/var/lib/tomcat6/webapps/pdnadmin/data/documents/2936

Assessing the costs of adaptation to climate change

First published by the International Institute for Environment and Development (UK)

and the Grantham Institute for Climate Change, Imperial College London (UK) in 2009
Copyright © International Institute for Environment and Development

ISBN:

978-1-84369-745-9

For a full list of publications or catalogue please contact:
International Institute for Environment and Development (IIED)

3 Endsleigh Street

London WC1H 0DD

United Kingdom

newpubs@iied.org

www.iied.org/pubs
A catalogue record for this book is available from the British Library.
Citation:

Martin Parry, Nigel Arnell, Pam Berry, David Dodman, Samuel Fankhauser,

Chris Hope, Sari Kovats, Robert Nicholls, David Satterthwaite, Richard Tiffin, Tim

Wheeler (2009) Assessing the Costs of Adaptation to Climate Change: A Review of the

UNFCCC and Other Recent Estimates, International Institute for Environment and

Development and Grantham Institute for Climate Change, London.
The views expressed in this report are those of the authors and do not necessarily reflect

the views of the International Institute for Environment and Development and the

Grantham Institute for Climate Change.
Design by:

McGillan Eves

Printed by:

Emtone, UK on Challenger offset

Please recycle

Assessing the costs of adaptation to climate change

Contributors

Nigel Arnell

, Walker Institute, University of Reading, UK

n.w.arnell@reading.ac.uk
Pam Berry,

Environmental Change Institute, University of Oxford, UK

pam.berry@eci.ox.ac.uk
David Dodman

, International Institute for Environment and Development, UK

david.dodman@iied.org
Samuel Fankhauser

, Grantham Research Institute and Centre for Climate Change

Economics and Policy, London School of Economics and Political Science, UK

s.fankhauser@lse.ac.uk
Clair Hanson

, School of Development Studies, University of East Anglia, UK

clair.hanson@uea.ac.uk
Chris Hope

, Judge Business School, University of Cambridge, UK

chris.hope@jbs.cam.ac.uk
Sari Kovats

, London School of Hygiene & Tropical Medicine, UK

sari.kovats@lshtm.ac.uk
Jason Lowe

, Meteorological Office, UK

jason.lowe@metoffice.gov.uk
Robert Nicholls

, School of Civil Engineering and the Environment

and the Tyndall Centre for Climate Change Research, University of Southampton, UK

R.J.Nicholls@soton.ac.uk
Martin Parry

, Grantham Institute for Climate Change and Centre for Environmental Policy,

Imperial College London, UK

martin@mlparry.com
David Satterthwaite

, International Institute for Environment and Development, UK

david.satterthwaite@iied.org
Richard Tiffin

, Walker Institute and Department of Agricultural and Food Economics,

School of Agriculture, Policy and Development, University of Reading, UK

j.r.tiffin@reading.ac.uk
Tim Wheeler

, Walker Institute and Department of Agricultural and Food Economics,

School of Agriculture, Policy and Development, University of Reading, UK

t.r.wheeler@reading.ac.uk

Summary

Several recent studies have reported adaptation costs for climate change, including for developing

countries. They have similar-sized estimates and have been influential in discussions on this issue.

However, the studies have a number of deficiencies which need to be transparent and addressed

more systematically in the future. A re-assessment of the UNFCCC estimates for 2030 suggests

that they are likely to be substantial under-estimates. The purpose of this report is to illustrate the

uncertainties in these estimates rather than to develop new cost estimates, which is a much larger

task than can be accomplished here. The main reasons for under-estimation are that: (i) some

sectors have not been included in an assessment of cost (e.g. ecosystems, energy, manufacturing,

retailing, and tourism); (ii) some of those sectors which have been included have been only

partially covered; and (iii) the additional costs of adaptation have sometimes been calculated as

‘climate mark-ups’ against low levels of assumed investment. In some parts of the world low levels

of investment have led to a current adaptation deficit, and this deficit will need to be made good

by full funding of development, without which the funding for adaptation will be insufficient.

Residual damages also need to be evaluated and reported because not all damages can be avoided

due to technical and economic constraints. There is an urgent need for more detailed assessments

of these costs, including case studies of costs of adaptation in specific places and sectors.

1 Introduction

This is an evaluation of estimates of the costs of adaptation made by the United Nations

Framework Convention on Climate Change (UNFCCC) in 2007 and by some preceding studies

(UNFCCC, 2007; Stern, 2006; World Bank, 2006; Oxfam, 2007; UNDP, 2007). The costs have

been used as the basis for discussion regarding the levels of investment needed for adaptation to

climate change. They have been influential in the debate concerning funding for climate change

and it is important, therefore, that such estimates of cost are as robust as possible. The purpose of

this report is to assess these estimates and consider ways to improve them in the future.
The UNFCCC report was based on a set of commissioned studies (UNFCCC background

papers, 2007). These took place over a short period dictated by the timescale of the UNFCCC

process and the need to report the results to the next Conference of the Parties, so there was no

time for independent review of a draft of the report.
It is important, therefore, to recall the objectives of the UNFCCC report and the caveats that the

authors ascribed to its conclusions. The study was a preliminary one of the funding, especially the

public funding, estimated to be needed in the year 2030 to meet the challenge of climate change.

It is not a study of the full cost of avoiding all damage. It does not cover some important activities,

and other activities are only partially covered. The authors suggest that their estimates are

probably under-estimates and that much more study is needed.
The purpose of this evaluation is to consider the relative strengths and weaknesses of the

UNFCCC study, so that we can determine what next steps can be taken to improve our

understanding of the issue. It is not our purpose here to develop a revised set of numbers for

the funding of adaptation to climate change, because we believe this requires detailed study.

Overview of conclusions

Assessing the costs of adaptation to climate change

But we do conclude that the UNFCCC study has probably under-estimated this cost, and we take

this under-estimate to be substantial. The next step needs a substantial study combining ‘bottom-

up’ local case studies with ‘top-down’ integration, which is more than is possible in this report.

We conclude with some recommendations for that study.

2 Studies before the UNFCCC study

The UNFCCC estimates of adaptation cost are broadly in line with preceding studies published

by the World Bank, Oxfam, UNDP, and in the Stern report. These have recently been

summarised by the OECD and are given in Table 1. Since these studies appear to support each

other, the conclusion has sometimes been made that there exists a comforting convergence of

evidence, but that would be misleading because: (i) none of these are substantive studies, (ii) they

are not independent studies but borrow heavily from each other, and (iii) they have not been

tested by peer review in the scientific or economics literature.

Table 1 Estimates of adaptation costs in developing countries, for 2010-2015

Source

US$ billion p.a.

Comments

World Bank (2006)

9-41

Cost of climate-proofing FDI, GDI and
ODA flows

Stern (2006)

4-37

Update, with slight modification of
World Bank (2006)

Oxfam (2007)

>50

Based on World Bank, plus extrapolation
of costs from NAPAs and NGO projects

UNDP (2007)

86-109

World Bank, plus costing of PRS targets,
better disaster response

Source: Agrawala and Fankhauser (2008)
Note: FDI = foreign direct investment, GDI = gross domestic investment, ODA = official development assistance,

NAPA = National Adaptation Programme of Action, PRS = poverty reduction strategy

Importantly, most of the precursors of the UNFCCC study were based on the same method, first

developed by the World Bank (2006). This takes a fraction of current investment that is climate-

sensitive and applies a ‘mark-up’ factor to this fraction to reflect the cost of ‘climate-proofing’

those investments. We shall consider below the weakness of this approach.

3 The UNFCCC study

The UNFCCC commissioned six studies which provided estimates of the cost of adaptation for

the year 2030, usually assuming a climate scenario similar to the IPCC’s SRES A1B and B1.

In summary these cover:
• Agriculture, forestry and fisheries.

The agriculture estimate (McCarl, 2007) consists of three

distinct cost items: extra capital investment at farm level, the need for better extension services at

country level and the cost of additional global research (e.g. on new cultivars).

Overview of conclusions

Assessing the costs of adaptation to climate change

• Water supply.

The water estimate (Kirshen, 2007) considers the effect of additional water

demand and changes on the supply side. Investment decisions are made in anticipation of 2050

water needs.

• Human health.

The health estimates (Ebi, 2007) are the extra prevention costs for three health

issues: malnutrition, malaria and diarrhoea. The health impacts are based on the Global Burden

of Disease study (McMichael et al., 2004).

• Coastal zones.

Coastal protection costs are based on the DIVA model (Nicholls, 2007), which

considers a limited set of adaptation options that are applied globally. Uniquely, the coastal

estimate considers both adaptation costs and residual damages. For long-life defence

infrastructure, investments are made in anticipation of sea-level rise in 2080.

• Infrastructure.

The infrastructure estimate adopts the World Bank (2006) methodology, using

insurance data to determine the share of climate-sensitive investment, and applying a percentage

increase on current infrastructural investment to suggest additional costs for climate-proofing

new infrastructure. (However, the background paper by Satterthwaite (2007) took a different

approach.)

• Ecosystems.

An indication of adaptation costs for ecosystems was derived from the costs of

increasing protected areas to at least 10% of the land area of each nation or ecosystem, although

it was not possible to split this into baseline costs of meeting current deficits and incremental

adaptation. See Berry (2007).

The UNFCCC report concluded that total funding need for adaptation by 2030 could amount to

$49 – 171 billion per annum globally, of which $27 – 66 billion would accrue in developing

countries (Table 2) (note that all references to dollars in this report are to US dollars unless

otherwise specified). By far the largest cost item is infrastructure investment, which for the upper-

bound estimate accounts for three-quarters of total costs. Costs are over and above what would

have to be invested in the baseline to renew the capital stock and accommodate income and

population growth. Note that the total excludes the estimate for ecosystem adaptation, for reasons

discussed below in this report.

Table 2 UNFCCC estimate of additional annual investment need and financial flow needed
by 2030 to cover costs of adaptation to climate change (billion dollars per year in present-
day values)

Sector

Global cost

Developed countries

Developing countries

Agriculture

Water

Human health

Not estimated

Coastal zones

Infrastructure

8 – 130

6 – 88

2 – 41

Total

49 – 171

22 – 105

27 – 66

Source: UNFCCC (2007)

Overview of conclusions

Assessing the costs of adaptation to climate change

4 Issues affecting the robustness

of the estimates

In reviewing the UNFCCC report (and others), the following issues emerge which require more

thorough treatment in future assessments.

The potential damages to be avoided by adaptation

The Fourth Assessment Report of the IPCC (IPCC, 2007) gives a summary of some impacts

likely to occur under varying amounts of global warming. Mapping onto this is the expected

warming range for 2030 which indicates the potential impacts that adaptation will need to

address. Figure 1 shows this for the A1B scenario assumed generally in the UNFCCC study, as

well as for 2050 and 2080 (used respectively in the water and coastal analyses of that study, on the

assumption that adaptation in 2030 will need to anticipate future warming due to the long-term

nature of investment needs in those sectors). One aspect emerging here is the substantial

magnitude of impacts that could occur even within the next few decades, and the scale of

damages that could be expected if adaptation is not fully successful in avoiding them.

Figure 1 Projected damages without adaptation

Source: This background table of impacts is from the Technical Summary of IPCC Working Group II (IPCC, 2007)
Note: The shaded columns show the 10 and 90 percentile uncertainty range for the scenarios assumed in the UNFCCC

estimation of funding needs for adaptation. For most sectors the A1B scenario was taken. For water and coastal protection
the scenarios were 2050 and 2080 respectively, due to the need for adaption to anticipate future climate change. Scenario
data from a simple Earth system model (Parry, Lowe and Hanson, 2009)

Overview of conclusions

Assessing the costs of adaptation to climate change

The scarcity of information on adaptation and its cost

Information is scarce about the scale of future potential impacts, and is even more scant for the

costs of avoiding them by adaptation, a point stressed in the UNFCCC report. Some sectors

such as mining and manufacturing, energy, retailing, and tourism, were not included in the

UNFCCC report. Cost estimations for ecosystems, although made, were left out of the final

table showing total costs (see Table 2 above), due to lack of sufficiently robust information.

Within some sectors that were examined, the funding needs estimated were clearly only partial.

In health for example, just three areas of impact, where there were sufficient estimates, were

considered: the effect of climate change on diarrhoeal diseases, malaria and malnutrition in low-

and middle-income countries. Adaptation costs for health effects in high-income countries were

not estimated.
A major problem is the absence of case studies to test the top-down form of UNFCCC analysis.

The few national figures available tend to suggest costs in excess of the UNFCCC estimates.

For example, agencies responsible for flood management in England and Wales have estimated

a need to spend (due to climate change) an additional $30 million annually in 2011, growing to

$720 million by 2035 (Environment Agency, 2009).

Applying a ‘climate mark-up’ against future investment trends

In most cases, the UNFCCC estimation of funding needs was derived from applying an increase

in cost to those areas of investment that are deemed to be climate-sensitive. In agriculture for

example, 2% of investment on infrastructure is taken to be climate-sensitive. In some sectors,

particularly the built environment, the investment flows are so large that even small changes in

this mark-up can change estimates significantly.

Investment needed to remove the ‘adaptation deficit’

In particular, applying a ‘climate mark-up’ is not appropriate when current investment flows are

well below what they should be. In several parts of the world, current levels of investment are

considered far from adequate, and lead to high current vulnerability to climate, including its

variability and extremes, the latter case being termed a current ‘adaptation deficit’ (Burton,

2004). This partly explains why impacts from climate change are expected to be greatest in low-

and middle- income countries (IPCC, 2007). To avoid these impacts the adaptation deficit

(which is largely a development deficit) will need to be made good. For good reason these costs

were not included in the UNFCCC estimate, which was aimed at identifying the additional cost

of climate change, but it needs to be stressed that without the adaptation deficit being made good,

the enhanced investment for adaptation will not be sufficient to avoid serious damage from

climate impacts. Dlugolecki (2007), in a background paper for the UNFCCC study, estimated (at

$200 billion per year) the costs of damage from present-day extreme weather and took this as a

reflection of the current scale of inadequate adaptation. The Millennium Development Goals

represent an attempt to make good some, but probably not all, of the adaptation deficit, and have

been costed at about $200 billion by 2015 (Sachs and McArthur, 2005). To make good the full

development deficit probably requires enhancing official development assistance to 0.7% of GDP

of OECD countries. Hence the issues of development and adaptation costs are intimately linked,

and this requires further exploration.

Overview of conclusions

Assessing the costs of adaptation to climate change

Adaptation costs in a world without an ‘adaptation deficit’

There follows from this the question as to whether the ‘climate mark-up’ should be against

investment levels that reflect current trends (which in many regions, of Africa for example, are

insufficient to remove high levels of vulnerability to climate), or to elevated levels that are needed

to attain the MDGs, or to even higher levels that would help achieve sustainable and equitable

development. The UNFCCC takes the former line, but this leads to estimations of cost that are

substantially lower than if one assumed a development pathway which protects the poor against

vulnerability to climate change. In this report we conclude that removing the housing and

infrastructure deficit in low- and middle-income countries will cost around $315 billion per year

(in today’s figures) over 20 years; while adapting this upgraded infrastructure specifically to meet

the challenge of climate change will cost an additional $16–63 billion per year.

How much impact is being avoided by adaptation?

It is not clear what proportion of expected damage would be avoided by the proposed UNFCCC

investment levels. Most impacts are projected to increase non-linearly with climate change, and

adaptation costs similarly with impacts (IPCC, 2007). Therefore it will probably be very

inexpensive to avoid some impacts but prohibitively expensive to avoid others; and some impacts

we cannot avoid even if funds were unlimited, because the technologies are not available (e.g. in

connection with ocean acidification). A simple schema of a generalised adaptation cost curve is

shown in Figure 2. The curve is likely to vary greatly between different sectors and places, but

probably common to most cases will be that adaptation to (say) the first 10% of damage will be

disproportionately cheaper than for 90% of damage. We need to be clear, then, about how much

we are willing to pay for adaptation to avoid damages. To illustrate, we might aim (in a scale of

reducing cost) to adapt to: (i) all those impacts that reduce human welfare, or (ii) all those that are

economically feasible (i.e. cheaper to adapt to than to be borne), or (iii) all those that are affordable

within a given budget constraint (for example, the size of the global Adaptation Fund).

Figure 2 Schematic of adaptation costs, avoided damages and residual damage compared:
a) at a point in time, and b) over time

Figure 2a

Adaptation costs

Residual damage

Avoided

damages

High cost to avoid damage

Ratio = 1 (incremental adaptation cost: avoided damage)

Low cost to avoid damage

Overview of conclusions

Assessing the costs of adaptation to climate change

Figure 2b

Avoided

damage

Residual damage
(not adapted to)

Gross benefit
of adaptation

Trend in damage
due to asset growth

2000

2030

2050

The costs of damage not adapted to, or ‘residual damage’

Implicit in the above (and illustrated in Figure 2) is that much damage will not be adapted to over

the longer term, because adaptation is either not economic or not feasible. We term this ‘residual

damage’. In the UNFCCC report it is not clear how much residual damage might be expected.

But it is very important that we start to consider this, because the amount may be significant and

is likely to increase over time. In the evaluation reported here, residual impacts are estimated at

about a fifth of all impacts in agriculture in 2030 (see Chapter 2 in this report) and, over the

longer term, may account for up to two-thirds of all potential impacts across all sectors, depending

on the amount of climate change not avoided by mitigation (see Chapter 8 in this report).

‘Soft adaptation’

The UNFCC study may have given insufficient weight to the value of ‘soft adaptation’. It is easier

analytically to cost out structural measures like the expansion of water supply systems, and the

UNFCCC study focused on these. In reality, it will often be cheaper to apply ‘soft adaptation’

options. Measures to use water more efficiently, for example, may obviate the need for expensive

new infrastructure. Conversely, the human health costs do not include changes in infrastructure

(‘hard adaptation’) which may be considerable (see Chapter 4 in this report).

How will adaptation costs change over time?

The UNFCCC estimate of adaptation costs is a ‘snapshot’ for 2030 at one point along the climate-

impact curve, and its authors note the importance of the question, ‘While the adaptation cost

curve seems quite gentle between now and 2030, how steeply will it grow thereafter?’ Some

believe it may rise steeply, possibly quadratically in some sectors (IPCC, 2007). It is very

important that this be analysed so we are sufficiently prepared for escalating adaptation costs

beyond 2030.

Overview of conclusions

Assessing the costs of adaptation to climate change

The level of under-estimation in the UNFCCC study

For a number of reasons discussed above, and given in more detail in the following section, we

believe that the UNFCCC estimate of investment needs is probably an under-estimate by a factor

of between 2 and 3 for the included sectors. It could be much more if other sectors are considered.

We conclude that for coastal protection the factor of under-estimation could be 2 to 3. For

infrastructure it may be several times higher, at the lower end of the cost range. For health the

‘intervention sets’ that were costed relate to a disease burden that is approximately 30–50% of the

anticipated total burden in low- and middle-income countries (and do not include interventions

in high-income countries). Including ecosystems protection could add a further $65–$300 billion

per year in costs. Furthermore, estimates are not made for sectors such as mining and

manufacturing, energy, the retail and financial sectors and tourism. This probably explains why

the investment levels proposed by the UNFCCC appear so small – roughly the annual cost of

running two or three Olympic Games. That this represents a doubling of current ODA only

highlights the very low current level of development assistance.

Recommendations for future studies

It is important that robust studies of adaptation cost are, in future, based upon case studies that

cover a wide range of places and sectors, and support top-down analyses of the kind evaluated

here. The World Bank and McKinsey will be shortly reporting on this (World Bank, 2009;

McKinsey, 2009). The time period and expected climate changes need specifying (as they were in

the UNFCCC study), and results for multiple timeframes would be useful. Non-climate trends

need careful portrayal, especially the future levels of non-climate investment. Costs of adapting to

varying amounts of impact should be analysed, thus providing a choice range for preparedness to

pay; and there needs to be some analysis of the residual impact that adaptation is not likely to

avoid, and the resulting damage costs that we need to anticipate.

5 Summary of conclusions

for each sector

This section presents summary conclusions from Chapters 2–7 below, considering costs of

adapting to climate changes in the six areas of agriculture, water, human health, coasts,

infrastructure and natural ecosystems.

Agriculture

The UNFCCC estimated the cost of adaptation of agriculture to climate change at $11.3–

12.6 billion for 2030. The basis for this is an assumption that the ‘climate mark-up’ will amount

to 10% of research and extension spend, and 2% of infrastructure spend. In the background

paper the authors state that these assumptions are uncertain and speculative given the limited

basis from which they were formed, but the UNFCCC report does not repeat these caveats nor

make clear the reasoning behind the mark-up levels that were adopted.

Overview of conclusions

Assessing the costs of adaptation to climate change

One outcome of the small ‘climate mark-up’ is the UNFCC conclusion that the cost of adapting

to climate change will be one-fortieth the cost of adapting to population change (i.e. meeting

increased demand). This contrasts with estimations for the water sector where the ratio is given as

1:3. These are important differences which deserve analysis.
The current adaptation deficit in agriculture is high. The number of people at risk from hunger

increased from around 300 million in 1990 to 700 million in 2007, and may exceed 1 billion in

2010 (FAO, 2009). A measure of the cost of making good this deficit is the cost of achieving the

relevant Millennium Development Goal, estimated at $40–60 billion per year. Without this non-

climate investment, the estimated levels of investment for adaptation to climate change will be

insufficient to avoid serious damage.
There are a few bottom-up case studies which indicate the magnitude of adaptation costs, and

these suggest that UNFCCC may be on the low side of adaptation costs in this sector. For

example, there is an estimate of $8 billion for adapting crop irrigation systems to climate change

by 2030, and of $14.5 billion for the year 2030 for the reduction in the value of global crop

outputs due to climate change.
For these reasons the present study concludes that the UNFCCC estimate of $11.3–12.6 billion is

a reasonable first approximation of adaptation costs in this sector, but we expect the estimate of

adaptation costs for agriculture, forestry and fisheries to increase as more detailed studies of

specific adaptation actions become available. Finally, the present study concludes that, with such

levels of adaptation, about 80% of the cost of potential impacts might be avoided, but about

20% might not.

Water

The UNFCCC estimated national water resource availability in relation to large-area projections

of national rainfall, and then compared availability with expected demand. It assumed that a

quarter of the total cost of adaptation due to changes in demand and supply would be due to

changes in supply resulting from climate change (i.e. $11 billion per year). This 1:3 ratio contrasts

with 1:40 given by the UNFCCC for agriculture.
The study worked at the national level only, assuming that water resources could be transferred

within a country from areas of surplus to areas of deficit. For small countries this would not be a

problem, but for large ones it is unrealistic and probably a source of under-estimation of true cost.

To illustrate, for a single basin in China (Huang Ho) the annual costs of adapting to climate

change could be $0.5 billion per year (see Chapter 3 in this report). Unfortunately, few such

studies are currently available and are insufficient for drawing reliable conclusions.
The UNFCCC costs include that of water provision, but not of adapting to altered flood risk in

river basins. These altered flood management costs may be very substantial (potentially $0.1–

0.2 billion annually in the Sacramento Basin in California alone), but there have been no

consistent inter-comparisons of costs in different parts of the water sector. The use of an average

climate change scenario rather than an ensemble which describes the range of possible impacts

has probably led the UNFCCC to under-estimate the costs of providing for the full range water-

storage need. In all, these costs omitted from the UNFCCC could be very substantial.

Overview of conclusions

Assessing the costs of adaptation to climate change

Human health

The UNFCCC estimates of costs of adaptation are the costs of the intervention set to prevent the

additional burden of disease due to climate change for three health outcomes in low- and middle-

income countries: diarrhoeal diseases, malaria and malnutrition. The estimates are in the range of

$4–12 billion per year in 2030.
These three outcomes are not the total projected burden on human health from climate change.

That total has yet to be assessed accurately, but authors of the WHO study of the global disease

burden estimate that these outcomes amount to 30–50% of the probable future total burden in

2030 in low- and middle-income countries (McMichael and Bertollini, 2009, personal

communication).
A potential source of under-estimation is that the UNFCCC considers a narrow range of

development futures. It takes a single median population projection in which population numbers

increase and cases of diarrhoea/malaria/malnutrition remain constant, i.e. there is steep relative

decline in incidence. The present study considers this to be an optimistic assumption.

Coasts

The costs reported by the UNFCCC study for coastal adaptation appear reasonable as a snapshot

cost for IPCC sea-level rise projections in 2030, and are more reliable than those for other sectors

since they are based on a model assessment rather than top-down assumptions. However, some

post-2007 IPCC assessments suggest significantly higher observed and projected rises in sea levels

than reported in the 2007 IPCC assessment. If these are assumed, then adaptation costs would be

roughly doubled.
If account is taken of the need to protect coastal landscapes for amenity or ecological reasons,

then the adaptation approach might change and protection costs could increase significantly in

most cases. Another deficiency of the UNFCCC study is the lack of consideration of other aspects

of climate change such as more intense storms. No detailed estimates are available, but in the

worst-case the necessary adaptation costs (and residual damage costs) could match those of

adapting to sea-level rise.
When combined with the uncertainties about sea-level rise, adaptation costs three times those

reported in the UNFCCC study are not implausible. Experience with the DIVA model since

Nicholls (2007) suggests that UNFCCC estimations for residual damages are overly conservative

and should be roughly doubled (to $2–$3billion per year).

Infrastructure

To estimate adaptation costs for infrastructure, the UNFCCC: (i) estimated global investment in

gross fixed capital formation in 2030 (around three times the global investment in 2000), as $22.2

trillion; (ii) multiplied this by the proportion that is vulnerable to the impact of climate change (based

on data for losses from weather disasters – 0.7% (Munich Re data) or 2.7% (ABI data)), resulting in

$153–650 billion a year; then (iii) assumed 5–20% of this total as the increase in capital costs

needed for adaptation, giving $8–31 billion (Munich Re) or $33–130 billion (ABI data).

Overview of conclusions

Assessing the costs of adaptation to climate change

The estimates based on the Munich Re data are likely to be substantial under-estimates of

damages from climate because only data from large events are included and (as the UNFCCC

study notes), this leaves out the cost of a high proportion of all extreme-weather disasters. The

climate-cost fraction of 0.7–2.7% was recognised as being low by the authors of the background

papers to the UNFCCC report, and in earlier work they adopted climate-cost fractions of 2–10%

for domestic investment and up to 40% for overseas development assistance.
Applying a climate mark-up to levels of infrastructure provision that are currently very low (e.g. in

most countries in Africa, many in Asia, and considerable parts of Latin America and the

Caribbean) yields low estimations of future cost. Infrastructure provision needs substantial

improvement to meet present-day needs and these are partly embraced in the Millennium

Development Goals. The present study concludes that removing the housing and infrastructure

deficit in low- and middle-income countries will cost around $315 billion per year (in today’s

figures) over 20 years. Adapting this upgraded infrastructure specifically to meet the challenge of

climate change will cost an additional $16–63 billion per year.
Investment in adaptation will not avoid all damages to infrastructure. The annual economic

damage caused by large extreme-weather disasters, 1996–2005, was over $50 billion a year. This

gives an indication of weather impacts that are currently not avoided by adaptation – even in

countries where the population is served by protective infrastructure and good-quality buildings.

Ecosystems

The UNFCCC methodology consisted of estimating: (i) the current global expenditure on

conservation in the form of protected areas (PAs), (ii) the shortfall in the PA network (PAN), (iii)

the level of additional expenditure needed for PAs to be adequate for climate-change adaptation,

and (iv) costing adaptation outside the protected area network. An estimate of $12–22 billion was

given by UNFCCC for the cost of expanding and protecting terrestrial PAN areas so that they

represent 10% of each country, but this was excluded from the ultimate list of needed investment.
The present study supports the conclusion of the background paper for the UNFCCC, which is

that $65–80 billion reflects the range of probable adaptation costs for PAN areas, including both

terrestrial and marine environments. Additionally, this study supports the conclusion in the same

background paper that adaptation costs for non-protected (non-PAN) areas should be included

and could amount to about $290 billion, although these involve key assumptions and have a

higher degree of uncertainty than estimates for some of the other fields. Since the UNFCCC

report on global costs of adaptation omitted the costs of protecting ecosystems and the services

they can provide for human society, the present study concludes that this is an important source

of under-estimation.

Overview of conclusions

Assessing the costs of adaptation to climate change

References

Agrawala, S. and Fankhauser, S. (eds) (2008). Economic Aspects of Adaptation to Climate Change. Costs, Benefits and Policy Instruments. OECD,

Paris.
Berry, P. (2007). Adaptation Options on Natural Ecosystems. A report to the UNFCCC Financial and Technical Support Division, Bonn (http://

unfccc.int/cooperation_and_support/financial_mechanism/ financial_mechanism_gef/items/4054.php).
Burton, I. (2004). ‘Climate change and the adaptation deficit’ in A. French, et al., Climate Change: Building the Adaptive Capacity,

Meteorological Service of Canada, Environment Canada, pp25–33.
Dlugolecki, A. (2007). The Cost of Extreme Events in 2030. A report to the UNFCCC Financial and Technical Support Division (http://unfccc.

int/cooperation_and_support/financial_mechanism/ financial_mechanism_gef/items/4054.php).
Ebi, K. (2007). Health Impacts of Climate Change. A report to the UNFCCC Financial and Technical Support Division (http://unfccc.int/

cooperation_and_support/financial_mechanism/ financial_mechanism_gef/items/4054.php).
Environment Agency (2009). Investing for the Future Flood and Coastal Risk Management in England: A Long-term Investment Strategy.

Environment Agency, Bristol (downloadable at http://publications.environment-agency.gov.uk/pdf/GEHO0609BQDF-E-E.pdf).
FAO (2009). ‘1.02 Billion People Hungry’ (www.fao.org/news/story/en/item/20568/icode/) and (forthcoming, 2009) The State of Food

Insecurity in the World.
IPCC (2007). Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of

the Intergovernmental Panel on Climate Change. M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson (eds),

Cambridge University Press, Cambridge, 976pp.
Kirshen, P. (2007). Adaptation Options and Costs in Water Supply. A report to the UNFCCC Financial and Technical Support Division (http://

unfccc.int/cooperation_and_support/ financial_mechanism/financial_mechanism_gef/items/4054.php).
McCarl, B. (2007). Adaptation Options for Agriculture, Forestry and Fisheries. A report to the UNFCCC Financial and Technical Support

Division (http://unfccc.int/ cooperation_and_support/financial_mechanism/financial_mechanism_gef/items/4054.php).
McKinsey (2009) (forthcoming). The Economics of Adaptation.
McMichael, A.J., Campbell-Lendrum, D., Kovats, R.S., Edwards, S., Wilkinson, P., Edmonds, N., Nicholls, N., Hales, S., Tanser, F.C., Le

Sueur, D., Schlesinger, M. and Andronova, N. (2004). ‘Climate Change’ in M. Ezzati et al. (eds), Comparative Quantification of Health Risks:

Global and Regional Burden of Disease due to Selected Major Risk Factors. Vol. 2. World Health Organization, Geneva, pp1543–1649.
Nicholls, R. (2007). Adaptation Options for Coastal Zones and Infrastructure. A report to the UNFCCC Financial and Technical Support

Division (http://unfccc.int/ cooperation_and_support/financial_mechanism/financial_mechanism_gef/items/4054.php).
Oxfam (2007). Adapting to Climate Change. What is Needed in Poor Countries and Who Should Pay? Oxfam Briefing Paper 104.
Parry, M.L., Lowe, J.A. and Hanson, C. (2009). ‘Overshoot, Adapt and Recover’, Nature 258(7242), pp1102–1103.
Sachs, J.D. and McArthur, J.W. (2005). ‘The Millenium Project: A Plan for Meeting the Millennium Development Goals’, Lancet 365, pp347–353.
Satterthwaite, D. (2007). Adaptation Options for Infrastructure in Developing Countries. A report to the UNFCCC Financial and Technical

Support Division (http://unfccc.int /cooperation_and_support/financial_mechanism/financial_mechanism_gef/items/4054.php).
Stern, N. (2006). Stern Review: Economics of Climate Change. Cambridge University Press, Cambridge.
UNDP (2007). Human Development Report 2007/08. Palgrave McMillan, New York.
UNFCCC (2007). Investment and Financial Flows to Address Climate Change. Climate Change Secretariat, Bonn.
UNFCCC background papers (2007): see individual authored items cited in this list.
World Bank (2006). Investment Framework for Clean Energy and Development. World Bank, Washington DC.
World Bank (2009) (forthcoming). The Economics of Adaptation to Climate Change.

Overview of conclusions

Summary

• Global adaptation cost estimates range from $4 billion a year to well over $100 billion. These

numbers refer to the annual cost of adapting to ‘median’ climate change over the next 20 years.

They are different from estimates of the social costs of climate change, which measure

adaptation costs plus residual impacts (that is, damages after adaptation) over the atmospheric

lifetime of greenhouse gases.

• The wide range is symptomatic of the poor state of knowledge, with most estimates indicative

and incomplete, but also of the analytical difficulty of defining adaptation. There is also a dearth

of independent studies using different estimation techniques.

• The assumptions made to derive these estimates create both negative and positive biases, but on

balance it is likely that current approaches under-estimate the true cost of adaptation.

1.1 Adaptation and the total costs

of climate change

To understand the costs of adaptation one has to look at adaptation in its larger context.

Adaptation is only one part of the overall response to (and therefore the costs of) climate change.

The total burden of climate change consists of three elements: the costs of mitigation (reducing

the extent of climate change), the costs of adaptation (reducing the impact of change), and the

residual impacts that can be neither mitigated nor adapted to.
For example, society may seek to limit the overall temperature increase to 2°C (mitigation), invest

in coastal protection to limit the negative impacts of 2°C warming (adaptation) and accept the

loss of certain coastlines because they cannot be defended at reasonable cost (residual damage). In

the simplest possible economic framework, society would fine-tune global mitigation and

adaptation efforts until the combined (opportunity) costs of mitigation, adaptation and residual

damage are minimised.

What is important to understand at this point is that this optimisation process is more complex

than simply comparing annual adaptation costs – as reviewed in this chapter – with estimates of

global mitigation costs. There are two main complications. First, adaptation will not reduce

impacts to zero. There may be substantial residual damages that have to be taken into account.

Second, greenhouse gases are stock pollutants with long atmospheric lifetimes. Cutting emissions

today reduces the need for costly adaptation (and residual impacts) not just today but over many

decades – as long as the gases would have remained in the atmosphere.

Assessing the costs of adaptation to climate change

1.2 Adaptation and development

Another important delineation is between adaptation and socio-economic development. Socio-

economic trends over the coming decades – population growth, economic expansion, the

deployment of new technologies – will affect our vulnerability to climate events and indeed may

be shaped by climate conditions. Human activity has always been influenced by the climate

conditions people find themselves in. It is therefore difficult to delineate where socio-economic

development ends and adaptation to anthropogenic climate change begins.
This is particularly the case for developing countries, where there is a well-documented adaptation

deficit – that is, insufficient adaptation to the current climate. Poor people and poor countries are

less well prepared to deal with current climate variability than rich people and rich countries.

There is evidence that higher measures of development indicators like per capita income, literacy

and institutional capacity are associated with lower vulnerability to climate events (Noy, 2009;

Bowen et al., 2009). This has led authors like Schelling (1992) to conclude that good development

is one of the best forms of adaptation. More subtly, McGray et al. (2007) identify a continuum of

measures that address, to varying degrees, both development and adaptation needs. They range

from measures that reduce vulnerability to stress more broadly (whether climate-related or not) to

the creation of ‘systems for problem solving’, the management of current climate risks, and

policies specifically addressing climate change.
Most cost estimates deliberately and understandably ignore the overlap between adaptation and

development and focus on incremental adaptation over and above a vaguely defined baseline that

includes climate-relevant development programmes. With these caveats in mind we turn to a

review of existing adaptation cost estimates.

1.3 Adaptation costs in the

impact literature

Research on adaptation costs has a surprisingly long history, starting with early attempts to

estimate the economic costs of climate change.

The objective at that time was not to measure

adaptation costs per se, but to refine our understanding of climate change impacts. Modellers

recognised that their impact estimates would be wrong if they did not include an adaptive

response and overcame the ‘dumb farmer hypothesis’ (the assumption that farmers and other

actors would not react to a change in climate).
In a survey of adaptation in impact models, Tol et al. (1998) concluded that many impact

categories covered in the economic cost literature were actually adaptation costs, in particular

coastal protection, space heating and cooling (an adaptive response to changing temperatures),

defensive expenditures against air pollution and in some cases migration (an adaptive response if

premeditated but arguably a residual damage in the case of climate refugees). Adaptation also

featured prominently in the agriculture literature and to a lesser extent in health, but the adaptive

measures considered there were rarely costed out. Overall, Tol et al. found that adaptation costs

amounted to 7–25% of total impacts, with residual damages accounting for the balance.

Assessing the costs of adaptation to climate change

Over the years, the treatment of adaptation in global impact models was refined in a series of

sector studies.

However, a recent survey found that beyond coastal protection our knowledge of

adaptation costs (and benefits) is still fairly limited (Agrawala and Fankhauser, 2008), as shown in

Table 1.1.

Table 1.1 The state of knowledge on adaptation costs and benefits

Analytical coverage

Cost estimates

Benefit estimates

Coastal zones

Comprehensive

√√√

Agriculture

Comprehensive

–

√√√

Water

Isolated case studies

√

Energy

N. America, Europe

√√

Infrastructure

Cross-cutting, partly
covered in other sectors

√√

–

Health

Selected impacts

√

–

Tourism

Winter tourism

√

–

Source: Agrawala and Fankhauser (2008)

1.4 Adaptation costs in country

case studies

Other than in sector studies, adaptation costs are increasingly studied at the country level. Cost

estimates usually emerge as part of a broader planning exercise to develop a country-level

adaptation strategy. Early examples include studies sponsored by the World Bank in Bangladesh

(Smith et al., 1998) and the Pacific (World Bank, 2000). A somewhat broader exercise is World

Bank (2008), a recently initiated series of case studies that will deepen our understanding of

adaptation costs in developing countries.
More significantly in terms of actual adaptation planning, the international community has

embarked on a series of adaptation studies for the most vulnerable countries of the world, called

the National Adaptation Programmes of Action (NAPAs). Around 40 NAPAs have so far been

completed (see http://unfccc.int/adaptation/napas /items/4583.php). The aim of the NAPAs is to

identify priority adaptations and initiate a process of planning, preparation and implementation

in vulnerable developing countries.
NAPAs vary in quality and scope, with cost estimates ranging from less than $4 million in

Madagascar, Comoros and the Central African Republic, to several hundred million dollars in

Ethiopia and The Gambia, the only two countries to include extensive infrastructure investments.

Elsewhere, NAPA priorities predominantly cover preparatory measures and capacity building,

mostly on agriculture and water. As such, NAPAs are a poor indicator of the ultimate adaptation

expenditures in vulnerable countries, although they can give a rough indication of what the initial

outlay (and sectoral priorities) may be as the global adaptation effort is ramped up.

Assessing the costs of adaptation to climate change

1.5 First-generation global estimates

Interest in global adaptation cost estimates increased sharply a few years ago when international

support for adaptation in developing countries emerged as a key aspect in the negotiations for a

successor to the Kyoto Protocol. In response to this demand a handful of aggregate adaptation

cost estimates emerged in quick succession (Table 1.2). Although they are often dubbed ‘global’

they in fact concern adaptation in only the developing world.

Table 1.2 Estimates of adaptation costs in developing countries

Cost (US$
billion per year) Comments

World Bank (2006)

9 – 41

Cost of climate-proofing FDI, GDI and ODA flows

Stern (2007)

4 – 37

Update, with slight modification of World Bank
(2006)

Oxfam (2007)

>50

Based on World Bank, plus extrapolation of
costs from NAPAs and NGO projects

UNDP (2007)

86-109

World Bank, plus costing of PRS targets,
better disaster response

Source: Agrawala and Fankhauser (2008)

Note: FDI = foreign direct investment, GDI = gross domestic investment, ODA = official development assistance,

NAPA = National Adaptation Programme of Action, PRS = poverty reduction strategy

Another important feature of these estimates is that they are not independent. They are all based

on the same method, first developed by the World Bank (2006). The Bank estimated the fraction

of current investment flows that is climate sensitive and then used a ‘mark up’ factor that reflects

the cost of ‘climate-proofing’ these investments. The Bank team assumed that 2–10% of gross

domestic investment (GDI, worth $1500 billion a year at the time), 10% of foreign direct

investment (FDI, $160 billion) and 40% of official development assistance (ODA, $100 billion)

would be sensitive to climate change. The assumed mark-up to climate-proof these investments

was 10–20%. Of these assumptions only the ODA figure had any empirical grounding. It was

derived from earlier OECD work (Agrawala, 2005) about climate risks in six developing countries

(Bangladesh, Egypt, Fiji, Nepal, Tanzania and Uruguay). The range found in that study was

considerably wider though (12–65%), and its authors cautioned against drawing general conclusions.
Nevertheless, subsequent studies all adopted the same approach. The Stern Review (Stern, 2007)

reduced the mark-up for climate-proofing from 10–20% to 5–20%, and the share of climate-

sensitive ODA from 40% to 20%, but made no further adjustments to the method. The changes

in assumptions were not explained, other than to say that they were derived ‘through discussions

with the World Bank’.

Assessing the costs of adaptation to climate change

Oxfam (2007) adopted the World Bank numbers but added additional cost items, such as the

extra cost of NGO work at community level and the cost of implementing a NAPA-style

programme. Both cost items are based on fairly strong assumptions. The costs of community-level

adaptation were extrapolated from just three projects, while the cost of early adaptation was

derived from the 13 NAPAs available at the time.
The 2007 Human Development Report (HDR) (UNDP, 2007) again adopted the World Bank

approach to costing infrastructure adaptation. The HDR used later, considerably higher

investment data and a different share for climate-sensitive ODA (17–33%) but otherwise adopted

the Stern assumptions. In addition, it included the costs of adapting poverty reduction strategies

($44 billion a year) and strengthening disaster response systems ($2 billion a year).
Although based on a common method, the studies resulted in a large range in estimates, with the

lowest number at $4 billion and the highest one at over $100 billion. This indicates a fundamental

problem with this estimation approach. There is little empirical information about the share of

climate-sensitive investments and the mark-ups required to ‘climate-proof’ them, which are likely

to be situation-specific. Yet, because investment flows are so large, even small changes in this

uncertain parameter can change results by up to an order of magnitude.

1.6 The UNFCCC estimates

Perhaps the best global adaptation cost estimate available to date is UNFCCC (2007).

The

Secretariat of the Framework Convention commissioned six sector studies to get a better idea of

adaptation costs both globally and in developing countries. The estimates were made for the year

2030, usually assuming SRES A1B and B1 or similar scenarios.
• Agriculture, forestry and fisheries.

The agriculture estimate consists of three distinct cost

items: extra capital investment along the production chain (farms, transport, processing, etc.),

the need for better extension services at country level and the cost of additional global research

(e.g. on new cultivars). (See McCarl, 2007.)

• Water supply.

The water estimate considers the effect of additional water demand and changes

on the supply side. Investment decisions are made in anticipation of 2050 water needs. (See

Kirshen, 2007.)

• Human health.

The health estimate includes the extra treatment costs for three health issues:

malnutrition, malaria and diarrhoea. Scenarios are based on the Global Burden of Disease study

(McMichael et al., 2004). (See Ebi, 2007.)

• Coastal zones.

Coastal protection costs are based on the DIVA model, which considers a

limited set of adaptation options, which are applied globally. Uniquely, the coastal estimate

considers both adaptation costs and residual damages. Investments are made in anticipation of

2080 sea level rise. (See Nicholls, 2007.)

• Infrastructure.

The infrastructure estimate adopts the World Bank (2006) methodology, using

insurance data to determine the share of climate-sensitive new investment. (See Satterthwaite, 2007.)

Assessing the costs of adaptation to climate change

• Ecosystems.

In the absence of firm data, an indication of adaptation costs for ecosystems was

derived from the costs of increasing protected areas by 10%, although it was not possible to

split this into baseline cost (e.g. prevention of land conversion) and incremental adaptation.

(See Berry, 2007.)

UNFCCC (2007) concluded that total adaptation costs by 2030 could amount to $49–171

billion per annum globally, of which $27–66 billion would accrue in developing countries (Table

1.3). By far the largest cost item is new infrastructure investment, which for the upper- bound

estimate accounts for three-quarters of total costs. Costs are over and above what would have to

be invested in the baseline to renew the capital stock and accommodate growth. Note that the total

excludes the estimate for ecosystem adaptation, which was not considered to be sufficiently robust.

Table 1.3 UNFCCC estimates of global investment costs for adaptation

Sector

Global cost
($bn per
annum)

of which

Developed
countries

Developing
countries

Residual
damage

Agriculture

–

Water

–

Human health

–

Coastal zones

1.5

Infrastructure

8 – 130

6 – 88

2 – 41

–

Total

49 – 171

22 – 105

27 – 66

1.5

Source: UNFCCC (2007)

1.7 The state of the art

Most authors readily admit that adaptation cost estimates are preliminary, incomplete and subject

to a number of caveats. A recent survey of cost estimates by the OECD came to the same

conclusion (Agrawala and Fankhauser, 2008). Important gaps remain in terms of:
• the scope of the analysis (whether all relevant impacts and countries are covered)
• the depth of analysis (whether, for a given impact/country, all relevant adaptation options and

needs are considered)

• the costing of measures (whether all relevant costs are included)
• the treatment of uncertainty (how uncertainty about future change affects costs).

Assessing the costs of adaptation to climate change

Many of these gaps are exacerbated by the lack of a clear operational definition of adaptation.

The question then is to what extent the prevailing knowledge gaps lead to an upward or a

downward bias in the existing estimates. The sign of the bias is not immediately obvious, since

there are omissions in both directions. However, it is likely that adaptation costs have been

under-estimated so far (Table 1.4).

Table 1.4 Biases and omissions in existing cost estimates

Shortcoming

Likely bias

•

Scope: Limited range of impacts considered

•

Depth: focus on hard adaptation, no cost-effectiveness test

•

Depth: focus on planned, public adaptation

•

Costing: ignore preparation, lifetime costs

•

Costing: ignore higher-order effects

•

Uncertainty: ignore need to plan for a range of outcomes

– – –

++
– –

–
–
?

Source: author’s own assessment

The main downward bias in existing studies has to do with their limited scope. Adaptation analysis is

still predominantly a domain for case studies. Only a handful of studies aspire to provide a

comprehensive global estimate, and the first generation of global studies (see Table 1.2 above)

measured adaptation costs in only one particular sector, the cost of climate-proofing new investments.
Even the most extensive study to date (UNFCCC, 2007) is limited to six sectors, of which five

were ultimately reported (agriculture, water, health, coastal zones, infrastructure). Some areas

with clear adaptation needs – such as energy and tourism – were omitted, as were some prominent

adaptation strategies that are likely to feature prominently, most notably migration.

Moreover, in

the areas that were considered, the analysis was not always comprehensive. For example, only

three health impacts were assessed (malnutrition, malaria and diarrhoea) while the water estimate

excludes the cost of flood control and the extra expenditures required to maintain water quality.
The effect of insufficient depth is more difficult to assess. Even in detailed case studies it is difficult

to consider systematically all available adaptation options. Adaptation is too broad and ‘nebulous’

(Agrawala and Fankhauser, 2008) a concept to capture analytically in a comprehensive way. The

problem is exacerbated in global studies, which have to simplify and ignore local circumstances.

This introduces at least two biases, one leading to an under-estimation and the other to an over-

estimation of true adaptation costs.
The first bias is a preference for ‘hard’ structural adaptation measures over ‘soft’ behavioural or

regulatory adaptations (agriculture is an exception). Hard adaptation measures, such as the

expansion of water supply systems, are relatively easy to capture and generalise, but they are also

potentially much more expensive than soft measures like changes in water demand (e.g., in

response to price incentives). More generally, the proposed adaptation options are rarely subjected

to a rigorous cost-effectiveness test, suggesting that other, more effective options might be

identified once concrete options are being considered.

Assessing the costs of adaptation to climate change

The second bias concerns the focus on public adaptation over private adaptation. This is again

primarily because public adaptations are easier to identify than are the plethora of autonomous

adaptations individuals and firms are likely to undertake. However, certainly in terms of quantity,

and perhaps also in terms of costs, private autonomous measures will dominate the adaptation

response as people adjust their buildings, change space-cooling and -heating preferences, reduce

water use, alter holiday destinations or even relocate. UNFCCC (2007) includes many measures

that will ultimately be the responsibility of private actors, in particular in health, agriculture and

to some extent infrastructure, but it is likely that this represents only the tip of the iceberg in

terms of private adaptation. Global cost estimates, including that of the UNFCCC, are primarily

about planned adaptation.
Although costing is almost always ad hoc, the resulting bias is more difficult to ascertain. A careful

assessment of adaptation costs would ideally consider the net present value of costs over the entire

lifetime of a project, including preparation costs, investment costs, operations and maintenance

costs and decommissioning costs. In reality, the focus is often on the initial capital expenditures

(in the case of UNFCCC (2007), the focus is deliberately on investment costs). Another potential

omission, not least in a developing country context, is institutional and administrative costs,

including the costs of building planning capacity. There is some speculation, but little analytical

evidence, about the importance of higher-order costs and inter-linkages, for example the effect of

a large-scale coastal defence programme on construction activity elsewhere or the interaction

between adaptation in water and agriculture.

Finally, all cost estimates assume that future climate change impacts are known with certainty. In

reality, adaptation planners will have to deal with considerable uncertainty about the likely extent

and even direction of change at the local level. In some (rare) cases, this may justify a delay in

adaptation until more information is available (thus reducing adaptation costs, but increasing

residual impacts). In other cases, it will force planners to extend the scope of adaptations so that

they can deal with a wider range of outcomes. This is likely to increase adaptation costs.
Readers interested in the overall costs of climate change will also have to remember that the scope

of adaptation estimates is deliberately narrow. Important costs that matter for this broader debate

were excluded to keep the problem tractable, including the costs of closing prevailing adaptation

gaps, the cost of residual impacts and the cost of mitigation. Many of these will be higher than the

cost of adaptation. Moreover, if the mitigation and development investments are not made, the

cost of adaptation will be considerably higher.

1 More complex frameworks would also consider reasons for concern other than aggregate costs, such as the unfair distribution of impacts, the

risk of tipping points, excessive climate variability and the threat to unique natural systems (Smith et al., 2001).

2 Chapter 8 in this report shows how the estimates of adaptation costs reviewed here translate into estimates of the social costs of carbon, that is

the cost emissions impose on society over their atmospheric lifetime.

3 Smith and Tirpak (1989), Cline (1992), Nordhaus (1994), Fankhauser (1995), Pearce et al. (1995) and Tol (1995). A recent survey is Tol (2005).
4 For example, Nicholls and Tol (2006) on coastal zones and Parry et al. (2004) on agriculture. See Parry et al. (2007) for an assessment.
5 More accurately, the UNFCCC aimed to estimate the investment and financial flows for adaptation (and mitigation).
6 Planned migration away from affected areas can be an effective adaptation strategy. In contrast, the plight of climate refugees may be better

classified as a residual damage as it reflects insufficient anticipatory adaptation. In addition, migration (e.g. a growing urbanisation) is also

part of ‘baseline’ development.

7 See Bosello et al. (2007) for a rare general equilibrium study.

Assessing the costs of adaptation to climate change

References

Agrawala, S. (2005). Bridge over Troubled Waters: Linking Climate Change and Development. Paris: OECD.
Agrawala, S. and S. Fankhauser (eds) (2008). Economic Aspects of Adaptation to Climate Change. Costs, Benefits and Policy Instruments. Paris: OECD.
Berry, P. (2007). Adaptation Options on Natural Ecosystems. A report to the UNFCCC Financial and Technical Support Division, Bonn.
Bosello, F., R. Ronson and R.S.J. Tol (2007). ‘Economy-wide Estimates of the Implications of Climate Change: Sea Level Rise’, Environmental and

Resource Economics 37: 549–571.
Bowen, A., B. Chatterjee and S. Fankhauser (2009). Adaptation and Economic Growth. Paper prepared for the UK Department for International

Development, Grantham Research Institute, London School of Economics, February.
Cline, W.R. (1992). The Economics of Global Warming. Washington DC: Institute for International Economics.
Ebi, K. (2007). Health Impacts of Climate Change. A report to the UNFCCC Financial and Technical Support Division, http://unfccc.int/cooperation_

and_support/financial_mechanism/ financial_mechanism_gef/items/4054.php.
Fankhauser, S. (1995). Valuing Climate Change. The Economics of the Greenhouse. London: Earthscan.
Kirshen, P. (2007). Adaptation Options and Costs in Water Supply. A report to the UNFCCC Financial and Technical Support Division, http://unfccc.int/

cooperation_and_support/ financial_mechanism/financial_mechanism_gef/items/4054.php.
McCarl, B. (2007). Adaptation Options for Agriculture, Forestry and Fisheries. A report to the UNFCCC Financial and Technical Support Division, http://

unfccc.int/ cooperation_and_support/financial_mechanism/financial_mechanism_gef/items/4054.php.
McGray, H., A. Hamill, R. Bradley, E.L. Schipper and J.-O. Parry (2007). Weathering the Storm. Options for Framing Adaptation and Development.

Washington DC: World Resources Institute.
McMichael, A. et al. (2004). ‘Climate Change’, in: Ezzati, M., A. Lopez, A. Rodgers and C. Murray (eds), Comparative Quantification of Health Risks:

Global and Regional Burden of Disease. Geneva: World Health Organization.
Nicholls, R. (2007). Adaptation Options for Coastal Zones and Infrastructure. A report to the UNFCCC Financial and Technical Support Division, http://

unfccc.int/ cooperation_and_support/financial_mechanism/financial_mechanism_gef/items/4054.php.
Nicholls, R. and R.S.J. Tol (2006). ‘Impacts and Responses to Sea Level Rise: A Global Analysis of the SRES Scenarios over the 21st Century,

Philosophical Transactions of the Royal Society A 363: 1073–1095.
Nordhaus, W.D. (1994). Managing the Global Commons. The Economics of Climate Change. Cambridge MA: MIT Press.
Noy, I. (2009). ‘The Macro-economic Consequences of Disaster’, Journal of Development Economics 88: 221–231.
Oxfam (2007). Adapting to Climate Change. What is Needed in Poor Countries and Who Should Pay? Oxfam Briefing Paper 104.
Parry, M.L. et al. (2007). Climate Change 2007. Impacts, Adaptation and Vulnerability. Working Group II Contribution to the Fourth Assessment Report of the

Intergovernmental Panel on Climate Change. Cambridge: CUP.
Parry, M.L. et al. (2004). ‘Effects of Climate Change on Global Food Production under SRES Emissions and Socio-economic Scenarios’, Global

Environmental Change 14(1): 53–67.
Pearce, D.W. et al. (1996). ‘The Social Costs of Climate Change: Greenhouse Damage and the Benefits of Control’ in: IPCC, Climate Change 1995.

Economic and Social Dimensions of Climate Change. Contribution of Working Group III to the Second Assessment Report of the IPCC. Cambridge: CUP.
Satterthwaite, D. (2007). Adaptation Options for Infrastructure in Developing Countries. A report to the UNFCCC Financial and Technical Support

Division, http://unfccc.int /cooperation_and_support/financial_mechanism/financial_mechanism_gef/items/4054.php.
Schelling, T., (1992). ‘Some Economics of Global Warming’, American Economic Review 82(1): 1–14.
Smith, J.B. et al. (2001). ‘Lines of Evidence for Vulnerability of Climate Change: A Synthesis’, in: IPCC, Climate Change: Impacts, Adaptation and

Vulnerability. Contribution of Working Group II to the Third Assessment Report of the IPCC. Cambridge: CUP.
Smith, J.B. and D. Tirpak (eds) (1989). The Potential Effects of Global Climate Change on the United States. Washington DC: US Environmental Protection

Agency.
Smith, J.B, A. Rahman, M.Q. Mirza, G.J. Kenny and G.C. Sims (1998). Considering Adaptation to Climate Change in the Sustainable Development of

Bangladesh. Washington DC: South Asia Region, World Bank.
Stern, N. (2007). The Economics of Climate Change: The Stern Review. Cambridge: CUP.
Tol, R.S.J. (2005). ‘The Marginal Damage Costs of Carbon Dioxide Emissions: An Assessment of the Uncertainties’, Energy Policy 33(16): 2064–2074.
Tol, R.S.J (1995). ‘The Damage Costs of Climate change – Toward More Comprehensive Calculations’, Environmental and Resource Economics 5: 353–374.
Tol, R.S.J, Fankhauser and J.B. Smith (1998). ‘The Scope for Adaptation to Climate Change: What Can We Learn from the Impact Literature?’, Global

Environmental Change 8(2): 109–123.
UNDP (2007). Human Development Report 2007/08. New York: Palgrave McMillan.
UNFCCC (2007). Investment and Financial Flows to Address Climate Change. Bonn: Climate Change Secretariat.
World Bank (2008). The Economics of Adaptation. Concept Note and Study Plan, April. www.worldbank.org/environment/eacc.
World Bank (2006). Investment Framework for Clean Energy and Development. Washington DC: World Bank.
World Bank (2000). Cities, Seas, and Storms. Managing Change in Pacific Island Economies. Volume IV: Adapting to Climate Change. Washington DC:

Papua New Guinea and Pacific Island Country Unit, World Bank.

Assessing the costs of adaptation to climate change

Summary

Production and activity within the agriculture, forestry and fisheries sector is inherently affected

by variability in climate. There is a tradition of coping with year-to-year changes in climate;

nevertheless, human-induced climate change is expected to push these managed ecosystems

beyond their natural climatic boundaries, requiring a greater rate and extent of adaptation than

previously needed. In this chapter we examine the methods and assumptions of costing this

adaptation to climate change within the agricultural, forestry and fisheries sector.
A range of methods exists for costing adaptation. The UNFCCC report Adaptation Options for

Agriculture, Forestry and Fisheries (McCarl, 2007) takes a top-down approach to costing

adaptation. A 10% increase in research and extension funding and a 2% increase in capital

infrastructure costs are assumed due to climate change using the A1B1 SRES scenario for the year

2030, termed ‘without mitigation’. The global marginal estimates for additional funding for

adaptation of the agricultural sector due to climate are $12.6 billion and $11.3 billion – without

and with mitigation, respectively, in the year 2030 (McCarl, 2007). The cost of efforts to keep up

with the demands from the sector with future changes in population are given as more than an

order of magnitude higher, at $520 billion. Alternative bottom-up approaches focus on costing

specific adaptation options, such as improved irrigation or developing new adapted crop varieties.

For example, interpretation of the study of Fischer et al. (2007) provides a cost of $8 billion for

adapting crop irrigation systems to climate change by 2030. A different approach by Cline

(2007), using simple crop growth models, provides an estimate of $14.5 billion for the year 2030

for the reduction in the value of global crop outputs due to climate change.
We conclude that estimating the cost of adaptation of the agriculture, forestry and fisheries sector

to climate change is a difficult task requiring further research. The McCarl (2007) estimate from

the UNFCCC takes a top-down approach that projects forward increases on current costs to

provide a figure of $11.3–12.6 billion for the year 2030. We conclude that this is a reasonable first

approximation of the additional costs of adaptation of agriculture, forestry and fisheries.

Nevertheless, on the basis of the limited independent evidence available, we expect bottom-up

approaches to reveal this to be on the low side of adaptation costs in this sector.

2.1 Assumptions

When measuring the economic cost of an activity it is important to understand the purpose for

which the cost is being evaluated. The usual reason for measuring cost is to compare the costs and

benefits of a project in order to judge whether it should be implemented. The issues associated

with such an evaluation are complex, and include a recognition that market prices may not

represent the true value of the project owing to distortions in the market, and the fact that many

benefits and costs are not subject to market transactions and therefore do not have readily

observed values. In many cases the cost-benefit analysis (CBA) is conducted from the perspective

of a single economic agent. For example, a firm may conduct a CBA of a particular investment

project, a government might analyse a scheme to protect a site of environmental value, and

consumers might be interested in the costs and benefits of replacing their car. In a broader

context, such as that under consideration in this chapter, the evaluation does not pertain to a

Assessing the costs of adaptation to climate change

single economic agent but to society as a whole. Moreover, the nature of the ‘project’ to be

evaluated needs careful consideration.
Climate change is an exogenous shock to the economy and adaptation and mitigation are

responses to that shock. There are qualitative differences between these responses. One aspect of

these qualitative differences is considered by Callaway (2004), who argues that adaptation to

climate change is essentially private, in contrast to mitigation. The consequences of the adaptation

action can be excluded and therefore accrue primarily to their instigator. For example, the sole

beneficiary of the decision by a farmer to adjust the planting date of his crop will be the farmer

himself. The distinction is important, because the existence of a public good will lead to a market

failure which in turn leads to an under-provision. In contrast, where a good is private, the market

can be expected to provide an efficient outcome if there are no distortions (a big if). One reason

for evaluating the costs of adaptation is to determine the optimal combination and level of

adaptation and mitigation.
While viewing adaptation to climate change as a private good serves to make a point, it is

probably an oversimplification. There will be aspects of adaptation that are also public goods, for

example the research and development of new varieties adapted to new climates. The other

important consideration relating to adaptation is its distributional consequences. While the

market might solve the climate problem through adaptation, the distribution of the market

benefits might be substantially different. For example, if prices rise as a result of the adaptation,

producers may benefit at the expense of consumers. Moreover, the distribution of the benefits

among producers may be spatially uneven with, for example, developed country farmers

benefiting at the expense of those in developing countries.
The distinction between public and private goods is important because the presence of the former

implies that the market will not provide a welfare-maximising response to the exogenous shock.

As a consequence, government intervention is justified, to improve welfare. In this context, the cost that

needs to be measured is that of the proposed government intervention, and this needs to be compared

with the welfare loss that results from the shock as a measure of the benefit from intervention.
The perspective may not be one of market failure, however. One might be interested simply in

how much money it is going to cost us to make the necessary adaptations to climate change. This

is essentially the focus of the McCarl (2007) paper. In this context it is important to recognise

that what we seek is a measure of the opportunity cost that is forgone as a result of us needing to

adapt to climate change. For example, we might have to spend money on developing new varieties

capable of withstanding drought stress. In the absence of climate change we might have equally

spent money on developing varieties that are better adapted to the unchanged climate. What

matters is the additional expenditure that is necessary to adapt. Once we have measured a cost of

this sort it is important to recognise its significance. Spending money is not of itself a bad thing

and one person’s cost is another person’s benefit. When we measure costs of this sort what we

often measure are transfers from one group to another. This is not to say that distributional

questions are not important. The problem is that the answer as to whether a move is good or bad is

much harder as it depends on us making value judgements about the importance of different

groups in society. From a social perspective, the only thing that is unambiguously bad is a market

failure which results in an overall reduction in output or welfare.

Assessing the costs of adaptation to climate change

The final conceptual issue which is particularly important in the context of climate change is the

distinction between static and dynamic costs. In a static approach the cost that is measured is one

that compares two discrete situations. In this case, one with climate change and one without.

There are costs that are associated with making the change which may be ignored in a static

approach. These are termed adjustment costs and are the focus of a dynamic approach. With

climate change these costs may include things such as migration and reduction. A population could

have an equally high level of welfare in a new location once it had established itself, but there are

likely to be substantial reductions in welfare that are the consequence of having to relocate.

2.2 Issues

There are a wide range of possible adaptation strategies to climate change within agriculture,

fisheries and forestry. These span a range of scales from farms to governments, and from relatively

simple autonomous changes in the way agricultural businesses are managed to complex

programmes of research and development lasting decades. Many authors provide lists and

categories for adaptation strategies for this sector, with much in common from one to the other.

The review by Howden et al. (2007) provides a good overview of adaptation in agriculture,

forestry and fisheries. It considers many adaptation options as either changes in management

decisions or changes in the decision environment. Changes in management within cropping

systems include altering crops or crop varieties, more efficient management of water, altering the

timing or location of cropping, and improving the effectiveness of crop protection measures.

Within livestock systems many adaptation options are connected with maintaining the

availability of fodder and feed and reducing heat stress from animal housing. Adaptation of

managed forests could involve changes in tree species composition, harvesting patterns, pest

control and location of managed woodland. Marine fisheries adaptation is less sensitive to

management changes, except for changes in catch size (Howden et al., 2007). Adaptation of the

sector through decision making may include policy changes, the development of infrastructure

and general changes to the decision-making environment.
It is also clear that there is an enormous range of actors within the sector as well: individuals

and collectives, private and public institutions. Given this diversity and complexity of possible

adaptation options and actors, how well can the spectrum of possible adaptation strategies

within agriculture, forestry and fisheries be costed and summarised in a single or a few global

headline figures?

2.3 Existing methods

The UNFCCC report Adaptation Options for Agriculture, Forestry and Fisheries (McCarl, 2007)

takes a top-down approach to costing adaptation. It splits adaptation costs for the agriculture,

forestry and fishery sectors taken as a whole into those needed for research, extension and physical

capital expenditure. It then projects forward the past trends in each of these, sourced from current

estimates and the literature, and imposes an additional increase due to climate change. A 10%

increase in research and extension funding and a 2% increase in capital infrastructure costs are

assumed due to climate change using the A1B1 SRES scenario for 2030, termed ‘without

mitigation’. These additional costs are then reduced for the SRES B1 ‘mitigation’ scenario by

Assessing the costs of adaptation to climate change

multiplying by the ratio of the projected global temperature change for A1B1 relative to B1;

that is 1.4/1.6.
The global estimates for additional funding for adaptation of the agricultural sector due to

climate change in the year 2030 are $12.6 billion and $11.3 billion, without and with mitigation

respectively (McCarl, 2007). The cost of efforts to keep up with the demands from the sector with

future changes in population are given as more than an order of magnitude higher than this at

$520 billion.

2.4 Critique of methods

Although the approach of McCarl (2007) seems just to take the current state and add a more or

less arbitrary amount to represent the additional costs of adaptation, it is very hard to judge whether

the magnitude of the figure is reasonable without trying to take a bottom-up approach to costing.

One useful outcome seems to be that there is already a substantial amount invested in research

and development within the sector and that adaptation is already well embedded. As a consequence,

it is clear that the incremental costs of adaptation are likely to be small (relatively speaking).
Can global estimates of the cost of adaptation in the agricultural sector be derived from the

bottom up as an alternative approach to that taken by McCarl (2007)? A number of options are

available. The first is to try explicitly to model impacts and different adaptation strategies and

then to cost them. For example, much progress has been made in assessing the impacts of global

climate change on crop production to date. Research is moving towards a consensus that predicts

a small increase in production of the major grain crops across northern, developed nations up to

2050, followed by a gradual decline to 2100, and a steady decline in yields in tropical regions over

this entire period (Parry et al., 2004; Cruz et al., 2007). The effects of climate change on world

cereal prices follow these trends in crop yields, with little consistent change in output price among

different studies until a warming of 2–3°C from current climates, beyond which prices start to

rise uniformly across different studies. These approaches can be extended to consider upstream

and downstream impacts on costs through the use of a general equilibrium model (for example,

Winters et al., 1998; Lewandrowski and Schimmelpfennig, 1999).
Second, a meta-analysis of all existing impacts and adaptation studies can be brought together on

a common axis as a surrogate for climate change. Such an exercise was undertaken by Easterling

et al. (2007) who, from many maize, rice and wheat studies, quantified the impacts of and

adaptation to a range of temperature warming on these crops. At mid- to high latitudes, where

the majority of global cereal production is found, a moderate warming of 1–3°C (expected in the

decades close to 2050) is thought to have a small beneficial effect on wheat yields of 5–10%,

but warming above 3°C reduces yield below current values. At low latitudes, the yield of wheat

declines steadily with any warming. Adaptation of cropping practices improves yield across the

temperature warming range, and hence pushes the onset of negative impacts further into the

future. For example, the potential benefits of adaptation of cropping practices to climate change

may be as much as an 18% improvement in the yield of temperate and tropical wheat systems.

The response of rice to temperature warming is broadly similar to that of wheat. For maize, there

is less benefit to yield of small temperature changes at mid- to high latitudes and a more dramatic

decline at warmer temperatures, even with adaptation. This meta-analysis is still one step away

Assessing the costs of adaptation to climate change

from providing adaptation costs, but does sample a lot of empirical evidence and defines the

response of adaptation to different magnitudes of climate change, as represented by global

temperature changes.
The use of hedonic modelling presents a positive alternative to the normative approaches that are

based on simulation through the use of integrated climate, crop and economic models. The use of

hedonic equations assumes that the impacts of climate change on profitability will be capitalised

into the value of land and yields. They carry the advantage of not being constrained by the

adaptation strategies that are deemed feasible by the parameterisation of the models used in a

simulation approach. An equation relating land values to climate variables is then estimated and

used to forecast the impact of changes in these variables on land values over a number of different

climate change scenarios. This approach is exemplified by Schlenker et al. (2005; 2006).

Deschenes and Greenstone (2007) note that the hedonic approach is subject to criticism because

unobserved factors which influence land values are excluded from the hedonic equation. As a

result it is probable that the estimates which result from this approach are biased. They therefore

propose an alternative approach in which profitability is directly related to climatic variables.

They note that this may also produce biased results because it fails to take into account the full

extent to which farms may adapt to climate change. They argue that because the direction of bias

is known (an under-estimate of the impacts), their method is superior to the hedonic approach.

2.5 Comparison with case studies

Compared with McCarl (2007), the examples of bottom-up approaches are more closely linked to

individual processes of adaptation within the agricultural sector. Can studies of the cost of single

adaptation strategies, or on a local or regional scale, provide an independent assessment of the

UNFCCC global cost? Potentially they could, through careful selection of representative sites,

enterprises and adaptation options, but costed studies on the local or regional scale are also scarce

in the literature.
Some studies provide costs of single adaptation options. For example, Fischer et al. (2007)

estimated adaptation costs through meeting future irrigation demands by 2080 to be $24–27

billion per year for an unmitigated scenario and $8–10 billion per year less for the mitigated B1

scenario. These adaptation costs of providing increased irrigation capacity can be compared with

the estimates of McCarl (2007) by bringing them to a common metric under the same emissions

scenario, where this is possible. The total additional costs of McCarl (2007) are $12.6 billion and

$11.3 billion of additional adaptation costs for the year 2030 for the A1B and B1 emissions

scenarios, respectively. The unmitigated scenario used by Fisher et al. (2007) was different from

that of McCarl (2007). Under the same mitigated B1 scenario the study of Fischer et al. (2007,

Figure 3) gives additional annual costs of $8 billion per year for increasing irrigation capacity for

2030. These additional adaptation costs to maintain irrigation capacity under climate change are

therefore about 65% of those of McCarl (2007) although we are comparing the cost of a single

adaptation option in the crops sector to costs across agriculture, forestry and fisheries. Of course

our confidence in such a comparison using this bottom-up approach to estimate the additional

costs of adaptation to climate change are highly uncertain using the limited evidence currently

available. Nevertheless, if we were to extend this approach with costings for other explicit

Assessing the costs of adaptation to climate change

adaptation measures, it could be that the additional adaptation costs for the agricultural sector

will exceed the estimates of McCarl (2007).
The cost of developing new crop varieties in public breeding programmes has been put at $2.1

million for oats and $2.8–3.0 million for wheat for a single variety (NRC, 2000). The Alliance for

a Green Revolution in Africa puts the cost of developing 200 new crop varieties better adapted to

local environments at $43 million. The development of Bt maize by Monsanto is thought to have

cost $10–25 million (NRC, 2000). These examples for developing new crop varieties show a wide

spread of costs for at least some components of global estimates for the sector, but are all tiny in

comparison with the total estimate for the sector of McCarl (2007). It is very difficult to see how

this bridge across the divide of spatial scale and process detail to a global scale could be made.

Perhaps it cannot, so that we should view the global estimates of adaptation cost together with

(unconnected) enterprise- or location-specific studies. It is also difficult to separate population-

related needs from needs related to climate change.
Within the economics literature two approaches that can be interpreted as the net costs of

adaptation exist for costing the impacts of climate change. These are referred to as the Ricardian

and the crop growth model methods. In the former, land values are modelled under an

assumption that they reflect the future profitability of farms (Mendelsohn et al., 1994). Studies

employing the Ricardian approach are country specific and show a diverse array of effects, with

some countries (e.g. the US) benefiting and others losing (Sanghi and Mendelsohn, 2008). In the

crop growth model approach, the impacts of climate change are simulated using crop growth

models and the value of the resulting change in output is taken as a measure of the economic

impact of climate change. Cline (2007) summarises these estimates for in excess of 70 countries.

He estimates that the overall impact of climate change on agriculture will be a reduction in the

value of output of the order of $38 billion dollars by 2080. This figure masks some wide variations

in the spatial impacts, however. For example, India alone is predicted to witness a reduction in its

output of the value of $38 billion dollars, which is offset by gains elsewhere including China ($14

billion) and the US ($8 billion). However, if we simply scale back this estimate to the year 2030,

assuming (probably incorrectly) a linear increase in cost over time, the Cline (2007) estimate

equates to $14.3 billion for the year 2030. This is very close to the estimate of McCarl (2007).

Nevertheless we need to bear in mind that the Cline (2007) study is for adaptation of crop

production alone, while McCarl (2007) is for the entire agriculture, forestry and fisheries sector.

2.6 Outstanding issues

We have argued above that the McCarl approach can be criticised from a theoretical perspective

as a result of being insufficiently clear regarding the objectives of measuring the cost. In this

section we examine some more detailed aspects of this approach.
Costs of research in the UNFCCC report are based on information from the Consultative Group

on International Agricultural Research (CGIAR) institutes. These omit national and private-

sector research efforts, both of which may have a different magnitude of costs and cost-benefit

ratios, particularly for agricultural research and development (R&D) in developed countries.

However, there is a close relationship between increases in agricultural productivity and

investment in R&D (World Bank, 2008), with an average rate of return on investment in

agriculture R&D found to be 43% across 1673 studies worldwide.

Assessing the costs of adaptation to climate change

As the variability of climate increases, farmers and agribusiness will have to adapt to different

patterns of rainfall and more extremes of rainfall and temperature than currently experienced.

Variability in climate can have important and dramatic impacts on the productivity of cropping

systems (Porter and Gawith, 1999; Wheeler et al., 2000). Conceptually, we could see this as a

challenge different from that of adapting to changes in mean climate conditions. To adapt to

increased climate variability, those managing farms, forests and fisheries will have to adjust their

risk-management strategies to try to maintain production and profits or, for some, to counter

their losses or better exploit good years. Deciding how to cost in what is effectively a probabilistic

adaptation response is difficult, but nevertheless important. It is likely to take a form quite

different from a simple linear enhancement of spending on adaptation. At a conceptual level, the

cost associated with increased variability should be measured as the additional resources that

society expends in avoiding the associated risk. In a practical sense this is perhaps relatively

straightforward and may, for example, be a matter of measuring the incremental cost of fertiliser,

irrigation and crop protection. Conceptually the cost is determined by the size of the increase in

variability and the attitude of the producer to risk. The issue is further complicated by the

recognition that the costs are likely to spill over into sectors of the economy beyond the primary

producers. Intermediaries and final consumers are likely to adopt strategies designed to mitigate

the impacts of risk on their profitability and welfare.
The development of agriculture occurs against a moving background of changing demand for

food and agricultural products, each with their own features specific to different locations,

different rates of change and timescales of responses. McCarl (2007) provides a useful contrast of

the magnitude of the costs of providing for an expanding population that is about 40 times

greater than the cost of adapting to climate change. This suggests that adaptation/development

costs within the sector will be dominated by changes driven not directly by climate. Nonetheless,

there are many intrinsic links between increases in productivity and resilience to climate that

confuse this discrimination as a basis for costing.
A clear and consistent conclusion from several decades of studies of the impacts of climate change

on agriculture is the uneven distribution of positive and negative impacts in different parts of the

world. The precise balance between positive and negative impacts depends on the type of

agricultural business and the degree of climate change, but can roughly be summarised as positive

impacts and opportunities for agriculture at the higher latitudes over the coming decades in

contrast to more immediate negative impacts at lower latitudes. Thus, a global cost of adaptation

needs to account for changes in the balance of benefits or opportunities and negative impacts in

different parts of the world and over a range of time projections – a complex situation.

Nevertheless, in the agricultural sector it is important to account for opportunities in some

regions where production may potentially benefit from a moderate amount of climate change.

Here, farmers could exploit short-term opportunities for their businesses with some investments

costs, but with little or no investment would miss these. Although such investments are unlikely

to be seen as adaptation to climate change, they could dominate spending in the short term in

some regions. In addition, there is the potential for adaptation funds to be directed to developing

countries within any global total. However, to date such adaptation funds have been

comparatively small, for example, $150–300 million per year (World Bank, 2008).

Assessing the costs of adaptation to climate change

Interaction between the degree of climate change, how sensitive agriculture in different parts of

the world is to climate, and the level of farming technology available locally will be important in

determining adaptive capacity. Part of this complex interaction is accounted for by considering, as

the UNFCCC report does, developed and developing nations separately. This is sensible, but

there are obvious exceptions. It could be that the exceptions outnumber the general rule once

further divisions are considered, thus providing more room for error.

2.7 Estimate of residual damage

Costs of the residual damage from climate change after adaptation of agriculture, forestry and

fisheries are difficult to find. An example estimate of the residual damage from adaptation of

agriculture can be provided by interpretation of the adaptation study of the wheat crop in

Australia. The effect of adaptation through changes in sowing dates and crop varieties on the

gross value of the national wheat crop was simulated by Howden and Jones (2004). Uncertainty

in response was represented using an ensemble of IPCC scenarios and spatial variability across

Australia. Without adaptation the maximum potential increase under climate change for 2080

was 10% of gross value of the national crop – $0.3 billion per year (Figure 2.1). With adaptation

there was a median increase of $158 million per year in crop value, but with a range of values

about this, many below zero, and a small chance of a 20% decrease in crop value ($0.6 billion per

year) with the uncertainty that was sampled (Figure 2.1). As a proportion, this represents a

residual damage of about 20%.

Figure 2.1 Changes in national gross values of wheat in Australia for 2070
(a) without adaptation; (b) with adaptation

-30

-40

-20

-10

0.02

0.04

0.06

0.08

0.1

-30

-40

-20

-10

0.02

0.04

0.06

0.08

0.1

(a)

(b)

Source: Howden and Jones (2004)

r
o

a
b

i
l
i
t
y

r
o

a
b

i
l
i
t
y

Change in crop value (%)

Assessing the costs of adaptation to climate change

2.8 Conclusions

We have concentrated in this paper on why it is difficult to estimate a single global cost for the

adaptation of agriculture, forestry and fisheries to climate change. The estimate of McCarl (2007)

is the only example of a global estimate specifically for this sector. However, we argue that this

approach is certainly difficult to verify. A bottom-up comparison with this global figure using

single adaptation options provides orders of magnitude smaller than (crop variety) or more than

half of the total of (irrigation) this global value for the entire sector. Whilst not providing a direct

link to the global figure, these single components of adaptation do suggest that $12.6 billion and

$11.3 billion without and with mitigation could be an under-estimate of the cost of global

adaptation of agriculture, fisheries and forestry.
Our interpretation of the crop growth model study of Cline (2007) is in good agreement with the

UNFCCC estimate of McCarl (2007), with the important difference that Cline (2007) studied

crop production alone whilst McCarl (2007) covered agriculture, forestry and fisheries. Therefore,

we conclude that the UNFCCC estimate of $11.3-12.6 billion is a reasonable first approximation

of adaptation costs in this sector, but we expect the estimate of adaptation costs for agriculture,

forestry and fisheries to increase as more detailed studies of specific adaptation actions become

available and as our understanding of the impacts of climate change matures.

Assessing the costs of adaptation to climate change

References

Callaway, JM (2004) ‘Adaptation Benefits and Costs: Are they Important in the Global Policy Picture and How Can We Estimate Them?’,

Global Environmental Change 14: 273–282.
Cline, WR (2007) Global Warming and Agriculture: Estimates by Country. Washington: Center for Global Development and Peterson Institute

for International Economics.
Cruz, RV, Harasawa, H, Lal, M, Wu, S, Anokhin, Y, Punsalmaa, B, Honda, B, Jafari, M, Li, C and Ninh, NH (2007) Climate Change 2007:

Impacts, Adaptation and Vulnerability: Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate

Change, ML Parry, OF Canziani, JP Palutikof, PJ van der Linden and CE Hanson (eds). Cambridge: Cambridge University Press, 469–506.
Deschenes, O and Greenstone, M (2007) ‘The Economic Impacts of Climate Change: Evidence from Agricultural Output and Random

Fluctuations in Weather’, American Economic Review 97: 354–385.
Easterling, WE, et al. (2007) ‘Food, Fibre and Forest Products’, Climate Change 2007: Impacts, Adaptation and Vulnerability: Contribution of

Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, ML Parry, OF Canziani, JP Palutikof,

PJ van der Linden and CE Hanson (eds). Cambridge: Cambridge University Press, 273–313.
Fischer, G, Tubiello, FN, van-Velthuizen, H and Wiberg, DA (2007) ‘Climate Change Impacts on Irrigation Water Requirements: Effects of

Mitigation 1990–2080’, Technological Forecasting and Social Change 74: 1083–1107.
Howden and Jones (2004), cited in Hennessy, KB, Fitzharris, BC, Mates, N, Harvey, SM, Howden, L, Hughes, L, Sainger, J and Warrick, R

(2007) ‘Australia and New Zealand’, Climate Change 2007: Impacts, Adaptation and Vulnerability: Contribution of Working Group II to the

Fourth Assessment Report of the Intergovernmental Panel on Climate Change, ML Parry, OF Canziani, JP Palutikof, PJ van der Linden and CE

Hanson (eds). Cambridge: Cambridge University Press, 507–540.
Howden, SM, Soussana, JF, Tubiello, FN, Chhetri, N, Dunlop, M and Meinke, H (2007) ‘Adapting Agriculture to Climate Change’,

Proceedings of the National Academy of Sciences 104: 19691–19696.
Lewandrowski, J and Schimmelpfennig, D (1999) ‘Economic Implications of Climate Change for US Agriculture: Assessing Recent Evidence’,

Land Economics 75: 39–57.
McCarl, BA (2007) Adaptation Options for Agriculture, Forestry and Fisheries. A Report to the UNFCCC Secretariat Financial and Technical

Support Division.
Mendelsohn, R, Nordhaus, WD and Shaw, D (1994) ‘The Impact of Global Warming on Agriculture: A Ricardian Analysis’, American

Economic Review 84(4): 753–771.
NRC (National Research Council) (2000) Genetically Modified Pest-protected Plants. Washington: National Academy Press.
Parry, ML, Rosenzweig, C, Iglesias, A, Livermore, M and Fischer, G (2004). ‘Effects of Climate Change on Global Food Production under

SRES Emissions and Socio-economic Scenarios’, Global Environmental Change 14: 53–67.
Porter, JR and Gawith, M (1999) ‘Temperatures and the Growth and Development of Wheat: A Review’, European Journal of Agronomy 10: 23–36.
Sanghi, A and Mendelsohn, R (2008) ‘The Impacts of Global Warming on Farmers in Brazil and India’, Global Environmental Change 18: 655–665.
Schlenker, W, Hanemann, WM and Fisher, AC (2005) ‘Will US Agriculture Really Benefit from Global Warming? Accounting for Irrigation

in the Hedonic Approach’, American Economic Review 95: 395–406.
Schlenker, W, Hanemann, WM and Fisher, AC (2006) ‘The Impact of Global Warming on US Agriculture: An Econometric Analysis of

Optimal Growing Conditions’, Review of Economics and Statistics 88: 113–125.
Wheeler, TR, Craufurd, PQ, Ellis, RH, Porter, JR and Vara Prasad, PV (2000) ‘Temperature Variability and the Yield of Annual Crops’,

Agriculture, Ecosystems and Environment 82: 159–167.
Winters, P, Murgai, R, Sadoulet, E, De Janvry, A and Frisvold, G (1998) ‘Economic and Welfare Impacts of Climate Change on Developing

Countries’, Environmental and Resource Economics 12: 1–24.
World Bank (2008) World Development Report; Agriculture for Development. Washington DC: The World Bank.

Assessing the costs of adaptation to climate change

3.1 Introduction

One of the largest impacts of climate change is likely to be on water resources and their

management. While there have been many studies of hydrological impacts at catchment, regional

and global scales, there have been very few published studies on the costs of potential adaptation

options at any scale.
The UNFCCC report (UNFCCC, 2007) provides the first global-scale estimates of the potential

costs of adaptation. The headline conclusion is that the additional investment and financial flows

required for adaptation to potential changes in the availability of water supplies would be

approximately $9–11 billion per year in 2030, approximately 85% of which would be needed in

non-Annex 1 Parties.

The UNFCCC report suggests that this is the same order of magnitude as

the additional investment required to meet the Millennium Development Goal targets for

sustainable access to safe drinking water and basic sanitation.
However, there are grounds (outlined below) for expecting that the UNFCCC figure is an under-

estimate of likely potential costs of adaptation to water supply. The figure does not include an

allowance for costs of adapting in other aspects of water resources management, such as managing

increased flood risk, maintaining water quality standards and supporting instream economic and

environmental uses. The figure does not include any additional costs of residual impacts of

climate change, caused by events above the design standard, and these costs may be very

substantial in practice because adaptation will be neither perfect nor completed on time. Also, the

figure of $9–11 billion/year represents the additional cost of maintaining a defined level of service

in the presence of climate change. It does not include the cost of providing this level of service

where it does not currently exist. The UNFCCC report estimates that the total cost (to 2030) of

providing a specific level of service globally – thus reducing the development deficit – varies

between $639 billion and $797 billion, depending on assumed economic and population growth,

corresponding to an annual average of approximately $32–40 billion (assuming investment is

over a 20-year period).
This chapter provides a critique of the UNFCCC estimates, but first examines the conceptual and

methodological issues around the estimation of adaptation costs and residual damages in the

water sector. It also reviews the (limited) literature.

3.2 Issues and assumptions

The ‘water sector’ is very diverse, and climate change is anticipated to impact upon many

activities. These activities include: the supply of safe water to domestic, industrial and agricultural

consumers (including for irrigation); the provision of sanitation and the removal and treatment of

effluent; support of navigation; management of flood hazard (from drains, rivers, groundwater,

overland flow and so on); measures to provide protection or reduce exposure; generation of

hydropower; and the management of river flows and water levels to support agriculture, recreation

and the provision of ecosystem services (such as support for instream and riverine ecosystems).
Estimates of the total current expenditure on water infrastructure are uncertain; one estimate

(Briscoe, 1999) gives a total annual expenditure of $65 billion in developing countries for

Assessing the costs of adaptation to climate change

hydropower ($15 billion), water supply and sanitation ($25 billion) and irrigation and drainage

($25 billion), but this excludes flood management costs. Table 3.1 summarises the potential

economic and non-economic consequences of the effect of climate change on the water sector,

assuming no explicit adaptive response to climate change.

Table 3.1 Potential consequences of climate change on the water sector, without adaptation

Activity

Potential economic consequences

Potential non-economic
consequences

Water supply

Domestic/
municipal

Cost of altered health
Cost of dealing with droughts

Disruption to established
patterns of use

Industrial

Cost of change in industrial productivity

Disruption and uncertainty

Agricultural
(including
irrigation)

Cost of change in agricultural
productivity

Uncertainty

Sanitation and
effluent removal

Cost of altered health
Cost of impacts on instream ecosystems
(e.g. fisheries)
Cost of dealing with pollution incidents

Navigation

Cost of altered navigation opportunities

Flood
management

Change in economic value of flood
damages (direct and indirect)
Change in economic value of injury
and ill health

Disruption and anxiety

Hydropower

Cost of change in generation potential

Recreation

Cost of changes in recreational
opportunities

Change to cultural value of
the water environment

Water level
and soil water
management

Change in habitat
characteristics

Ecosystem
services

Change in instream and
riverine habitats and species

Assessing the costs of adaptation to climate change

In practice, of course, water users and managers will adapt to a changing climate – possibly

inappropriately and probably reactively – and so the actual cost of climate change will not be

equal to the simple costs of impacts outlined in Table 3.1. For example, water supply companies

are likely to introduce a variety of measures to try to maintain the security of supplies to

customers, and consumers exposed to water shortage will seek alternative sources of supply.

Managers of a wetland will alter pumping regimes to seek to maintain target water levels.
Measures taken to respond to other pressures on the water environment will also affect the

impacts of climate change. If these other pressures and the responses to them are ignored for the

moment, then the ‘cost’ of climate change for a given activity in the water sector is:

Cost

cost of explicit adaptation measures

residual impacts of climate change

transaction costs of implementing adaptation measures

The cost of explicit adaptation measures here represents the cost of measures explicitly introduced,

by all interested parties (consumers and providers), to cope with climate change, and does not

include the cost of measures introduced to meet other challenges which incidentally help

adaptation to climate change. In practice, it is likely that in many cases adaptation measures will

be developed to meet a variety of challenges, and it may be difficult to separate out the portion

associated explicitly with climate change.
Adaptation will not remove all the consequences of climate change, and there will be residual

impacts. These impacts will arise because adaptation may lag behind the changing climate, or

because the adaptation measures introduced do not cope with the change in climate that actually

occurs (due to imperfect knowledge). Impacts will also occur, as in the absence of climate change,

in events which exceed the design standards or provisions of the adaptation measure; the net effect

of climate change is the difference between such residual impacts with and without climate

change. These residual impacts will be difficult to quantify, as they vary across different parts of

the water sector. They are relatively easy to characterise in flood management – they are the costs

incurred in damaging events which are not prevented – but in other parts of the sector the residual

impacts take the form of lost productivity (agriculture, industry and power generation),

inconvenience and exposure to water-related disease and ill health. Residual impacts also include

the cost of emergency actions taken in response to an event, including the cost of implementing

drought management plans and measures.
Transaction costs are the costs associated with making changes to policies and practices in the

face of potential climate change. They include R&D costs and the costs of refining policies or

reviewing decisions; these costs will be incurred even if decisions are subsequently made not to

adapt to climate change.
The objectives of adaptation will influence the costs of adaptation and the residual impacts. In the

most general terms, adaptation can aim at: (i) maintaining a given standard of service, (ii)

achieving a new ‘optimum’ standard of service, or (iii) meeting some new service standard. This

new service standard could be higher – because for example the threat of climate change increases

risk aversion – or could be lower because of financial or feasibility constraints. These objectives

Assessing the costs of adaptation to climate change

will vary from place to place and activity to activity. In many parts of the water sector, standards

are set by some form of regulation (design standards for flood defence, for example, or reliability

standards for the provision of drinking water or the discharge of effluent), but in other parts ‘levels

of service’ are determined locally.
In practice, many parts of the world do not enjoy ‘acceptable’ standards of service, even in the

absence of climate change. For example, a billion people currently lack access to safe drinking

water, and 2.4 billion lack access to basic sanitation. A key conceptual question in assessing the

costs of adaptation is therefore: should the costs of adaptation to climate change be (i) the cost of

maintaining current standards of service, even where they are currently inadequate, or (ii) the cost

of providing a defined standard of service? Under the first rule, costs of adaptation would be

estimated to be extremely low where, for various historical reasons, standards of service are low (the

‘development deficit’ is high); under the second rule, the costs could be very high. The resolution of

this conundrum requires an evaluation of ethical issues associated with the adaptation deficit and

historical responsibilities both for climate change and for economic development.

3.3 Methodologies: approaches

to costing

The previous section highlighted the issues associated with estimating the actual and potential

costs of adaptation in the water sector. In principle, the equation above can be used both to

characterise the actual (realised) cost of impacts and adaptation in the water sector, and to

estimate in advance what these costs may be. In practice, of course, analysts are most interested in

projections of future adaptation and impact costs.
It is relatively easy (in principle) to estimate future adaptation and residual damage costs at the

water management system or scheme level. The simplest approach is first to estimate (over the next

few decades) capital and operating costs (over all actors) plus residual damages, in the absence of

climate change (but including other changes, such as the effects of changes in demand or exposure),

and then repeat the calculations assuming a climate change trend and a management strategy

specifically designed to cope with this trend. In circumstances where a development deficit is

projected to persist, the reference ‘no climate change’ case can represent continued inadequate

adaptation to climatic variability; alternatively, the reference case can include improvements to

meet enhanced standards. The choice does not affect the estimation methodology.
This simplest approach assumes perfect knowledge: the water managers know exactly how climate

will change, and plan in advance, keeping pace with the changing climate. A variation on this

approach assumes ‘perfect, but delayed’ knowledge, whereby there is a lag between change in

climate and the management response, with subsequent rescheduling of adaptation costs but an

increase in residual impacts.
A more complicated approach recognises that water managers do not know how climate will

change, and therefore cannot plan perfectly. One way of addressing this is to assume that a

specific adaptation decision is made (perhaps after a lag), and then estimate costs and residual

impacts under a range of possible future climates. By assigning likelihoods to these possible

climate outcomes,

it is then possible to estimate the expected costs and impacts under the defined

Assessing the costs of adaptation to climate change

adaptation action. By repeating the analysis with alternative adaptation decisions it is possible to

identify a range of possible costs and residual impacts (and also identify the action which copes

‘best’, using some defined criteria, with climate change uncertainty). However, while it is possible

to construct an experimental design which enables the estimation of the costs of adaptation and

residual impacts at the scheme or management-unit level (assuming varying degrees of perfect

knowledge), it is rather more complicated to estimate costs across a region or indeed the globe.
Such costs, aggregated across a large global domain, could in principle be estimated by: (i)

extrapolating from a small number of case studies (bottom-up), or (ii) taking a (top-down) large-

scale perspective and estimating at the same time costs and impacts across multiple locations.

Unfortunately, adaptation is very locally specific. In most cases there are multiple adaptation

options, which vary with local geographical, financial, institutional and socio-economic

circumstances. Extrapolation from a small number of case studies could therefore be very

misleading (and the studies would be likely to use different climate scenarios), and it is impossible

in practice to take a realistic top-down approach. Extrapolation has been used, however, to

construct climate response functions relating impact (and implicitly adaptation costs) to climate

forcing (e.g. Mendelsohn et al., 2000).
Estimation of the cost of adaptation and residual damages across a large geographic domain

therefore needs to adopt a different approach from the estimation of costs and impacts at the

system scale. One approach could be to base estimates on current and projected expenditure on

management infrastructure, assuming that an additional fraction represents the additional cost

of climate change. However, such an approach under-estimates the cost of adaptation where

expenditure is currently low (where the adaptation deficit is highest), and does not account for

the residual damages.
An alternative is therefore to take a top-down approach using a suite of generic indicative

adaptation options, and estimating changes in hydrological characteristics using some form of

macro-scale hydrological model. For example, the costs of adaptation to changes in water supply

availability due to climate change could be indexed by the costs of providing additional storage

capacity to maintain supply reliability. The costs of adapting to altered flood risk could be

characterised by estimating the costs of providing flood protection to a target standard of service.

In both cases, costs would be estimated by applying generalised cost functions (dollars per

megalitre of storage, for example, or dollars per metre of flood protection embankment). Such an

approach would give an indication of the potential magnitude of adaptation and residual damage

costs in a consistent way, although would not be precisely accurate.

3.4 Case studies at the system scale

Very few estimates of the cost of adaptation to climate change in the water sector at the system

level have been published. Many estimates are likely to have been made – as they have been for

water supply in England and Wales, for example – but in most cases reports are confidential for

business reasons.
Among the few published examples, Zhu et al. (2007) used a dynamic programming method to

estimate the optimum floodplain adaptation strategy in Sacramento, California, under different

Assessing the costs of adaptation to climate change

climate and urbanisation assumptions. With no urbanisation, they estimated the present value

(over 150 years) of flood control costs and residual damages to be $392 million in the absence of

climate change, and $485 million under a climate change scenario, giving an additional cost

(present value) of $93 million due to climate change. With a high urbanisation rate, the

corresponding costs are $828 million without and $1031 million with climate change, giving a

cost of $203 million due to climate change.

The difference between the two sets of estimates

emphasises the role of additional drivers in influencing the costs of adaptation. Zhu et al. (2007)

noted that their study assumed perfect foreknowledge of climate change trends.
Medellin-Azuara et al. (2008) used a similar dynamic programming method to identify optimum

water supply management measures under current and future climatic conditions (as

characterised by one climate model), assuming perfect foreknowledge. They found that the

scenario considered did not require any major capital investments, but led to increases by 2050 in

annual operating costs of $369 million per year (essentially the cost of adapting pumping and

treatment regimes to altered volumes and timing of flows). They also calculated that water scarcity

costs – primarily in terms of agricultural impacts – would increase by $121 million per year; these

are the residual impacts left over after adaptation. In this example, the residual impacts are

therefore approximately a third of the adaptation costs, although it is not self-evident how general

this ratio is.
These studies did not attempt to extrapolate from the system that was investigated. A study of the

potential costs of adaptation to altered water quality and stormwater flood risks in England and

Wales (ICF, 2007) sought to estimate national costs using a small number of case studies

(bottom-up generalisation, as outlined in Section 3.3 above). This concluded that the costs of

adapting to altered stormwater flood risks – as characterised by increasing storage requirements –

ranged between £0.9 billion and £1.1 billion per year (at discount rates of 5–6% and over 40

years), equivalent to approximately 25% of current stormwater drainage expenditure. On the

other hand, the costs of increasing effluent quality to maintain water quality standards was only

between £4 million and £25 million per year, depending on standards and change in river flows.
Kirshen et al. (2005) described a methodology to estimate the costs of changes in water supply in

China under climate change, using a variation on the top-down methodology introduced in the

previous section. Runoff was simulated in each major basin in China under current and two

future climates, and the costs of additional storage and groundwater development necessary to

maintain target yields were estimated using generalised cost functions relating dollars to units of

storage. The study assumed perfect foreknowledge. Although summary cost data were not

presented, in the Huang Ho River it was calculated that annualised costs (50-year time horizon

and 3% discount rate) of meeting present demand would increase from approximately $200

million by the 2050s with no climate change, to approximately $700 million under one climate

change scenario (a climate change effect of approximately $500 million); under another scenario,

it proved impossible to meet present demand.

Assessing the costs of adaptation to climate change

3.5 Estimates of the costs of

adaptation at the global scale

There have been two global-scale assessments of the potential cost of climate change in the water

sector, one examining water supply to domestic, industrial and agricultural consumers (Kirshen,

2007, and as subsequently revised to provide the basis for the estimate in the UNFCCC report

(UNFCCC, 2007)), and one focusing on the altered costs of providing irrigation water (Fischer et

al., 2007).
The Kirshen/UNFCCC study assumed a generic set of adaptation responses to water shortage,

and used generalised cost functions to estimate the costs of these adaptation responses. Estimates

of the need for adaptation were made by changing national water resource availability in accord

with large-scale scenarios for change in national rainfall, derived from a set of scenarios based on

climate models used in the IPCC AR4 (they did not use hydrological models to estimate change

in generic reservoir performance). The study compared resource availability with demand for

water from domestic, municipal, industrial and agricultural consumers, plus aquatic ecosystems.

Resource availability included both surface and groundwater resources, and the study used a

simple prioritisation procedure to determine, for each country, a feasible combination of supply-

side measures (new reservoirs or groundwater abstraction, and desalination) and demand-side

measures (increased water-use efficiency). If it was not possible to meet demands from all possible

supplies, then it was assumed irrigation demands were left unmet – a measure of residual damage.

The study concluded that the cost of adaptation to both population and climate change in the

water sector, across the globe, totalled $531 billion to 2030 (under an A1b scenario), including

both economic and climate change. This figure was increased to $898 billion in the UNFCCC

report to account for use of more expensive sites and unmet irrigation demands. Approximately a

quarter of this is assumed to be specifically for climate change (UNFCCC, 2007) – based on a

case study in West Africa which explicitly calculated costs with and without climate change –

giving a cost of approximately $225 billion to 2030, or $11 billion per year.
This study represents a coherent attempt to estimate costs of increasing water supply capacity,

using best-available (at the time) estimates of the cost of generic adaptation measures; it takes

the top-down approach outlined in the previous section. However, the total is likely to be a

considerable under-estimate, for several reasons (most acknowledged in Kirshen, 2007). The

study worked at the national level, and assumed that water resources would be transferred within

a country from areas of surplus to areas of deficit. For small countries, this is not problematic; for

large and diverse countries, however, this is unrealistic. For example, although water resource

availability at the national scale in China may not vary much with climate change, large areas of

China are likely to suffer reduced availability. As noted above, Kirshen et al. (2005) estimated

that annual costs of adapting to climate change in just the Huang Ho basin could be $0.5 billion

by 2050 – in other words, costs in just one river basin account for approximately 5% of the

estimated global costs. The study used an empirical relationship between annual runoff and its

variability, and reservoir storage capacity (McMahon et al., 2007). The relationship cannot

incorporate the effects of changes in the timing of runoff through the year due to climate change,

which would also be expected to alter the storage capacity required to maintain a target yield

and reliability.

Assessing the costs of adaptation to climate change

The use of an average climate-change scenario, rather than individual scenarios, may give a biased

– probably downwards – estimate of the costs of climate change. The mean response from several

climate scenarios is likely to be higher than the response from the mean of the scenarios, because

the loss function is asymmetrical: a reduction in availability is worth more than an increase in

availability. The study also did not consider the residual damages: it assumed perfect adaptation to

maintain a notional standard of service, so under the study assumptions arguably these residual

damages could be assumed to remain approximately constant. However, because adaptation will

be imperfect, residual damages would not be zero in practice.
Changes in operating costs may also be substantial (in the California example above, annual

operating costs increased by $0.4 billion/year due to climate change), and are not included

because the focus of the study was on investment needs. In the opposite direction, the study

focused on the costs of augmenting supplies; demand-side measures to maintain a reliable

supply–demand balance are typically cheaper than supply-side measures. Finally, the Kirshen

study provided an estimate of the public investment costs (by water suppliers, not users) to meet

socio-economic and climate changes by 2050. The UNFCCC report assumed that 25% of these

costs represents the fraction attributable to climate change alone; the basis for this assumption is

not clear.
It is not possible to assess the extent of potential under-estimation of the costs of maintaining

water supplies without undertaking a more detailed analysis taking into account multiple

scenarios and working at the finer spatial scale. It is also important to emphasise that the

adaptation-cost figure characterises only the cost of adapting to water shortage. It does not

include costs of flood management, storm drainage, water quality enhancement, hydropower

generation, navigation or maintaining ecosystem services.
The irrigation study (Fischer et al., 2007) simulated future irrigation demands without climate

change and under two climate models and two emissions scenarios (representing ‘no policy’

(SRES A2) and ‘mitigation’ (SRES B1)), using the FAO agro-ecological zones model. Again, they

used a top-down methodology similar in principle to that outlined above. By 2080, unmitigated

climate change would increase the cost of providing additional irrigation – in terms of capital

infrastructure and operating costs – by $24–27 billion per year; the mitigation scenario reduced

these costs by $8–10 billion per year. The ‘unmitigated’ adaptation costs are considerably higher

than the approximately $11 billion calculated by Kirshen/UNFCCC for water supply as a whole,

but the Fischer et al. figure relates to 2080 rather than spend to 2030.
There is currently no global-scale information on the relative costs of adaptation among the

different components of water resources management. However, it is highly likely that costs of

storm and river flood management will be very significant (globally), even if costs of water quality

management, navigation enhancement and ecosystem protection are not large relative to the costs

of maintaining water supplies.

Assessing the costs of adaptation to climate change

3.6 Conclusions

The UNFCCC estimates apply a broadly robust methodology to estimate the potential costs of

meeting the challenges to water supplies posed by climate change. However, the application of the

methodology – given the time available for the study – necessitated a few approximations

(outlined above) which are likely to mean that the headline global figure of $9–11 billion/year

additional public investment requirements to cope with climate change is an under-estimate of

the costs of adaptation. The residual impacts not covered by adaptation are likely to be high,

although again unquantified, largely because adaptation will be imperfect and lagged. The figure

also represents the cost of adapting to climate change, assuming no adaptation deficit.

1 Annex I countries are the 36 industrialised countries and Economies in Transition that have accepted emission ‘caps’ under the UNFCCC

that limit their total greenhouse-gas emissions within a designated timeframe.

2 This is non-trivial, and potentially problematic. There is the practical problem that it can be difficult to define credibly the probability

distribution of future impacts, because this distribution may be very much influenced by assumptions about distributions of the driving

causes of uncertainty. Conceptual problems arise because it is not possible to assign likelihoods to some aspects of uncertainty (such as future

emissions).

3 They do not publish the breakdown between adaptation costs and residual damages.

Assessing the costs of adaptation to climate change

References

Agrawala, S., Crick, F., Jette-Nantal, S. and Tepes, A. (2008) ‘Empirical estimates of adaptation costs and benefits: a critical assessment’. In

Agrawala, S. and Fankhauser, S. (eds) Economic Aspects of Adaptation to Climate Change. OECD.
Briscoe, J. (1999) ‘The financing of hydropower, irrigation and water supply infrastructure in developing countries’, International Journal of

Water Resources Development 15 (4): 459–491.
Fischer, G. et al. (2007) ‘Climate change impacts on irrigation water requirements: effects of mitigation, 1990–2080’, Technological Forecasting

and Social Change 74(7): 1083–1107.
ICF (2007) The Potential Costs of Climate Change Adaptation for the Water Industry. Report to Environment Agency.
Kirshen, P. (2007) Adaptation Options and Cost in Water Supply. Report to UNFCCC Secretariat Financial and Technical Support Division.
Kirshen, P., McCluskey, M., Vogel, R. and Strzepek, K. (2005) ‘Global analysis of changes in water supply yields and costs under climate

change: a case study in China’, Climatic Change 68(3): 303–330.
McMahon, T.A., Pegram, G.G.S., Vogel, R.M. and Peel, M.C. (2007) ‘Revisiting reservoir storage-yield relationships using a global

streamflow database’, Advances in Water Resources 30: 1858–1872.
Medellin-Azuara, J. et al. (2008) ‘Adaptability and adaptations of California’s water supply system to dry climate warming’, Climatic Change

87: S75–S90.
Mendelsohn, R., Morrison, W., Schlesinger, M.E. and Andronova, N.G. (2000) ‘Country-specific market impacts of climate change’, Climatic

Change 45: 553–569.
UNFCCC (2007) Investment and Financial Flows to Address Climate Change. Climate Change Secretariat, Bonn.
Zhu, T. et al. (2007) ‘Climate change, urbanisation and optimal long-term floodplain protection’, Water Resources Research 43, W06421, doi:

10.1029/2004WR003516.

Assessing the costs of adaptation to climate change

Summary

• All countries will need to implement measures to reduce or avoid the additional impacts on

health due to global climate change.

• The magnitudes of costs of adaptation in the health sector are unknown but potentially large.

Not implementing additional adaptation will be even more expensive in terms of the burden of

additional injuries, illnesses and deaths on society.

• The UNFCCC has estimated costs of $4–12 billion for adaptation in the health sector in

developing countries. These costs represent the costs of preventing additional cases of

malnutrition, malaria and diarrhoeal disease due to climate change by 2030.

• The UNFCCC costs do not consider the full range of climate futures or the full range of

disease/health outcomes that will be affected by climate change. On balance, the UNFCCC

costs are likely to be an under-estimate of the full costs of adaptation in the health sector in

developing countries.

• High-income countries will also bear costs to adapt to climate change, including the high costs

associated with adaptation of health systems infrastructure.

• The current lack of investment in public health is a considerable barrier to adaptation. However,

investment alone is not sufficient to improve health in developing countries, as many other

barriers remain, such as poor governance, inequality and low adaptive capacity.

4.1 Introduction and scope

The potential impacts of climate change on population health include a wide range of diseases and

health outcomes, from infectious diseases to malnutrition and disaster-related injuries

(Confalonieri et al., 2007). Adaptation, broadly defined, would include all activities that reduce

or prevent these ‘additional’ cases or deaths. Several reviews of such adaptation strategies, policies

and measures have now been published (Menne and Ebi, 2005; Ebi, Kovats and Menne, 2006).

Further, initiatives are now underway to support adaptation in the health sector in low- and

middle-income countries.
This chapter provides an overview of the current literature on adaptation costs for the ‘health

sector’. The health sector is here limited to conventional public health activities, although it is well

recognised that adaptation in other sectors is probably more important for reducing the health

impacts of climate change (through disaster mitigation, food and water security, and providing

decent infrastructure). At the time of writing (July 2009), there is only one set of comprehensive

adaptation costs for health and these are provided in the UNFCCC report and related journal

paper (UNFCCC, 2007; Ebi, 2008).

Assessing the costs of adaptation to climate change

Health ‘adaptation costs’ include:
• costs of improving or modifying health protection systems to address climate change, for

example, expanding health or vector surveillance systems – this includes the costs associated

with building new infrastructure, training new health care workers, and increasing laboratory

and other capacities

• costs of introducing novel health interventions (e.g. heat-wave warning systems)
• additional costs for meeting environmental and health regulatory standards (e.g. air quality

standards, water quality standards)

• costs of improving or modifying health systems infrastructure, for example adapting hospitals

to hotter summers

• occupational health costs, for example measures to prevent the adverse impacts of increased heat

load on the health and productivity of workers

• costs of health research on reducing the impact of climate change, for example evaluation studies
• costs of preventing the additional cases of disease due to climate change as estimated by

scenario-driven impact models.

This last type could also be considered as ‘damage’ or impact costs – rather than adaptation costs

(OECD, 2008). The health costs presented in the UNFCCC report are of this type (UNFCCC,

2007). Given that many of the relevant preventive activities are outside the remit of the health

sector, and are currently not implemented, a key role for the health sector will be to simply prevent

or treat the additional disease cases caused by climate change.

4.2 The UNFCCC estimates for health

adaptation costs

The UNFCCC report estimated the adaptation costs for the health sector to be in the range of

$4–12 billion per year in 2030 (UNFCCC, 2007). The adaptation costs are for preventing the

additional climate-change-related cases of diarrhoeal disease, malnutrition and malaria in 2030.

(These are discussed in more detail below.) The total costs were estimated by taking the number of

additional cases and multiplying them by the costs of prevention per child, obtained from the

Disease Control Priorities in Developing Countries project (World Bank, 2006). Thus,

uncertainties relate to both the estimates of the future burden and the method for costing.
The prevention activities were based on currently deployed interventions and are listed in Table

4.1. The costs did not include the cost of implementing programmes, including health care

personnel costs or infrastructure costs. The costs of initiating programmes in new areas can be

significant. Further, such costs would occur with a shift in disease distribution, even if there was

no net increase in the number of cases.

Assessing the costs of adaptation to climate change

Table 4.1 Health ‘intervention sets’ included in the UNFCCC report

Outcome

Intervention sets

Average cost per
child ($), 2001

Diarrhoeal disease

Breastfeeding promotion, rotavirus and
cholera immunisation

15.09

Improvement in water supply and sanitation

53.00

Malaria

Insecticide-treated bed-nets plus case
management with artemesin-based
combination therapy

88.50

Indoor residual insecticide spraying plus above

123.05

Malnutrition

Breastfeeding promotion, child survival
programmes, nutritional programmes,
growth monitoring

17.40 to 23.09

As Table 4.1 shows, the costs of interventions for child health, while requiring large investments

to meet the MDGs, result in costs per child that are relatively low. Water and sanitation

interventions are highly cost-effective (Hutton, Haller and Bartram, 2007). Further,

improvements in water supply and sanitation, and other environmental interventions, have

multiple benefits (e.g. for education and welfare), not all of which are accounted for.
The costs of the nutrition interventions are likely to be under-estimated as they address only a

narrow range of public health measures. The annual per capita cost of providing food to improve

child health in Africa has been estimated to be much higher (Edejer et al., 2005). Ebi (2008)

estimated that including such costs would increase the adaptation costs more than 10-fold. The

conservative approach used also avoids any double counting with the agriculture sector

adaptation costs that include the cost of feeding more people (see Chapter 2, Wheeler and Tiffin,

in this report).

Assessing the costs of adaptation to climate change

Table 4.2 Projected costs in 2030 to manage climate change-related cases of diarrhoeal
diseases, malnutrition, and malaria for three climate scenarios (million US$)

Emissions
scenario

Diarrhoeal

diseases

Malnutrition

Malaria

Middle

High

Middle

High

Middle

High

Stabilisation
of emissions
at 550ppm
CO

equivalent

1706

6024

53.9 – 71.5

112.9 – 149.9

1573 – 2145

3236 – 4515

Stabilisation
of emissions
at 750ppm
CO

equivalent

1983

6814

81.3 – 107.9 162.5 – 215.6

1928 – 2691

3994 – 5573

Unmitigated
emissions

2731

9010

62.2 – 82.6

125.2 – 166.2 3059 – 4269

6293 – 8781

Note: the three climate scenarios are relative to baseline climate. High, middle and low costs (low estimates not shown)

reflect uncertainty around the relative risk of increased cases due to climate change

Source: Ebi (2008)

The analysis makes a number of necessary but unlikely assumptions. The number of annual cases

of diarrhoeal disease, malaria and malnutrition is assumed to remain the same over time.

Population growth is projected to increase under the medium variant from 6.1 billion in 2000 to

8.3 billion in 2030. Using the current number of cases in the analysis assumes that incidence will

decrease as population increases, without attribution of the possible reasons for such a decline. If

disease incidence rates remain constant, or decline less rapidly, until 2030, then the number of

cases attributed to climate change would increase in these estimates (and therefore the costs

would increase).
The estimates also assume that the cost of the interventions remain constant from 2000 to 2030.

There are further uncertainties about the unit cost, which varies significantly between countries

(World Bank, 2006). It should also be noted that new interventions may be developed in the next

few years (e.g. a vaccine for malaria) or current ones may become ineffective (e.g. through

widespread insecticide resistance in mosquitoes). The health interventions in Table 4.1 are not

100% effective (based on research on both efficacy and compliance) (Edejer et al, 2005).

Therefore, they will not prevent all the additional cases attributed to climate change, and there

will be some ‘residual’ health impacts.

Assessing the costs of adaptation to climate change

4.3 The burden of disease due to

climate change

The estimated additional cases are derived from the WHO Global Burden of Disease (GBD)

project (Lopez et al., 2006; McMichael et al., 2004). The limited number of health outcomes in

this assessment was due to the lack of exposure-response relationships and other information

needed to estimate future health burdens due to climate change.
The estimates of future health impacts are dependent upon the future worlds (development

pathways), which makes presentation of costs difficult as they have to be contextualised within

specific sets of scenarios and assumptions. The GBD estimates were derived from three emissions

scenarios (Table 4.2), that included two stabilisation scenarios (Arnell et al., 2002) and one

business-as-usual scenario that corresponds to the IPCC IS92a emissions scenario.
The estimates for additional cases of diarrhoeal disease were based on the published evidence for

the direct effect of temperature on hospital admissions for diarrhoeal disease. No estimates were

available for the impact of changes in rainfall patterns on diarrhoeal disease because of the lack of

published evidence – and the complexity of the system. Similarly, the estimates for malnutrition

cases are based on large-scale crop modelling that does not include extreme events such as severe

drought (Parry et al., 1999). The malnutrition cases make the biggest contribution to the climate-

change-attributable burden of disease, but are also the most uncertain.
The estimates for additional cases incorporate some assumptions regarding adaptation

(Campbell-Lendrum and Woodruff, 2006; McMichael et al., 2004). For example, the estimates

for malnutrition incorporate the assumptions for technological adaptation in the agricultural

sector, and the estimates for diarrhoeal diseases assume that the disease is reduced as countries get

richer. The original authors of the WHO study estimated that the three outcomes represent

approximately 30–50% of the total burden attributable to climate change in low- and middle-

income countries in 2030 (McMichael, Bertollini, 2009, personal communication).

4.4 The ‘adaptation deficit’

The current burden of climate-sensitive disease is high (Ezzati et al., 2005; Mills and Shillcutt,

2004). The ‘adaptation deficit’ is therefore a key issue in health impacts as the greatest burden of

climate-sensitive diseases is in low-income countries. The current ‘level of health service provision’

is inadequate in many countries. The target levels of provision are either seen as what is feasible

(e.g. Millennium Development Goals) or what is ideal (e.g. Health for All). As well as the moral

arguments for improving health, there are also economic ones. The WHO established its

Commission on Macroeconomics and Health to provide the evidence that poor health impedes

economic development (Sachs, 2001). Despite recent improvements in bilateral aid, the

development of targeted programmes (e.g. Roll Back Malaria) and the new philanthropic

ventures (the Global Fund, Bill and Melinda Gates Foundation), many countries still lack the

investment needed to achieve the health-related Millennium Development Goals (Mills and

Shillcutt, 2004).

Assessing the costs of adaptation to climate change

As discussed in Chapter 1 of this report (Fankhauser), there is often no clear distinction between

strategies for development and strategies for adaptation. Many current public health and disease

control measures, if implemented successfully, will reduce the impact of climate change on

human health. This message has recently been reinforced by the World Health Organization

which recommended that countries should strengthen their health systems as well as integrating

health measures into plans for adaptation to climate change (WHO Executive Board, 2008).

4.5 Costing health: methods

The ‘burden of disease’ is the term used to describe the total impact of disease or health condition

in a population, including deaths, cases and years lived with disability (for chronic diseases). The

metrics used in environment and health decision making include: deaths, DALYS (disability-

adjusted life years), QALYs (quality-adjusted life years) (Hofstetter and Hammitt, 2001; Mathers

et al., 2003). Methods of cost-benefit analysis (CBA) and cost-effectiveness analysis (CEA) are not

generally applied to health issues at the global scale but are used to assess the benefits of alternative

policy choices at the local scale. The main criticism of the costing method in the UNFCCC report

is that micro-methods (which are very context-specific) are being applied inappropriately to a

global problem.
Health has been incorporated into some integrated assessment models (IAMs) as a damage

function or an impairment to economic productivity (Bosello, Roson and Tol, 2006; Tol, 2002).

Health costs, when estimated, make a considerable contribution to overall damage costs, as the

statistical value of a life is high. In general, the IAMs have incorporated into their damage

estimates only the very basic health models for heat/cold effects on mortality and malaria.
There have been several reviews of the costs of health interventions, where cost-effectiveness is the

main method used. The Disease Control Priorities in Developing Countries project has evaluated

the scientific and economic evidence on the available interventions for all the major infectious and

chronic (non-communicable) diseases (World Bank, 2006). Thus, there have been extensive

reviews of the cost-effectiveness of health interventions relevant to climate change in the

following areas:
• malaria

• malnutrition

• diarrhoeal disease

• food safety.
However, there is still considerable uncertainty around the cost-effectiveness of these

interventions. Some have been relatively well researched, e.g. insecticide-treated bed-nets, but

for some there is less information (e.g. indoor residual insecticide spraying) (Mills and Shillcutt,

2004). For some interventions, such as heat-health warning systems, there is practically no robust

information on their effectiveness (Kovats and Hajat, 2008). It should also be noted that health

adaptation costs in high-income countries are not included in the UNFCCC costs. Such costs

would be higher due to the higher infrastructure and labour costs. In addition, the thresholds

for intervention (costs per case prevented) are higher.

Assessing the costs of adaptation to climate change

4.6 Case studies

There have been increasing numbers of studies that quantify climate change impacts at the

national or local level, but very few estimate the costs of these health effects, either as direct health

costs (deaths, welfare costs) or in terms of health service usage. However, in most assessments

some consideration of adaptation is included. For example, the PESETA study monetised the

impact of climate change on heat- and cold-related mortality and on food poisoning in Europe,

assuming some acclimatisation (Watkiss et al., 2007). At the time of writing (July 2009) there are

no comprehensive ‘global’ studies of the health impacts or adaptation costs.
No case studies of health adaptation costs have been published in high-income countries.

However, many countries, including the UK, have implemented health adaptation measures in

the form of heat-wave plans (Department of Health, 2008). The costs of these plans can be

estimated. Heat early-warning systems are relatively inexpensive, unless they include active

measures that are implemented following an alert. The reported costs for European countries

range from €200,000 to €6 million per year (WHO Regional Office for Europe, 2008).

Structural interventions are more expensive. France spent more than €150 million in 2004 on

providing additional staff and cool rooms in residential homes for the elderly (Michelon, Magne

and Simon-Delaville, 2005). Mortality in this group doubled during the 2003 heat wave.

Estimates for the costs of adapting dwellings in London to hotter summers are discussed in

Section 6.5 in this report.
In low- and middle-income countries, a recent review for the World Health Organization found

very few examples of studies estimating the costs of adaptation (Markandya and Chiabai, 2008).

An unpublished study in South Africa quantified adaptation costs as the prevention costs of the

additional burden of malaria cases due to climate change to 2025 (van Rensburg and Blignaut,

2002). However, several such case studies are currently being undertaken in African and

Asian countries.
Health projects are now being included in the National Adaptation Programmes of Action

(NAPAs) but they typically address current disease control issues, usually for malaria. Financial

requests for specific projects range from $22,000 to $600,000 for disease-control, early-warning

information systems to $7 million for a major project in Bangladesh. Environmental projects such

as rainwater harvesting and improving food security are not included as health projects although

they will have health benefits. Oxfam selected some of the NAPA projects (including one form

Samoa on developing climate-based health early-warning systems) in order to scale up the

adaptation costs (Oxfam, 2007). As discussed in Chapter 1 of this report (Fankhauser), the

NAPA projects capture only a very small part of adaptation needs and adaptation costs.

Assessing the costs of adaptation to climate change

4.7 Conclusions

In conclusion, the UNFCCC costs are an under-estimation of the total ‘health sector’ costs

because of all the activities, diseases and countries that are not included in these estimates. The

UNFCCC costs are for impacts that consider incremental improvements in health and economic

development, but not explicit measures that might be implemented in response to the threat of

climate change.
The methods used to generate the UNFCCC health costs were based on the best available

information at the time for developing countries. This evidence base is still very limited and it is

difficult to see how a more comprehensive study could have been undertaken in the short

timeframe available. There are many uncertainties in the UNFCCC health costs, and the most

important of these are summarised in Table 4.3.

Table 4.3 Summary of uncertainties in the UNFCCC adaptation costs of health in
developing countries in 2030

Issue

Comments

Likely effect on
total final costs

Burden of disease limited
to 3 health outcomes –
diarrhoeal disease,
malaria, malnutrition

Climate change is likely to affect
other infectious diseases but the
magnitude is uncertain

Under-estimate

Number of cases assumes
a decline in incidence

Highly uncertain as incidence
is declining in some countries,
but many will fail to meet their
MDG targets.

Unknown

Choice of intervention sets

Limited range of options, could
include a wider range of measures

Under-estimate

Costs of interventions,
and the assumption of no
changes to cost over time

Costs vary between countries
and are likely to change over time.

Under or over-estimate

Acknowledgements

I would like to thank the following for helpful review comments:
• Diarmid Campbell-Lendrum, World Health Organization, Geneva

• Guy Hutton, World Bank

• Kristie Ebi, IPCC Working Group II Technical Support Unit

• Anil Markandya, Aline Chiabai, Elena Ojea, Julia Martin-Ortega, Basque Climate Change Centre

• Tony McMichael, Australian National University.

Assessing the costs of adaptation to climate change

References

Arnell, N.W., Cannell, M.G., Hulme, M., Kovats, R.S., Mitchell, J.F., Nicholls, R.J., Parry, M., Livermore, M.T. and White, A. (2002), ‘The

consequences of CO

stabilization for the impacts of climate change’, Climatic Change, vol. 53, pp. 413–446.

Bosello, F., Roson, R. and Tol, R.S. 2006, ‘Economy-wide estimates of the implications of climate change: human health’, Ecological Economics,

vol. 58, pp. 579–591.
Campbell-Lendrum, D. and Woodruff, R.E. (2006), ‘Comparative risk assessment of the burden of disease from climate change’, Environmental

Health Perspectives, vol. 114, no. 12, pp. 935–1941.
Confalonieri, U., Menne, B., Akhtar, R., Ebi, K.L., Hauengue, M., Kovats, R.S., Revich, B., and Woodward, A. (2007) ‘Human health’ in M.L.

Parry et al. (eds) Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of

the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, pp. 391–431.
Department of Health (2008) Heatwave Plan for England – Protecting health and reducing harm from extreme heat and heatwaves. Department of

Health, London.
Ebi, K.L. (2008) ‘Adaptation costs for climate change-related cases of diarrhoeal disease, malnutrition, and malaria in 2030’, Globalization and

Health, vol. 4, no. 9.
Ebi, K.L., Kovats, R.S. and Menne, B. (2006) ‘An approach for assessing human health vulnerability and public health interventions to adapt to

climate change’, Environmental Health Perspectives, vol. 114, no. 12, pp. 1930–1934.
Edejer, T.T.T., Alkins, M., Black, R., Wolfson, L., Hutubessy, R. and Evans, D.B. (2005) ‘Cost effectiveness analysis of strategies for child health

in developing countries’, British Medical Journal, vol. 331.
Ezzati, M., Utzinger, J., Cairncross, S., Cohen, A.J. and Singer, B.H. (2005), ‘Environmental risks in the developing world: exposure indicators

for evaluating interventions, programmes, and policies’, Journal of Epidemiology and Community Health, vol. 59, pp. 15–22.
Hofstetter, P. and Hammitt, J.K. (2001) Human Health Metrics for Environmental Decision US EPA Office of Research and Development Support

Tools: Lessons from Health Economics and Decision Analysis. EPA/600/R-01/104, Environmental Protection Agency, Washington DC.
Hutton, G., Haller, L., and Bartram, J. (2007) ‘Global cost-benefit analysis of water supply and sanitation interventions’, Journal of Water and

Health, vol. 5, no. 4, pp. 481–502.
Kovats, R.S. and Hajat, S. (2008) ‘Heat stress and public health: a critical review’, Annual Review of Public Health, vol. 29, pp. 41–55.
Lopez, A.D., Mathers, C.D., Ezzati, M., Jamison, D.T. and Murray, C.J. (2006) ‘Global and regional burden of disease and risk factors, 2001:

systematic analysis of population health data’, Lancet, vol. 367, pp. 1747–1757.
Markandya, A. and Chiabai, A. (2008) Assessment of the Likely Financial Costs Necessary for Health Protection for Climate Change. Technical Report

for WHO Global consultation on ‘Guiding Research to improve health protection from climate change’. World Health Organization, Geneva.
Mathers, C.D., Murray, C.J., Ezzati, M., Gakidou, E., Salomon, J.A. and Stein, C. (2003), ‘Population health metrics: crucial inputs to the

development of evidence for health policy’ (editorial), BioMed Central.
McMichael, A.J., Campbell-Lendrum, D., Kovats, R.S., Edwards, S., Wilkinson, P., Edmonds, N., Nicholls, N., Hales, S., Tanser, F.C., Le

Sueur, D., Schlesinger, M. and Andronova, N. (2004), ‘Climate change’ in M. Ezzati et al. (eds) Comparative Quantification of Health Risks:

Global and Regional Burden of Disease due to Selected Major Risk Factors. Vol.2. World Health Organization, Geneva, pp. 1543–1649.
Menne, B. and Ebi, K.L. (2005) Climate Change and Adaptation Strategies for Human Health. Steinkopff Verlag, Darmstadt, Heidelburg.
Michelon, T., Magne, P. and Simon-Delaville, F. (2005) ‘Lessons of the 2003 heat wave in France and action taken to limit the effects of future

heat waves,’ in W. Kirch, B. Menne and R. Bertollini (eds) Extreme Weather Events and Public Health Responses. Springer-Verlag, pp. 131–140.
Mills, A. and Shillcutt, S. (2004) ‘Communicable diseases’ in B. Lomborg (ed.) Global Crisis, Global Solutions. Cambridge University Press,

Cambridge, pp. 62–114.
OECD (2008) Economic Aspects of Adaptation to Climate Change: Costs, Benefits and Policy Instruments. OECD, New York.
Oxfam (2007) Adapting to Climate Change: What’s Needed in Poor Countries, and Who should Pay. Oxfam, Oxford.
Parry, M.L., Rosenzweig, C., Iglesias, A., Fischer, G. and Livermore, M.T. (1999), ‘Climate change and world food security: a new assessment’,

Global Environmental Change, vol. 9, pp. S51–S67.
van Rensburg, J.J.J. and Blignaut, J.N. (2002) ‘The economic impact of an increasing health risk due to global warming’, Forum for Economics

and the Environment – First Conference Proceedings. Forum for Economics and the Environment, Cape Town, pp. 17–127
Sachs, J. (2001), Macroeconomics and Health: Investing in Health for Economic Development. Report of the Commission on MacroEconomics and

Health. World Health Organization, Geneva.
Tol, R.S. (2002), ‘Estimates of the damage costs of climate change. Part I. Benchmark estimates’, Environmental and Resource Economics, vol. 21, pp. 47–73.
UNFCCC (2007), Background Paper on Analysis of Existing and Planned Investment and Financial Flows Relevant to the Development of Effective

and Appropriate International Response to Climate Change. UNFCCC, Bonn.
Watkiss, P., Horrocks, L., Pye, S., Searl, A. and Hunt, A. (2007) Projection of Economic Impacts of Climate Change in Sectors of Europe based on

Bottom up Analysis (PESETA). Human Health. Summary for Policy Makers: Report to JRC / IPTS.
WHO Executive Board (2008) Climate Change and Health: Report by the Secretariat. WHO, Geneva, EB122/4.
WHO Regional Office for Europe (2008), Improving Public Health Responses to Extreme Weather/Heat-waves - EuroHEAT. Technical Summary.

WHO Regional Office for Europe, Copenhagen.
World Bank (2006) Disease Control Priorities in Developing Countries (2nd edn). The World Bank/Oxford University Press, Washington DC.

Assessing the costs of adaptation to climate change

Summary

• The UNFCCC study (Nicholls, 2007) is one of several that have attempted to estimate the costs

of adaptation due to sea-level rise, starting with IPCC CZMS (1990).

• All of these studies draw on the extensive experience of ‘hold the line’ with traditional coastal

engineering responses (dykes and beach nourishment). Given the long experience of coastal

engineering measures, these costs are useful guides to the required investment, as long as the

residual damage is considered (not everywhere will be protected).

• Other aspects of climate change were not considered, such as protection against more intense

tropical storms and hurricanes. Response costs would rise significantly under this scenario.

Aspects of disaster preparedness and institutional capacity to respond to these extreme events

will also need to be considered.

• For coastal ecosystems, dyke construction and upgrade will enhance losses via coastal squeeze,

and sustaining the environmental functions and human safety in coastal zones represents a

significant challenge to coastal management which is not included in the UNFCCC

cost estimates.

• The UNFCCC estimate of additional costs of $4–11 billion/year is a reasonable estimate of the

coastal engineering protection measures that would be required, assuming a 50-year planning

horizon and that there is no adaptation deficit.

• However, the costs are under-estimates if we consider responses to high-end sea-level rise

scenarios, and other climate change such as more intense storms – as an indicative estimate,

costs three times greater than those reported by the UNFCCC might result.

• Residual damage is included in the UNFCCC report in terms of sea flood and land loss

estimates of $1–2 billion/year. However, this is incomplete as sea-level rise produces other

damages, including significant environmental damages. Countering these additional damages

is feasible, but will further raise adaptation costs.

5.1 Introduction

Sea-level rise is one of the issues that brought human-induced climate change to the fore due to

the large concentration of settlements and economic activity in low-lying coastal areas. The issue

has been extensively assessed since the 1980s with the spectre of millions of environmental

refugees as a worst-case impact. Adaptation needs and costs were considered from the beginning,

drawing on the extensive experience of coastal engineering and management, including on

subsiding coasts. The global costs of protecting developed coasts against sea-level rise were first

estimated by IPCC CZMS (1990). There have been updates of these costs based on several

different methodologies as outlined below. However, other dimensions of climate change in

coastal areas are less assessed and could substantially raise costs, for example in the case of more

intense hurricanes, or coastal ecosystem changes such as coral reef degradation due to rising sea

surface temperatures and falling pH of ocean waters (Nicholls et al., 2007a).