

• RCP8.56

• IPCC Scenario

CIAM Model (DIVA database)

−0.09% of world GDP

Coast type, flood, wetland, inundation, retreat, protection

Note: Wee comments on SRES scenarios. EC (2011) values have been converted from euros to USD.

Table A6. Damage Estimates—Extreme Events

Study

Region

Year

Temperature increase/ppm

Method

Estimates

Factors included

Nordhaus and Boyer (2000)

Global

2100

2.5°C

• DICE99

• RICE99

• −1.02% of world GDP

• −1.05% of world GDP

Catastrophic impacts

Global

2100

6°C

• DICE99

• RICE99

• −6.94% of world GDP

• −7.12% of world GDP

Narita et al. (2009)

Global

2100

2.5°C

FUND Model

−0.0074% of world GDP7

Tropical cyclones

Narita et al. (2009)

Global

2100

2.5°C

FUND Model

−0.001% of world GDP8

Extra-tropical storms

EC (2011)

EU

2050

• Non-mitigation scenario

• 2 °C

LISFLOOD Model

• −66.3 Bn. USD

• −60.1 Bn. USD

River floods (socioeconomic changes included)

EU

2080

• Non-mitigation scenario

• 2 °C

LISFLOOD Model

• −141 Bn. USD

• −98 Bn. USD)

Bosello et al. (2012)

EU27

2050

1.92 °C

ICES Model

Negligible

River floods due to sea level rise

DARA (2012)

Global

2010

Enumerative

−45 Bn. USD (2010 USD)

Drought, floods & landslides, storms, wildfires

Global

2030

IPCC SRES A1B

Enumerative

−307 Bn. USD loss

Ciscar et al. (2014)

EU

2080s (2071-2100)

IPCC SRES A1B

GCM—RCM—GEM-E3—DIVA

• −0% of EU GDP

• (−15.9 Bn. USD)

Residential buildings damages; production activities losses

Note: See comments on SRES scenarios. EC (2011) values have been converted from euros to USD.

Table A7. Damage Estimates—Tourism

Study

Region

Year

Temperature increase/ppm

Method

Estimates

Factors included

Bigano et al. (2007)

Global

2100

IPCC SRES A1B

HTM Model

0% of total tourism expenditures

Length of stay; average expenditure per day

Bosello et al. (2009)

Global

2050

3.1°C

ICES Model

−0.5% (ca.)

Market services demand; expenditure flows

Bosello et al. (2012)

Global

2050

1.92°C

ICES Model

−0.1% (ca.) of world GDP

Tourist flows

Ciscar et al. (2014)

EU

2080s (2071–2100)

IPCC SRES A1B

GCM—RCM—GEM-E3—DIVA

−21.6 Bn. USD

Tourism expenditure

EU

2080s (2071–2100)

2°C

GCM—RCM—GEM-E3—DIVA

−21.6 Bn. USD

Tourism expenditure

OECD (2015)

Global

2060

3 °C

−0.1% of world GDP

Changes in tourism flows and services

Note: See comments on SRES scenarios. EC (2011) values have been converted from euros to USD.

Table A8. Damage Estimates—Ecosystems

 Study Region Year Temperature increase/ppm Method Estimates Factors included DARA (2012) Global 2010 Enumerative −115 Bn. USD (2010 USD) Biodiversity, desertification, permafrost Global 2030 IPCC SRES A1B Enumerative −570 Bn. USD (2010 USD)

Table A9. Damage Estimates—Industry and Infrastructure

Study

Region

Year

Temperature increase/ppm

Method

Estimates

Factors included

DARA (2012)

Global

2010

Enumerative

• −15 Bn. USD (2010 USD)

• −1 Bn. USD (2010 USD)

Fisheries Transport

Global

2030

IPCC SRES A1B

Enumerative

• −150 Bn. USD

• −5 Bn. USD

Fisheries Transport

Ciscar et al. (2014)

EU

2080s (2071–2100)

IPCC SRES A1B

GCM—RCM—GEM-E3—DIVA

−1.3 Bn. USD

Transport infrastructure

EU

2080s (2071–2100)

2 °C

GCM—RCM—GEM-E3—DIVA

−1.1 Bn. USD

Transport infrastructure

World Bank (2010)

Global

2010–2050

Two scenarios considered—wet and dry

Enumerative

−13 to 27 Bn. USD

Transport and buildings

## Notes:

(1.) See Ortiz and Markandya (2009) for a detailed literature review of previous versions of IAMs for climate change analysis with damage functions mentioned here.

(2.) Equilibrium climate sensitivity (ECS, sometimes referred to simply as climate sensitivity) is a key parameter of climate models. In one version it gives the equilibrium increase in temperature resulting from a doubling of GHG concentrations in the atmosphere. The IPCC Fifth Assessment Report(AR5) states: “there is high confidence that ECS is extremely unlikely less than 1°C and medium confidence that the ECS is likely between 1.5°C and 4.5°C and very unlikely greater than 6°C.”

(3.) Equity weightings refer to the practice of modifying damages resulting from climate change to reflect the relative income status of the affected individuals. Thus damages in poor countries can be raised to account for the fact that money value based on market prices will be lower than social values. This is particularly important in valuing loss of life, which has been a controversial area in damage estimation when lower values were applied to losses in poor countries. Equity weighting aims to correct for this kind of undervaluation.

(4.) Not all impacts can be modeled in the CGE framework. In particular, the quantitative analysis does not deal with livestock mortality and morbidity from heat and cold exposure, changes in aquaculture productivity, non-market impacts in coastal zones, changes in forest plantation yields, changes in energy supply, changes in availability of drinking water to end users including households, civil conflict, or human migration.

(9.) Link and Tol (2004), Narita et al. (2009, 2010), and Tol (2002a, 2002b).

(5.) The point is particularly important, as stressed by IPCC (2012, p. 9), according to which in most cases “loss estimates are lower bound estimates because many impacts, such as loss of human lives, cultural heritage, and ecosystem services, are difficult to value and monetize, and thus they are poorly reflected in estimates of losses.”

(10.) DARA (2012) also provides estimates at country level for 184 countries. SRES A1B is a medium emissions scenario, resulting in a temperature increase of around 2.8°C by 2100.

(11.) Nordhaus and Boyer (2000) also provide estimates for 13 regions.

(12.) Global circulation model and regional climate model.

(6.) RCP8.5 is one of the four Representative Concentration Pathways (RCPs) adopted by the IPCC for its fifth Assessment Report (AR5) in 2014 (Moss et al., 2008), describing greenhouse gas concentration (not emissions) trajectories. In correspondence to RCP8.5, global mean temperature is projected to rise by about 3.7°C (in a range from 2.6 to 4.8°C) by the end of the century.

(7.) Changes in (1) income elasticities of storm damage and mortality, (2) the sensitivity factor of ocean warming on wind speed increase, and (3) the exponent for wind speed causing damage lead to a loss between 0.0054% and 0.0262% of global GDP.

(8.) Changes in (1) income elasticities of storm damage and mortality, (2) the sensitivity factor of ocean warming on wind speed increase, and (3) the exponent for wind speed causing damage lead to a change between 0.0006% and −0.0034% of global GDP.