vii visioning, regulating, cultural and supporting services. This step also ...... This introduces uncertainty on the analysis of adaptation cost in the en- ...... Moral-Carcedo J. and J. Vicens-Otero (2005), “Modeling the non-linear response of Spanish electric- ...... d'évaluation des services délivrés, Paris, Economica, 37-62.
Impacts of Climate Change and Biodiversity Effects European Investment Bank, University Research Sponsorship Programme __________________________________________________________________________________
by Carlo Carraro Scientific Coordinator, Department of Economics, Ca’ Foscari University of Venice
Peter Carter Project Coordinator, European Investment Bank
Francesco Bosello Research Leader, Department of Economics, Ca’ Foscari University of Venice
Paulo A.L.D. Nunes Research Leader, Department of Economics, Ca’ Foscari University of Venice and
Mattia Cai, Aline Chiabai Enrica De Cian, Helen Ding Fabio Eboli, Andrea Ghermandi Emanuele Lugato, Giulia Macagno Ruslana Palatnik, Ramiro Parrado Renato N. Rosa, Silvia Silvestri Department of Economics, Ca’ Foscari University of Venice
Final Report Venice, Italy • November 2009
Table of Contents
Acknowlegments Excutive summary
i
1 1.1 1.2
1 1 2
Introduction Setting the scene Structure of the report
PART I Macro economic assessment of climate change impacts: a regional and sectoral perspective 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7
Assessing climate change impacts using computable general equilibrium model Introduction to the methodology Sea-level rise Tourism Agriculture Energy demand Health Results from the direct impact assessment exercises
7 7 10 13 17 19 24 27
3 3.1 3.2 3.3
Assessing climate change cost for the economy: A general equilibrium perspective Economic model and benchmark Simulation and results Conclusions
37 37 38 45
PART II Climate change impacts on biodiversity/ecosystem services: a partial equilibrium economic valuation approach
49
4 4.1 4.2 4.3 4.4 4.5
Valuation of climate change effects on forest ecosystem services Introduction A hybrid ecosystem-based approach to the monetisation of climate change impacts A Assessing bio-physical flows of ecosystems goods and services under climate change An Economic valuation of European forest ecosystems: results Conclusions
51 51 53 61 65 78
5 Valuation of the linkages between climate change, biodiversity and the productivity of agricultural systems 5.1 Introduction 5.2 A roadmap to the monetisation of climate change impacts on agro-ecosystems 5.3 Assessing the impact of climate change on the provisioning services of agro-ecosystems 5.4 Assessing the impact of climate on the regulating services of agro-ecosystems 5.5 Conclusions
87 87 88 90 97 101
a
6 6.1 6.2 6.3 6.4 6.5
The value of European freshwater ecosystems under climate change: a meta-analysis Introduction Materials and methods Meta-regression results of current values Evaluation of climate change scenarios Conclusions
105 105 106 115 120 130
7 7.1 7.2 7.3 7.4 7.5
The recreational value of european coastal and marine ecosystems Introduction Materials and methods Meta-regression results of current values Scaling up of values at a European scale and evaluation of climate change Conclusions
135 135 136 141 145 152
PART III The integration of climate change impacts on biodiversity into a general equilibrium valuation approach 8.1 8.2 8.3 8.4
Introduction Including ecosystem services in a CGE assessment Valuation results Conclusions
155 156 158 161
PART IV Overall conclusions and future research 9 Conclusions 165 9.1 The macro-economic implications of selected climate change impacts on the EU national economies 166 9.2 Economic assessment of climate change-caused impacts in biodiversity services 169 9.3 The integration of climate change impacts on biodiversity into a general equilibrium valuation approach 176 9.4 Gaps and future research 178
b
Acknowledgements: This report is the final output of the research project CLIBIO (Impacts of CLImate change and BIOdiversity effects), carried out under the University Research Sponsorship Programme (Line of research: Financial and Economic Valuation of Environmental Impacts) financed by the European Investment Bank (EIB). The authors and the whole research team at the Department of Economics of the University of Venice are grateful to Peter Carter, Maria Luisa Ferreira and Mateu Turrò for their valuable comments and suggestions. They are also grateful to Dagmar Schroter and the research team Advanced Terrestrial Ecosystem Analysis and Modelling at the Potsdam Institute for Climate Impact Research, as well as to Jeannette Eggers at the European Forest Institute, for having provided important data and modelling tools. This report also benefited from comments by participants at the midterm CLIBIO Conference, held in Venice in April 2008, ad particularly by Anil Markandya, Pushpam Kumar, Juan Carlos Ciscar and Paul Watkiss. Helpful comments were also provided by participants at the 1st and 2nd Annual Meeting on the EIB-Universities Research Action, at several seminars at the University of Venice, at the 17th European Association of Environmental and Resource Economists (EAERE) Annual Conference in Amsterdam, and at the EIB - Department of Economics session on the “Economic costs of climate related ecosystem services losses and the consequent macro-economic impact” held at the 11th BIO.ECON conference, Venice, September 2009. The research team is also grateful to Martina Marian for having carefully managed the whole project and Sonja Teelucksingh for proofreading the whole report.
Executive Summary 1. Introduction This research provides an economic assessment of climate change impacts from a global perspective and with a focus on the EU, proposing some highly innovative features with respect to what is currently available in the impact literature. Firstly, economic consequences of a wide set of climate-change impacts for major world regional economic systems are not confined to direct costs - i.e. the “pricing” of a positive or negative quantity change - but also incorporate social and economic reactions triggered within economic systems by those impacts. This approach depicts the world economy as a system of markets interacting through exchanges of inputs, goods and services responding to changes in relative prices induced by climate shocks. In other words, market-driven or autonomous social-economic adaptation is explicitly described, the mechanisms through which it is likely to operate are highlighted, and the interaction of impacts is stressed. These research activities are at the core of workpackage one (WP1) entitled the macro-economic implications of selected climate change impacts on the EU national economies. Secondly, the impacts of climate change on biodiversity and ecosystem services are evaluated. The present approach builds upon the Millennium Ecosystem Assessment (MA) conceptual framework, considering biodiversity as the underpinning of ecosystems and ecosystem services, which in turn contribute to human well-being. In other words, the MA proposes an assessment of the status of ecosystems and ecosystem services (“the benefits people obtain from ecosystems) because of their contribution to human well-being. The proposed economic valuation analysis follows a three-step approach. The first step is the determination of the role of biodiversity in the creation of relevant ecosystem services. The second step is the calculation of the reduced quantity and quality of these ecosystem services resulting in losses to human welfare under alternative climate scenarios. In the current analysis, we make use of the four different storylines used by the International Panel on Climate Change (IPCC), i.e. A1, A2, B1 and B2 scenarios. The third step is the (monetary) valuation of those losses. These research activities are at the core of workpackage two (WP2) entitled the economic assessment of climate-change-caused impacts in biodiversity services.
Thirdly, these two streams of research are integrated through the development of a partialgeneral equilibrium valuation framework. This is characterized by the introduction of the i
‘ecosystem’ sector into the “market-based”, general equilibrium assessment.. To our knowledge, this exercise constitutes an original procedure, at a global level of analysis, in the economic welfare assessment of biodiversity/ ecosystem impacts induced by climate change.
2. The macro-economic implications of selected climate change impacts on the EU national economies (WP1) 2.1 WP1 methodology A macro economic general equilibrium climate change impact assessment is not independent of other disciplines; in particular it comes into play only after climatic changes have been translated into physical consequences (impacts) that induce changes in human activities. The present study therefore begins with an assessment of the physical implications of an extended set of climate-change impacts through a comprehensive survey and meta analysis of the available literature – see Table 1.
Table 1: Climate-change impact analyzed within this assessment Supply- side impacts Impact on labour quantity (change in mortality – health effect of climate change) Impacts on labour productivity (change in morbidity – health effect of climate change) Impacts on land quantity (land loss due to sea level rise) Impacts on land productivity (Yield changes due to temperature and CO2 concentration changes) Demand-side impacts Impacts on energy demand (change in households energy consumption patterns for heating and cooling purposes) Impacts on recreational services demand (change in tourism flows induced by changes in climatic conditions) Impacts on health care expenditure
Following this, we make unconventional use of a computable general equilibrium (CGE) model ICES (developed specifically for this research) in order to model and quantify marketdriven adaptation. ICES is a recursive dynamic CGE model, running from 2001 to 2050. It is calibrated to replicate regional GDP growth paths consistent with the A2 IPCC scenario and is then used to assess climate change economic impacts for 1.2 and 3.1 °C increases in 2050 with respect to 2000 (the likely temperature range associated with that scenario). The regional and sectoral details of the model are reported in Table 2.
ii
Table 2: Regional and sectoral disaggregation of the ICES model ICES Model Region (this study) USA: United States Med_Europe: Mediterranean Europe North_Europe: Northern Europe East_Europe: Eastern Europe FSU: Former Soviet Union KOSAU: Korea, S. Africa, Australia CAJANZ: Canada, Japan, New Zealand NAF: North Africa MDE: Middle East SSA: Sub Saharan Africa SASIA: India and South Asia CHINA: China EASIA: East Asia LACA: Latin and Central America
ICES Model Sectors (this study) Rice Other industries Wheat Market Services Other Cereal Non-Market Services Vegetable Fruits Animals Forestry Fishing Coal Oil Gas Oil Products Electricity Water Energy Intensive industries
2.2 WP1 Results Table 3 summarizes the physical impacts assessment. They are highly differentiated regionally and not all negative. This said, larger negative impacts are clearly concentrated in developing countries. This highlights their greater vulnerability to climate change with respect to developed economies, a vulnerability that results from a combination of higher degrees of exposure and sensitivity. The economic implications of these impacts are summarized by Figure 1. When we consider the combined impacts at a global level, we see that these can, depending on the temperature increase scenario, impose a cost ranging from 0.3% to the 1% of GDP in 2050. These global figures hide important regional differences. While developed regions lose slightly, or even gain as in the case of Northern Europe, developing regions can lose considerably more. For a temperature increase of 3.1° C with respect to 2000, for instance, South East Asia, South Asia, Sub Saharan Africa and Northern Africa can experience a GDP contraction of 4%, 3%, 2.6%, 2.4% respectively. The bulk of losses in developing countries is due to negative impacts on GDP driven by the dynamics in the agricultural and tourism markets, while for developed countries climate change impacts on tourism activities, affecting the service sector, are of paramount importance1.
1
The negligible impact on GDP exerted by land and capital losses due to sea level rise is worthy of note. This results from the fact that GDP measures the flow-value of goods and services produced within a region, and
iii
Table 3: Climate change impacts as inputs for the ICES model (% change wrt baseline, reference year 2050 reference temperature +1.2, 3.1 °C wrt 2000) Region
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA Region
Health Labour Public Private Productivity Expenditure Expenditure 1.2°C 3.1°C 1.2°C 3.1°C 1.2°C 3.1°C -0.06 -0.18 -0.15 -0.28 -0.02 -0.03 0.01 0.01 -0.10 -0.18 0.00 -0.01 0.06 0.16 -0.35 -0.88 -0.01 -0.03 0.09 0.23 -0.47 -1.18 -0.01 -0.02 0.16 0.40 -0.41 -1.03 -0.01 -0.03 -0.43 -1.14 0.57 1.62 0.04 0.11 0.09 0.22 0.03 0.24 0.00 0.00 -0.28 -0.69 2.02 4.42 0.10 0.23 -0.22 -0.34 1.34 1.82 0.10 0.14 -0.31 -0.84 0.47 1.34 0.07 0.19 -0.11 -0.30 0.28 0.76 0.06 0.17 0.14 0.37 0.65 1.80 0.06 0.17 -0.11 -0.32 1.05 2.96 0.06 0.17 -0.14 -0.39 0.68 1.98 0.07 0.19 Sea-Level Rise Tourism Market ServiExpenditure Land Losses ces Demand Flows 1.2°C 3.1°C 1.2°C 3.1°C 1.2°C 3.1°C -0.03 -0.05 -0.68 -1.76 -0.04 -0.11 -0.01 -0.01 -1.86 -4.82 -0.02 -0.07 -0.02 -0.04 7.54 19.47 0.18 0.48 -0.02 -0.05 -2.46 -6.37 -0.006 -0.02 -0.01 -0.01 0.00 -0.01 0.00001 0.00003 -0.01 -0.01 -1.31 -3.39 -0.007 -0.02 0.00 -0.01 5.54 14.30 0.13 0.35 -0.02 -0.04 -2.52 -6.52 -0.005 -0.01 0.00 -0.01 -4.67 -12.07 -0.50 -0.13 -0.07 -0.14 -4.43 -11.46 -0.008 -0.02 -0.20 -0.43 -1.21 -3.12 -0.007 -0.02 -0.05 -0.09 -4.99 -12.90 -0.076 -0.20 -0.32 -0.66 -4.69 -12.11 -0.03 -0.07 -0.02 -0.05 -2.68 -6.92 -0.06 -0.16
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA Notes: nss non statistically significant. Expenditure flows in US$ trillions
Land Productivity Other Cereal Crops 1.2°C 1.2°C 3.1°C 3.1°C 3.1°C -18.89 -6.19 -20.37 -8.18 -25.15 -8.33 -4.62 -18.94 -2.00 -11.84 -7.74 -5.90 -26.01 50.00 107.82 -10.50 -2.64 -13.57 -4.60 -18.35 -21.92 -7.47 -24.64 -9.73 -30.10 -17.00 -2.90 -7.41 -3.11 -7.38 -12.33 -1.87 -14.31 -2.24 -15.17 -42.14 -10.78 -41.00 -12.62 -45.97 -32.40 -11.73 -38.52 -13.60 -43.12 -26.33 -7.17 -21.43 -8.81 -25.36 -14.92 -4.89 -18.89 -6.61 -22.99 -2.30 0.50 -3.61 -1.42 -8.25 -0.54 0.34 -4.98 -1.15 -8.50 -21.71 -6.61 -23.38 -8.25 -25.78 Households' Energy Demand
Wheat 1.2°C -5.66 -1.14 1.50 -1.13 -6.12 -7.78 -0.74 -12.81 -8.40 -9.89 -2.96 0.93 2.45 -6.69
Natural Gas 1.2°C -13.67 -12.68 -13.75 -12.93 -13.02 nss -5.05 -8.60 -13.12 nss nss nss nss nss
3.1°C -35.31 -32.76 -35.51 -33.41 -33.65 nss -13.04 -22.22 -33.89 nss nss nss nss nss
Rice
Oil Products 1.2°C -18.52 -15.84 -15.52 -17.39 -17.39 -13.03 -12.63 -13.25 -17.39 -6.51 nss nss nss nss
3.1°C -47.84 -40.91 -40.09 -44.92 -44.92 -33.66 -32.63 -34.22 -44.92 -16.83 Nss Nss Nss Nss
Electricity 1.2°C 0.76 0.76 -2.20 0.76 0.75 12.31 -4.80 5.95 0.74 16.35 20.38 20.38 20.38 21.37
3.1°C 1.96 1.96 -5.68 1.97 1.94 31.81 -12.40 15.37 1.92 42.23 52.65 52.65 52.66 55.20
In addition, the general equilibrium economic assessment confirms that climate change imposes important equity and redistributional issues. It should also be noted that when the detail of the investigation increases, high losses which are otherwise hidden within a global therefore does not directly measure stock losses. These are recorded “indirectly” as long as they change the region’s ability to produce those good and services.
iv
assessment emerge. This stresses the need to carefully tailor the scope of any climate change impact assessment - the impacts are scale-dependent.
Climate Change Impacts: Summary 4.0 3.0
USA Med_Europe
2.0
North_Europe
% of GDP
1.0
East_Europe FSU
0.0
KOSAU CAJANZ
-1.0
NAF
-2.0
MDE SSA
-3.0
SASIA -4.0
CHINA EASIA
-5.0 1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
LACA
Agriculture
Energy Demand
Health
Sea Level Rise
Tourism
All Impacts
World
Tem perature increase
Figure 1. Final climate change impact: % change in regional GDP wrt no climate change baseline (ref. year 2050) Market-driven adaptation mechanisms are crucial in determining the final cost of climate change as shown by Figure 52 which compares a “direct cost” with a “welfare” assessment. Social economic systems show a great capacity to “smooth” initial impacts. In general, when a factor of production, good or service becomes scarcer and thus more costly or less productive, it tends to be substituted with others that are cheaper or more productive. Thus, marketdriven adaptation operates as a partial buffer of initial shocks. Particularly enlightening is the case of Mediterranean Europe where initial negative impacts are eventually turned into gains. There, negative direct impacts are in fact counterbalanced by terms of trade improvements. This is due to the decreased prices of energy imports, driven by a decreased energy demand that is consequent upon world GDP contraction, and by the regional ability to attract foreign investments offering relatively higher rates of return to capital. However and more importantly: what is proposed here can be taken only as the lowest possible bound for climate change costs. This is due to the following considerations: (1) the GDP impacts shown are calculated only for a sub set of potential adverse effects of climate change (for example, possible consequences of increased intensity and frequency of extreme weather events and of biodiversity losses are not included), (2) irreversibilities or abrupt climate and catastrophic changes to which adaptation can be only limited are neglected, (3) the current v
assessment assumes costless and instantaneous market driven adjustments, and (4) the world is currently moving on an emission path leading to a higher temperature increase than that consistent with the A2 scenario. The main implication is that despite its impact smoothing potential, market-driven adaptation cannot be the solution to the climate change problem. Its distributional and scale consequences need to be addressed with policy-driven mitigation and adaptation strategies.
% 2050 GDP
8
3
Dire ct e conom ic cos t -2 Final cos t for the e conom ic s ys te m (w ith m ark et driven adaptation)
-7
SS A SA S IA C H IN A EA S IA L A C A W o rld
N A F
M D E
FS U K O S A U C A JA N Z
U S A M ed _E N U or th _E U E as t_ E U
-12
Figure 2. Direct vs final climate change costs (+3.1°C wrt. 2000 ref. year 2050) Note: Direct economic costs are expressed as a % of 2050 GDP; final costs are expressed as % changes in 2050 GDP wrt the no climate change baseline value
3. Economic assessment of climate change-caused impacts in biodiversity services (WP2) 3.1 WP2 methodology The valuation methodology used in the WP2 is an integrated, hybrid valuation approach: integrated because it is characterized by the use of both of bio-physical and economic valuation model and hybrid because it is characterized by an integrated use of alternative economic valuation methodologies. Figure 3 depicts the road map within which this integrated valuation approach operates. We begin with the mapping of the different European countries according to their latitude/geo-climatic area. The ecosystem study, which is built up on the MA approach, is applied to each of the geographical groups and respective individual countries under consideration. The MA approach provides a practical, tractable, and sufficiently flexible classification for the categorisation of the various types of ecosystem goods and services (EGS). In this context, all EGS can generally be classified into four main categories, i.e. provi
visioning, regulating, cultural and supporting services. This step also encompasses the determination of the role of biodiversity in the creation of relevant ecosystem services. 2
Figure 3: Valuation of ecosystem goods and services in the context of climate hange Against this background, climate change is introduced here by an analysis of the different IPCC storylines, embracing a 2050 time horizon and with specific attributes in terms of population growth, CO2 concentration, degree of temperature changes, and precipitation changes in Europe. Efforts have been placed on the development of a general circulation model – HadCM33 – so as to directly relate socioeconomic changes to climatic changes through greenhouse gas concentration, and to relate land use changes to climatic and socioeconomic drivers such as the demand for food (Schröter D. et al. 2004). As a consequence, the IPCC is able to present four brief storylines differently developed in economic, technical, environmental and social dimensions (Nakicenovic and Swart, 2000). At this stage, we are able to make use of the bio-physical valuation models, building upon the HadCM3, in order to evaluate the impact of climate change on biodiversity, with the calculations of the reduced quantity and quality of respective ecosystem services under alternative scenarios. In particular the bio-physical quantitative assessment is provided by land use
2 Comprehensive economic assessment covering all ecosystem services is in principle meaningless since all
human activity is fully dependent on basic ecosystem services such as, for example, photosynthesis. The only sound approach is marginal analysis of selected service flows (see bold EGS in Figure ). 3 HadCM3, Hadley Centre Couplet Model Version 3 is a coupled atmosphere-ocean GCM developed at the Hadley Centre and described by Gordon et al. (2000).
vii
models including the ATEAM4 and IMAGE 2.2 models. These are responsible for modelling and predicting changes in the forest area, forest land, wood and agricultural products across the four IPCC storylines. Finally, in order to deliver a (monetary) valuation of the reduced quantity and quality of the ecosystem services under consideration, we adopt a hybrid valuation approach that is characterized by an integrated use of different economic valuation methodologies. The different characteristics of the economic benefits of each type of ecosystem goods and services shall be the basis for the selection of the most appropriate valuation methodologies. We therefore work with market prices analysis when valuing provisioning services, damage cost assessments when valuing carbon sequestration services and meta-analytical valuation methods for the assessment of cultural services – see Figure 4.
Figure 4: Hybrid approach to the economic valuation of ecosystems goods and services (illustration with the forest ecosystem goods and services) This exercise is applied to the valuation of biodiversity, and its impact on the provision of ecosystem goods and services, for different ecosystems – see Table 3 for an overview. The main valuation results are presented, and discussed, in the following section.
4 Advanced Terrestrial Ecosystem Analysis and Modelling (ATEAM) is an integrated project funded by the 5th Framework Programme of the European Commission that focuses on the assessment of the vulnerability of ecosystem services with respect to global change (Schröter D. et al. 2004).
viii
Table 3: Synthesis Ecosystem
Provisioning services
Regulation services
Cultural services
Forest
Agriculture
Freshwater
Coastal
3.2 WP2 Results Forest Ecosystem goods and services Figure 4 summarizes the economic valuation results from three different types of ecosystem goods and services provided by forest ecosystems in Europe across four IPCC scenarios. As we can see, the impact of climate change on biodiversity, and its welfare evaluation in terms of the respective changes in the provision of forest ecosystem goods and services, is multifaceted. Firstly, it depends on the nature of the forest good and service under consideration. For example, cultural values reveal a greater sensitivity to climate change in comparison to other values, while in contrast the wood forest products reveal a greater resilience to climate change. Secondly, the distributional impacts of climate change on the provision of these goods and services also depend upon the geo-climatic regions under consideration. In other words, the specific compositions of ecosystem benefits in the local economies may result in different economic impacts of climate change. Taking B1 scenario as an example, the Central European countries are ranked the highest in the provision of both wood forest products (about 41.4 billion$) and carbon sequestration services (about 190.8 billion$), but ranked second in terms of cultural services provision (about 4.3 billion$); conversely Mediterranean Europe demonstrates exactly the opposite. The Scandinavian region occupies the second best position in terms of provisioning wood forest products in Europe (accounting for 31.8 billion$), and is valued much lower in terms of carbon sequestration (46.3 billion$) and cultural services (about 3.0 billion$). The different trends of value changes across geo-climatic regions demonstrate that each region may have varying resiliencies to climate change, depending upon the major economic component provided by the respective forest ecosystems
ix
Figure 1: Forest ecosystem: valuation results
Agriculture Ecosystem goods and services With respect to the agricultural sector, we shall evaluate the role of biodiversity on human welfare by exploring two value transmission mechanisms. The first refers to the effect of bio-
x
diversity on crop productivities.5 The second refers to the assessment of carbon sequestration services provided by agricultural land. As far as the economic valuation of the carbon sequestration is concerned, we subscribe to a similar methodology adopted for the forest ecosystem. We distinguish cropland and grassland areas since these two ecosystems show different marginal carbon sequestration productivities (see Smith et al. 2005 – Global Change Biology). Figure 5 shows the economic valuation results of this service.
Economic value of stocked carbon in Croplands, 2050
6,537
17,685
2050 B2 (M$)
53,202
36,995 5,565
21,810
2050 B1 (M$)
59,503
39,220
Scandinavian Europe Northern Europe
5,120
15,905
2050 A2 (M$) 4,512
"Mediterranean Europe"
20,959
2050 A1 (M$)
55,054
36,696 0
10,000
20,000
30,000
Central-North Europe
50,599
37,570
40,000
50,000
60,000
70,000
Economic value of stocked carbon in Grasslands, 2050
936
10,275
2050 B2 (M$)
23,302
12,678 1,345
14,745
2050 B1 (M$)
33,210
14,237
Scandinavian Europe Northern Europe
577
8,019
2050 A2 (M$) 575
"Mediterranean Europe"
10,331
2050 A1 (M$)
23,507
8,983 0
5,000
10,000
Central-North Europe
20,581
10,458
15,000
20,000
25,000
30,000
35,000
Figure 5: Agricultural ecosystem: valuation results
Freshwater and wetland ecosystems goods and services The provision of goods and services by freshwater wetlands, lacustrine and palustrine ecosystems in 17 West European countries is investigated by means of meta-analysis. A data set 5
Bearing in mind the limited time resources available to the current research project (three years), we decided to preclude cultural values from the analysis, including agricultural landscape values. This can be an interesting topic for further future research.
xi
containing 236 independent observations from 103 valuation studies and reflecting 123 distinct freshwater ecosystems worldwide is created, and the average flow of values for provisioning, regulating and cultural services is estimated on the basis of a series of study, site and context specific variables. Freshwater ecosystem values are found to be subject to income and substitution effects. Biodiversity richness and population density positively influence benefits, while high temperature is negatively correlated with values. Mean values per hectare are high in countries with relative scarcity of freshwater ecosystems (e.g., Portugal and Italy), high population density (e.g., Belgium), and high Gross Domestic Product per capita (e.g., Luxembourg). Aggregation of the values over total freshwater ecosystem area in each country reveals that the highest benefits are experienced where values per hectare are high (e.g., Italy, 46,725 million US$/annum) or total ecosystem area is very large (e.g., Sweden, 42,096 million US$/annum). The impact of climate change on the baseline value estimates is assessed based on the A1, A2, B1 and B2 IPCC storylines. Scenario-specific variations in real Gross Domestic Product per capita, population density, biodiversity and maximum yearly temperature in the year 2050 are included in the model. The values of wetlands and freshwater ecosystems are predicted to decrease in 2050 as a consequence of climate change. Scenarios B2 and A1 are the least favourable, while under the most favorable conditions of scenario B1 a decrease of 9% is predicted. The comparative analysis of IPCC scenarios shows that with respect to welfare gains, scenario B1 is ranked higher than A1, A2, and B2 in all the countries considered. The absolute value difference in scenario B1 with respect to the baseline scenario A2 is estimated in 24,773 million US$/year at the European scale – see Table 5. Table 5. Projection of Total Benefits of Services of Freshwater and Wetlands
Benchmark A2 Scenario Absolute value difference (Million$, 2005)
Percentage Change
A1vs.A2 B1vs.A2 B2vs.A2 A1vs.A2 B1vs.A2 B2vs.A2
Mediterranean Europe (N35-45) -3,420 12,412 -4,612 -3.63% 13.19% -4.90%
xii
Central Europe (N4555) 971 2,505 -1,289 3.47% 8.96% -4.61%
Northern Europe (N55-65) 598 644 -172 9.10% 9.79% -2.61%
Scandinavian Europe (N65-71) -2,863 9,212 -4,373 -7.20% 24.00% -11.30%
Europe -4,713 24,773 -10,446 -2.83% 14.85% -6.26%
Coastal ecosystems goods and services The average consumer surplus or willingness to pay per person per year observed in 315 valuations of recreational services in 89 coastal and marine ecosystems worldwide is used as the endogenous variable of a meta-regression model. Values are regressed against a series of variables that include biodiversity levels and climatic conditions. The calibration of the metaanalytical model on the current provision of ecosystem services reveals that the highest values per person per year are in Mediterranean countries, Greece and Italy in particular, where high temperatures encourage sea, sun and sand recreational activities. Average values range between 78.9 US$/person/annum in Finland and 399.8 US$/person/annum in Greece. Individual values are aggregated at country level based on the number of coastal tourists per year. High aggregated values are found in Mediterranean countries where individual values are high and the tourism industry is particularly developed, and in the United Kingdom. The impact of climate change on the baseline value estimates is investigated through variations in real Gross Domestic Product per capita, population density, biodiversity, minimum monthly temperature and maximum monthly temperature in accordance with the four IPCC storylines (A1, A2, B1 and B2) and for the year 2050. For all scenarios, it follows that individual values in the considered countries in 2050 will generally increase . The largest percentage increase in individual values will be concentrated in Northern European and Scandinavian countries. Total values aggregated at country level are similarly expected to grow, mainly due to an increase in the number of coastal tourists. A comparison across the four scenarios reveals that the highest increase in values is predicted for the two economic oriented scenarios, in particular for scenario A1 – see results Table 6.
Table 6. Projection of Total Benefits of Recreational Services of Coastal Ecosystems Benchmark A2 Scenario Absolute value differ- A1vs.A2 ence B1vs.A2 (Million US$, 2003) B2vs.A2 A1vs.A2 B1vs.A2 Percentage Change B2vs.A2
Mediterranean Europe (N35-45) 86,526 -38,083 -30,682 21.98% -9.67% -7.79%
xiii
Central Europe (N45-55) 32,808 -13,771 -9,745 22.84% -9.59% -6.78%
Northern Europe (N55-65) 39,186 -21,902 -33,628 14.39% -8.04% -12.35%
Scandinavian Europe (N65-71) 45,674 -21,344 -18,881 33.72% -15.76% -13.94%
Europe 204,194 -95,101 -92,936 21.61% -10.06% -9.83%
4. Evaluating the cost for the EU economic system of climate change impacts on ecosystem services 4.1 Methodology: a partial-general equilibrium analysis To assess the general equilibrium implications of climate change impacts on ecosystems services, each of these is first translated into a marketable item and then into changes in the appropriate economic variables that act as inputs into or constitute outputs from the CGE model Climate-change induced changes in EU forest-timber production, representing forest ecosystem provisioning services, are modelled as changes in the productivity of the natural resource inputs used by the EU timber industries; similarly, changes in agricultural productivity due to climate-change impacts on biodiversity are modelled as changes in the productivity of the land inputs for EU agricultural industries. The general equilibrium assessment of changes in European forests’ carbon sequestration services follows a different path. Changes in forest based carbon sequestration alter the GHG’s balance between land sinks and the atmosphere, thereby causing a temperature change that can be defined over a period of time. Taking this into consideration, we use a global warming approach which, to the best of our knowledge, has never been performed under a CGE modelling framework. This exercise consists of the formulation of a scenario where the carbon sequestration services from European forests are affected by climate change, thereby producing a different CO2 concentration level in the atmosphere and a corresponding variation in temperature. That change in temperature in its turn impacts the economy at various levels that can then be assessed through the use of the CGE methodology developed within WP1. Accordingly, we first compute a temperature equivalent induced by the higher release of CO2 emissions in the atmosphere resulting from climate change. On the basis of this new information, we then re-estimate all of the climate change impacts considered within WP1 and recalculate new macro-regional GDP effects. The differences between climate-change impacts on GDP considering the original and the new carbon sequestration levels are used as an approximation of the general equilibrium value of the changes in EU forest carbon sequestration service.
4.2 Results Table 7 summarizes, for different scenarios of temperature increase estimated by WP2, the impact on agricultural land and forest timber productivity. Agricultural productivity declines for all Europe, while forest timber productivity declines in the Mediterranean but increases in xiv
other EU areas, in particular the Northern Europe. These changes are due to biodiversity/ecosystem effects and are used to adjust the information originally utilised in WP1. Table 7: Climate change impacts on ecosystem services (% change wrt 2000, reference year 2050) +1.2°C T wrt 2000 Agricultural Land Productivity
+3.1°C T wrt 2000
Forest productivity (timber)
Agricultural Land Productivity
Forest productivity (timber)
Med_Europe
-2.30
-6.08
-5.94
-15.70
North_Europe
-0.93
15.09
-2.39
38.97
East_Europe
-1.42
4.48
-3.67
11.56
Productivity losses in agriculture dominate in Mediterranean and Eastern Europe, resulting in GDP decreases that are particularly pronounced in the latter case. While Mediterranean EU is the more exposed region in terms of direct negative impacts, there are still gains experienced (albeit at small levels) due to positive terms of trade effects and decreased energy imports. Northern Europe’s GDP is largely unaffected since the losses in agriculture are more or less compensated for by gains in the timber industry. Lastly, the discounted difference of GDP between the simulations with and without climate change impacts on ecosystem services provides a general equilibrium assessment as shown in table 8. Calculated over the period 2000-2050 (at a 3% discount rate), and depending on the temperature increase scenario, these results imply a loss (that is, a lower gain) for Mediterranean EU ranging from 9.7 to 32.5 billion US$, a higher loss for Eastern EU ranging from 7.2 to 22 US$ billion, and a slight gain for Northern EU ranging from 2 to 5.6 US$ billion. Table 8. Climate change impact on GDP without (1) and with (2) impacts on ecosystem/biodiversity Climate Change indirect impact NPV 2001-2050 (dr=3%) Million US$ + 1.2°C T wrt 2000 + 3.1°C T wrt 2000 Difference Difference Region Part I Part I Part I (biodiversity Part I (biodiversity and Part II and Part II (1) (1) effect) effect) (2) (2) (2) – (1) (2) – (1) Med_Europe -33,979 -43,733 -9,754 -65,084 -97,631 -32,548 North_Europe 488,420 490,350 1,929 1,360,399 1,366,058 5,659 East_Europe -20,808 -28,046 -7,238 -101,529 -123,787 -22,258 Total 433,633 418,571 -15,063 -166,613 -221,418 -49,147 Nota: % change wrt no climate change baseline (ref. year 2050, 3% discount rate)
xv
With respect to EU forests’ ecosystem carbon sequestration services, their temperature smoothing potential can be estimated in 0.018°C taking 2050 as the reference point. Recomputing the impact assessment exercise of WP1 increasng by this value the temperature scenarios, higher GDP losses and higher gains are experienced by those regions respectively damaged and advantaged by climate change.
Table 9: Estimated value of EU forests carbon sequestration service
Region
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA Europe World
+ 1.2°C T wrt 2000 Climate Change impacts on GDP NPV 2001-2050 (dr=3%) Million US$ Year. av. Part I and Part I Difference (2001-2050) Part II (1) (2) – (1) (2) -266,294 -270,566 -4,273 -87 -33,979 -34,476 -497 -10 488,420 496,059 7,639 156 -20,808 -21,189 -381 -8 -21,482 -22,422 -941 -19 -71,135 -72,260 -1,125 -23 102,803 104,473 1,670 34 -50,229 -51,229 -1,001 -20 -221,033 -224,571 -3,537 -72 -52,729 -53,895 -1,167 -24 -368,147 -375,246 -7,099 -145 -431,586 -438,733 -7,147 -146 -212,334 -215,812 -3,478 -71 -332,006 -337,790 -5,784 -118 433,633 440,394 6,761 138 -1,490,538 -1,517,658 -27,120 -553
+ 3.1°C T wrt 2000 Climate Change impacts on GDP NPV 2001-2050 (dr=3%) Million US$ Part I and Part I Difference Part II (1) (2) – (1) (2) -631,392 -635,746 -4,354 -65,084 -63,792 1,292 1,360,399 1,372,541 12,142 -101,529 -103,035 -1,506 -214,426 -222,225 -7,799 -172,240 -173,401 -1,160 361,249 366,294 5,044 -210,749 -215,451 -4,702 -620,101 -626,561 -6,460 -218,737 -222,748 -4,010 -1,474,608 -1,503,348 -28,740 -1,863,000 -1,887,020 -24,020 -730,920 -739,675 -8,755 -995,229 -1,007,254 -12,025 1,193,786 1,205,714 11,928 -5,576,367 -5,661,421 -85,054
Year. av. (20012050)
When carbon sequestration services from the European forests, croplands and grasslands is reduced by climate change, climate change impacts themselves become larger., Table 9 shows that at a global level, and depending upon the climate change scenario, the damage imposed by climate change on carbon sequestration services provided by EU ecosystems can cost on average 553 to 1736 million US$ per year. These figures monetize the negative GDP performances of all the economies considered due to the higher temperature increases consequent upon the lower CO2 sequestered by EU forests. Alternatively, if we focus our attention on Europe, Table 9 shows that a reduced carbon sequestration service provided by EU ecosystems implies a welfare gain that ranges from 138 to 243 million US$ on a yearly base. Although for Meditarranean and Eastern Europe the net welfare effect of the carbon sequestra-
xvi
-89 26 248 -31 -159 -24 103 -96 -132 -82 -587 -490 -179 -245 243 -1,736
tion services provided by ecosystems is positive as higher temperature are “bad” for them, it is negative for Northern Europe which ultimately gains from climate change.
5. Conclusions The present research addresses the economic assessment of climate change impacts from three innovative angles, that to our knowledge have never before been proposed. Firstly, as opposed to the traditional direct costing approach, we estimate the final economic implications of an extended set of climate impacts both globally and for the EU economic systems, thus describing and quantifying the autonomous adaptation processes at play. Secondly, we propose an economic evaluation of climate change impacts on EU forestry, agricultural, freshwater, wetland and coastal ecosystems by means of a hybrid economic valuation model approach. Finally we merge the two research angles by proposing the integration of ecosystem services, including regulating and cultural values, into the original macro economic valuation tool. We label this a partial-general-equilibrium valuation model approach. Main results highlight the substantive impact-reducing effect of autonomous adaptation. At the global level, the impacts amount to about 1% of world GDP in 2050. Neither can autonomous adaptation reverse the adverse distributive implications of climate change. Even in terms of final impacts on economic activity, the developing world is more severely affected than the developed one. These results lead to the main conclusion that autonomous adaptation cannot be invoked as “the” solution to climate change which therefore needs to be addressed with proper mitigation and planned adaptation strategies.
xvii
xviii
1 Introduction
1.1
Setting the scene
The worldwide decay in environmental quality and the gradual depletion of natural resources, sometimes referred to as the ‘new scarcity’, has prompted intense scientific attention in both the natural and social sciences. The global interest in climate-environmental-economic matters is partly caused by the increased pressures that mounting population and increased temperature exert on the earth’s natural resource base. In addition, as personal incomes rise and leisure time becomes more freely available in the developed world, concern for more immediate human needs is accompanied by interest in nature preservation and conservation for future generations. Although climate and environmental issues can manifest on local or regional scales, they are both part of a globally interwoven ecosystem. Consequently, the ‘new scarcity’ has spatial and temporal horizons that extend far beyond the current level of thinking and acting. Human well-being is dependent upon "ecosystem services" provided freely by nature, such as water and air purification, fisheries, timber and nutrient cycling. These are predominantly public goods with no markets and no prices. As a result, their loss is often not detected by our current economic incentive system and can continue unabated. A variety of pressures resulting from population growth, changing diets, urbanisation, and climate change are causing continuous ecosystem degradation and a resultant biodiversity decline. Europe, along with many other countries, experiences these pressures. In this current context we therefore contribute to the ongoing investigations in the study of the linkages between biodiversity and human well being. In addition, this investigation takes into account the additional element of global climate change impacts. No longer ‘exclusive’ to the academic and research arena, this topic is now highlighted in current policy agendas. For example, at the recent meeting of environment ministers of the G8 countries and the five major newly industrialising countries in Potsdam in March 2007, the German government proposed a study on 'The economic significance of the global loss of biological diversity' as part of the so-called 'Potsdam Initiative' for biodiversity. The following wording was agreed upon at Potsdam: 'In a global study we will initiate the process of analysing the global economic benefit of biological diversity, the costs of the loss of biodiversity and the failure to take protective measures versus the costs of effective conservation.' This proposal was endorsed by G8+5 leaders at the Heiligendamm Summit on 6-8 1
June 2007. With this in mind, the German Federal Ministry for the Environment and the European Commission, with the support of several other partners, have jointly initiated preparatory work for this global study, named 'The Economics of Ecosystems & Biodiversity (TEEB)'. An interim report was presented at the IUCN World Conservation Congress. Barcelona, 5-14 October 2008. Against this background, the present report contributes to the ongoing debate by addressing key issues. In particular, we assess the economic costs of climate related biodiversity losses and the consequent macro-economic impacts through the following. Firstly, we depict the world economy as a system of markets interacting through exchanges of inputs, goods and services responding to changes in relative prices induced by climate shocks, explicitly describing market-driven or autonomous social-economic adaptation. These research activities are at the core of the first workpackage entitled the macro-economic implications of selected climate change impacts on the EU national economies. Secondly, we evaluate the impacts of climate change on biodiversity and ecosystem services, building upon the Millennium Ecosystem Assessment (MA) conceptual framework, considering biodiversity as the underpinning of ecosystems and ecosystem services, which in turn contribute to human well-being. The proposed economic valuation analysis follows a three-step approach. The first step is the determination of the role of biodiversity in the creation of relevant ecosystem services. The second step is the calculation of the reduced quantity and quality of these ecosystem services resulting in losses to human welfare under alternative climate scenarios. The third step is the (monetary) valuation of those losses. These research activities are at the core of workpackage two entitled the economic assessment of climatechange-caused impacts in biodiversity services. Thirdly, we integrate these two streams of research through the development of a partialgeneral equilibrium valuation framework. This is characterized by the introduction of the ‘ecosystem’ sector into the “market-based”, general equilibrium assessment which we label the partial-general equilibrium valuation model. To our knowledge, this exercise constitutes an original procedure, at a global level of analysis, in the economic welfare assessment of biodiversity/ecosystem impacts induced by climate change.
1.2
Structure of the report
The sequence of report chapters follows the structure of the EIBURS research proposal and respective working packages. This report is composed of three main parts addressing different objectives. 2
The specific objectives of Part I of the report are threefold. Firstly, we evaluate the dynamics of economic damages induced by (future) climate change by the development of a recursive-dynamic version of the CGE model. Secondly, we study the biophysical impacts of climate change on societal and economic systems. Thirdly, climate change impacts are quantified by a translation into changes in key economic variables such as GDP variations. The objectives of Part II of the report are multi-disciplinary in nature, with a specific focus on forest ecosystems, agricultural ecosystems (both croplands and grasslands), freshwater and wetland ecosystems, and coastal ecosystems. From the natural science perspective, we link climate change to biodiversity and ecosystem services via the exploration of general circulation models such as HADCM3. From the economics perspective, we identify the contribution of biodiversity to the ecosystem provision of goods and services and value these magnitudes in monetary terms. We endorse a hybrid economic valuation methodology by the integration of alternative valuation methods that are best suited to the ecosystem services under consideration. Finally, Part III of the report addresses the economic assessment of climate change impacts from the innovative angle of the development of a partial-general equilibrium valuation framework. This involves the introduction of the ‘ecosystem’ sector into the “market-based”, general equilibrium assessment, with simulations of both carbon sequestration and provisioning services. The difference between GDP performances from the baseline (from the results of Part I) to the perturbed scenario (from the results of Parts I and II) quantifies the general equilibrium effects of climate change impacts on ecosystem services. This value, expressed in monetary terms, embeds all of the macro economic adjustments at play within the system. The sequence of chapters is as follows. Chapter 2 introduces the CGE modelling methodology. Chapter 3 undertakes an assessment of climate change impacts on the categories of sea level rise, tourism, agriculture, energy demand and health. Chapter 4 quantifies these costs by the application of the general equilibrium model. Chapters 5-8 each focus on the valuation of climate change effects and biodiversity on the provisioning of specific European ecosystem services. Chapter 5 considers forest ecosystems and monetises impacts via a hybrid ecosystem-based approach. Chapter 6 values the linkages between climate change, biodiversity and the productivity of agricultural ecosystems (both croplands and grasslands). Chapter 7 uses a meta-analytical approach to evaluate freshwater ecosystems. Chapter 8 focuses on the recreational values of coastal and marine ecosytems, again with the use of a meta-analytical methodology. Chapter 9 presents an innovative integration of the previous results by the inclusion
3
of the considered ecosystem services in a CGE assessment. Chapter 10 concludes and suggests directions for future research. The information offered in this book is not only meant to be of interest to students and teachers in environmental and resource economics, but also researchers in other fields related to ecological economics. This study aims to be a source of reference for all those concerned about the future of biodiversity.
4
PART I Macro economic assessment of climate change impacts: a regional and sectoral perspective
5
6
2 Assessing climate change impacts using computable general equilibrium model
2.1 Introduction to the methodology The final aim of this part of the research is to assess the economic consequences of climate change impacts for major world regional economic systems, not just restricted to direct costs, but considering, in addition, social and economic reactions triggered within economic systems by those impacts.
Table 1: Methodological approaches in the economic assessment of climate change . Market value based assessments
Non market values based assessments
Partial Equilibrium
“Direct Costing” and Partial Equilibrium
“Direct Costing” + Non Market Evaluation
General Equilibrium
“CGE Modelling”
“CGE Approach” + Non Market Evaluation
As shown in Table 1, which classifies available approaches proposed by the impact literature, both a partial and general equilibrium perspective can be identified. The first, offers an assessment of costs which does not take into account the feedback that an economic perturbation into a sector or activity exerts on the rest of the systems. Albeit the many differences in direct costing techniques, cost estimation is generally confined to direct impacts and its final result can be described with the process below:
(Economic cost of climate change) = (“Quantity with Climate Change” – “Quantity without Climate Change”) x (“Price”). In this perspective for instance, the direct cost of sea-level rise induced land loss can be expressed as the quantity of land lost times its price. These methodologies are largely diffused in the impact literature (see e.g. Fankhauser, (1994); Yohe et al. (1996); Yohe and Schlesinger, (1998); Dennis et al., (1995); Volonte and Nicholls, (1995); Gambarelli and Goria (2004) for sea-level rise; Hamilton et al., (2005a,b); Hamilton and Tol, (2007); Amelung et al., (2007), Elsasser and Burki, (2002); Scott et al., (2004);(2007); OECD,(2007) for tourism; Aldy and Viscusi (2003) for health). Their strength is their relatively easier applica7
bility and perhaps a smaller degree of uncertainty in the estimated values, as a more limited number of simplifying assumptions on economic dynamics are necessary, compared to a holistic approach. Their major shortcoming is that they cannot measure possible rebounds on costs that a changing economic context can impose. In fact, sectors within the economy are not isolated and markets are linked by flows of input, goods and services domestically, and internationally. Thus, economic perturbations like those induced by a changing climate can spread their effects outside the area initially impacted and, more importantly, can induce a reallocation of resources which can smooth out or amplify the initial effect. To capture these processes a systemic perspective is necessary. This concept is made operational by Computable General Equilibrium (CGE) models. In the beginning, CGE models were developed mainly to analyze international trade policies and to a lesser extent, public sector policies. Soon however, due to their vast flexibility, they started to be applied to environmental taxation and climate change impact assessment The peculiar feature of CGE models is market interdependence. All markets are linked, as factors of production are mobile between sectors and internationally, each change in relative prices induces a cost-minimizing input reallocation in the entire economic system. This is also true for the demand side: responding to a scarcity signal in one market, utility-maximizing consumers readjust their entire consumption mix. As a consequence, CGE models can capture and describe the propagation mechanism induced by a localized shock onto the global context via price and quantity changes, and vice versa. Moreover, they are able to assess the “systemic” effect of these shocks, or more specifically, the final welfare or general equilibrium outcome which is determined after all the adjustment mechanisms at play in the economic system operated. The final impact on national GDPs summarizes these “higher order” effects which are usually very different from the initial impacts; this last difference quantifies market-driven adaptation. In this research phase we follow this approach and we quantify market-driven adaptation, applying a computable general equilibrium (CGE) model, ICES, developed for this purpose. ICES is a recursive dynamic CGE model, running from 2001 to 2050. It has been calibrated to replicate regional GDP growth paths consistent with the A2 IPCC scenario and has then been used to assess climate change economic impacts for a 1.2 and a 3.2 °C increase in 2050 wrt 2000, which is the likely temperature range associated to that scenario. The aveage world temperature data has been downscaled to ICES regions through data elaboration from Giorgi and Mearns (2002). The regional and sectoral detail of ICES adopted within this exercise are re8
ported in Table 2. Table 2. Regional and sectoral disaggregation of the ICES model REGIONAL DISAGGREGATION OF THE ICES MODEL (this study) USA: United States Med_Europe: Mediterranean Europe North_Europe: Northern Europe East_Europe: Eastern Europe FSU: Former Soviet Union KOSAU: Korea, S. Africa, Australia CAJANZ: Canada, Japan, New Zealand NAF: North Africa MDE: Middle East SSA: Sub Saharan Africa SASIA: India and South Asia CHINA: China EASIA: East Asia LACA: Latin and Central America SECTORAL DISAGGREGATION OF THE ICES MODEL (this study) Rice Gas Wheat Oil Products Other Cereal Crops Electricity Vegetable Fruits Water Animals Energy Intensive industries Forestry Other industries Fishing Market Services Coal Non-Market Services Oil
It is evident that economics is not independent from other disciplines; it comes into play only after climatic changes have been translated into physical consequences (impacts) which are then able to induce a change in human activities (see Figure 1).
Figure 1. The structure of the integrated impact assessment exercise
9
Thus, the present study initially assessed the physical implications of an extended set of climate change impacts through a comprehensive survey and meta analysis of the available literature. It then proceeded to transform them into appropriate changes in key economic variables, suitable to be used as an input to the ICES model. This was done representing climatic impacts as changes in productivity, supply or demand for different inputs and/or outputs of the model, as reported in Table 3.
Table 3: Climate-change impact analyzed within this assessment Supply- side impacts Impact on labour quantity (change in mortality – health effect of climate change) Impacts on labour productivity (change in morbidity – health effect of climate change) Impacts on land quantity (land loss due to sea level rise) Impacts on land productivity (Yield changes due to temperature and CO2 concentration changes) Demand-side impacts Impacts on energy demand (change in households energy consumption patterns for heating and cooling purposes) Impacts on recreational services demand (change in tourism flows induced by changes in climatic conditions) Impacts on health care expenditure
In what follows section 3 will describe the impact assessments studies by category, section 4 the general equilibrium assessment, sections 3.6 and 4.3 will conclude the respective chapters.
2.2
Sea-level rise
2.2.1 Introduction and background The rise in sea level is often seen as one of the most threatening impacts of climate change. Coastal erosion and sea floods, impacts on often densely populated and infrastructure rich river deltas, the destruction of entire islands and island nations (Nichols et al., 2007) make it one of the most prominent assessments in the climate change impact literature. In 1991 the IPCC already proposed methodologies and estimates of the cost of rising sea levels and the benefits of coastal protection (IPCC CZMS, 1991). The following and very large body of literature was dominated by engineering research and by Geographical Information Systems (GIS) used to determine areas, people and activities at risk, to which an eco10
nomic value was attached. Studies in this vein include investigation at the world level with macro regional and country detail (see e.g. Hoozemans et al., (1993); Fankhauser, (1998), Tol, (2002); (2006)), at the macro-regional level (see e.g. Fankhauser, (1994); Yohe et al. (1996); Yohe and Schlesinger, (1998) for the USA); Nicholls and Klein, (2003); CEC, (2007), for Europe), at the country level (see e.g. Dennis et al., (1995) for Senegal, Volonte and Nicholls, (1995) for Uruguay, Volonte and Arismendi, (1995) for Venezuela, Morisugi et al. (1995) for Japan, Zeider (1997) for Poland) and at the site level (see e.g. Gambarelli and Goria (2004) for the Fondi plane in Italy, Breil et al. (2005) for the city of Venice, Smith and Lazo (2001) analyzing among others, the Estonian cities of Tallin and Pärnu, and the Zhujian Delta in China, Saizar (1997) for Montevideo). This vast literature concentrates on the direct costs of the rise in sea levels and of possible adaptation options. The main result of these studies is that the cost of sea-level rise (albeit in some cases a small fraction of GDP) can be considerably high in absolute terms. As an example, US$ 0.06 billion is the estimated annuitized cost of 50 cm. of sea-level rise for the US according to Yohe et al. (1996). US$ 3.4 billion is Morisugi et al, (1995) the estimate for Japan. A yearly cost ranging from Euros 4.4 to 42.5 billion is the evaluation proposed for Europe by CEC (2007) for a sea-level increase of 22cm. and 96 cm., respectively. The Netherlands, Germany and Poland are expected to suffer a cumulated undiscounted capital loss of US$ 186, 410 and 22 billion for 1 meter of sea-level rise according to Nicholls and Klein (2003). Against this background, coastal protection seems to be not only effective, but also efficient in most cases. This is, for instance, confirmed for Europe as a whole (CEC, 2007), for the Netherlands, Germany (Nicholls and Klein, 2003), Poland (Zeider, 1997; Nicholls and Klein, 2003), for Japan (Morisugi et al., 1995) and Senegal (Dennis et al. 1995). Tol (2007), showed that high levels of coastal protection (>70% of the threatened coast) would be optimal for the majority of the world’s regions. However, for some countries or sites the efficient level of coastal protection is likely to be low or even zero (e.g. the case of Dar es Salaam and the entire populated coastline of Tanzania (Smit and Lazo, 2001), Uruguay (Volonte and Nicholls, 1995) and Venezuela (Volonte and Arismendi, 1995)), pointing out the importance of carefully evaluating benefits and costs of different options for sea level rise adaptation. The cited studies are based on a direct costing approach: they basically evaluate costs by multiplying a quantity loss (land or capital) or “displaced” (people), by the unitary “price” of the item lost or of the displacement. By contrast, the present research attempts to assess the “higher-order” impact of a rise in sea level.
11
A similar approach has been followed by Deke et al. (2002), Darwin and Tol (2001), Bosello et al. (2004) and Bigano et al., (2008). Deke et al. (2002) use a recursive dynamic CGE model, they restrict the study to the costs of coastal protection, ignoring land loss and its wider economic consequences. The costs of coastal protection are subtracted from investment, this essentially reduces the capital stock, and hence economic output. They estimate a direct protection cost against the 13 cm. of sealevel rise in 2030 ranging from 0.001% of GDP in Latin America to 0.035% in India and a final GDP loss ranging from 0.3% of India to 0.006% of Western Europe with respect to the no protection case. Darwin and Tol (2001), Bosello et al. (2007) and Bigano et al. (2008) all use static CGE models for their assessment. The three studies, in accordance with Deke et al. (2002), highlight limited impacts on GDP, but considerable in absolute value, from sea level rise. In the no protection case, Darwin and Tol, (2001) estimate the annuitized total cost for a 50 cm. sealevel rise in 2100 of nearly US$ 66 billion. The highest losses among OECD countries are the nearly US$ 7 billion of Europe. Asian economies as a whole would lose US$ 42 billion. With an optimal protection policy, direct costs are US$ 4.4 billion for the world as a whole. In developed regions they are fairly small, they reach the highest level in the China-South KoreaTaiwan-Hong Kong region where they amount to 0.1% of 1990 expenditure. Bosello et al. (2007) and Bigano et al. (2008) highlight the highest GDP loss for South East Asia amounting to (-) 0.1% in 2050.
2.2.2 Assessing the impact of climate change on sea level rise induced land losses Estimates of potential dryland loss, in the absence of any protection intervention for each region analyzed are reported in Table 4. As described in Bosello et al. (2007), areas at risk of inundation for all coastal countries are extracted from the Global Vulnerability Analysis (Hoozemans et al., 1993). To obtain land losses, the starting point is Bijlsma et al. (1996) where information is reported for 18 selected countries. The exponent of the geometric mean of the ratio between areaat-risk and land loss for the 18 countries was then used to derive land loss for all other countries. Combined, these data specify, per country, the amount of land lost due to a sea-level rise of one metre. Land loss is assumed to be linear in sea-level rise thus results can be parameterized to any other sea-level rise scenario. This overly simplified procedure introduces obvious impre-
12
cisions in land loss estimates, however better estimates would have required the complex use of land elevation maps at the world level, not feasible within the present exercise. Moreover, it is also worth stressing that presently there is still high uncertainty concerning, on the one hand sea-level rise processes themselves, particularly at the local scale; for instance, scientists are still debating if and how the Mediterranean responded to global warming in the last few decades as apparently its level is not rising as its increasing salinity seems to compensate thermal expansion. On the other hand, the measure and forecast of their effects on coastal areas are determined by complex and not yet fully understood interactions between eustatism, geological and anthropic subsidence, tectonic movements, pressure differential of water and ice masses.
Table 4: Estimated land losses to sea-level rise in 2050 (Km2)
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA
2.3
+1.2°C 2388 130 395 226 1566 488 417 997 210 15161 9751 4223 20886 5041
+3.1°C 5000 271 826 474 3279 1022 874 2088 440 31745 20416 8842 43732 10555
Tourism
2.3.1 Introduction and background Climate change plays an obvious role in tourist destination choice. Although climate is by no means the only determinant in the choice of holiday destination (Crouch, 1995; Witt and Witt, 1995; Gossling and Hall, 2006a,b; Bigano et al., 2006a; Rosello et al., 2005), the “amenity of climate” is recognized as one of the major determinants of tourism flows (Maddison, 2001; Lise and Tol, 2002; Bigano et al., 2006b). The Mediterranean in particular, benefits from this 13
determinant, being close to the main holidaymakers of Europe’s wealthy, but cool and rainy, Northwest. Tropical islands are another example, where in the recipe of a dream holiday their “perfect” climate is a fundamental ingredient. Climate change would alter that, as tourists are particularly footloose. The currently popular holiday destinations may become too hot, and destinations that are currently too cool would see a surge in their popularity (Hamilton et al., 2005a,b; Hamilton and Tol, 2007; Amelung et al., 2007). Low ski resorts and winter tourism may be particularly vulnerable (Elsasser and Burki, 2002; Scott et al., 2004; 2007; OECD, 2007). This could have a major impact on some economies. Just consider that about 10% of world GDP is now spent on recreation and tourism, and that recent contributions highlight the importance of tourism in stimulating economic growth (Lee and Chang, 2008).
2.3.2 Assessing the impact of climate change on tourism expenditure This part of the research aims to assess the economic implications of climate change impacts on tourism activity; the necessary first step is thus, to define how tourism activity could be affected by a changing climate. In the present approach we proxy “tourism activity” with “tourism expenditure”, this in turn, is made directly dependent on the net number of tourists visiting a given location. As in Berrittella et al., (2006) and Bigano et al. (2008) estimations of tourism flows by region are obtained from version 1.2 of the Hamburg Tourism Model (HTM), an econometric simulation model of tourism flows in and between 207 countries. This were run on the temperature change scenarios chosen as reference: +1.2°C and +3.1°C with respect to 2000 in 20506. In the HTM, as described by Bigano et al. (2005)7 and Hamilton et al. (2005,a,b) to which the interested reader is addressed for a detailed description, tourists in a given country stem from a compounded calculation considering domestic tourists and international tourists’ arrivals and departures.
6
HTM is a dynamic model, for computational tractability in the following CGE exercise, we use the model statically, computing climate change impacts in 2050 and then interpolating linearly from 2001 to 2050. 7 Previous versions of this model were version 1.0, described by Hamilton et al. (2005a) and version 1.1, described by Hamilton et al. (2005b). The econometrics are inspired by Maddison (2001), Lise and Tol (2002) and Hamilton (2003), while the data are as in Bigano et al. (2005b). Further details, including papers and model code, can be found at http://www.uni-hamburg.de/Wiss/FB/15/Sustainability/htm.htm
14
Three econometrically estimated equations, for arrivals (Equation (1)) and departures (Equations (2) and (3)) define the core of the model. In these equations the variables are, respectively:
Total arrivals per year Land area (km2) Annual average temperature (C°) Length of coastline (km) Per capita income Total departures per year Population (in thousands) The number of countries with shared land borders Total domestic tourist trips per year The destination country The origin country
A G T C Y D P B H D O
Arrivals are defined by:
ln Ad = 5.97+ 2.05⋅10−7 Gd + 0.22 Td − 7.91⋅10−3 Td2 + 7.15⋅10−5 Cd + 0.80 ln Yd 0.97
(1)
2 adj
N = 139; R
0.96
0.07
2.21
3.03
0.09
= 0.54
Departures are determined following a two-step procedure: firstly the number of tourists “generated” by a country is defined, then it is subdivided between domestic tourists and those who travel abroad. The number of tourists that a country generates depends on the size of the population and on average income. The share of domestic tourists in total tourism depends on the climate in the home country and on per capita income. Missing observations are completed using two regressions. Total tourist numbers, D+H, where H is the number of domestic tourists are interpolated using
ln
(2)
Do + H o = −1.67+ 0.93ln Yo 0.83 0.10 Po
2 N = 63; Radj = 0.60
The ratio of domestic to total holidays was interpolated using ln
(
Ho = −3.75+ 0.83⋅10 −1 ln Go + 0.93⋅10−1 ln Co + 0.16⋅10 −1To − 0.29⋅10−3 To2 1.19 0.42 0.30 0.32 1.11 Do + H o
)
(3) + 0.16− 4.43⋅10−7 Yo ln Yo 0.12
1.24
2 N = 63; Radj = 0.36
15
The model is calibrated to 1995 data. Climate is proxied by the annual mean temperature. A number of other variables, such as country size, were included in the estimation, but these factors are held constant in the simulation. International tourists are allocated to all other countries on the basis of a general attractiveness index, climate, per capita income in the destination countries, and the distance between origin and destination. Again, other explanatory variables were included in the regression for reasons of estimation efficiency, but these are held constant in the simulation. The number of international tourists to a country is the sum of international tourists from the other 206 countries. Total tourism expenditure is then calculated multiplying the number of tourists times an estimated value of the average individual expenditure. Table 5 reports the absolute values of the changes in tourism expenditure in respect to the “no climate change” baseline so calculated and aggregated according to ICES regional detail.
Table 5: % changes in tourism expenditure in 2050 compared to no climate change baseline(US$ trillion)
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA
1.2°C -0.44 -0.26 1.86 -0.06 0.00 -0.07 1.36 -0.05 -0.50 -0.08 -0.07 -0.76 -0.29 -0.64
3.1°C -0.11 -0.07 0.48 -0.02 0.00 -0.02 0.35 -0.01 -0.13 -0.02 -0.02 -0.20 -0.07 -0.16
The model shows that regions at higher latitudes and altitudes will become more attractive to tourists, (both domestic tourists and those from abroad) and thus, will also experience an increase in expenditures on tourist services 8.
8 Tourists from the north west of Europe currently dominate international tourism, – the Germans and the British together account for 25% of the international tourist market – which implies that the world total of inter-
16
2.4
Agriculture
2.4.1 Introduction and background The relationships between climate change and agriculture are complex and manifold. They involve climatic and environmental aspects and social and economic responses. The latter can either take the form of autonomous reactions or of planned economic or technological policies. This picture is complicated further: climate change and agriculture interdependencies evolve dynamically over time, they often span over a large time and space scale and are still surrounded by large uncertainties. Climate change impacts on crop productivity can be substantial. According to the meta analysis proposed by Easterling et al. (2007) summarizing the extended literature on this topic, a moderate degree of warming can be beneficial to crops mainly due to a positive CO2 fertilization effect. However, when a given threshold is trespassed, impacts become negative and rapidly worsen. This pattern is common to main cereal cultivations worldwide, but much more pronounced when concerning the lower latitudes where the majority of developing countries are located. In these regions a positive effect on rice and wheat yields occurs roughly at +1°C with an increased productivity of approximately 5%, maize production appears negatively affected. A rise in temperature, above +1.5°C, negatively affects yields: for instance with a temperature increase scenario of +5°C, rice yields can decrease by 20% while maize and wheat yield can decrease by more than 50%. At mid-high latitudes, positive effects are estimated up until +3.5°C for wheat and rice and +2°C for maize. Wheat, rice and maize yield increases can peak by 7%, 10% and 5% at the “optimum” temperature increases of +1.8°C, +2.2°C and +0.8°C, respectively. For higher temperature increases, yields drop everywhere: 20% for wheat and rice and 8% for maize when temperature increases by 5°C. The literature assessing the economic impact of these changes presents quite small figures. Global studies report for the world as a whole, a loss ranging from –0.05% to a slight gain of 0.01% in terms of welfare in case of a doubling CO2 concentration or between -1.5% and +2.6% of global agricultural GDP by 2080 (Fisher et al. 2002). Regional studies show higher
national tourist numbers initially falls because of climate change. The model also shows that the effect of climate change is much smaller than the combined effects of population and economic growth, at least for most countries.
17
and more dispersed estimations: welfare changes range between –5.5% in China (Kane, 1992) and, -5% in India (Kainuma et al. 2003) to +2.7% in the USA (Adams et al. 1999). Economic losses are concentrated in developing countries belonging to the Asian regions and to SubSaharan Africa that is likely to surpass Asia as the most food-insecure regionby the middle of the century. Negative impacts on some types of crop productivity are also expected in the developed world: Australia, New Zealand and in Mediterranean Europe, however, these regions are more able to adapt technologically, institutionally and socio economically and are thus, likely to suffer negligibly. Accordingly, it is worth stressing that the final economic impact of climate change on agriculture is crucially driven by how natural and human adaptation mechanisms are considered. In general, strong negative impacts on yields highlighted by exercises neglecting adaptation turn into much smaller losses or even slight gains when proper adaptation options are modelled. Examining the rather extended literature on the subject (see: Fisher et al., (1993); Reilly et al., (1994); Rosenzweigh and Parry, (1994); Rosenzweigh and Hillel, (1998); Antle et al., (2001); Bindi and Moriondo (2005); Easterling et al. (2007) for a non exhaustive list) it is possible to grasp some rough estimates of damage reducing power of adaptation in agriculture which ranges from 40% to 98% of total climatic damage for a doubling of CO2 concentration. Needless to say, modelling adaptation in agriculture is a process which is prone to large uncertainty, especially the assumption made by the majority of authors regarding the negligible cost of farm-level adaptation, and is an open issue. Should adaptation be less effective or more costly than hypothesized, much stronger adverse consequences of climate change on agricultural production and welfare cannot be excluded.
2.4.2 Assessing the impact of climate change on land productivity To assess climate change impacts on agriculture we fed our temperature scenarios into a simple agricultural productivity module developed by Tol (2002). This module calibrates a reduced-form function linking temperature, CO2 concentration and yield from a meta analysis of the relevant literature among which data from Rosenzweig and Hillel, (1998) are predominantly important. This last study is quite dated, but remains one of the few which reports detailed results from an internally consistent set of crop modeling studies for 12 world regions and 6 crop varieties, thus allowing a reasonable degree of comparability.
18
We partly updated the Rosenzweig and Hillel,(1998) figures using more recent and detailed information from Bindi and Moriondo (2005) for the North African and Mediterranean Europe regions. In our estimates the role of the CO2 fertilization effect is explicitly taken into account, but we do not consider the role of farm-level adaptation. This was for two reasons: firstly, as mentioned, the role of farm level adaptation is highly uncertain. Thus many assumptions on this ground could have been (un)justifiable. More importantly, we wanted to capture marketdriven adaptation by means of the following exercise performed with the CGE model. More specifically, we want to isolate the effect on agricultural production triggered by changes in relative prices. The estimated impact on agricultural land productivity is reported in table 9.
2.5
Energy demand
2.5.1 Introduction and background It is reasonable to assume that an increased mean temperature will increase the amount of energy demanded for cooling purposes and decrease that for heating purposes9. It is also reasonable to assume that with a low level of temperature change the increased spending on cooling should be dominated by savings from reduced heating expenditure, the situation however reverses at some point, when levels of temperature change are higher. It can thus be expected that the relationship between energy demand and mean global temperature might be “U” shaped. An important question is whether we are already to the right–of-the-minimum of such a curve, in which case global energy consumption will rise with a higher global mean temperature, or whether we are still on the left-to-the-minimum portion of the curve, in which case global energy consumption will first decline and then eventually rise as global mean temperature increases (Hitz and Smits, 2004). An important role in this analysis is played by wealth or income effects. Indeed, if - as it seems - air conditioning is strongly correlated with income, even if we had the same climate in future years the demand pattern between heating and cooling would change because of changes in income. This introduces uncertainty on the analysis of adaptation cost in the en9
Although adaptation to warmer temperature does not necessarily have to be through extra energy consumption, i.e. it can be partially implemented through passive building cooling, design, behavioural change, etc.
19
ergy sectors, and greater care must be taken in assuming on which branch of the curve we are placed, based on current energy consumption. To our knowledge there are only two studies that have estimated the effects of climate change on the demand for global energy: Tol (2002) and Bigano et al. (2006). Tol based his extrapolations on a UK-specific model that relates the energy used for heating or cooling to degree days, per capita income, and energy efficiency. Climatic change is likely to affect the consumption of energy via decreases in the demand for heating space and increases in demand for cooling and Tol (2002), hypothesized that both relationships are linear. Economic impacts were derived from energy price scenarios and extrapolated to the rest of the world. Energy efficiency is assumed to increase, lessening costs. According to Tol’s (2002) best guess parameters, by 2100, benefits (reduced heating) are about 0.75% of GDP and damages (increased cooling) are approximately 0.45%. The global savings from reduced demand for heating remain below 1% of GDP through to 2200. However, by the 22nd century, they begin to level off due to increased energy efficiency. For cooling, the additional amount spent rises to just above 0.6% of GDP by 2200. Thus, throughout the next two centuries the net energy demand decreases, suggesting we are on the downward sloping part of the energy demand- temperature curve. These finding are somehow confirmed by Bigano et al. (2006a). They conducted a dynamic panel data econometric estimation of the demand for coal, gas, electricity, oil and oil products by residential, commercial and industrial users in OECD and (a few) non-OECD countries, to derive long-run elasticities for temperature. The main findings highlighted that residential demand responds negatively to temperature increases, (except in the case of coal), pointing at a prevalence of heating needs in determining residential demand. By contrast, industrial demand is insensitive to temperature increases. In the case of the service sector, only electricity demand displays a mildly significant negative elasticity to temperature changes. The estimated elasticities range from –0.6 for electricity to -3 for oil products. Transposed to a scenario of a 1°C increase in mean global temperature, this may configure a decrease in households’ demand ranging from -1.5% to -7% for electricity and from -6% to 40% for oil products, depending on the region. However, studies for the US provide mixed results. With the exception of Rosenthal et al. (1995), Cline (1992) and Fankhauser (1995) found a net increase in electricity expenditure for a 2.5°C increase in mean global temperature amounting to 9900 and 6900 US $ billion, respectively whilst more recently, Mendelsohn (2001) showed that energy costs will increase 20
even with an approximate 1°C increase. Since the United States consumes about one fourth of global energy, this may be an indication that global energy demand will immediately increase as temperatures rise. Concluding, the still limited quantitative evidence does not allow us to assess with certainty the impact of climate change on energy demand, and consequently the cost of the associated demand shifts. However, it can be noted that when these costs are positive, they appear to be a tiny percentage of GDP. All the cited studies however, have been obtained by a direct cost methodology or in partial equilibrium. In other words the rebound effects on the overall economic context of the recomposition of demand have not been taken into account. These effects can potentially be very large and need to be carefully assessed. This is the aim of the present research.
2.5.2 Assessing the direct impact of climate change on energy demand To assess climate change impacts on energy demand the present research elaborates on the results proposed by De Cian et al. (2007). They estimate household energy demand on a macro dynamic panel dataset spanning from 1978 to 2000, for 31 countries. In this approach, pioneered by Balestra and Nerlove (1966), energy demand is modelled as a dynamic process, depending on prices, income and temperatures, but also on the lagged value of energy demand. This is in order to capture the relative influence of short-run and long-run changes. Separate regressions are specified for electricity, gas and oil products demand, each defined as an autoregressive process. Another important issue is capturing weather variability. Degree days are particularly popular in the studies dealing with residential demand of space heating energy (Madlener and Alt, 1996; Parti and Parti, 1980) however, given the aim to determine the sensitivity of energy demand with respect to temperature variations, De Cian et al., (2007) include temperatures directly (see also Moral -Carcedo et al., (2005); Mansur et al. (2004); Henley and Peirson, (1998); Asadoorian et al., (2006) for similar approaches) and segment them using seasonal (spring, fall, summer and winter) data. An interesting feature of the exercise is the introduction and treatment of regional differences as temperature effects are expected to vary, especially between warm and cold countries.
21
A clustering algorithm has been used (see Kaufman and Rousseeuw (2005), Hartigan (1975) and Everitt (1974)) to split the sample into partitions, which become finer and finer. The metric used is the Euclidean distance measured on the annual average, maximum and minimum temperature. Countries have then been grouped into the three main clusters and identified as: Mild Countries: Austria, Belgium, Denmark, France, Germany, Ireland, Luxembourg, Netherlands, New Zealand, Switzerland, Greece, Hungary, Italy, Japan, Korea, Portugal, South Africa, Spain, Turkey, United Kingdom, United States. Hot Countries: Australia, India, Indonesia, Mexico, Thailand, Venezuela. Cold Countries: Canada, Finland, Norway, Sweden. Group dummies for the different temperature groups are also introduced into the regression. The use of dummies in the level allows us to capture the different effects of temperature increases between groups on the intercept. By interacting all the covariates with the two dummies we capture the different effects on the slope. Table 6 reports the estimation of the model. The results reveal the presence of a cooling and heating effect. Summer temperature leads to higher annual electricity demand to feed a higher usage of air conditioners; the other fuels instead tend to respond negatively to temperature increases, especially when occurring in fall, spring or winter. Summer temperature is relevant only in the model for electricity demand, whereas gas, oil products respond significantly only to temperature variations that take place in winter, fall or spring. Within each model specification a further distinction between cold, hot and mild countries emerges. Consider for example the demand for electricity. The effect of summer temperature is significant in all groups, but with a different sign. In very cold countries (d0 = 1) an increase in summer temperature of 1% reduces annual demand by 0.508%. In very hot countries (d2 = 1) it increases electricity demand by 1.695 %. In mild countries (d1 = 1), which is the largest group, the increase in electricity demand is lower and it equals 0.542 %. Own price and income long-run elasticities, computed are well between the ranges reported by the relevant literature (see De Cian et al. (2007) for details) .
22
Table 6: Full sample regression, Within Estimator Electricity
Dependent Variable: yit
yit-1 gdppc pi pj yit-1d1 gdppcd1 pid2 pjd2 yit-1d1 gdppcd1 pid2 pjd2 Summer Summerd1 Summerd2 Winter Winterd1 Winterd2 Spring Springd1 Springd2 Fall Falld1 Falld2 OBS T N R-sq AIC
Gas Coefficient (t-statistic) 0.705 (5)*** -0.199 (-0.85) -0.306 (-2.34)** -0.310 (-1.36) 0.233 (1.63) 0.324 (1.66)* 0.172 (1.28) 0.370 (1.6) -0.094 (-0.38) 0.637 (1.87)** 0.019 (0.08) 0.550 (2.05)**
0.892 (32.22)*** 0.034 (0.91) -0.052 (-1.57) 0.013 (0.41) 0.049 (1.3) 0.037 (1.1) -0.079 (-1.25) 0.117 (2.31)** 0.034 (0.99) -0.508 (-2.36)** 0.542 (2.31)** 1.659 (3.89)*** -0.029 (-0.69) -0.060 (-0.91) 0.406 (1.26) -0.172 (-1.11) -0.169 (-0.86) 0.935 (2.1)**
-0.243 (-1.15) -0.082 (-0.33) 0.370 (0.5) 0.256 (0.49) -1.285 (-2.24)** -0.588 (-0.54)
550 22 25 0.9875 -2219.949
418 19 22 0.936 -702.2341
Oil Products 0.854 (10.75)*** -0.292 (-1.01) -0.218 (-1.67)* -0.004 (-0.05) 0.052 (0.57) 0.414 (1.69)* 0.179 (1.25) -0.019 (-0.22) 0.145 (0.89) 0.230 (1.01) 0.073 (0.3) -0.309 (-1.23)
0.080 (0.84) -0.571 (-2.61)*** 0.511 (0.45) -0.223 (-0.63) -0.938 (-1.88)* -1.276 (-0.75) -0.858 (-1.56) 0.784 (1.41) 0.821 (1.49) 418 19 22 0.8748 -378.0887
*** significant at 1% ** significant at 5% * significant at 10%
Source: de Cian et al. (2007)
Table 7 reports the long run temperature demand elasticities of different energy vectors, when they are statistically significant. Table 7 : Long-run temperature elasticities (only significant elasticities have been reported) Electricity Group Cold Group Mild
-4.704 0.315
Gas
Oil ducts
Pro-
3.911
Electricity
-4.356
Group Hot 15.676 Source: De Cian et al. (2007)
8.657
23
Gas
6.425
Oil Products
The pattern is the same as in the short run, but with a bigger magnitude. In the long run, when the stock of equipment and appliances can also be adjusted, the effect of temperature, as well of income and prices, is larger. These elasticities have been used to compute the variation of energy demand in respect to the temperature increase scenarios used as reference.
2.6
Health
2.6.1 Introduction and background Of the many impacts of climate change, those on human health are often placed amongst the most worrying (e.g., Smith et al., 2001). The impacts of climate change on human health are many and complex. Global warming would increase heat-related health problems, which mostly affect people with pre-established cardiovascular and respiratory disorders. On the other hand, global warming would reduce cold-related health problems, again most prevalent in people with cardiovascular disorders. Climate change would affect the range and abundance of species carrying diseases, and would also affect the virulence of those same diseases. Malaria, in particular, is generally thought to increase because of climate change. Other vector-borne diseases may increase or decrease, but currently cause much less victims than malaria does. Climate change would allow diseases to invade immunologically naïve populations with unprepared medical systems. Climate change would affect water-borne diseases too, with cholera and diarrhoea being potentially the most problematic (McMichael et al., 2001). Human health therefore figures prominently in assessments of the impacts of climate change, however the quantification of these impacts is extremely difficult., it is hard to record changes in health status which have actually occurred in response to observed trends in climate over recent past. Some studies estimated that in 2000, climate change caused the loss of over 150,000 lives and 5,500,000 DALYs (Campbell-Lendrum et al., 2003; Ezzati et al., 2004; McMichael, 2004). Secondly, when future estimations are involved, there is the crucial complication to include correctly possible autonomous acclimatisation processes, be they physiological, behavioural or driven by social economic conditions, which are also fundamental in determining final vulnerability (Tol et al., 2007). This said, according to the IPCC (2007) AR4 projected trends in climate-change-related exposures of importance to human health will: increase the number of people suffering from 24
death, disease and injury from heat waves, floods, storms, fires and droughts; continue to change the range of some infectious disease vectors; have mixed effects on malaria depending on the geographical context; increase the burden of diarrhoeal diseases; increase cardiorespiratory morbidity and mortality associated with ground-level ozone; increase the number of people at risk of dengue; finally bring some benefits to health, including fewer deaths from cold. The welfare implication of health impacts rely on the methodology of direct costs, namely, damage equals price times quantity. In the case of human health, the price is typically equal to the value of a statistical life, which is based on estimates of the willingness to pay to reduce the risk of death or diseases, or the willingness to accept compensation for increased risk (see Viscusi and Aldy, 2003, for a review). The global economic value of loss of life due to climate change ranges between around US$6 billion and US$88 billion, in 1990 dollar prices (IPCC (2007)). Few studies try to assess higher order costs. Jorgenson et al. (2004) do this for the USA, but health effects on labour supply are not disentangled from other impacts; Bosello et al. (2006) perform a world estimate, using a static CGE model, they find that health impacts on economic activity are negative in the developing world, but that GDP is only marginally affected with a 0.1% loss for a 1°C increase in temperature. 2.6.2 Assessing the direct impacts of climate change on human health We evaluate the impacts on human health changes - i.e. in mortality and morbidity associated with malaria, schistosomiasis, dengue, diarrhoea, cardiovascular and respiratory diseases in the thirteen regions of ICES applying the same methodology of Bosello et al. (2006). Estimates of the change in mortality due to vector-borne diseases (viz., malaria, schistosomiasis, dengue fever) as the result of a one degree increase in the global mean temperature are taken from Tol (2002). The estimates result from overlaying the model-studies of Martens et al. (1995, 1997), Martin and Lefebvre (1995), and Morita et al. (1994) with mortality and morbidity figures of the WHO (Murray and Lopez, 1996). These studies suggest that the relationship between global warming and malaria is linear. This relationship is assumed to apply to schistosomiasis and dengue fever as well. We follow the same methodology here. To account for changes in vulnerability possibly induced by improved access to health care facilities associated to improvement in living standards (read GDP growth) we use the relationship between per capita income and disease incidence developed by Tol and 25
Dowlatabadi (2001),10 using the projected per capita income growth of the ICES regions for the countries within those regions. For diarrhoea, we follow Link and Tol (2004), who report the estimated relationship between mortality and morbidity on the one hand, and temperature and per capita income on the other, using the WHO Global Burden of Disease data (Murray and Lopez, 1996). Martens (1998) reports the results of a meta-analysis of the change in cardiovascular and respiratory mortality for 17 countries. Tol (2002a) extrapolates these findings to all other countries, using the current climate as the main predictor. Cold-related cardiovascular, heat-related cardiovascular, and (heat-related) respiratory mortality are specified separately, as are the cardiovascular impacts on the population aged below 65 and above. Heat-related mortality is assumed to only affect the urban population. Scenarios for urbanization and aging are based on Tol (1996, 1997).11 We use this model directly on a country basis, before aggregating to the regions of ICES. Besides the changes in labour productivity, changes in health care expenditures are also estimated. The literature on the costs of diseases is modest and few papers can be used as reference. Kiiskinen et al. (1997) report the average costs of cardiovascular diseases, at $21,000 per case, for Finland. Blomqvist and Carter (1997), Gbesemete and Gerdtham (1992), Gerdtham and Jönnson (1991), Getzen (2000), Govindaraj et al. (1997), Hitires and Posnett (1992), and di Matteo and di Matteo (1998) estimate the income elasticity of health expenditures for countries in the OECD, Latin America and in Africa for the period 1960-1991. The average is 1.3. We use this to extrapolate the Finnish costs of cardiovascular diseases to other countries. Weiss et al. (2000) report the costs of asthma for the USA. The direct costs12 amount to $430 per case, or $40,000 per year diseased.13 We assume that asthma is representative for all respiratory diseases, and again extrapolate this to other countries using an income elasticity of 1.3. The costs of vector borne diseases are taken from Chima et al. (2003), who report the expenditure on prevention and treatment costs per person per month. Their data suggest the following relationships 10
Vulnerability to vector-borne diseases strongly depends on basic health care and the ability to purchase medicine. Tol and Dowlatabadi (2001) suggest a linear relationship between per capita income and health. In this analysis, vector-borne diseases have an income elasticity of –2.7 (Tol and Heinzow, 2003). 11 The income elasticity of the share of the population over 65 is 0.25. Urbanisation follows
U (t ) = U (1995)
0.031Y (t ) − 0.011PD(t ) 1 + 0.031Y (1995) − 0.011PD(1995) 1 + 0.031Y (t ) − 0.011PD(t ) 0.031Y (1995) − 0.011PD(1995)
where U is the level of urbanisation, Y is per capita income and PD is population density. Weiss et al. (2000) also estimate the indirect costs to the economy. 13 The average treatment for asthma lasts 4 days. 12
26
(1) P = 0.1406+ 0.0026 Y (0.3103)
(0.0008)
(2) T = −0.4646+ 0.0053Y (0.8217)
(0.0018)
where P is monthly prevention costs ($/capita), T is monthly treatment costs ($/cap) and Y is income per capita ($/cap). We scale this up with the increase in mortality. The resulting changes in national mortality and morbidity aggregated to the ICES regions are reported in table 8. Table 8: Additional number of deceases due to climate change (1000 people, reference year 2050)
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA
2.7
Vector borne and enteric diseases 1.2°C 3.1°C 12 31 2 5 2 6 0 1 1 3 173 450 0 0 16 41 5 12 782 2029 54 141 2 5 14 36 37 97
Cardio Vascular diseases 1.2°C 3.1°C -170 -431 -73 -183 -115 -292 -54 -136 -281 -718 -21 -53 -95 -240 -19 -48 -50 -126 -19 -40 -142 -345 -966 -2463 -26 -60 -18 -37
Respiratory diseases 1.2°C 3.1°C 4 15 3 10 0 0 0 0 5 13 7 20 5 15 30 73 48 79 99 271 204 557 4 12 46 129 42 120
Total 1.2°C -154 -67 -113 -53 -275 159 -90 27 2 861 116 -960 33 62
3.1°C -385 -167 -286 -135 -702 417 -225 67 -34 2260 353 -2446 106 180
Results from the direct impact assessment exercises
To determine the economic consequences of the different impacts assessed with the ICES model, these need first to be translated into changes in economic variables existing in the model. Accordingly: land losses due to sea level rise have been modelled as percent decreases in the stock of productive land by region; changes in mortality/morbidity have been modelled as changes in regional labour productivity; changes in land productivity by crop is already in the format of suitable input as ICES includes factor specific productivity as an exogenous parameter. In these three cases, variables exogenous to the model are involved and their modification is straightforward. Changes in tourists expenditure are modelled as changes in demand addressing the “market services sector” which includes recreational services; changes in health care expenditures 27
are translated into changes in the public and private demand for the “non market services” sector which includes health services; changes in regional demand for oil, gas and electricity are modelled as changes in the demand for the output of respective industries. In these last cases variables which are endogenous to the model are concerned. To model their changes we adopted the following procedure; the computed percentage variations in the demands have been imposed as exogenous shifts in the respective demand equations. The implicit assumption is that the starting information refers to partial equilibrium assessment thus with all prices and income levels constant. The model is then left free to determine the final demand adjustments. Modification in demand structure however, imposes to comply with the budget constraint, so we compensated the higher level of public consumption with a lower lever of aggregate private consumption and, within the latter, we compensated the changed consumption of healthcare and tourism services and energy with opposite changes in expenditure shares for all other sectors. Table 9 summarizes the results of the direct impact assessment exercises (we report only year 2050 for exemplification) once they have been translated in suitable inputs for the ICES model. Table 9: Climate change impacts as inputs for the ICES model (% change wrt baseline, reference year 2050 reference temperature +1.2, 3.1 °C wrt 2000) Region
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA
Health Labour Public Productivity Expenditure 1.2°C 3.1°C 1.2°C 3.1°C -0.06 -0.18 -0.15 -0.28 0.01 0.01 -0.10 -0.18 0.06 0.16 -0.35 -0.88 0.09 0.23 -0.47 -1.18 0.16 0.40 -0.41 -1.03 -0.43 -1.14 0.57 1.62 0.09 0.22 0.03 0.24 -0.28 -0.69 2.02 4.42 -0.22 -0.34 1.34 1.82 -0.31 -0.84 0.47 1.34 -0.11 -0.30 0.28 0.76 0.14 0.37 0.65 1.80 -0.11 -0.32 1.05 2.96 -0.14 -0.39 0.68 1.98
Land Productivity Private Expenditure 1.2°C 3.1°C -0.02 -0.03 0.00 -0.01 -0.01 -0.03 -0.01 -0.02 -0.01 -0.03 0.04 0.11 0.00 0.00 0.10 0.23 0.10 0.14 0.07 0.19 0.06 0.17 0.06 0.17 0.06 0.17 0.07 0.19
28
Wheat 1.2°C -5.66 -1.14 1.50 -1.13 -6.12 -7.78 -0.74 -12.81 -8.40 -9.89 -2.96 0.93 2.45 -6.69
3.1°C -18.89 -8.33 -7.74 -10.50 -21.92 -17.00 -12.33 -42.14 -32.40 -26.33 -14.92 -2.30 -0.54 -21.71
Rice 1.2°C -6.19 -4.62 -5.90 -2.64 -7.47 -2.90 -1.87 -10.78 -11.73 -7.17 -4.89 0.50 0.34 -6.61
3.1°C -20.37 -18.94 -26.01 -13.57 -24.64 -7.41 -14.31 -41.00 -38.52 -21.43 -18.89 -3.61 -4.98 -23.38
Other Cereal Crops 1.2°C 3.1°C -8.18 -25.15 -2.00 -11.84 50.00 107.82 -4.60 -18.35 -9.73 -30.10 -3.11 -7.38 -2.24 -15.17 -12.62 -45.97 -13.60 -43.12 -8.81 -25.36 -6.61 -22.99 -1.42 -8.25 -1.15 -8.50 -8.25 -25.78
Table 9: Climate change impacts as inputs for the ICES model (% change wrt baseline, reference year 2050 reference temperature +1.2, 3.1 °C wrt 2000) (cont.) Sea-Level Rise Region
Land Losses 1.2°C -0.03 -0.01 -0.02 -0.02 -0.01 -0.01 0.00 -0.02 0.00 -0.07 -0.20 -0.05 -0.32 -0.02
3.1°C -0.05 -0.01 -0.04 -0.05 -0.01 -0.01 -0.01 -0.04 -0.01 -0.14 -0.43 -0.09 -0.66 -0.05
USA Med_Europe North_Europe East_Europe FSU KOSAU CAJANZ NAF MDE SSA SASIA CHINA EASIA LACA Notes: nss non statistically significant. Expenditure flows in US$ trillions
Tourism Market ServiExpenditure ces Demand Flows 1.2°C -0.68 -1.86 7.54 -2.46 0.00 -1.31 5.54 -2.52 -4.67 -4.43 -1.21 -4.99 -4.69 -2.68
Households' Energy Demand Natural Gas
1.2°C 3.1°C 1.2°C 3.1°C -1.76 -0.04 -0.11 -13.67 -4.82 -0.02 -0.07 -12.68 19.47 0.18 0.48 -13.75 -6.37 -0.006 -0.02 -12.93 -0.01 0.00001 0.00003 -13.02 -3.39 -0.007 -0.02 nss 14.30 0.13 0.35 -5.05 -6.52 -0.005 -0.01 -8.60 -12.07 -0.50 -0.13 -13.12 -11.46 -0.008 -0.02 nss -3.12 -0.007 -0.02 nss -12.90 -0.076 -0.20 nss -12.11 -0.03 -0.07 nss -6.92 -0.06 -0.16 nss
3.1°C -35.31 -32.76 -35.51 -33.41 -33.65 nss -13.04 -22.22 -33.89 nss nss nss nss nss
Oil Products 1.2°C -18.52 -15.84 -15.52 -17.39 -17.39 -13.03 -12.63 -13.25 -17.39 -6.51 nss nss nss nss
3.1°C -47.84 -40.91 -40.09 -44.92 -44.92 -33.66 -32.63 -34.22 -44.92 -16.83 Nss Nss Nss Nss
Electricity 1.2°C 0.76 0.76 -2.20 0.76 0.75 12.31 -4.80 5.95 0.74 16.35 20.38 20.38 20.38 21.37
3.1°C 1.96 1.96 -5.68 1.97 1.94 31.81 -12.40 15.37 1.92 42.23 52.65 52.65 52.66 55.20
It is evident that impacts are highly differentiated across regions and by typology. First, except for the case of land losses to sea level rise, they are not all necessarily negative. For instance, labour and land productivity could decrease in some regions, whilst increasing in others responding to regionally differentiated climatic stimuli and to different health characteristics of the labour force or crop and land characteristics. Consistently with a consolidated literature, land productivity tends to increase in mid to high latitude countries and decrease in low latitude countries. Labour productivity decreases where vector borne diseases dominate (the developing world) and tend to increase elsewhere where reduced cold related mortality overcompensates the increased heat related mortality. Second, impacts concern both the supply and the demand side of the economic system. In the first case they can be defined as “good” or “bad” quite unambiguously 14: for instance a decrease in labour productivity due to adverse health impacts or in the availability of productive land due to sea-level rise are sure initial losses for the economic system. In the second case, when agents’ preferences change, determining the “quality” of the impact is more difficult. Indeed, when the demand for a given good or service (say e.g. energy demand) decreases, consumer expenditure is typically reallocated towards all the other goods and services. Consequently, the final impact on utility 14
In this statement we disregard distributional implications across income groups or classes within the same country.
29
cannot be determined a priori, but only at the end of a fully fledged general equilibrium exercise. As said, demand-side impacts involve changes in demand for market services, changes in households’ energy demand and changes in non market services demand. The first two are somewhat larger than other impacts and affect sectors of the economy which generate high value added. Consistently with changes in tourism flows, the demand for market services increases in colder regions whose climatic attractiveness also increases. The result of the CAJANZ aggregate is dominated by the Canada effect. Market services’ demand decreases elsewhere; note particularly the negative impact on the “hotter” Mediterranean Europe. The use of gas and oil products declines everywhere as less products are necessary for heating purposes, while electricity demand increases especially in hotter regions reflecting a higher use of air conditioning. This general picture points out that negative impacts are clearly concentrated in developing countries. This highlights their greater vulnerability to climate change with respect to developed economies which is a combination of a higher exposure and sensitivity.
30
References Adams, R. M., McCarl, B. A., Segerson, K., Rosenzweig, C., Bryant, K. J., Dixon, B. L., Conner, R., Evenson, R. E., & Ojima, D. (1999), ' The Economic Effects of Climate Change on U.S. Agriculture, ' in The Impact of Climate Change on the United States Economy, R. O. Mendelsohn & J. E. Neumann, eds. (eds.), Cambridge University Press, Cambridge, pp. 18-54. Amelung, B., S.Nicholls, and D.Viner (2007), “Implications of Global Climate Change for Tourism Flows and Seasonality”, Journal of Tourism Research 45: 285-297. Antle, J., Apps M., Beamish R., Chapin, T. , Cramer, W., Frang, J., Laine,J., Lin Erda, J. Magnuson, I. Noble, J. Price, T. Prowse , T. Root,E. Schulze, O. Sirotenko, B. Sohngen, J. Soussana (2001), “Ecosystems and Their Goods and Services”, in Climate Change 2001: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, J. McCarty, O.F. Canziani, N.Leary, D.J. Dokken and K.S. White, Eds., Cambridge University Press, Cambridge, UK, 237-315. Asadoorian O. M., R. Eckaus and C.A. Schlosser (2006), “Modeling Climate feedbacks to Energy Demand: The Case of China”, MIT Joint Program on the Science and Policy of Global Change, Report No. 135. Balestra, P. and M. Nerlove (1966), “Pooling cross-section and time-series data in the estimation of a dynamic model: The demand for natural gas”, Econometrica, 31: 585-612. Berrittella, M., Bigano, A., Roson, R. and R.S.J. Tol (2006), “A general equilibrium analysis of climate change impacts on tourism”, Tourism Management vol 25. Bigano, A., Bosello, F., Roson, R. and R.S.J. Tol (2008) “Economy-wide impacts of climate change: a joint analysis for sea level rise and tourism", Mitigation and Adaptation Strategies for Global Change, Vol. 13, n. 8. Bigano, A., F. Bosello and G. Marano (2006a) “Energy Demand and Temperature: a Dynamic Panel Analysis”, Fondazione ENI Enrico Mattei Working Paper No. 112.06. Bigano, A., J.M.Hamilton, D.J.Maddison, and R.S.J.Tol (2006b) “Predicting Tourism Flows under Climate Change -- An Editorial Comment on Goessling and Hall (2006)”, Climatic Change. 79: 175-180. Bigano, A., J.M. Hamilton and R.S.J. Tol (2006c) “The Impact of Climate on Holiday Destination Choice”, Climatic Change 76 (3-4): 389-406. Bijlsma, L., Ehler, C. N., Klein, R. J. T., Kulshrestha, S. M., McLean, R. F., Mimura, N., Nicholls, R. J., Nurse, L. A., Perez Nieto, H., Stakhiv, E. Z., Turner, R. K., & Warrick, R. A. (1996) Coastal Zones and Small Islands. In: R. T. Watson, M. C. Zinyowera, & R. H. Moss, eds. (eds.) Climate Change 1995: Impacts, Adaptations and Mitigation of Climate Change: Scientific-Technical Analyses -- Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Climate Change, 1 edn., Cambridge University Press, Cambridge: 289-324. Bindi, M. and M. Moriondo, (2005), “Impact of a 2 °C global temperature rise on the Mediterranean region: Agriculture analysis assessment”, in Climate change impacts in the Mediterranean resulting from a 2 °C global temperature rise, Giannakopoulos, C., Bindi, M., Moriondo, M., and T. Tin, (Eds.), WWF , 54–66. Blomqvist, A. G. and Carter, R. A. L. (1997), 'Is health care really a luxury?', Journal of Health Economics, 16, 207-229. Bosello, F., Lazzarin, M., Roson, R. and R.S.J. Tol (2007), “Economy-wide estimates of climate change implications: sea-level rise”, Environment and Development Economics, 37:549–571 31
Bosello, F., Roson, R. and R.S.J. Tol (2006), “ Economy wide estimates of the implications of climate change: human health” Ecological Economics, 58, 579-591 Breil, M., Gambarelli, G., and P.A.L.D. Nunes (2005), “Economic valuation of on site material damages of high water on economic activities based in the city of Venice: results from a doseresponse-expert-based valuation approach”, FEEM Note di Lavoro 53.05 Burniaux J-M., Truong, T.P., (2002) GTAP-E: An Energy-Environmental Version of the GTAP Model. GTAP Technical Paper n.16 (www.gtap.org). Campbell-Lendrum, D., A. Pruss-Ustun and C. Corvalan (2003), “How much disease could climate change cause?” Climate Change and HumanHealth: Risks and Responses, A. McMichael, D. Campbell-Lendrum, C. Corvalan, K. Ebi, A. Githeko, J. Scheraga and A.Woodward, (eds.),WHO/WMO/UNEP, Geneva, 133-159. CEC (2007) Limiting Global Climate Change to 2 degrees Celsius The way ahead for 2020 and beyond. Commission Staff Working Document. Brussels Chima, R. I., Goodman, C. A., and Mills, A (2003), 'The economic impact of malaria in Africa: a critical review of the evidence', Health Policy, 63, 17-36. Cline, W.R. (1992), “The Economics of Climate Change”, Institute for International Economics, Washington DC Crouch, G.I. (1995): “A Meta-Analysis of Tourism Demand”, Annals of Tourism Research, 22 (1): 103-118. Darwin, R. F. and Tol, R. S. J. (2001), ' Estimates of the Economic Effects of Sea Level Rise, ' Environmental and Resource Economics, 19, 113-129. De Cian E., Lanzi, E. And R. Roson (2007), ”The Impact of Temperature Change on Energy Demand: A Dynamic Panel Analysis”, FEEM Note di Lavoro, 46.2007 Deke, O., Hooss, K. G., Kasten, C., Klepper, G., & Springer, K. (2001) Economic Impact of Climate Change: Simulations with a Regionalized Climate-Economy Model. Kiel Institute of World Economics, Kiel, 1065. Dennis, K.C., Niang Diop, I. and R.J. Nicholls (1995) Sea Level Rise and Senegal: Potential Impacts and Consequences. J. of Coast. Res. 14: 243-261. Di Matteo, L. and di Matteo, R. (1998), 'Evidence on the determinants of Canadian provincial government health expenditures: 1965-1991', Journal of Health Economics, 17, 211-228. Dixon, P. and Rimmer, M., (2002) Dynamic General Equilibrium Modeling for Forecasting and Policy. North Holland. Easterling, W.E., P.K. Aggarwal, P. Batima, K.M. Brander, L. Erda, S.M. Howden, A. Kirilenko, J. Morton, J.-F. Soussana, J. Schmidhuber and F.N. Tubiello (2007), “Food, fibre and forest products” in 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, UK, 273-313. Elsasser, H. and R.Bürki (2002), “Climate change as a threat to tourism in the Alps”, Climate Resarch, 20: 253-257. Everitt, B. (1974). Cluster Analysis. London: Heinemann Education Books. Ezzati,M., Lopez, A., Rodgers A. and C. Murray, (2004), “Comparative Quantification of Health Risks: Global and Regional Burden of Disease due to Selected Major Risk Factors”, Vols. 1 and 2.World Health Organization, Geneva, 2235 pp. Fankhauser, S. (1994), “Protection vs. Retreat -- The Economic Costs of Sea Level Rise”, Environment and Planning, A, 27: 299-319. 32
Fankhauser, S. (1995), “Valuing Climate change. The Economics of the Greenhouse”, Earthscan, London. Fankhauser, S., Smith, J.B., and Tol, R.S.J. (1999), “Weathering Climate Change: Some Simple Rules to Guide Adaptation Decisions” Ecological Economics 30 : 67-78 Fischer, G., Frohberg, K., Parry, M. L., & Rosenzweig, C. (1993), “Climate Change and World Food Supply, Demand and Trade”, in Costs, Impacts, and Benefits of CO2 Mitigation, Y. Kaya et al., eds. (eds.), pp. 133-152. Fischer, G., M. Shah and H. van Velthuizen, (2002), “Climate change and agricultural vulnerability”, IIASA Special Report commissioned by the UN for the World Summit on Sustainable Development, Johannesburg 2002. International Institute forApplied SystemsAnalysis, Laxenburg, Austria, Gambarelli, G. and A. Goria (2004), “Economic evaluation of climate change impacts and adaptation in Italy”, FEEM Note di Lavoro 103.04 Gbesemete, K. P. and Gerdtham, U-G. (1992), 'Determinants of Health Care Expenditure in Africa: A Cross-Sectional Study', World Development, 20 (2), 303-308. Gerdtham, U-G. and Jönsson, B. (1991), 'Conversion factor instability in international comparisons of health care expenditure', Journal of Health Economics, 10, 227-234. Getzen, T. E. (2000), 'Health care is an individual necessity and a national luxury: applying multilevel decision models to the analysis of health care expenditures', Journal of Health Economics, 19, 259-270. Giorgi, F. and L.O. Mearns (2001), ”Calculation of Average, Uncertainty Range, and Reliability of Regional Climate Changes from AOGCM Simulations via the "Reliability Ensemble Averaging" (REA) Method”, Journal of Climate, 15, 1141-1158. Gossling and Hall, (eds) (2006), “Tourists and Global Environmental Change: A Possible Scenario in Relation to Nature and Authenticity”, Routledge Govindaraj, R., Chellaraj, G., and Murray, C. J. L. (1997), 'Health expenditures in Latin America and the Caribbean', Social Science and Medicine, 44 (2), 157-169. Hamilton, J.M. and R.S.J. Tol (2007), “The Impact of Climate Change on Tourism in Germany, the UK, and Ireland: A Simulation Study” Regional Environmental Change, 7 (3): 161-172. Hamilton, J.M., D.J. Maddison and R.S.J. Tol (2005a), “Climate Change and International Tourism: A Simulation Study”, Global Environmental Change 15 (3): 253-266. Hamilton, J.M., D.J. Maddison and R.S.J. Tol (2005b), “The Effects of Climate Change on International Tourism”. Climatic Research, 29: 255-268. Hartigan, J. A. (1975). Clustering Algorithms. New York: Wiley. Henley, A. and J. Peirson (1998), “Residential energy demand and the interaction of price temperature: British experimental evidence”, Energy Economics, 20:157-171. Hertel, T.W. (1996) Global Trade Analysis: Modeling and applications. Cambridge University Press. Cambridge. Hitiris, T. and Posnett, J. (1992), 'The determinants and effects of health expenditure in developed countries”, Journal of Health Economics, 11, 173-181. Hitz, S., and Smith, J. (2004), “Estimating Global Impacts From Climate Change”, Global Environmental Change, 14 (3), 201-218. Hoozemans, F. M. J., Marchand, M., & Pennekamp, H. A. (1993) A Global Vulnerability Analysis: Vulnerability Assessment for Population, Coastal Wetlands and Rice Production and a Global Scale (second, revised edition). Delft Hydraulics. Delft. 33
IMAGE (2001), “The IMAGE 2.2 Implementation of the SRES Scenarios”, RIVM CD-ROM Publication 481508018. Bilthoven. The Netherlands. IPCC CZMS (1991) Common Methodology for Assessing Vulnerability to Sea-Level Rise. Ministry of Transport, Public Works and Water Management, The Hague. Jorgenson, D.W., Goettle, R.J., Hurd, B.H. and Smith, J.B. (2004), US Market Consequences of Global Climate Change, Pew Center on Global Climate Change, Washington, D.C. Kainuma, M., Matsuoka, Y. and Morita, T. (2003), (eds), Climate Policy Assessment Asia-Pacific Integrated Modeling, Springer-Verlag. Kane, S., Reilly, J. M., and J. Tobey (1992), “An Empirical Study of the Economic Effects of Climate Change on World Agriculture”, Climatic Change, 21, 17-35. Kaufman, L. and P.J. Rousseeuw (2005). Finding Group in Data: an Intro to Cluster Analysis, Wiley Series in Probability and Statistics. Kiiskinen, U., Vartiainen, E., Pekurinen, M, and Puska, P. (1997), “Does Prevention of Cardiovascular Diseases Lead to Decreased Cost of Illness? Twenty Years of Experience from Finland”, Preventive Medicine, 26, 220-226. Lee, C.-C. and C.-P. Chang (2008), “Tourism Development and Economic Growth: A Closer Look at Panels”, Tourism Management, 29: 180-192. Link, P.M. and R.S.J. Tol (2004), “Possible economic impacts of a shutdown of the thermohaline circulation: an application of FUND”, Portuguese Economic Journal, 3:99–114. Lise, W. and R.S.J. Tol (2002), “The Impact of Climate on Tourism Demand”, Climatic Change 55 (4): 429-449. Maddison, D.J. (2001), “In search of warmer climates? The impact of climate change on flows of British tourists”, Climatic Change, 49: 193-208. Madlener R. and R. Alt (1996), ”Residential Energy Demand Analysis: An Empirical Application of the Closure Test Principle”, Empirical Economics, 21:203-220. Mansur E.T., R. Mendelsohn, and W. Morrison (2004), “A Discrete-Continuous Choice Model of Climate Change Impacts on Energy”, Yale SOM Working Paper No. ES-43. Martens, W. J. M., Jetten, T. H., and Focks, D. A. (1997), “Sensitivity of Malaria, Schistosomiasis and Dengue to Global Warming”, Climatic Change, 35, 145-156. Martens, W. J. M., Jetten, T. H., Rotmans, J., and Niessen, L. W. (1995), “Climate Change and Vector-Borne Diseases -- A Global Modelling Perspective”, Global Environmental Change, 5 (3), 195-209. Martin, P. H. and Lefebvre, M. G. (1995), “Malaria and Climate: Sensitivity of Malaria Potential Transmission to Climate”, Ambio, 24 (4), 200-207. McMichael, A., (2004), “Climate change. Comparative Quantification of Health Risks: Global and Regional Burden of Disease due to Selected Major Risk Factors,” Vol. 2 , M. Ezzati, A. Lopez, A. Rodgers and C.Murray, Eds.,World Health Organization, Geneva, 1543-1649. McMichael, A., A. Githeko, R. Akhtar, P. Carcavallo, D. Gubler, A. Haines, R.S.Kovats, P.Martens, J. Patz, and A. Sasaki (2001), “Human health”, in J.J.McCarthy, O.F. Canziani,N.A. Leary, D.J. Dokken, and K.S.White (eds), Climate Change 2001: Impacts, Adaptation, and Vulnerability – Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge:Cambridge University Press. Mendelsohn, R. (2001), “Global Warming and the American Economy: a Regional Assessment of Climate Change Impacts” Edward Elgar, Northampton, MA. Moral-Carcedo J. and J. Vicens-Otero (2005), “Modeling the non-linear response of Spanish electricity demand to temperature variations”, Energy Economics, 27: 477-494. 34
Morisugi, H., Ohno, E., Hoshi, K., Takagi, A., and Y. Takahashi (1995), “Definition and measurement of a household’s damage cost caused by an increase in storm surge frequency due to sea level rise”, Journal of Global Environmental Engineering, 1: 127-136. Morita, T., Kainuma, M., Harasawa, H., Kai, K., & Matsuoka, Y. (1994), An Estimation of Climatic Change Effects on Malaria, National Institute for Environmental Studies, Tsukuba. Murray, C. J. L. & Lopez, A. D. (1996), Global Health Statistics Harvard School of Public Health, Cambridge. Nicholls, Robert J. and Richard J.T. Klein (2003) “Climate change and coastal management on Europe's Coast”, EVA W. P. No.3. OECD (2007) Climate Change in the European Alps: Adapting Winter Tourism and Natural Hazards Management. Agrawala, S. (ed.). Paris. France Parti P. and C. Parti (1980), “The Total and Appliance-Specific Conditional Demand for Electricity in the Household Sector”, The Bell Journal of Economics, 11: 309-321. Reilly, J. M., Hohmann, N., and Kane, S. (1994), “Climate Change and Agricultural Trade: Who Benefits, Who Loses?”, Global Environmental Change, 4 (1): 24-36. Rosenthal, D. H. and H. K. Gruenspecht (1995), “Effects of global warming on energy use for space heating and cooling in the United States”, Energy Journal, 16(2):. 77-20. Rosenzweig, C. and Parry, M. L. (1994). 'Potential Impact of Climate Change on World Food Supply, ' Nature, 367: 133-138. Rosenzweig, C., and Hillel, D.(1998), “Climate Change and the Global Harvest: Potential Impacts of the Greenhouse Effect on Agriculture”, Oxford University Press. New York, N.Y.. Rosselló, J., E.Aguiló, and A. Riera (2005), “Modelling Tourism Demand Dynamics”, Journal of Travel Research, 44: 111-116. Saizar, A. (1997), “Assessment of a potential sea level rise on the coast of Montevideo, Uruguay”, Climatic Research, 9: 73-79. Scott, D., G.McBoyle (2007), “Climate change adaptation and the ski industry”, Mitigation and Adaptation Strategy For Global Change, 12: 1411-1431. Scott, D., G.McBoyle, and M.Schwartzentruber (2004), “Climate change and the distribution of climatic resources for tourism in North America”, Climatic Research, 27: 105-117. Smit, B., O. Pilifosova, I. Burton, B. Challenger, S. Huq, R.J.T. Klein, and G. Yohe (2001), “Adaptation to climate change in the context of sustainable development and equity”, in J.J. McCarthy, O.F. Canziani, N.A. Leary, D.J. Dokken, and K.S. White (eds), Climate Change 2001: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge: Cambridge University Press. Smith, J. B. and Lazo, J. K. (2001), “A summary of climate change impact assessments from the US Country Studies Programme”, Climatic Change, 50: 1-29. Tol, R. S. J. (1996), 'The Damage Costs of Climate Change Towards a Dynamic Representation', Ecological Economics, 19, 67-90. Tol, R. S. J. (1997), 'On the Optimal Control of Carbon Dioxide Emissions: An Application of FUND', Environmental Modeling and Assessment, 2, 151-163. Tol, R. S. J. (2007), “The Double Trade Off Between Adaptation and Mitigation for Sea Level Rise: An Application of FUND”, Mitigation and Adaptation Strategy for Global Change, 5(12): 741-753 Tol, R.S.J. (2002a), ‘New Estimates of the Damage Costs of Climate Change, Part I: Benchmark Estimates’, Environmental and Resource Economics, 21 (1), 47-73. 35
Tol, R.S.J. (2002b), ‘New Estimates of the Damage Costs of Climate Change, Part II: Dynamic Estimates’, Environmental and Resource Economics, 21 (1), 135-160. Tol, R.S.J. and H. Dowlatabadi (2001), “Vector-borne diseases, climate change, and economic growth”, Integrated Assessment 2: 173–181. Tol. R.S.J., Ebi, K.L. and G. Yohe, (2007), “Infectious disease, development, and climate change: a scenario analysis”, Environment and Development Economics, 12: 687–706 UNPD (2008), “World population prospects. The 2008 revision”, on line material available at: http://esa.un.org/unpp/index.asp Viscusi, W.K. and Aldy, J.E. (2003), ‘The value of a statistical life: A critical review of market estimates throughout the world’, Journal of Risk and Uncertainty, 27 (1), 5-76. Volonte, C.R. and J. Arismendi, (1995), “Sea-level Rise and Venezuela: Potential Impacts and Responses”, Journal of Climate Research, 14: 285-302. Volonte, C.R. and R.J. Nicholls, (1995), “Uruguay and Sea Level Rise: Potential Impacts and Responses”, Journal of Climate Research, 14: 262-284. Weiss, K. B., Sullivan, S. D., and Lyttle, C. S (2000), 'Trends in the cost of illness for asthma in the United States, 1985-1994', J Allergy Clin Immunol, 106 (3), 493-499. Witt, S.F. and C.A.Witt (1995), “Forecasting Tourism Demand: A Review of Empirical Research”, International Journal of Forecasts, 11: 447-475. Yohe, G. and M. Schlesinger (1998), “Sea Level Change: The Expected Economic Cost of Protection or Abandonment in the United States”, Climatic Change, 38: 447-472. Yohe, G., J. Neumann, P. Marshall, and H. Ameden (1996), “The Economic Cost of Greenhouse Induced Sea Level Rise for Developed Property in the United States”, Climatic Change, 32: 387-410. Zeider, R.B. (1997), “Climate change vulnerability and response strategies for the coastal zone of Poland”, Climatic Change, 36, 151-173.
36
3 Assessing climate change cost for the economy: A general equilibrium perspective 3.1
Economic model and benchmark
The economic implications of climate change have been determined using a multi-country world CGE model: ICES. Based on the Global Trade Analysis Project (GTAP) database version 6 and core model (Hertel, 1996), ICES develops a recursive-dynamic growth engine where a sequence of static equilibria are linked by the process of capital accumulation driven by endogenous investment decisions. The specification of the supply side of the model follows Burniaux and Truong (2002), which increase the detail in the description of energy production. The model runs from 2001 to 2050 in one year time steps. As a first step, the benchmark for the period 2001-2050 has been built. Investment choices and thus capital stock are determined endogenously and other key economic variables in the calibration dataset of the model have been exogenously updated, to identify a hypothetical general equilibrium state in the future (this methodology is described in Dixon and Rimmer (2002)). Our chosen reference is the social economic scenario A2 of the IPCC. Since we are working on the medium to long term, we focused primarily on the supply side variables projecting changes in the national endowments of population/labour, land, natural resources, as well as variations in factor-specific and multi-factor productivity. We obtained estimates of future regional labour stocks from UNDP (2008) whereas estimates of land endowments and agricultural land productivity have been obtained from the IMAGE model version 2.2 (IMAGE, 2001). A rather specific methodology was adopted to get estimates for the natural resources stock variables. Due to the uncertainty in the determination of their “true” amount we preferred to exogenously fix the price of the natural resources, making it a variable over time, in line with the GDP deflator, while allowing the model to endogenously compute the stock levels. In the specific case of oil, coal and gas we set their price evolution to mimic what was forecasted by EIA (2009). By changing the calibration values for these variables, the CGE model has been used to simulate a general equilibrium state for the future world economy. This is the benchmark for all subsequent exercises. Therefore, this benchmark corresponds to the case in which no economic impacts of climate change have taken place, whereas the counterfactual scenarios consider the effects generated by one or more impacts. 37
3.2
Simulation and results
The economic implications of the impacts calculated in Table 9 are reported in Figures 2, 3 (for the +1.2°C and +3.1°C warming scenarios respectively) and summarized for 2050 in Figure 4, which also shows the relevance of each single impact category. For the world as a whole, all the impacts jointly considered, can impose an additional cost ranging from 0.3% to 1% of GDP in 2050. However, these global figures hide important regional differences. While developed regions lose slightly, or even gain, as in the case of Europe and especially its northern part, developing regions can lose considerably more. For a temperature increase of 3.1° C wrt 2000 for instance, South East Asia, South Asia, Sub Saharan Africa and Northern Africa can experience a GDP contraction of 4%, 3%, 2.6%, 2.4%, respectively. This final effect can be decomposed in its different determinants. For instance, it is interesting to note that the bulk of losses in developing countries is due to negative impacts on GDP driven by the dynamics in agricultural and tourism markets, while for developed countries climate change impacts on tourism activity, affecting the service sector, are of paramount importance. It is also interesting to note the time pattern of GDP impacts. In the case of Mediterranean Europe, eventually gaining from climate change, GDP performances with climate change are lower than those of the benchmark until 2035. They are higher only afterwards when positive terms of trade effects and international capital inflows counterbalance negative impacts (see below). It is worth commenting on the negligible impact on GDP exerted by land losses due to sealevel rise and those related to the health impact. This depends mainly on the fact that GDP measures the flow-value of goods and services produced within a region, and accordingly does not directly measure endowment (stock) losses. These are recorded only “indirectly” in GDP as long as they change the region’s ability to produce goods and services.
38
CC (1.2° C) vs baseline: Real GDP (% change)
USA Med_Europe
4
North_Europe
3
East_Europe
2
FSU
1
KOSAU CAJANZ
0
NAF
-1
MDE
-2
SSA
-3
SASIA CHINA
-4
50
47
20
44
20
41
20
38
20
35
20
32
20
29
20
26
20
23
20
20
20
17
20
14
20
11
20
08
20
20
20
05
World
20
LACA
02
EASIA
-5
Figure 2. Final climate change impact +1.2°C wrt 2000: GDP
CC (3.1° C) vs baseline: Real GDP (% change)
USA Med_Europe
4
North_Europe
3
East_Europe
2
FSU
1
KOSAU CAJANZ
0
NAF
-1
MDE
-2
SSA
-3
SASIA CHINA
-4
50
47
20
44
20
41
20
38
20
35
20
32
20
29
20
26
20
23
20
20
20
17
20
14
20
11
20
08
20
20
20
05
World
20
LACA
02
EASIA
-5
Figure 3. Final climate change impact +3.1°C wrt 2000
39
Climate Change Impacts: Summary 4.0 3.0
USA Med_Europe
2.0
North_Europe
% of GDP
1.0
East_Europe FSU
0.0
KOSAU CAJANZ
-1.0
NAF
-2.0
MDE SSA
-3.0
SASIA -4.0
CHINA EASIA
-5.0 1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
1.2 °C 3.1 °C
LACA
Agriculture
Energy Demand
Health
Sea Level Rise
Tourism
All Impacts
World
Temperature increase
Figure 4. Final climate change impact: % change in regional GDP wrt no climate change baseline (ref. year 2050)
This is why, for instance, catastrophic events deploying their effect on property values typically determine negligible impacts if measured in GDP changes. In addition, our assessment cannot capture other important cost determinants, for instance, displacement costs and not to mention the value of human life and capital losses are only partially considered15. As said, these could, at least in principle, be evaluated by a direct costing approach. Thus, the cost of climate induced sea level rise can be measured by multiplying the quantity of land (and/or capital and/or population “dwelling” that land) loss, times its “value”; health impact of climate change can be economically assessed multiplying disability adjusted life years (DALY) by a “value” of life. With a general equilibrium assessment costs are instead quantified by the different performance that the economic system can attain because of those initial losses. However, a systemic assessment can capture the economic system’s ability to adapt to the initial loss. Market-driven adaptation mechanisms are indeed crucial in determining the final
15
Indeed the simulation does include also some effect derived from capital losses. These however are simply modelled imposing the same percent of land stock losses to the capital stock. This was ment to capture some capital deterioration due to sea-level rise. We restrained from estimating more detailed capital losses from sealevel rise as any calculation is highy controversial involving often the evaluation of damages to urban, higly populated areas and rich of cultural values.
40
cost of climate change as shown in Figure 5 which compares a “direct” with a “welfare” assessment.
% 2050 GDP
8
3
Direct econom ic cost -2 Final cos t for the econom ic s ys tem (w ith m ark et driven adaptation)
-7
IA C H IN A EA S IA LA C A W or ld
A
SA S
SS
N A F M D E
FS U K O SA U C A JA N Z
U SA M ed _E N U or th _E U Ea st _E U
-12
Figure 5. Direct vs final climate change costs (+3.1°C wrt. 2000 ref. year 2050) Note: Direct economic costs are expressed as a % of 2050 GDP; final costs are expressed as % changes in 2050 GDP wrt the no climate change baseline value
Social economic systems show a substantial capacity to smooth out initial impacts. In general, when a factor of production becomes scarcer and thus more costly or less productive, it tends to be substituted with others which are cheaper or more productive. This also happens with goods and services in consumers’ preferences. Thus, following market signals, market-driven adaptation operates as a partial buffer of initial shocks. This mechanisms is however “neutral” as it works indifferently with positive and negative shocks as can be appreciated in Northern Europe and CAJANZ. In addition to factor and good substitution processes which always smooth out the initial impacts, our exercise captures another two adaptive mechanisms which can smooth or amplify the initial impact: terms of trade effects and international capital movements.
41
C C ( 3 .1 °C ) v s b a s e lin e : w o r ld p r ic e s ( % c h a n g e s )
R ic e W heat
15
Ce r Cr o p s V e g Fr u its
10
A n im a ls W o o d Pr o d
5
Fis h in g 0
Co a l O il
-5
Gas O il_ Pc ts
-10
Ele c tr ic ity W a te r
-15
En _ In t_ in d O th _ in d 2050
2047
2044
2041
2038
2035
2032
2029
2026
2023
2020
2017
2014
2011
2008
2005
2002
-20
MS e rv NMS e r v
Figure 6. Climate change impact on world prices CC (3.1°C) vs Baseline: term s of trade (% change s) 15
USA Med_Europe
10
North_Europe East_Europe
5
FSU KOSAU
0
CAJANZ -5
NAF MDE
-10
SSA SASIA
-15
CHINA 2050
2047
2044
2041
2038
2035
2032
2029
2026
2023
2020
2017
2014
2011
2008
2005
2002
-20
Figure 7. Climate change impact on terms of trade
42
EASIA LACA
CC (3.1°C) vs Baseline: term s of trade (% change s) 15
USA Med_Europe
10
North_Europe East_Europe
5
FSU KOSAU
0
CAJANZ -5
NAF MDE
-10
SSA SASIA
-15
CHINA 2050
2047
2044
2041
2038
2035
2032
2029
2026
2023
2020
2017
2014
2011
2008
2005
2002
-20
EASIA LACA
Figure 8. Climate change impact on international capital flows As shown in Figure 6, compared to the benchmark, climate change impacts induce higher food prices, driven by the net decrease in agricultural productivity and lower fossil fuel prices, determined by the decrease in world energy demand in consequence of lower production. On international markets, this tends to favour developed regions which, can be defined as energy importers and sometimes food exporters as well, and to penalize energy exporters and more generally, developing countries. Indeed, terms of trade effects are particularly negative for the Middle east and North Africa, and positive for CAJANZ, USA, the three European aggregates and Kosau (Figure 7). The different penalization of economic activity also determines a different ability in attracting foreign investment (Figure 8). In the model these flows originate from the equalization of returns to capital which are higher in those regions where aggregated demand, and thus, prices increase (decrease), more(less). This is again the case of Northern Europe, CAJANZ and Mediterranean Europe which experience a net capital inflow. Thus, developed regions, which have already experienced lower direct negative impacts than developing countries and as a result lower direct costs (if not gains), can also benefit from improved terms of trade and capital inflows showing a greater autonomous adaptive capacity. The implications of this can be grasped in the particularly enlightening case of Mediterranean Europe where initial negative impacts (of around 4% of GDP in 2050) are even turned into gains. 43
What are the implications at the sectoral level? This is another issue that can be addressed with a CGE exercise. As a general trend, consistently with the GDP contraction world wide, negative signs prevail. More specifically, sectoral production is adversely affected in agriculture and in energy sectors – see Table 10. The first are hit by decreased crop productivity and the second by the worldwide reduction in energy demand. Note however, that in “hot regions” such as Mediterranean Europe and the developing world, electricity production increases. This is fostered by the increased space cooling demand from households. Market services production tends to decrease as a response to decreased recreational services’ demand in those regions such as Mediterranean Europe and other hotter countries which become less attractive climatically and as tourism destinations. The opposite happens in Northern EU and in Canada whose climatic attractiveness increases.
Table 10: Climate change impacts on sectoral production: + 3.1°C wrt 2000, ref. year 2050 Sector
USA
Med_Eu
North_Eu
East_Eu
FSU
Rice Wheat CerCrops VegFruits Animals Wood Products Fishing Coal Oil Gas Oil_Pcts Electricity Water En_Int_ind Oth_ind MServ NMServ Int. Inv. Flows
KOSAU CAJANZ
NAF
MDE
-3.46 -10.27 -9.85 -10.08 -3.96
3.18 -1.12 -1.05 0.79 2.56
1.14 6.28 48.39 21.76 -2.33
0.21 0.85 -2.14 -0.11 -0.01
-0.99 -3.11 -4.53 -4.28 -3.18
-0.55 -0.10 1.61 0.16 -1.19
-10.09 3.03 -3.01 -3.14 -7.89
-9.07 6.18 -3.77 -5.69 5.10 -3.57 -7.35 -1.77 -1.60 -4.89 3.79 -0.99 -5.54 3.33 -4.56
-5.39 -1.38 -3.73 -4.17 -6.14
-1.11 -0.99 0.58 1.65 -1.61
-1.75 0.04 -1.45 0.62 -1.47
2.10 -1.84 -3.20 -2.76 1.07
0.71 2.11 -0.15 -0.54 -4.93 -2.28 -0.18 0.84 0.10 1.80 -0.74 0.36 -0.67
2.16 8.77 0.06 -0.61 -14.51 0.72 1.68 5.92 3.54 5.11 -0.21 -1.20 0.72
-14.50 -8.09 -2.13 -0.72 -13.84 -0.77 -16.16 -11.66 -10.46 -12.22 10.70 -2.21 6.30
0.87 1.53 -0.69 -0.75 -16.62 -1.10 -2.58 0.71 0.26 2.03 -1.83 -0.24 -2.37
1.29 1.89 -0.19 -0.31 -5.84 -7.58 -1.47 -0.64 0.01 -1.79 0.33 -2.20 -4.87
-0.94 2.12 -0.08 -0.55 -1.71 -0.63 3.39 1.51 -1.27 -0.14 -1.95 1.31 -1.47
-20.36 -18.75 -1.24 -0.83 -3.99 0.97 -13.45 -14.18 -13.05 -14.87 7.25 -1.26 2.41
-0.05 -0.99 0.38 -0.25 -1.49 -7.13 4.01 -0.77 2.47 -4.45 -2.85 0.29 -4.66
-4.07 -4.45 0.56 -0.29 -0.43 1.28 7.39 -0.79 -3.52 -6.22 -4.13 2.60 -5.79
-1.01 -0.55 0.64 -0.20 2.46 3.55 4.73 0.30 -0.75 -0.98 -4.64 2.59 -4.30
-2.86 0.52 0.58 -0.27 0.55 6.42 13.83 1.21 -2.46 -1.71 -5.43 1.90 -5.22
2.72 3.31 0.58 -0.30 2.28 4.64 16.44 2.84 4.05 2.55 -5.82 0.85 -5.28
8.29 7.54 0.62 -0.22 -2.43 -9.19 3.78 2.75 13.00 9.68 -7.17 -2.67 -5.84
SSA
-0.83 -3.08 0.65 -0.25 -0.26 -6.16 10.95 -1.96 2.44 -3.91 -4.56 -0.83 -4.93
SASIA CHINA EASIA LACA
To conclude, we address the issue of impact interactions. Basically: what is the value added of conducting a joint impact assessment respect to performing a set of disjoint impact assessment exercises and then sum the individual results to get the total effect? This is quantified in Figure 9, showing that there are detectable differences between the GDP effects of all impacts together and the sum of the GDP effect associated to each single impact. In particular it can be noted that impacts jointly considered produce a higher cost estimate than those derived by summing the single GDP effects. Our exercise thus highlights an interaction among impacts which, if neglected, tends to underestimate climate change costs. 44
50 40
% difference
30 20 10 0 -10
A LA C
IA EA S
IN A C H
SA
S
IA
A SS
E M D
N A F
N Z A JA
C
SA U K O
FS U
U SA M ed _E ur op N or e th _E ur op Ea e st _E ur op e
-20
Figure 9. Role of Impact interaction: % difference between GDP costs of all climatic impacts implemented jointly and the sum of GDP costs associated to each impact implemented individually
3.3
Conclusions
The proposed exercise conveys some clear and robust messages. Firsttly, In the light of the still limited set of climatic impacts considered, climate change raises important distributional and equity issues. Higher negative impacts are felt in developing regions which are poorer and already facing severe challenges for their development. The bigger threat seems to concern food availability either directly in terms of lower agricultural production, and indirectly in terms of higher food prices. High income countries are hit less severely or even gain slightly from climatic change. This evidence justifies support from the richer countries to the poorer ones. Secondly, the comparison of GDP implications of climate change at the world level with those at the macro-regional level highlights another important aspect of the issue: when the detail of the investigation increases high losses emerge which are otherwise hidden within a global assessment. This emphasizes the need to carefully tailor the scope of any climate change impact assessment as results are scale-dependent. Thirdly, The inclusion of marketdriven adaptation dramatically changes the economic impact assessment of climate change, thus any impact and consequently cost benefit assessment of climate change policies should not neglect this aspect. The risk is to define unnecessary or un-effective strategies likely to 45
lead to maladaptive practices. Finally, and most importantly, some fundamental qualification of our results are in order. GDP impacts shown are calculated only on a sub set of potential adverse effects of climate change (possible consequences of increased intensity and frequency of extreme weather events and of biodiversity losses for instance are not included); irreversibilities or abrupt climate and catastrophic changes to which adaptation by autonomous mechanisms can be only limited, are neglected; the present assessment assumes costless and instantaneous market driven adjustments; finally the world is currently moving on an emission path leading to a higher temperature increase than that consistent with the A2 scenario. In the light of all that, what is proposed here, should be taken as the lowest possible limit for climate change costs and it is still far from negligible. The main message we would like to convey thus, is that albeit its impact smoothing potential, market-driven adaptation cannot be the solution to the climate change problem: its distributional and scale consequences need to be addressed with policy-driven mitigation and adaptation strategies.
46
References: Burniaux J-M., Truong, T.P., (2002) GTAP-E: An Energy-Environmental Version of the GTAP Model. GTAP Technical Paper n.16 (www.gtap.org). Dixon, P. and Rimmer, M., (2002) Dynamic General Equilibrium Modeling for Forecasting and Policy. North Holland. Energy Information Aministration (2009), “International Energy Outlook 2009”, Report n° DOE/EIA0484(2009). On line material available at: http://www.eia.doe.gov/oiaf/ieo/index.html Giorgi, F. and L.O. Mearns (2001), ”Calculation of Average, Uncertainty Range, and Reliability of Regional Climate Changes from AOGCM Simulations via the "Reliability Ensemble Averaging" (REA) Method”, Journal of Climate, 15, 1141-1158. Hertel, T.W. (1996) Global Trade Analysis: Modeling and applications. Cambridge University Press. Cambridge. IMAGE (2001), “The IMAGE 2.2 Implementation of the SRES Scenarios”, RIVM CD-ROM Publication 481508018. Bilthoven. The Netherlands. UNPD (2008), “World population prospects. The 2008 revision”, on line material available at: http://esa.un.org/unpp/index.asp
47
48
PART II Climate change impacts on biodiversity/ecosystem services: a partial equilibrium economic valuation approach
49
50
4 Valuation of climate change effects on forest ecosystem services 4.1
Introduction
It has been proven that climate change, due to the increasing in temperature and concentration of greenhouse gases emission, has significant impacts on the natural environment and human health (MEA, 2005). This, in turn, has led to an increasing number of scientific studies focusing on the mapping and identifying the scale of the impacts of the climate change on ecosystem performance and the respective provisioning of ecosystem goods and services. More recently, accompanying studies on the assessment of the role of the ecosystems with respect to their contribution to the economy and human wellbeing were made popular by the Millennium Ecosystem Assessment (MEA). However, to the authors’ knowledge, few studies have put an emphasis on the estimation of human welfare losses related to climate-driven changes of biodiversity and ecosystems. In the literature, one can find that the economic costs of climate change mitigation have been relatively well studied by an aggregation of data from the sectors and industries most likely to be affected by mitigation policies and measures (e.g. IPCC, 2007). Yet the costs of climate change impacts on biodiversity are not well mapped due to the complex (and not fully understood) interactions between climate change, ecosystems, and the respective impacts on human well-being (both in utility and productivity/employment terms). For these reasons, the present paper attempts to contribute to this line of research by undertaking an empirical analysis of the European forest ecosystems, addressing the role of biodiversity as it “forms the foundation of the vast array of ecosystem services that critically contribute to human well-being” (MEA, 2005. p.p.18). The results of the present research shall be integrated into the cost-benefit analysis of alternative policy options (e.g. mitigation and adaptation policies) against global warming. To better understand the question at stake, a conceptual DPSIR (OECD, 1999) framework is applied to capturing the causal relationship between climate change, biodiversity, forest ecosystems and human well-being (see Figure 1). Today, scientific evidence has demonstrated with high certainty that climate change is one of the main drivers that directly alter ecosystem functioning and cause biodiversity losses. The shift of climate conditions can change species distribution, population sizes, the timing of reproduction or migration events, and increase the frequency of pest and disease outbreaks (MEA 2005, p.p. 10). As a consequence, increases in global temperature and greenhouse gases concentrations may be detrimental to the health of forest ecosystems and ultimately human well-being, both through the 51
disturbance of existing biodiversity and through a negative influence on the ability of ecosystem to deliver goods and services, both linked to human well-being. These are damages directly caused by climate change, and are therefore associated with particular costs to human society. Yet it is important to note that forest ecosystems also engender feedback effects to climate change due to their important contributions to the stock of CO2 emissions. These are important benefits that ecosystems provide to society. Therefore, monetarizing the respective costs and benefits associated with climate change impacts on ecosystems has practical sense in guiding cost-effective climate-change policy. Moreover, while mitigation and adaptation policy measures can reduce the associated costs, they also themselves imply economic costs. Both need to be considered in the evaluation of the pros and cons of each single policy measure.
Figure 1. Conceptual model for the climate change, forest biodiversity and human well-being interactions
Against this background, an economic valuation of climate changes impacts on biodiversity and forest ecosystems requires a three-step approach. The first step is the characterisation of the climate role in the provision of relevant forest ecosystem services. The second step is the calculation of the reduced quantity and quality of these ecosystem services that result in a 52
loss to human welfare under alternative IPCC scenarios. Finally, the third step is the (monetary) valuation of that loss. Following these steps, the paper is organized as follows. Section 4.2 contains a set of comprehensive valuation strategies for the quantification of climate change impacts on forest ecosystem in monetary terms, based on the projections of physical changes in the flows of EGS under the IPCC storylines. Section 4.3 presents the current status of European Forests as well as the projections of their future trends, in terms of (1) the current forest areas, (2) the total quantity of provisioning services, and (3) the total stored carbon in Europe. In Section 4.4, we apply specific economic valuation methods with respect to each type of ecosystem service, and present the respective economic estimates. Section 4.5 concludes.
4.2 4.2.1
A hybrid ecosystem-based approach to the monetisation of climate change impacts The present status of European forests
Before proceeding with the proposed three-step valuation approach, we focus first on a systematic mapping of the geo-climatic regions of the European countries of interest that correspond to a classification of forest ecosystems located in every region. We shall adopt the regions defined in the European Forest Sector Outlook Study 1960-2000-2020 main report (UNECE/FAO, 2005), covering 3416 European countries located in Western Europe and Eastern Europe sub-regions. Furthermore, we regroup the same countries into four subgroups, i.e. (1) Mediterranean Europe (Latitude N35-45°), (2) Central-Northern Europe (Latitude N45-55°), (3) Northern Europe (Latitude N55-65°) and (4) Scandinavian Europe (Latitude N65-71°), in terms of their climatic-geographical locations in the respective latitude intervals. This new geographical grouping is presented in Table 1. The underlying idea of this grouping is based on the assumption that particular types of forests in each country are closely determined by the specific, and common, climate conditions. This allows us to identify the predominant tree species as well as the respective contributions to the local economy at both the national and the larger ,regional scales. From an ecological view point, different tree species can play different roles in ecosystem regulation and life supporting functions, which will ultimately influence the provision of forest ecosystem 16
Three EFSOS sub-regions are presented in the Appendix. Note that in this paper, we exclude the CIS subregion (i.e. Belarus, Republic of Moldova, Russian Federation and Ukraine) in our study for we fear the large forest area and the relative low prices of the forest in these countries may bias our valuation result for the whole Europe.
53
goods and services. Alternatively, from an economic perspective, different tree species may deliver very different flows of ecosystem goods and services, which relate to various levels of economic importance and therefore to respective welfare impacts. Finally, and from a geoclimatic perspective, this classification may also allow us to explore the sensitivity of different tree species to climate changes, in particular increases in temperature and precipitation rates in the countries under consideration.
Table 1. Geographical grouping of the 34 European countries Geographical groupings
Latitude tion
classifica-
Countries included
Mediterranean Europe
Latitude N35-45°
Greece, Italy, Portugal, Spain, Albania, Bosnia and Herzegovina, Bulgaria, Serbia and Montenegro, Turkey, TFRY Macedonia
Central-Northern Europe
Latitude N45-55°
Austria, Belgium, France, Germany, Ireland, Luxembourg, Netherlands, Switzerland, Croatia, Czech Republic, Hungary, Poland, Romania, Slovakia, Slovenia
Northern Europe
Latitude N55-65°
Denmark, United Kingdom, Estonia, Latvia, Lithuania
Scandinavian Europe
Latitude N65-71°
Finland, Iceland, Norway, Sweden
Based on data from 34 European countries, forests cover a surface of about 185 million ha (FAO, 2005), which accounts for about 32.7% of the territory. By classifying forest areas in terms of their latitude, it is easy to see that European forests are not uniformly distributed in the four climatic-geographical regions that we have classified above. For instance, in Mediterranean Europe, most of the forests are coniferous and broadleaved evergreen forests, which account for 30% of the total forest area in the three regions. The Central-Northern and Northern European regions are home to most of the temperate forests, which account for 35% and 19% of the total forests, respectively. Finally, in Scandinavian Europe, forest area accounts for the remaining 16% of total forest, in which the identical forest biomes are mainly boreal see Figure 2.
54
60 ART - Artic
50
40
Figure 2 Classification of European forests
Due to the diverse climatic conditions across latitudes, species diversity and dynamics of forest ecosystems differ considerably throughout Europe, as reflected in the numbers and composition of tree species. For instance, the Ministerial Conference on the Protection of Forests in Europe MCPFE (2007) reported that about 70% of the forests in Europe are dominated by mixed forest consisting of two or several tree species, and the remaining 30% are dominated by one tree species alone, mainly conifers. In addition to the natural conditions, the current European forest structures, in particular forest species compositions, have been heavily influenced by anthropophagic interventions, such as past land use and management (Ellenberg, 1986). In particular, the forest protective management strategy in Europe has resulted in a 1.0 percent annual expanding in the area of mixed forests over the last 15-year period (MCPFE 2007); this may partly be because of the widely acknowledged scientific evidence that mixed forests composed of several tree species are usually richer in biodiversity than forests dominated by one tree species. With respect to the sensitivity of tree species to temperature changes, this has been studied in terms of specific forest types located in different geographical regions in Europe. In Mediterranean Europe, most forests consist of sclerophyllous and some deciduous species that are adapted to summer soil water deficit. Temperature changes may allow the expansion of some thermophilous tree species (e.g. quercus pyrenaica) when water availability is sufficient (IPCC, 2001). Similarly, Garcia-Gonzalo et al. (2007) find that in Scandinavian Europe, 55
where growth of boreal forests is currently limited by a short growing season, low summer temperature and short supply of nitrogen, climate change can be associated to an increase forest productivity namely the an increase to for stock carbon. This is because an increase in temperature can prolong the growing season, enhance the decomposition of soil organic matter and thus increase the supply of nitrogen. In turn, these changes may have positive impacts on forest growth, timber yield and the accumulation of carbon in the boreal forests (Melillo et al. 1993; Lloyd and Taylor 1994; Giardian and Ryan 2000; Jarvis and Linder 2000; luo et al. 2001; Strömgre 2001). The main features of forest ecosystems in our study area support the use of the current geo-climatic grouping structure. With respect to economic valuation, we can now proceed to a detailed description of each of the three steps as follows: Step 1 the mapping of the ecosystem goods and services provided by European forests; Step 2 the calculation of the reduced quantity and quality of these ecosystem services that result in a loss to human welfare due to climate change impacts; and, Step 3 the monetary valuation of that loss.
4.2.2
Mapping of the ecosystem goods and services provided by European forests
A concise mapping of ecosystem goods and services (EGS) is the basis of high quality ecosystem assessment studies. For this reason, we adopt the MA approach (MEA, 2003), which provides a practical, tractable, and sufficiently flexible classification for the categorisation of the various types of ecosystem goods and services (EGS). In this context, all EGS can be generally classified into four main categories, i.e. provisioning, regulating, cultural and supporting services – see Table 2.
Table 2. A general classification of Ecosystem Goods and Services for European Forests Types of Ecosystem Services
Supporting Services
Examples
Provisioning Services
Food, Fiber (e.g. timber, wood fuel), ornamental resources, etc.
Regulating Services
Climate regulation, water regulation, erosion regulation, etc.
Cultural Services
Recreation and ecotourism, aesthetic values, spiritual and religious values, cultural heritage values, etc.
Source: adapted from MEA 2003
56
Provisioning Services In this category, we classificy forest products into seven main groups: industrial roundwood, wood pulp, recovered paper sawnwood, wood-based panels, paper and paper board, and wood fuel. For all products, quantity information on the total annual removal from forests is available on the FAOSTAT-Forestry. We first collect quantity information for all 34 European countries under consideration, and then summed up the total quantities for four individual latitude groupings - see Table 3. These figures, in turn, are at the basis of the economic valuation exercise.17 Table 3. Applied MEA framework for European forest ecosystem Provisioning Services-(1) Total Removal of Wood Forest Products (WFPs) in 2005 Latitude Classification
Industrial Roundwood (Million m3/yr)
Wood pulp (Million t/yr)
Recovered paper (Million t/yr)
Sawnwood (Million m3/yr)
N35-45° N45-55° N55-65° N65-71° Total Europe
7.40 48.12 13.75 6.33 75.60
0.75 3.20 0.41 25.70 30.06
11.85 6.32 8.38 2.62 29.17
15.38 18.18 10.98 32.60 77.14
Woodbased panels (Million m3/yr) 17.86 12.48 4.98 3.31 38.63
Paper and paper board (Million t/yr) 19.60 11.87 6.88 26.35 64.70
Wood fuel (Million M3/yr)
20.24 14.25 4.96 12.66 75.60
Source: FAOSTAT, year of reference 2005
Regulating Services Here two types of ecosystem services provided by European forests are of particular importance: (1) climate regulation (i.e. carbon sequestration) and (2) water and erosion regulation (i.e. watershed protection). It is important to note that here we focus only on the carbon service due to a lack of data. In any event, the role of forest ecosystems in the mitigation of climate change through forest and soil carbon storage has also been more studied and understood. Given a better understanding of the complex relationship between watershed protection and climate change, the present work can be further elaborated and improved upon in the future.
17 The data report from FAOSTAT does not provide an efficient collection of data on non-wood forest products, for this reason, our figures of the forest provisioning services will not embed this provisioning service. We acknowledge that our estimation is underestimated compare to other studies in the literature, if there is less evidence to link the provision of with non-wood forest products climate change (e.g. Merlo and Croitoru, 2005).
57
Cultural Services In many European countries, forests are of particular importance in terms of cultural services. Among all others, cultural services represents the most important value (MCPFE 2007), including recreational hunting and fishing, natural park visiting, forest landscapes and other spiritual uses. Some of these services can involve both consumptive (e.g. consumption of animal meat) and non-consumptive (e.g. enjoyment derived from hunting activities and forest landscape) uses. To avoid potential double counting, we limit cultural services provided by European forests to the non-consumptive uses only. In addition, the passive use value of forests plays an essential role in the assessment of some particular forest areas, especially when these are linked to strong spiritual and religious values. It should be noted that the direct linkage of cultural service with climate change is rather complex to convey. Even though some of the existing literature in general equilibrium modelling have put considerable efforts into the analysis of climate-driven changes in tourism demands (Berrittella et al., 2006; Bigano et al., 2008), few studies – if any – are able to embrace non-consumptive, recreational use of forests (including forest recreation activities) or passive use values in their analysis. For this reason, in the assessment of welfare changes due to changes in the provision of cultural services, we use as key variables those forests areas that are designed for recreational and protective purposes, (as described by the Global Forest Resources Assessment 2005 (FRA, 2005)), Supporting Services Finally, with respect to supporting services, indicators for the measurement of the respective forest ecosystem changes in response to climate change are not well developed, and therefore quantity data to measure these are not readily available (MEA 2005). For this reason, we will not directly tackle the valuation study of this service category. However, it is important to realize that the relevant values are implicitly reflected in the valuation of all other three categories of forest ecosystem goods and services.
4.2.3
Estimation of the physical changes of ecosystem services due to climate change
Over the last 30 years, the world has experienced significant temperature increases, particularly in the northern high latitudes (IPCC, 2001). The research results of the International Panel on Climate Change (IPCC) show that the average temperature in Europe will increase from 2.1 to 4.4°C by 2050, varying across latitudes, with the strongest warming consistently in the higher latitudes. In addition, model simulations also suggest a decrease in precipitation 58
in the south of Europe, particularly in the summer, and an increase in precipitation over much of northern Europe (Schöter et al., 2005). In order to quantify the climate change impacts on forest ecosystems, both quantitative and qualitative data are needed to describe the ability of the ecosystems to provide the necessarygoods and services , both in the present time period and in future scenarios of climate change. Moreover, to specify these scenarios, we adopt the four major storylines that are developed by IPCC, coupling the circulation models (e.g HadCM) with socio-economic storylines18 (Nakicenovic and Swart 2000; Schöter et al. 2004; Schöter et al. 2005). Finally, this enables us to describe the change of flows of ecosystem services under different future scenarios, i.e. A1FI, A2, B1 and B2.
4.2.4
The Monetary valuation of forests ecosystem goods and services
The monetarisation of the loss of environmental services provided by forests under climate change scenarios is the main concern of the CLIBIO project. This requires the application of state-of-the-art economic theory and different valuation methodologies. Within the welfare economics framework, two main streams of economic theory are widely applied in the area of climate economics: (1) partial equilibrium theory – to estimate the impacts of climate change on a single market or economic sector and (2) general equilibrium theory – to estimate the influence of climate change over a larger scale economy through the changes of individual markets/sectors, see Chapter 3. The former requires the investigation of appropriate microeconomic valuation techniques, including market-based economic valuation tools (e.g. Market price analysis) as well as non-market valuation tools (such as contingent valuation methods, travel costs methods, meta-analysis and scale-up value transfer. On the other hand, the latter largely relies upon the advancements in computer technology, which have been intensively used in climate economics in the simulation of the larger scale economic damages under climate change scenarios of the future. This distinction of economic theories therefore clearly indicates that the present study is anchored in a partial equilibrium analysis, as forest ecosystems only contribute to a portion of entire economy. Moreover, the socio-economic valuation is anchored in the assessment of changes in the productivity of the economic sectors under concern and/or the respective consumer’s utility – see Figure 3. Once the physical change is identified and assessed, the economic value should reflect the change on the individuals
18
IPCC experts identify and characterize four storylines, i.e. A1FI, A2, B1, and B2, combined with a general circulation model HadCM3, developed Schöter et al. (2004), that are directly related socioeconomic changes to climatic changes through greenhouse gas concentration and to land use change through climatic and socioeconomic derivers, such as demand for food.
59
whose welfare has been affected, or the average welfare change of the individuals in a population (Nunes et al. 2003).
Figure 3 A Framework for the valuation of climate change impacts in human welfare Source: Australian Greenhouse Office report (2004), adapted.
Bearing in mind the MA classification of ecosystem goods and services – see Table 2 – it is not difficult to agree that no single valuation method will deliver a the full range of the forest value components under consideration, i.e. wood forest products’ values, carbon sequestration values and cultural values. Therefore, a flexible, integrated and generally straightforward approach is needed to estimate the costs of climate change through each of the above-mentioned value components. We refer this as an hybrid economic valuation methodology – see Figure 4. As we can see this will involve the use of market price analysis methods, cost assessments methods and valuation methods based on meta-analysis. These techniques are most appropriately applied in the context of regional or national scale climate change impacts, disaggregated by sector or markets. In addition, the use of the techniques in isolation (sometimes referred to as ‘bottom-up’ studies) is predicated on an assumption that any incremental damage due to climate change will not have large, indirect (non-marginal) impacts, affecting the prices of a range of goods and services that flow through the macro-economy.
60
Figure 4: An hybrid economic valuation methodology
4.3
Assessing change
4.3.1
bio-physical flows of ecosystems goods and services under climate
The advantage of IPCC future scenarios for climate change impact assessment
Given the underlying idea of the projection of future trends of European forest by the year 2050, this section will elaborate on how to model and map the impacts of climate change on forest area, production of wood forest products and carbon storage across the different IPCC storylines, i.e. A1FI, A2, B1 and B2 storylines. These storylines, as reported by the Special Report on Emission Scenarios, have specific attributes in terms of population growth, CO2 concentration, degree of temperature changes, and change of precipitation in Europe (Nakicenovic and Swart, 2000) – see summary in Table 4. As a consequence, the IPCC is able to present four brief “future stories” differently developed in economic, technical, environmental and social dimensions. Furthermore, efforts have been placed on the development of a general circulation model – HadCM3 19 – so as to directly relate socioeconomic changes to climatic changes through greenhouse gas concentration, and to relate land use changes through climatic and socioeco19 HadCM3, Hadley Centre Couplet Model Version 3 is a coupled atmosphere-ocean GCM developed at the Hadley Centre and described by Gordon et al. (2000).
61
nomic drivers such as the demand for food (Schröter D. et al. 2004). According to the IPCC specifications, A1FI, A2, B1 and B2 storylines are distinguished in terms of four future development paths, i.e. ‘global economic’ oriented, ‘regional economic’ oriented, ‘global environmental’ oriented, and ‘regional environmental’ oriented, respectively.
Table 4. The specifications of the four IPCC storylines Climatic model - HadCM3 Indicator
(Scenarios by 2050) Storyline
Storyline A2
A1FI 6
Storyline B1
Storyline B2
Population (10 )
376
419
376
398
CO2 concentration (ppm)
779
709
518
567
∆ Temperature (°C)
4,4
2,8
3,1
2,1
-0,5
0,5
4,8
2,7
High savings and high rate of investments and innovation
Uneven economic growth, high per capita income
High investment in resource efficiency
Human welfare, equality, and environmental protection
∆ Precipitation Europe (%) Socio-economic dimensions
(Source: Schröter et al., 2005; IPCC, 2001)
The two economic oriented scenarios (A1FI and A2) focus on ‘material consumption’, but A1 scenarios also consider different combinations of fuel, which is expressed as A1FI. The two environmental oriented scenarios (B1 and B2) mainly concentrate on the concepts of ‘sustainability, equity and environment’. It is important to point out that, among all others, the storyline A2 describes a very heterogeneous world which is characterized by high population growth, regional oriented economic development, and fragmented and slow per capita economic growth and technology, (this in fact mirrors the current socio-economic development pattern). For this reason, A2 is frequently used by the European Commission as the baseline scenario, with the remaining scenario analyses conducted vis a vis to this storyline. In particular, our focus is mainly on the comparison of A1 vs. A2, in an assessment of movements to a more economically focused world. Alternatively, we may also consider B1, and B2, vs. A2, in an assessment of movements to a more sustainably orientated world.
62
4.3.2
The status of European forests across the different 2050 IPCC scenarios
In order to project the quantitative changes of forest area and wood products in terms of climate change, and its distribution across the diferrent IPCC scenarios, we directly adopt the simulation results derived from the Advanced Terrestrial Ecosystem Analysis and Modelling (ATEAM) project. This project was funded by the 5th Framework Programme of the European Commission, with a specific emphasis on the assessment of the vulnerability of human sectors relying on ecosystem services with respect to global change (Schröter et al. 2004). In its delivered software, the percentage changes of forest area and wood products are projected regarding the four IPCC storylines, but only for EU-17. For the remaining 17 European countries, the respective forest areas are projected on the basis of IMAGE 2.2 program (IMAGE 2001). The values are in reference to the period 2050. The results of our projection for: forest area, wood products and carbon services will be discussed in detail in following paragraphs.
Forest area In the A1FI and A2 scenarios, forest areas decrease by about 21% and 9% by 2050, respectively - see Table A1a in Appendix for more details. The A1FI scenario shows the biggest impact because of the no-migration assumption and most severe climate change, with ∆ temperature (C°) equal to 4.4 degree (Thuiller et al., 2005). Both B1 and B2 scenarios present an increase in forest area, of about 6% for the former and 10% for the latter. The higher increasing rate of forest area in scenarios B2 highlights major change due to the hypothesis of afforestation that is associated with the higher levels of precipitation in this same scenario (Schöter et al., 2005). As we can see from Table A1, the impacts of climate change on forest land use vary significantly across latitudes. For example, Mediterranean Europe (N35-45°) is facing a general negative forest growth in scenario A1FI and A2, but a significant expansion in scenario B1 and B2. The Central-Northern Europe (N45-55°) and Northern Europe (N55-65°) regions present negative growth only in the A1FI scenario, in correspondence with the more severe climatic conditions. Scandinavian Europe (N+65°) always presents a decrease in forest growth. Finally, the A2 scenario shows the future projections by taking into account the historical trend: forest area tends to increase in the countries located at latitudes below 65 degrees. This implies that Scandinavian Europe, both under currents conditions and under influence of climate change, reduces its extension. We alsoinvestigateforest areas designated for recreational and conservation use, which corresponds to 7.8% and 10.2%, respectively of the total area - Table A1b in Appendix for more details. As we shall see, this data shall be of cru63
cial relevance when computing the economic value of cultural services, including recreation and passive use values provided by forest ecosystems.
Wood forest products As previously indicated, we assess the climate impact on the bio-physical levels of production of the wood forest products, including wood pulp, industrial roundwood, recovered paper sawnwood, wood-based panels, paper and paper board, and wood fuel. – see Tables A2a to A2g in Appendix for more details. Given different socio-economic and climatic assumptions for the IPCC storylines (as listed in Table 4), the projection of the quantitative changes of wood forest products varies across different IPCC scenarios in the next 50 years. Considering these tables together, it is clear that the impacts of climate change are unevenly distributed across European forests, with impacts dependent upon the regions where the forests are located, the types of forest products, and the scenarios in which either socio-economic or environmental policy is the focus. Overall, our results do show some significant trends of climate change impacts on the classified regions. For example, the productivity of most of the wood products in Mediterranean and Scandinavian Europe will be negatively affected by climate change, but the magnitudes of the impacts are subject to the assumptions of climate policies. However by comparing the quantitative assessment results among the four scenarios, there may be a policy option to mitigate the climate change impact through forest ecosystems in these two regions under the B-type scenarios. Furthermore, for some of the forest products we may observe some slightly positive impacts of climate change in Mediterranean Europe. Moreover, with respect to the countries located in both Central and Northern Europe, the direction of climate change impact is ambiguous to interpret. Generally speaking, the production of most of forest products will be increased in A2 and B2 scenarios as a result of the joint effects of both climate change and socio-economic changes in the future. In other words, climatic influence may in part affect the natural growth rate of forests in those two regions, but the existing policies may also play an important role in terms of their influence on current land use patterns.
Carbon Storage A Carbon cycle connects forests and climate change. Total carbon stored in forests has a very important role in determining any climate stabilization pathIn fact, the quantity of carbon stocked in trees biomass approximately corresponds to 77% of the carbon contained in the 64
global vegetation, while forest soil stores 42% of the global 1m top soil carbon (Bolin et al., 2000). Forests exchange large quantities of carbon in photosynthesis and respiration, contributing to the global carbon cycle as a source of carbon when they are disturbed, and as a sink when in recovery and regrowth after disturbances. In turn, climate changes may also influence the future carbon-storage capacity of forest ecosystems. Against this background, we construct projections for carbon sequestration in forests for all the European countries across the four IPCC storylines – see Table A3 in Appendix for more details. Our findings show that the average carbon stock tends to increase in all scenarios, but the respective magnitudes are different. For instance, in the A1FI scenario, which represents a world oriented towards ‘global economic’ growth together with the highest CO2 concentration and temperature, the total carbon sequestrated by forests appears to be the lowest when compared with the other three scenarios. This result is consistent with results reported by Schröter et al. (2005), who highlighted that for most ecosystem services the A1FI produces the strongest negative impacts. On the other hand, B-type storylines, which are sustainable development oriented, contribute to an increase in forest area and a consequently large quantity of carbon stock. These figures, in turn, will be at the basis of the economic valuation exercise, which shall be discussed in detail in the following sections.
4.4 4.4.1
An Economic valuation of European forest ecosystems: results Overview
Following our hybrid economic valuation model framework, different economic valuation methods are exercised in the capture of the values of the three types of ecosystem services under consideration. First of all, for the provisioning services provided by European forests, we can infer that the economic values are the direct use values obtained from trading wood forest products in the market. Therefore, market prices are used to value this ecosystem service, and this information is derived from Food and Agriculture Organization of the United Nations (FAO) database20 on forests. Secondly, in order to evaluate the welfare changes associated with the carbon regulation, we shall use the avoided damage cost methods that were undertaken by the recent EC funded project, CASES21 to estimate the marginal damage cost 20
http://faostat.fao.org/site/381/default.aspx CASES stands for “Cost Assessment of Sustainable Energy Systems” for EU countries and the selected nonEU countries, including Turkey, Brazil, India and China. The study aimed at providing a comprehensive and dynamic assessment of the full costs of electricity generation based on the state-of-the art methodologies, taking into account both geographical and temporal extend of the impacts and social economic impacts, such as health
21
65
of per additional unit of CO2 emission. Economic theory tells us that the optimal emission level is determined by the intersection of the marginal damage cost of emissions and the marginal benefit from damage mitigation (or marginal abatement costs). Thus the crossing point corresponds to the unit value of carbon sequestration, which gives rise to the optimal policy to incentivize the necessary abatement in the achievement of the global carbon stabilization goal, and can be used to calculate the total economic value of carbon stored in forests. Finally, with respect to the cultural services, meta-analyses and value transfer methods are jointly used. These two methods are anchored in non-market valuation methodologies and rely on the existing databases22 of non-market valuation studies for forests in Europe. All values are estimated under four IPCC scenarios in 2050 and expressed in 2005 US$. However, the specific nature and availability of data as well as the different valuation procedures embraced for each of the ecosystems services under consideration will merit a separate discussion.
4.4.2
The Economic Valuation of Provisioning Services
The valuation framework undertaken for forest provisioning services consists of the following two steps: i) The calculation of the current productivity value of forests in terms of the provisioning of the 7 types of Wood Forest Products (WFPs), i.e. industrial roundwood, wood pulp, and recovered paper, other processed wood products, sawnwood, wood-based panels, paper and paperboard and wood fuel; N
7
provisioning service ProductivityValuegeo-climatic region = ∑ ∑ ExportValuein n=1 i=1
N
∑ ForestArea
n
(1)
n=1
where i and n represent the type of forest product and the number of countries located in each geo-climatic region. The export values used here are published by FAOSTAT in year 2005. The values are first collected and aggregated up across all the 7 forestry sectors under consideration at country level and then divided by the quantity of each type of WFPs in the country so as to get the respective market prices of each single commodity. Furthermore, we aggregate the total values of WFPs at the scale of geo-climatic groupings. By dividing those and safety, economic production and consumption, recreation, and environmental and natural assets caused by climate change. 22 The popular databases for non-market valuation study include: Environmental Valuation Reference Inventory (EVRI), Envalue, and the Ecosystem Services Database.
66
values by the forest size located in the same area, we can therefore ì compare the productivity values (in $/ha term) of the forest biomes in terms of the profits associated with the types of WFPs they can deliver to the market (see Table 5). Therefore, the productive values can vary among the 4 geographical groupings as they reflect the different contributions of various forest biomes to the local economy.
Table 5. Projection of Total Productivity Value of WFPs (US$/ha/yr, measured in 2005) Scenarios
Latitude 35-45
latitude 45-55
latitude 55-65
latitude 65-71
A1 2050 A2 2050 B1 2050 B2 2050
168 (+5.3%) 139 (-12.8%) 134 (-16.1%) 141 (-11.9%)
824 (+0.6%) 777 (-5.1%) 584 (-28.7%) 633 (-22.7%)
749 (+60.8%) 682 (+46.4%) 401 (-13.9%) 503 (+8.0%)
749 (+64.2%) 730 (+60.0%) 668 (+46.4%) 701(+53.6%)
NB: Percentage variation from initial benchmark 2005 are showed in parentheses.
ii) The estimation of the future 2050 values of forest productivity To project the future trends of real wood prices in 2050, we refer to two studies (Clark, 2001; Hoover and Preston, 2006) that analyze long-term historical data.
Clark (2001) offers a theoretical analysis and an empirical examination of wood prices, based on aggregated global wood market data over the last three decades. Hoover and Preston (2006) analyses trends of prices of Indiana (USA) forest products using statistical data from 1957 to 2005. Although different in spatial scale, both papers lead to a similar conclusion that: there is no evidence of increase in real prices for wood in the near future. We therefore assume that real prices of wood products will remain stable in the next 50 years, while allowing different prices to exist across countries and continents. As a consequence, we can get the future total value of WFPs under different IPCC storylines by multiplying the real price of each wood product by the projected quantities of WFPs in 2050. These values are finally summed up over all the WFPs commodities and countries located at each geo-climatic regions, expressed in 2005 US$. The computation is expressed by Equation (2). 7
TVn = ∑ pin ⋅ QinS S
(2)
i =1
where TV is the total value of WFPs in Country n under IPCC Scenario S.
67
Estimations results are summarized in Table 6 – see also Table 4a-4g in Appendix for more details on the spatial distribution. Table 6 shows that climate change impacts on the productivity value of WFPs vary, depending upon the respective geo-climatic groupings as well the IPCC scenario under consideration. For example, among all others, Mediterranean Europe has a lower sensitivity to climate change in terms of the total productivity value. In other words, the lowest variations in total productivity value of WFPs are registered in Mediterranean Europe, while the highest variations are reported in Northern and Scandinavian Europe.
Table 6. Projection of Total Value of WFPs for European Forests (Million$, 2005) IPCC scenarios
Mediterranean Europe
Central Europe
Northern Europe
Scandinavian Europe
Europe
A1 2050 A2 2050 B1 2050 B2 2050
6,413 6,453 8,018 8,736
41,250 47,556 41,441 48,742
5,413 7,215 4,712 6,810
35,540 33,943 31,772 31,943
88,616 95,167 85,943 96,231
Another important finding is that the total productivity values of WFPs are generally higher in the scenarios A1 and A2 (“material consumption” specific scenarios) than in the scenarios B1 and B2 (“sustainability, equity and environment” specific) in all latitudes – see Table 7. Table 7. Comparison of Total Value of WFPs for European Forests
Benchmark A2 Scenario Absolute value difference (Million$, 2005) Percentage change
A1vs.A2 B1vs.A2 B2vs.A2 A1vs.A2 B1vs.A2 B2vs.A2
Mediterranean Europe (N35-45)
Central Europe (N45-55)
Northern Europe (N55-65)
Scandinavian Europe (N65-71)
Europe
-40 1,565 2,283 -0.6% 24.3% 35.4%
-6,306 -6,115 1,186 -13.3% -12.9% 2.5%
-1,802 -2,503 -405 -25.0% -34.7% -5.6%
1,597 -2,171 -1,999 4.7% -6.4% -5.9%
-6,551 -9,223 1,065 -6.9% -9.7% 1.1%
As we can see, the A1 scenario, with a higher concentration of CO2 and increasing temperature (ºC), will result in a welfare loss to all European countries, except Scandinavian Europe. But in the B-type scenarios, the enhanced consciousness of sustainable development 68
and environmental protection may lead to a reduction in the total extracted forest resources for WFPs for sale, and thus a decrease in total benefits. Moreover, our valuation results also suggest that a local or national oriented sustainable development strategy (i.e. the B2 scenario) may have a positive impact on social welfare, as this scenario shows an average higher welfare gain in almost all geo-climatic regions as compared to the B1 scenario.
4.4.3
The Economic Valuation of Regulating Services
Forest conservation or the prevention of deforestation in order to stabilize Green House Gas (GHG) emissions – questions not originally included in the Kyoto Protocol – have been officially recognized in COP13 in Bali in December 2007 as important mitigation tool/policy instrument. The estimation of the economic value of climate regulating services (i.e. carbon storage) provided by forest ecosystem is therefore considered to have a very important impact on policy making for CO2 stabilization in Europe.23 As we have shown that the carbon stocks in forests are projected to increase on average in Europe under all 4 IPCC storylines (see section 4) in the next 50 years, we may therefore expect to obtain some benefits from forest regulating services. However, the magnitudes of those benefits may vary across different forest biomes. The methodological framework for the valuation of the regulating services consists of two steps: we first compute the marginal value of carbon storage in forests (2005US$/tC), which are then used to estimate the total economic values that can be obtained in different geoclimate regions under the IPCC scenarios. First of all, the marginal value of carbon storage refers to the benefits from avoided damages24 caused by incremental CO2, or CO2-equivalent GHG emissions, in the atmosphere due to the carbon sequestration functions of forest ecosys-
23
However, it is important to note that our economic value estimates for regulating services in the present paper are underestimated, as we do not undertake a valuation of the other regulating services, e.g. watershed protection and soil nutrient cycling, due to the limited knowledge about how to quantify those services in physical terms, both with respect to climate change impacts as well as in terms of a projection of the respective future changes. 24 The avoided damage costs assessment method has been widely used in the literature (see Cline, 1992; Nordhaus, 1993a,b; Merlo&Croitoru, 2005; CASES, 2008) to calculate indirectly the benefits from carbon sequestrated in forests, but it is important to note that the concept is different from the market price of carbon (obtained via emission trading scheme) and the marginal abatement cost (involves the costs of technological R&D for facilitating the emission abatement), although under certain restrictive assumptions the three measures would be broadly equal, at the margin (DEFRA, 2007).
69
tem. In the present paper, we build our analysis upon an existing project, “Cost Assessment for Sustainable Energy Systems” - CASES25, a worldwide study funded by the EU. One of the main features of CASES is that it is built upon the Integrated Assessment Models (IAMs), which by definition combine the dynamics of global economic growth with the dynamics of geophysical climate dynamics, to estimate the cost of GHG emissions under different energy evolution paths in 2020, 2030 and 2050. The existing literature on IAM has been intensively reviewed under the project and various available estimates in the recent years were taken into account in its final value estimates. Among all others, the value of social costs of carbon estimated by the UK’s Department for Environment, Food and Rural Affairs (DEFRA 2005) was adopted, for it reflects the policy context in which the values are used, and it combines the results of a number of IAM’s in a transparent matter. As a consequence, the CASES project was able to obtain three levels of estimates of marginal damage costs, i.e. lower, upper and central estimates26, respectively. For example, as reported in the CASES final report, the lower estimates of marginal damage costs evolve from € 4/tCO2 in 2000 to € 8/tCO2 in 2030; the upper estimates evolve from € 53/tCO2 in 2000 to € 110/tCO2 in 2030; and the central estimate evolves from € 23/tCO2 in 2000 to € 41/tCO2 in 2030. In the present analysis, we adopt a value estimate of 96,1 Euro/tC from the CASES report, referring to the central estimate of the avoided cost of 1 ton of carbon in 2050. The value is first adjusted to our paper by discounting to the real Euro value in 2005, using a 3% discount rate, and then converted to 2005US$ taking into account the real exchange rate and the Purchasing Power Parity (PPP). Finally, future economic benefits (measured in 2005 US$) of carbon stocks in each country’s forests are calculated by multiplying the US$/tC value by the projected quantity of carbon totally stored in the same forests in 2050 (see section 4), for each of the IPCC storylines, and then aggregated to compute the regional total benefits for the four large geo-climatic groupings. The results of the our valuation are presented in Table 8. These suggest that, in addition to the forest area, the predominant tree species may play a significant role in the determination of the carbon sequestration capacity in a geographical region, and therefore on the value of 25 CASES, Project No.518294 SES6, (2006-2008). Project official website: http://www.feem-Project.net/cases/ 26
The values are based on full Monte Carlo runs of the FUND and PAGE models, in which all parameters varied to reflect the uncertainty surrounding the central parameter values in both models. The lower and upper bounds are the 5% and 95% probability values of the PAGE model, while the central guidance value is based on the average of the mean values of the FUND and PAGE models. A declining discount rates is use as suggested by the UK Government ‘Green Book’. The equity weighting of damages in different regions is applied to aggregate the regional damage costs to global damages, in other words, damages in richer regions receive lower weights and damages in poorer regions receive higher weights.
70
the forest’s regulating services. For example, the forests in Central Europe contribute to the largest portion of benefits from the carbon regulating services in Europe. But this does not only depend on the fact that this area occupies the largest forest areas in Europe, but also because the type of forests in this area may has tolerance and capacity in terms of carbon sequestration. Table 8. Projection of Total Benefits of Carbon Storage in European Forests (Million$, 2005) Scenarios
Mediterranean Europe
Central Europe
Northern Europe
Scandinavian Europe
Europe
A1 2050 A2 2050 B1 2050 B2 2050
37,176 45,790 66,575 63,609
117,241 159,453 190,755 190,341
11,489 17,362 22,679 23,546
32,817 32,605 46,310 35,733
198,722 255,210 326,320 313,229
In addition, the productivity value of climate regulating services ($/ha) is also calculated based on the projected forest areas under different future scenarios, see Table 9 (and Table 6 in Appendix for for disaggregated data). The results show clearly the marginal benefit of carbon regulating services provided by different forest lands. Moreover, different forest management scheme may also influence these values. For instance, ceteris paribus, the B1 scenario shows the highest marginal value of regulating services provided by European forests. Table 9. Projection of the Productivity Value of Carbon Sequestration (US$/ha/yr, measured in 2005) Scenarios
Mediterranean Europe
Central Europe
Northern Europe
Scandinavian Europe
Europe
A1 2050 A2 2050 B1 2050 B2 2050
927 950 1,093 990
2,712 2,795 2,879 2,684
1,563 1,625 1,913 1,720
748 763 992 836
927 950 1,093 990
To better interpret the results, we undertake a comparative study among all four IPCC scenarios. Table 10 shows the comparative results of three IPCC scenarios (i.e. A1, B1 and B2) with respect to the A2 (BAU) storyline. Our results suggest a loss of benefits of carbon storks from forests to the whole of Europe in the A1 scenario, compared to the A2 scenario. This may be the result of intensive harvesting of forest products to meet the rapid progress of economic development that is the A1 scenario. In contrast, a focus on sustainable development and environmental protection in the B-type scenarios may lead to the extension of protected forest 71
area and thus consequent welfare gains in most of the geo-climatic regions. As shown in Table 10, in the B1 scenario, the worldwide efforts for sustainable development result in high welfare gains in all regions; whereas in the B2 scenario, these effects are unevenly distributed in different latitudes as local planning may play a more essential role here. Table 10. Projection of Total Benefits of Carbon Storage in European Forests
Benchmark A2 Scenario Absolute value difference (Million$, 2005) Percentage Change
4.4.4
A1vs.A2 B1vs.A2 B2vs.A2 A1vs.A2 B1vs.A2 B2vs.A2
Mediterranean Europe (N3545)
Central Europe (N45-55)
Northern Europe (N55-65)
Scandinavian Europe (N65-71)
Europe
-8,614 20,785 17,819 -18.8% 45.4% 38.9%
-42,212 31,303 30,888 -26.5% 19.6% 19.4%
-5,874 5,317 6,183 -33.8% 30.6% 35.6%
212 13,705 3,128 0.6% 42.0% 9.6%
-56,489 71,109 58,018 -22.1% 27.9% 22.7%
The Economic Valuation of Cultural Services
The cultural services provided by forest ecosystems consist of two components in our analysis: recreational use (e.g. nature-based recreation in forests) and passive use (e.g. existence and bequest value of forests and biodiversity). Not being traded in regular markets, recreation and passive use values are usually measured as willingness to pay (WTP) figures using nonmarket valuation approaches (namely: travel cost method, contingent valuation and choice experiments). According to state-of-the-art literature on the economic valuation of cultural values, an expected utility model specification can be used to describe how individuals are willing to trade individual income for changes in the levels of provision of forest cultural services, under the assumptions that the estimated marginal value of the service decreases with an increase in the area size of the forest site, and increases with an increase of the income level of the country where the forest is located (e.g., Hammitt, 2000; Markandya et al., 2008). The driving force of changes in future forest area is considered to be climate change in the present paper, therefore the expressed WTP estimate for a trade-off of forest resources can also reflect the fact that an individual’s preference to enjoy a certain kind of culture service may shift from one forest to another driven by the changed future climate conditions. Due to the large scale of the current study, it is impossible to conduct new original studies for all 34 countries under consideration. Therefore, a meta-analysis based value transfer 72
method is preferred to up-scale the marginal cultural values of forest situated in each geoclimatic region. Future changes of these values driven by climate change are projected according to the change in forest areas, in GDP and population under different IPCC storylines. The change in demand for recreation in forests driven by climate change is not considered in the present analysis due to the lack of information and relevant studies in the literature. This leaves us
with a focus on the valuation of the average WTP estimates (expressed in
2005$/ha) for obtaining cultural services (either recreational use or passive use) from forests in each geo-climatic region. For each region, we assume that one major forest biome can be identified as a representative forest type that survives the local climate. The main advantage of such an assumption is that we can select a few original non-market valuation studies that have been conducted in any country located in the same geo-climate region to undertake the value transfer within the same region. The meta-analysis enables us to explain the variance of the available WTPs (WillingnessTo-Pay) as a function of a few statistically significant explanatory variables (see Nunes et al. 2009). In particular, main explanatory factors for forest recreation and passive use are: i) size of recreational forest sites; and for passive use, size of forest areas designated to biodiversity conservation; and ii) income level in the study area. 27 The utility model can be expressed by: (1)
V = f (S , I )
where: V is the marginal value of a given forest site designated for recreation or conservation of
biodiversity. S is the size of the forest area designated for recreation or conservation (hectares).
I is the income level of the country where the forest is located (measure as PPPGDP).
By running the regression function expressed by equation (1): (2)
log V = α + β log S + γ log I
27
A similar approach is used by the authors in COPI – ‘Costs of Policy Action’, a recent European project supporting TEEB (The Economics of Ecosystems and Biodiversity). Here the focus was on a worldwide valuation of forest ecosystems in the context of policy inaction rather climate change (see Markandya et al. 2008 for more details).
73
we estimate the marginal effect on V of the forest size ( β ) and the income level of the country where the site is located ( γ ). The WTP figures included in the regression are selected from an extensive literature review process focusing on all existing valuation studies. The estimated coefficients are then used for the geographical value transfer (in different geoclimate regions) as well as for the inter-temporal value-transfer under different IPCC scenarios. For the geographical value transfer, a few representative studies are selected in each European geo-climate region (tables 11 and 12).
Table 11. Selected studies on recreational use for geographical value-transfer Country
Reference study
Forest biome
Geo-climatic region
United Kingdom
Scarpa, R., S. M. Chilton, W. G. Hutchinson, J. Buongiorno (2000)
Temperate broadleaf and mixed forests
Northern Europe
The Netherlands
Scarpa, R., S. M. Chilton, W. G. Hutchinson, J. Buongiorno (2000)
Temperate broadleaf and mixed forests
Central-Northern Europe
Finland
Bostedt, G. and L. Mattsson (2005)
Boreal
Scandinavian Europe
Italy
Bellu, L. G. and Cistulli V. (1994)
Mediterranean and Temperate Broadleaf
Mediterranean Europe
Table 12. Selected studies on passive use for geographical value-transfer Country United Kingdom
Reference study Garrod, G.D. and Willis, K. G. (1997)
Forest biome
Geo-climatic region
Temperate
Northern and centralnorthern Europe
Boreal
Scandinavian Europe
Mediterranean
Mediterranean Europe
Hanley, N., Willis, K, Powe, N, Anderson, M. (2002) ERM Report to UK Forestry Commission (1996)
Finland
Kniivila, M., Ovaskainen, V. and Saastamoinen, O. (2002) Siikamaki, Juha (2007)
Spain
Mogas, J., Riera, P. and Bennett, J. (2006)
74
The WTP figures selected from these studies28 are then scaled up to the corresponding higher geo-climatic region and forest biome, by taking into account the effect of the size of the forest area under valuation, β, according to the following formula:
(3)
VEU ,l
S = Vi ,l i ,l S EU ,l
β
where VEU,l = estimated WTP/ha for Europe by geo-climatic region l Vi,l
= WTP/ha of country i by geo-climatic l (from representative case studies)
Si,l
= forest area designated for recreation or conservation in country i by geo-climatic
region l SEU,l = forest area designated for recreation or conservation in Europe by geo-climatic region l i = country l
= geo-climatic region Data on forest areas designated for recreation and biodiversity conservation by country
are taken from FAO/FRA2005. This procedure allows us to estimate marginal values corresponding to the main identified geo-climatic regions in Europe. For the inter-temporal value transfer the estimated marginal values in 2005 are projected in 2050 using population and PPPGDP growth rates, and taking into account the effect of forest size29, under different IPCC scenario, as illustrated below:
(4)
Vi ,T1
H i ,T1 = Vi ,T0 H i ,T 0
S i ,T0 S i ,T 1
β
PPPGDPi ,T1 PPPGDPi ,T 0
γ
where: Vi,T1 = estimated value/ha/year for country i in year T1 V*i,00 = estimated value/ha/year for country i in year T0 T1 = year 2050 T0 = baseline year 2005 i = country
28 When several representative case studies and values are available, the mean marginal value is used. 29 We assume no variation over time in the percentage of forest area designated to recreation or conservation.
75
Finally, by multiplying the WTP estimates V($/ha) for recreational or passive use of forests by the sizes of forest area S that have been designated for recreation or conservation following the different climate change scenarios (See Appendix-Table 10 for the computation results), we can obtain the total recreational or passive use value for each region under each IPCC storyline. For each individual IPCC storyline, the total cultural value of a geo-climatic region is the sum of the respective recreational and passive use value of the forests. The results of the meta-analyses confirm our expectations both for forest recreation and passive use values: income level and size of forest areas are the main statistically significant factors explaining variation in WTP estimates for changes in forest cultural services (Table 13). The β coefficient on forest recreation size (logSIZE) is negative and significant for both recreation and passive use, showing that the marginal value of these services decreases with a marginal increase in forest area. The coefficient on income γ (logINCOME) is positive and significant, revealing a negative correlation of marginal values and income. The coefficients on passive use values are higher when compared with those of recreation, showing a higher sensitivity of forest size and income on marginal values.
Table 13. Results of the meta-regression function for recreational and passive use values Dependent variable
Recreation use
Passive use
Coefficient (std.error)
t-value
Coefficient (std.error)
T-value
constant
3.274 (3.698)
0.89
3. 972 (2.835)
1.40
LogSIZE
-0.445 (0.073)
-6.14
-0.603 (0.079)
-7.58
LogINCOME
0.599 (0.352)
1.70
0.889 (0.255)
3.49
Nobs
59
23
R
2
0.452
0.797
Adj R
2
0.433
0.797
LogWTP
Explanatory factors:
Final results on cultural services show that marginal values may differ widely according to the latitude (or geo-climatic region) where the forest is located (Tables 14 and 15). For recreational values, the highest estimates can be seen in Northern Europe followed by CentralNorthern Europe, most likely due to the facilities provided for forest recreation in these countries. The lowest values are registered in the Scandinavian countries. Conversely, for passive use values, the highest estimates are registered in the Mediterranean countries, which have a 76
higher potential for biodiversity and ecosystem conservation. As regards the projected total cultural economic values, Mediterranean Europe appears to have the highest values, followed by Central and Scandinavian Europe (Table 16). Within the same geo-climatic region, climate change might have a different impact on the cultural services provided in the local economy. By comparing the different IPCC scenarios, we can see that total values are generally higher for the environmentally-oriented scenarios B1 and B2 , than for the economic oriented scenarios (A1 and A2).
Table 14. Projections of marginal recreational values of European forests (US$/ha/yr, measured in 2005). Scenarios
Initial 2000 A1 2050 A2 2050 B1 2050 B2 2050
Mediterranean Europe
Central-Northern Europe
Northern Europe
Scandinavian Europe
1.06-3.06 1.25-7.87 1.26-7.91 1.20-9.24 1.03-6.77
0.43-2.61 1.07-8.15 0.68-5.17 0.81-8.08 0.65-4.83
1.88-7.10 4.17-99.92 4.03-96.55 3.97-124.34 2.97-62.55
0.16-1.05 0.23-0.53 0.23-0.54 0.27-0.73 0.22-0.44
Table 15. Projections of marginal passive use values of European forests (US$/ha/yr, measured in 2005).
Scenarios
Mediterranean Europe
Northern and CentralNorthern Europe
Scandinavian Europe
Initial 2000 A1 2050 A2 2050 B1 2050 B2 2050
356-615 898-1,552 902-1,558 748-1,292 678-1,171
123-182 361-534 344-509 342-506 230-340
123-255 219-454 220-457 262-543 203-421
Table 16. Projection of Cultural Values of European Forest Ecosystem (Million$, 2005) Scenarios
Mediterranean Europe
Central Europe
Northern Europe
Scandinavian Europe
Europe
A1 2050 A2 2050 B1 2050 B2 2050
3,988 4,850 9,006 8,457
2,123 2,475 4,270 3,108
305 425 818 608
1,204 1,185 2,993 2,223
7,620 8,936 17,088 14,396
77
Finally, we compare the total values of forest cultural services among the different IPCC scenarios, using scenario A2 as a benchmark for the analysis (Table 17). This scenario is characterized by the largest population and the highest GDP per capita. By comparing the remaining scenarios with the benchmark, we can capture the costs associated with a change from one scenario to another, and from environmental oriented scenarios towards economically oriented ones. Table 17. Comparison of Total Value of Cultural Values for European Forests
Benchmark A2 Scenario Absolute vaA1vs.A2 lue difference B1vs.A2 (Million$, 2005) B2vs.A2 A1vs.A2 B1vs.A2 Change in % B2vs.A2
Mediterranean Europe -862 4,156 3,607 -17.8% 85.7% 74.4%
Central Europe -352 1,795 633 -14.2% 72.5% 25.6%
Northern Europe -121 393 182 -28.3% 92.3% 42.9%
Scandinavian Europe 18 1,808 1,038 1.5% 152.5% 87.5%
Europe -1,317 8,152 5,460 -14.7% 91.2% 61.1%
Our comparative analysis of IPCC scenarios shows results which are consistent with our previous findings. For instance, as far as biodiversity and ecosystem conservation are concerned, the A1 scenario is worse off when compared with the A2 scenarios, an opposite result to the ones obtained for the provisioning services. This is due to the fact that the harvesting of the forest resources for WFPs production may result in a reduction of forests available for other uses, such as recreational or educational use of the forests. On the contrary, in all B-type scenarios climate change has positive impacts on the social economy as the management efforts to sustainable development and environmental protection? may halt or compensate the negative impacts of climate change. This finding therefore suggests that moving from B-type scenarios to A2 scenario will involve costs of policy inaction, since the economic oriented policy may reduce the welfare gain from forest cultural services, such as the enjoyment of natural environment and the knowledge of existence of biodiversity in the forests.
4.5
Conclusions
This section reports an original economic valuation of climate change impacts on forest ecosystem goods and services and biodiversity. On one hand, we provide a comprehensive classification, and mapping, of the different European countries according to their contribution in the supply of forest goods and services. The analysis anchors in the well-known classification proposed by the MA Approach. On the other hand, we investigate in detail the role of each 78
country in the provision of forest provisioning services, regulating services and cultural services. In order to value the climate change impact, we first identify four different climate scenarios that we refer to the A1FI, A2, B1 and B2 scenarios, corresponding to the four IPCC storylines and evaluated here for the year 2050. Secondly, we proceed with the analysis and evaluation of climate change impacts on the total forest area (for each country) as well as on the provisioning quantities (in bio-physical terms) across all the forests goods and services under consideration. The projections of future trends of forest areas and the provision of wood forest products in 2050, in terms of four IPCC storylines, are constructed by exploring the use of global climate models, including HADCM3, and simulating the response of the global climate system to increasing greenhouse gas concentrations. Moreover, considerable impacts of differentiated latitudes on the variability of forest EGS are taken into account by carefully regrouping the 34 selected countries located in different latitude intervals. As a consequence, we are able not only to identify the respective forest productivity related to predominantly forest types situated in each latitude interval, but also to assess, and compare, the sensitivity of the differentiated forest types in response to climate change impacts. Both of the two aspects have been included in the projection of the future trends of forest areas and forest products flows by 2050, in terms of the four IPCC storylines – see Appendix-Table 7 for a summary of the results. Finally, we apply various economic valuation methods (including market and non-market valuation methods, primary and value transfers methods) to estimate the values of the three MA service categories under concern, i.e. the provisioning services, regulating services and cultural services provided by European forests. This characterizes the hybrid valuation approach. Figures 5a-5c summarize the economic valuation results from three different types of ecosystem goods and services provided by forest ecosystem in Europe across four IPCC scenarios. As we can see, scenario B’s are associated with the highest levels of provision in all the ecosystem services under consideration, i.e. wood products, carbon sequestration and cultural services. As far as the carbon sequestration services are concerned, we can see that the stock of carbon that is stored in the European regions varies from 37.2 to 45.8 billion dollars in the Mediterranean countries in the A1 and A2 scenarios respectively, to 63.6 billion, in the B2 scenario, and 66.6 billion in the B1 scenario. Therefore, the B1 scenario is ranked as the one with the highest level of provision. The same ranking holds for the Central-North Europe and Northern Europe, where the B1 scenario is associated with the provision of 190.3 and 23.5 billion dollars, respectively. Finally, for the Scandinavian group of countries, B1 is ranked 79
with the highest level of provision of carbon sequestration services, amounting to 46.3 billion dollars. In addition, we can see that cultural services provided by forest ecosystems have their highest levels in the Mediterranean countries, ranging from 8.4 to 9.0 million dollars, respectively in the B2 and B1 scenarios, to 3.9 to 4.8 million dollars, in the A1 and A2 scenarios. For the Scandinavian group of countries, B1 is also ranked with the highest level of provision of carbon sequestration services, amounting to 2.9 million dollars, followed by the B2 scenario, which is tagged with a total cultural value of 2.2 million dollars. Finally, we can see that the total value of wood forest products ranges between 41.2 and 47.5 million dollars for Central Europe to 5.4 and 7.2 million dollars in the Northern Europe, for the A1 and A2 scenarios respectively. For this service, the Mediterranean Europe provides a relatively weak role in the provision with values ranging from 6.4 million dollars in A1 scenario to 8.7 million dollars in the B2. Moreover, we can conclude that the magnitude of the values of forest ecosystem goods and services varies according to the nature of service under consideration, with the carbon sequestration being ranked among the most valuable services. Furthermore, the impact of climate change on biodiversity, and its welfare evaluation in terms of the respective changes on the provision of forest ecosystem goods and services, is multifaceted. Firstly, it depends on the nature of the forest good and service under consideration. For example, cultural values are revealed to be more sensitive to the four IPCC scenarios than the remaining ones, with the wood forest products being the more resilient to climate change. Secondly, the distributional impacts of climate change on the provision of these goods and services also depend on the geo-climatic regions under consideration. In other words these impacts are not distributed in a uniform way across the European countries under consideration. This evidence is particularly clear from the analysis of Table 18.
Figure 5a: Forest wood products value 80
Figure 5b: Forest carbon sequestration values
Figure 5c: Forest cultural values
Table 18 depicts the welfare changes associated with a potential deviation from the A2 scenario, which is characterized by a high population, strong economic growth and high income per capita. This scenario is often interpreted by the European Commission as the benchmark scenario and translated by assuming no intentional action in response to global warming. For these reasons, we propose to evaluate the (comparative) welfare changes due to climate change with this scenario as a eference point. In this context, one can clearly see that the countries within the Mediterranean Europe (Greece, Italy, Portugal, Spain, Albania, Bosnia and Herzegovina, Bulgaria, Serbia and Montenegro, Turkey and Yugoslav) will benefit from the highest welfare gain in a movement towards the B1 or B2 storyline. In fact, this geoclimatic zone can experience a welfare gain amounting to a 86% increase in the value of cultural values when moving from an A2 towards a B2 scenario. This is followed by an increase of 45% in the value of the carbon sequestration services and a 24% increase in the value of the wood provision services. In other words, the “no adoption” of a B2 storyline, and a movement towards an A2 scenario, will be associated with a high welfare loss in the Mediter81
ranean Europe due the reduced quantity and quality of the forest ecosystem services under consideration. Alternatively, moving from an A2 towards an A1 scenario will always involve a welfare loss for the Mediterranean Europe. In short, for Mediterranean Europe As scenarios will always be associated to reduced quantity and quality of forest ecosystem services, resulting in lower levels of human welfare. On the other hand, storyline B1 is ranked as the most preferred scenario for this geo-climatic area. The region of Scandinavian Europe (including Finland, Norway and Sweden) presents mixed results. Firstly, moving from an A2 towards an A1 will not involve any welfare loss; on the contrary small welfare gains can be registered, even if not statistically significant from zero. Furthermore, the adoption of any B type scenario will always be associated with a welfare loss with the consideration of the provision of wood products. Table 18 Comparison of Total Value of Forest Ecosystem Goods and Services in Europe across the four IPCC storylines Absolute value difference Geographical regions Mediterranean Europe (N35-45) Central Europe (N45-55) Northern Europe (N55-65) Scandinavian Europe (N65-71)
EGS
(Million$, 2005)
Change in %
A1vs.A2
B1vs.A2
B2vs.A2
A1vs.A2
B1vs.A2
B2vs.A2
WFPs Provision
-40
1,565
2,283
-1%
24%
35%
Carbon Stock
-8,614
20,785
17,819
-19%
45%
39%
Culture Service
-862
4,156
3,607
-18%
86%
74%
WFPs Provision
-6,306
-6,115
1,186
-13%
-13%
2%
Carbon Stock
-42,212
31,303
30,888
-26%
20%
19%
Culture Service
-352
1,795
633
-14%
73%
26%
WFPs Provision
-1,802
-2,503
-405
-25%
-35%
-6%
Carbon Stock
-5,874
5,317
6,183
-34%
31%
36%
Culture Service
-121
393
182
-28%
92%
43%
WFPs Provision
1,597
-2,171
-1,999
5%
-6%
-6%
Carbon Stock
212
13,705
3,128
1%
42%
10%
Culture Service
18
1,808
1,038
2%
153%
88%
Furthermore, Scandinavian Europe will also experience significant welfare gains in the provision of the cultural and carbon sequestration services when moving towards a B type scenario. The respective welfare gains are, however, much lower when compared to the Mediterranean Europe, ceteris paribus. Having the Mediterranean and the Scandinavian Europe as two ‘corner situations’, we can observe that Central Europe and Northern Europe each present an intermediate state of affairs. In any case, it is important to note that a movement from an A2 to an A1 scenario will be always associated with high welfare losses in all the three services under consideration, with the highest losses registered among the Northern Europe countries (Denmark, United Kingdom, Estonia, Latvia and Lithuania). As opposed to the 82
Mediterranean and Scandinavian countries, for Central Europe it is the B type scenario that presents mixed results on climate change-caused changes in wood provision services. In contrast, when compared to A2, these scenarios always provide lower values on wood provision services for Northern Europe, which is a comparable situation to that of the Scandinavian countries. Finally, both Central Europe and Northern Europe show a similar profile in terms of carbon sequestration and cultural values: any B type scenario is characterized by a welfare gain from the perspective of these two ecosystem services, welfare impacts that are in accordance to what is also registered in the Mediterranean and Scandinavian Europe. Finally, and in conclusion, to the authors’ knowledge the current paper represents the first systematic attempt to estimate human well-being losses with respect to changes in biodiversity and forest ecosystems services that are directly driven by climate change. However, we acknowledge the complexity in mapping, modeling and estimating the relationships between climate change, biodiversity, ecosystem functioning, ecosystems services and human welfare.
83
References
Australian Government (2004) “Australian Greenhouse Office – Annual Report 2003-2004)” Barbier, E.B. (1994) “Valuing Environmental Functions: Tropical Wetlands”, Land Economics, 70(2), pp. 155-173 Bigano, A., Bosello, F., Roson, R. and R.S.J. Tol (2008) “Economy-wide impacts of climate change: a joint analysis for sea level rise and tourism", Mitigation and Adaptation Strategies for Global Change, Vol. 13, n. 8. Berrittella, M., Bigano, A., Roson, R. and R.S.J. Tol (2006), “A general equilibrium analysis of climate change impacts on tourism”, Tourism Management vol 25. Bolin B., Sukumar R. et al., 2000. Global perspective. Land use, land-use change, and forestry. In: Watson, R. T. et al. (Eds.). A Special Report of the IPCC. Cambridge University Press, 23-51. Braden, J. B. and C. D. Kolstad (eds) (1991) “Measuring the Demand for Environment Quality”, Elsevier Science Publishers, North-Holland. Carson, R. T., L. Wilks and D. Imber (1994) “Valuing the Preservation of Australia's Kakadu Conservation Zone”, Oxford Economic Papers, Oxford, 46(5), pp. 727-749. Clark, J (2001) “the global wood market, prices and plantation investment: an examination drawing on the Australian experience”, Environmental Conservation 28 (1): 53-64 Cline, W.R. (1992) “The Economics of Global Warming”, Institute for International Economics, Washington, DC. Costanza Robert, d'Arge Ralph, de Groot Rudolf, Farber Stephen, Grasso Monica, Hannon Bruce, Limburg Karin, Naeem Shahid, O'Neill Robert V., Paruelo Jose, Raskin Robert G., Sutton Paul, van den Belt Marjan (1997) “The value of the world's ecosystem services and natural capital”, Ecological Economics, 25, pp. 3-15. DEFRA (2007) “The Social Cost of Carbon and The Shadow Price of Carbon: What They Are, and How to Use Then in Economic Appraisal in the UK”, Economics Group of the Department for Environment, Food and Rural Affairs, UK. Available online at: http://www.opsi.uk/click-use/value-added-licence-information/index.htm FAO/FRA 2005 (2006), Global Forest Resources Assessment 2005: Progress towards sustainable forest management, FAO Forestry Paper no.147, available on website: ftp://ftp.fao.org/docrep/fao/008/A0400E/A0400E00.pdf Garcia-Gonzalo J., Peltoa H., Gerendiain A. Z., Kellomäki S., 2007. Impacts of forest landscape structure and management on timber production and carbon stocks in the boreal forest ecosystem under changing climate, Forest Ecology and Management, 241: 243-257. GBA (1995) Global Biodiversity Assessment, published for the United Nation Environment Programme, Cambridge University Press Giardina, C.P., and M.G., Ryan (2000), ‘Evidence that decomposition rates of organic carbon in mineral soil do not vary with temperature’, Nature 404: 858-861 Hammitt, J.K. (2000). Valuing mortality risk: theory and practice. Environmental Science and Technology 34: 1394–1400. IEEP – Institute for European Environmental Policy (2006) “Value of Biodiversity: Documenting EU Examples Where Biodiversity Loss Has Led to the Loss of Ecosystem Services”, ENV.G.1/FRA/2004/0081, Final Report IMAGE (2001) Integrated Model to Assess the Global Environment, Netherlands Environmental Assessment Agency – RIVM, available at http://www.rivm.nl/image/ 84
IPCC, 2007. Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. IPPC, 2001. Intergovernmental Panel on Climate Change, Climate Change 2001: The Scientific Basis. Cambridge University Press, Cambridge, UK, pp. 881 Jarvis, P.G. and S., Linder (2000), ‘Botany-constraints to growth of boreal forests’, Nature 405: 904905 Lloyd, J. and J.A. Taylor (1994), ‘On the temperature dependence of soil respiration’, Funct.Ecol.8: 315-323 Luo, Y.Q., Wan, S.Q., Hui, D.F., Wallance, L.L. (2001), ‘Acclimatization of soil respiration to warming in a tall grass prairie’, Nature 413: 622-625 Markandya, A., Chiabai, A., Ding, H., Nunes P.A.L.D and C. Travisi (2008) " Economic Valuation of Forest Ecosystem Services: Methodology and Monetary Estimates", in “Cost of Policy Inaction” Final report for the European Commission, Brussels, Belgium. Mäler, K-B. (1988) “Production function approach in developing countries”, in Valuing Environmental Benefits in Developing Countries, Vincent et al. (eds), Special Report, Michigan State University, pp. 11-32. MCPFE and UNECE/FAO (2003) State of Europe’s Forests 2003 - Report on Sustainable Forest Management in Europe, Vienna, Austria. MCPFE (2007), State of Europe’s forests 2007: The MCPFE report on sustainable forest management in Europe, Ministerial Conference on the Protection of Forests in Europe, Liaison Unit Warsaw, Poland. ISBN: 83-922396-8-7 or 978-83-922396-8-0 MEA (2003) “Ecosystems and Human Well-being: A Framework for Assessment”, World Resources Institute, Washington, D.C. MEA (2005) “Ecosystems and Human Well-being: Biodiversity Synthesis”, World Resources Institute, Ashington, D.C. Melillo, J.M., McGuire, A.D,, Kicklighter,D.W., Moore III, B., Vorosmarty, C.J., Schloss, A.L. (1993), ‘Global climate change and terrestrial net primary production’, Nature 363: 234-239 Merlo, M. and L. Croitoru (2005) Valuing Mediterranean Forests: Towards Total Economic Value, CABI Publishing-CAB International, Wallingford, Oxfordshire, UK. Mitchell, R.C. and R.T. Carson (1989), Using Surveys to Value Public Goods: The Contingent Valuation Method (Washington, DC: Resources for the Future). Nakicenovic N., Swart R., 2000. IPCC Special Report on Emission Scenarios (Cambridge University Press, Cambridge. NOAA - National Oceanic and Atmospheric Administration (1993) “Report of the NOAA Panel on Contingent Valuation”, Federal Register, 58(10), pp. 4601-4614, US. Nordhaus, W.D (1991a) “A Sketch of the Economics of the Greenhouse Effect”, American Economic Review, Papers and Proceedings 81: 146-150. Nordhaus, W.D (1991b) “To Show or Not to Slow: the Economics of the Greenhouse Effect”, Economic Journal 101: 920-937. Nordhaus, W.D (1993a) “Optimal Greenhouse Gas Reductions and Tax Policy in the ‘DICE’ model”, American Economic Review, Papers and Proceedings 83: 313-317. Nordhaus, W.D (1993b) “Rolling the ‘DICE’: an Optimal Transition Path for Controlling Greenhouse Gases”, Resources and Energy Economics 15: 27-50.
85
Nunes, P.A.L.D., J.C.J.M. van den Bergh and P. Nijkamp (2003) “The Ecological Economics of Biodiversity: Methods and Policy Applications”, Edward Elgar Publishing (UK).
Nunes, P., Ojea, E. & Loureiro, M.L., 2009. Mapping of Forest Biodiversity Values: A Plural Perspective. Fondazioni Eni Enrico Mattei - Note di Lavoro, 04.09, Fondazioni Eni Enrico Mattei, Milan, Italy. OECD (1999) “Environmental Indicators for Agriculture: Volume 1 Concepts and Frameworks”, Organization for Economic Co-operation and Development, Pairs. Pearce, D. and D. Moran (1994) “The Economic Value of Biodiversity”, Earthscan Publications Limited, London, UK. Perrings, C.A., Mäler, D.-G., Folke, C., Holling, C.S. and B.-O. Jansson (1995) Biodiversity Conservation, Kluwer Academic Publishers, The Netherlands Riou-Nivert P., 2005. Les temps changent, la forêt doit s’adapter, Forêt-entreprise, 162: 12. Schöter D. et al., 2004. ATEAM (Advanced Terrestrial Ecosystem Analyses and Modelling) final report (Potsdam Institute for Climate Impact Research, 2004). Schöter D., Cramer W., Leemans R., Prentice I. C., Araùjo M. B., Arnell N. W., Bondeau A., Bugmann H., Carter T. R., Gracia C. A.,. de la Vega-Leinert A. C, Erhard M., Ewert F., Glendining M., House J. I., Kankaanpää S., Klein R. J. T., Lavorel S., Lindner M., Metzger M. J., Meyer J., Mitchell T. D., Reginster I., Rounsevell M., Sabaté S., Sitch S., Smith B., Smith J., Smith P., Sykes M. T., Thonicke K., Thuiller W., Tuck G., Sönke Zaehle, Bärbel Z., 2005. Ecosystem Service Supply and Vulnerability to global change in Europe. Science, 310: 13331337. Tavoni, M. B. Sohngen, et al. (2007). “Forestry and Carbon Market Market Response to Stablize Climate”. Climate Change Modeling and Policy Working Papers. Fondazione Eni Enrico Mattei (FEEM):17. Thuiller W., Lavorel S., Araùjo M. B., Sykes M. T., Prentice I. C., 2005. Climate change threats to plant diversity in Europe. PNAS, vol. 102, n° 23: 8245-8250. UNECE/FAO, 2005. European Forest Sector Outlook Study, 1960 – 2000 – 2020, Main report, United Nation Publications, pp. 265 Walker B. H., Steffen W. L., Canadell J., IngramJ. S. I. (eds), 1999. The terrestrial biosphere and global change: implication for natural and managed ecosystems. Synthesis volume. IGBP Book Series 4, Cambridge University Press, pp. 450
86
5 Valuation of the linkages between climate change, biodiversity and the productivity of agricultural systems
5.1
Introduction
In the 21st Century, the agricultural sector will be radically altered by both natural disasters and anthropogenic factors, including climate change, changing world economies, and potential changes in the Common Agricultural Policy (CAP) and the subsidies currently paid to farmers and land managers. Both climate change and socio-economic drivers will affect crop productivities and agricultural land use patterns. The work of Rounsevell et al. (2005) shows that climatic impacts on agriculture vary across different climate scenarios and land use changes will also influence future land management scenarios.. This chapter aims at analyzing the potential effects of climate change on the agricultural sector in Europe in terms of the changes in land use patterns, crop productivity and carbon sequestration. Our analysis will focus on the depiction of different future scenarios of the agricultural sector in the next 50 years following the four IPCC storylines, i.e. A1FI, A2, B1 and B2. Before entering into a specific discussion on the agricultural system, it is important to note that the methodological framework in this chapter follows a similar pattern as the one developed in the chapter on forests. We shall therefore report the assessment results for the same countries grouped in the same latitude classifications as in that chapter. In addition, the results shall be reported separately in terms of categorized ecosystem goods and services. Finally, an economic valuation exercise will be performed in order to monetize the climate change impacts for the purposes of policy-making. The chapter is organized as follows. Section 5.2 discusses a roadmap to the monetisation of climate change impacts on agro-ecosystem services, exploring the role of the two agrosystems of croplands and grasslands respectively. Section 5.3 focuses on the assessment of climate change impacts on provisioning services, with particular attention paid to the role of biodiversity. Section 5.4 provides an economic valuation of the carbon sequestration services provided by crop and grasslands. Section 5.5 concludes.
87
5.2 A roadmap to the monetisation of climate change impacts on agro-ecosystems 5.2.1
Ecosystem goods and services provided by agro-ecosystems
Natural and modified ecosystems provide many services and goods that are essential for humankind (Matson et al., 1997). Simultaneously, modern agriculture has both substantially changed agro-ecosystems and severely impacted the environment; these impacts include reductions in biodiversity and a degradation of soil quality (Solbrig, 1991). The present section focuses on cultivated ecosystems (also known as agro-ecosystems), their link to biodiversity, and how this is impacted upon by global changes such as climate and technological ones, Building upon the Millennium Ecosystem Assessment (2005), we are able to identify the following ecosystem services: food, feed, and fibre; soil erosion control; maintenance of the genetic diversity essential for successful crop and animal breeding; nutrient cycles; biological control of pests and diseases; erosion control and sediment retention; and water regulation. These are the local benefits that agro-ecosystems can provide to local communities. In addition, there are also global benefits to human wellbeing from agro-ecosystems in terms of regulating services such as carbon sequestration (Swift et al., 2004; Allen & Vandever 2003; MEA, 2005). Moreover, we also distinguish between croplands and the grasslands due to the very different types of ecosystem goods and services that these two distinct agro-systems provide .
5.2.2
Croplands and grasslands
We discuss croplands and grasslands in detail for two main reasons. Firstly, croplands and grasslands provide different goods for human consumption. Secondly, these two agricultural systems are characterised by different profiles with respect to the supply of regulating services, with particular reference to their capacity for carbon sequestration. In terms of provisioning services, croplands provide three kinds of natural products, including food, nonfood, and bio-energy30 (see Table 1 for examples), whereas grasslands are cultivated primarily for grazing. With respect to regulating services, both croplands and grasslands play significant (though magnitudinally different) roles in the sequestration of carbon. The distinction 30 Food includes crops destined for human consumption, such as sugar crops, nuts, cereals, fruits, oils crops, pulses, root and tubers, vegetables. “non-food” includes provisioning services non-destined for human consumption, such as latex, pharmaceuticals and agro-chemicals product. On the other hand, bio-energy includes crops for energy production, such as oilcrop for biodiesel and cereals for ethanol.
88
between croplands and grasslands is therefore essential to the quantitative projections of ecosystem goods and services under the climate change scenarios, and ultimately to the economic valuation exercise.
Table 1 – Agricultural ecosystem goods and services (source: adapted from Swift et al., 2004) Cropland
Grassland
Food, Non-Food, Bio-energy Provisioning services
Food, fibre, latex, pharmaceuticals and agro-chemicals. Different crop types for food production, for animal feeding and energy production
Grazing
Supporting services
Genetic library
Genetic library
Cultural services
Agricultural landscape and agri-tourism
Agricultural landscape and agri-tourism
Regulating services
Nutrient cycling, regulation of water flow and storage, regulation of soil and sediment movement, regulation of biological population including diseases and pests
Nutrient cycling, regulation of water flow and storage, regulation of soil and sediment movement, regulation of biological population including diseases and pests
5.2.3
Biodiversity indicators in the agriculture system
The multiple dimensions of biodiversity in cultivated systems make it difficult to categorize production systems into ‘‘high’’ or ‘‘low’’ biodiversity systems, especially at spatial and temporal scales. In the agro-ecosystems a distinction has been made between ‘planned’ and ‘associated’ diversity (Swift et al., 1996, 2004; GCTE, 1997). ‘Planned’ diversity refers to plants and livestock deliberately, imported, stocked and managed by farmers. The term ‘associated’ refers to the nature of the biota (plant, animal and microbial), associated with the planned diversity and influenced by its composition and diversity. Farmers play a dominant role in the context of agricultural biodiversity by the selection of the present biodiversity stock, by the modification of the abiotic environment and by interventions aimed at the regulation of specific populations (‘weeds’, ‘pests’, ‘diseases’ and their vectors, alternate hosts and antagonists). It is widely recognized that the relationship between cultivated systems and biodiversity is complex (see Macagno and Nunes 2009). Firstly, biodiversity is an input into cultivated systems (e.g. genetic resources for food and agriculture). Secondly, biodiversity supports the functioning of cultivated systems (e.g. the balancing of the nutrient cycle). Thirdly, cultivated 89
systems also host important wildlife species which, though they may not always play a functional role in land productivity, nonetheless constitute important sources of landscape amenities. Finally, cultivated systems can have an effect on biodiversity in the surrounding areas outside the cultivated fields, for example habitat fragmentation impacts. More recently, studies of intensive agro-ecosystems have pointed out that permanent grasslands represent “hot spots” of biodiversity (see Gardi et al., 2002, Giardi et al., 2002; Anger et al., 2002; Bignal and McCracken, 1996; de Miguel & de Miguel, 1999; Nagy, 2002; EEA, 2007). Furthermore, the quality of soil is also higher in permanent grasslands with respect to arable lands as confirmed by the many soil quality indicators (organic carbon, aggregate stability) as well as species and habitat diversity, as reported in Natura2000 (see Macagno and Nunes 2009) . Against this background, the ratio between cropland and grassland can be employed as a proxy indicator for the measurement of the levels biodiversity in agro-systems. This, in turn, can be tested to determine if a significant role is played in the levels of supply of provisioning services. In other words, we can investigate whether this indicator affects the productivity of croplands. Furthermore, we propose to evaluate this link in the context of global climate change through a methodological framework that is discussed in the following section.
5.3
Assessing the impact of climate change on the provisioning services of agroecosystems
5.3.1
A methodological framework
To understand the interface between climate change and the provisioning services of agroecosystems, a graphical presentation is given in Figure 1 below. First of all, crop productivity is affected by physical climatic variables (CC) including temperature and precipitation, and by the level of technology (T). In turn, both are anchored in the specific IPCC scenario under consideration ranging from AIF1 to B2. In addition, a biodiversity variable (Bio) is also assumed to impact crop productivity. Formally, we propose to estimate the β ’s of the following equation:
CrP = β 0 + [ β1T + β 2T 2 + β 3 P + β 4 P 2 ] + [ β 5 F + β 6Tr ] + [ β 7 GR / CL]
(Equation 1)
where CrP is the crop productivity of harvested product, measured in t/ha, β 0 is the intercept, 90
T is the average annual temperature (°C), P denotes the annual precipitation (mm), F is the total fertilizer consumption per hectare (Mt), Tr refers to the total tractors used per hectare, and GR/CL is the ratio of grassland to cropland. As expressed by the equation, crop productivity is a function of physical variables (T and P), technological level (F and Tr) and a proxy of biodiversity (GR/CL)31.
Climate (CC) Ä T° 4.4 -2.1 °C Ä Prec. -0.5 – 4.8 (%) CO2 518-779 ppm
IPCCScenarios Climate change
AIF1 A2 B1 B2
Socio-economic orientation (T) Economicenvironmental Global-local
Crop Productivity
Provisioning services Land use
Biodiversity (Bio)
Food
Bioenergy
Figure 1 –Methodological framework for the evaluation of IPCC story lines on agricultural provisioning services
Before presenting, and discussing, the estimation results, we shall focus on cropland and grassland data and its projections across the different IPCC story lines.
5.3.2
The grassland, and cropland land-use data
Quantitative data of present cropland and grassland areas and the respective crop products in Europe are collected from the FAO 2005 database at national levels (FAOStat, http://faostat.fao.org/). In the present report, we consider over 153 million hectares of croplands in Europe – see first column in Table 1, Appendix, and 92,5 million hectares of grassland – see first column in Table 4, Appendix. A large proportion is dedicated to cereal 31
GR/CL is considered to be a good proxy for biodiversity at the European scale due to the fact that grasslands have been demonstrated to be biodiversity ‘hot spots’ within the intensive agro-ecosystems and are therefore very important in the maintenance of associated biodiversity values (Baglioni et al 2009a, Baglioni et al 2009b).
91
crops – see Table 2, Appendix. With respect to production, crop yields of each of the selected crop categories are derived from the FAO database in terms of weighted average yield (i.e. t/ha, harvested production per hectare) – see Table 3, Appendix. By multiplying the weighted average yield of a crop product by each country’s cropland area, we can calculate the total harvesting of this specific type of crop for this country, see the first column of Tables 5 to 12, Appendix. If for example, the cereals area in Italy, for 2005, was 3.965 million ha and the average yield of 5.4 t/ha, also measured in 2005, then total production of cereals produced by Italy in that year was 3.965 Mha x 5.4 t/ha = 21 million tons, again as reported in the first column of Table 5, Appendix. The calculation of the actual land devoted to bio-energy crops is based on the EEA technical report No 12/2005, which shows that approximately 4.6 million hectares of agricultural land in the EU-25 is directly devoted to biomass production for energy use, see Table 13, Appendix. As an illustration, in Italy, the total land area for bio-energy production is estimated to be 355,000 ha in 2005, about 3.6% of total cropland area. The majority of the land area for bio-energy production, about 83 per cent, is devoted to oil crops (used for biodiesel), and the remaining 11 per cent is used to cultivate ethanol crops. Bearing in mind the lack of data at the individual country level on the distribution between these two land uses, we assume the same proportions to calculate the oil crops and cereals used for biodiesel and ethanol production at country level, respectively. With respect to the remaining non-EU countries, the distribution is based on the average estimate of relative area devoted to bioenergy of the EU member states located at the same latitude. Moreover, we assume that the quantity of oil crops and cereals used for bio-energy production equals that of food crops – see last column in Table 3, Appendix. This assumption enables us to calculate the total production of bio-energy – see Tables 14-15 in Appendix. Again, taking Italy as an example, our calculation shows that about 1 million tons of oil crops and more than 167,000 tons of cereals are used for bio-energy production. Finally, we need to estimate the agricultural areas assigned for cropland, grassland and bio-energy production in each country in 2050. Here we adopt two approaches. The primary approach is to base our calculation on the results of ATEAM model (Schröter et al. 2004, Schröter et al. 2005), which provides downscaled projections at country level using IPCC SERS circulation model. The results obtained are consistent with that of the IPCC report. Again, taking Italy as an example, our estimation shows that the country’s cropland area in 2050 will range between 5.9 and 8 Mha depending on the scenario – see last columns of Table 1, Appendix. These figures indicate a general contraction of cultivated areas. However, the limitation of the ATEAM model is that it covers 92
only 17 developed European countries. Therefore we need to use an additional approach, referring to an IMAGE 2.2 Integrated Assessment Model (IMAGE team, 2001) to calculate the required information for the 17 remaining countries of interest. This is therefore done based on a global projection of land use changes. Final results are presented in Tables 1 and 4 respectively for croplands and grasslands. Projections of land productivities for all four IPCC scenarios is the focus of the next section.
5.3.3
Land productivities under different IPCC scenarios: results
As seen in Figure 1, the estimation of the future crop yield takes into account the impacts of advancements in technology (T), climate effects (CC) and biodiversity contributions (Bio). With respect to the technology factor (T), the parameter value was derived from Ewert et al. (2005) which provides a mean coefficient for Europe - see Table 2. For instance, in the global economic scenarios (A1 and A2) show higher technological impacts on crop productivity when compared to the B’s scenarios. As an illustration, the actual cereals yield in Italy may increase from present 5.4 t/ha to 6.8 t/ha in 2050 in the scenario B2, using the parameters of relative change in crop productivity presented in Table 2. Table 2 – Estimated relative change in crop productivity due to technology factor on 2050 A1FI
A2
B1
B2
1.87
1.81
1.63
1.28
Source: Ewert et al., (2005).
In addition, with respect to climate change impacts, the coefficient (CC) was calculated on the basis of a study developed by Tor (2007), which estimates the relative wheat yield changes in 2050 for the European Environmental Zones under different IPCC scenarios. The information regarding the percentage of each environmental zone within the EU countries is used to calculate a weighted average for an estimation of the relative wheat yield changes for all 33 European countries of interest. Moreover, since wheat is the most cultivated crop in Europe it can be interpreted as a representative crop in Europe and considered the most representative of net primary production (NNP) variation and can therefore be an important crop to be studied in terms of the consequences of changing climatic parameters (such as temperature, precipitation and CO2). All of the calculated CC coefficients are reported in Table 3.
93
Table 3 – Estimated relative changes in crop productivity (2050) as affected by changes in climatic conditions (CC ) technology (T) and biodiversity (Bio) for the different scenarios Country Greece Italy Portugal Spain Albania Bosnia and Herz. Bulgaria Serbia and Mont. Turkey TFR of Yugoslavia Austria Belgium France Germany Ireland Luxembourg Netherlands Switzerland Croatia Czech Republic Hungary Poland Romania Slovakia Slovenia Denmark United Kingdom Estonia Latvia Lithuania Finland Norway Sweden
CC A1FI 0.91 0.94 0.91 0.92 0.95 1.05 1.01 1.03 1.03 1.02 1.07 0.98 0.95 1.01 0.95 0.98 0.96 1.08 0.99 1.05 1.01 1.03 1.04 1.03 1.08 1.00 0.97 1.06 1.04 1.02 1.12 1.20 1.12
A2 0.93 0.95 0.92 0.93 0.96 1.04 1.01 1.03 1.03 1.02 1.06 0.98 0.96 1.01 0.96 0.99 0.97 1.07 0.99 1.04 1.01 1.03 1.03 1.03 1.07 1.00 0.97 1.05 1.04 1.01 1.10 1.17 1.10
B1 0.98 0.98 0.98 0.98 0.99 1.01 1.00 1.01 1.01 1.01 1.02 0.99 0.99 1.00 0.99 1.00 0.99 1.02 1.00 1.01 1.00 1.01 1.01 1.01 1.02 1.00 0.99 1.02 1.01 1.00 1.03 1.05 1.03
B2 0.96 0.97 0.96 0.96 0.98 1.03 1.01 1.02 1.02 1.01 1.04 0.99 0.98 1.01 0.98 0.99 0.98 1.04 0.99 1.02 1.01 1.02 1.02 1.02 1.04 1.00 0.98 1.03 1.02 1.01 1.06 1.10 1.06
YC (CC+T) A1FI A2 1.78 1.74 1.81 1.76 1.78 1.73 1.79 1.74 1.82 1.77 1.92 1.85 1.88 1.82 1.90 1.84 1.90 1.84 1.89 1.83 1.94 1.87 1.85 1.79 1.82 1.77 1.88 1.82 1.82 1.77 1.85 1.80 1.83 1.78 1.95 1.88 1.86 1.80 1.92 1.85 1.88 1.82 1.90 1.84 1.91 1.84 1.90 1.84 1.95 1.88 1.87 1.81 1.84 1.78 1.93 1.86 1.91 1.85 1.89 1.82 1.99 1.91 2.07 1.98 1.99 1.91
B1 1.61 1.61 1.61 1.61 1.62 1.64 1.63 1.64 1.64 1.64 1.65 1.62 1.62 1.63 1.62 1.63 1.62 1.65 1.63 1.64 1.63 1.64 1.64 1.64 1.65 1.63 1.62 1.65 1.64 1.63 1.66 1.68 1.66
B2 1.24 1.25 1.24 1.24 1.26 1.31 1.29 1.30 1.30 1.29 1.32 1.27 1.26 1.29 1.26 1.27 1.26 1.32 1.27 1.30 1.29 1.30 1.30 1.30 1.32 1.28 1.26 1.31 1.30 1.29 1.34 1.38 1.34
Bio A1FI 1.14 1.00 0.94 1.05 0.92 0.91 0.94 0.94 0.91 0.88 0.94 1.00 0.99 0.99 0.98 1.00 1.01 0.92 0.91 0.97 0.98 0.97 0.92 0.92 0.93 0.99 0.98 0.98 0.92 0.99 1.03 1.00 1.00
A2 0.98 0.99 0.87 0.97 0.94 0.93 0.96 0.96 0.94 0.91 0.92 0.99 0.99 0.99 0.99 0.99 0.98 0.90 0.94 0.98 0.98 0.97 0.94 0.93 0.95 0.99 0.98 1.00 0.94 1.00 1.02 1.00 1.00
94
B1 1.20 0.99 0.90 1.09 0.92 0.91 0.94 0.94 0.92 0.88 0.98 1.00 1.00 1.00 1.01 1.00 1.00 0.96 0.92 0.98 0.98 0.97 0.92 0.92 0.93 1.00 1.01 0.98 0.92 0.99 1.05 1.02 1.01
B2 1.00 0.99 0.86 1.00 0.94 0.93 0.96 0.95 0.94 0.90 0.93 0.99 0.99 0.99 0.99 0.99 0.98 0.92 0.93 0.98 0.98 0.97 0.94 0.93 0.94 0.99 0.98 0.99 0.94 1.00 1.02 1.00 1.00
Bio+ CC A1FI A2 1.05 0.91 0.94 0.94 0.85 0.79 0.97 0.90 0.87 0.90 0.96 0.97 0.95 0.97 0.97 0.99 0.94 0.97 0.90 0.93 1.01 0.98 0.98 0.97 0.94 0.95 1.00 1.00 0.93 0.95 0.98 0.98 0.97 0.95 1.00 0.97 0.90 0.93 1.02 1.02 0.99 0.99 1.00 1.00 0.96 0.97 0.95 0.96 1.01 1.02 0.99 0.99 0.95 0.95 1.04 1.05 0.96 0.98 1.01 1.01 1.15 1.12 1.20 1.17 1.12 1.10
B1 1.18 0.97 0.88 1.07 0.91 0.92 0.94 0.95 0.93 0.89 1.00 0.99 0.99 1.00 1.00 1.00 0.99 0.98 0.92 0.99 0.98 0.98 0.93 0.93 0.95 1.00 1.00 1.00 0.93 0.99 1.08 1.07 1.04
B2 0.96 0.96 0.82 0.96 0.92 0.96 0.97 0.97 0.96 0.91 0.97 0.98 0.97 1.00 0.97 0.98 0.96 0.96 0.92 1.00 0.99 0.99 0.96 0.95 0.98 0.99 0.96 1.02 0.96 1.01 1.08 1.10 1.06
Bio + CC+ T A1FI A2 1.92 1.72 1.81 1.75 1.72 1.60 1.84 1.71 1.74 1.71 1.83 1.78 1.82 1.78 1.84 1.80 1.81 1.78 1.77 1.74 1.88 1.79 1.85 1.78 1.81 1.76 1.87 1.81 1.80 1.76 1.85 1.79 1.84 1.76 1.87 1.78 1.77 1.74 1.89 1.83 1.86 1.80 1.87 1.81 1.83 1.78 1.82 1.77 1.88 1.83 1.86 1.80 1.82 1.76 1.91 1.86 1.83 1.79 1.88 1.82 2.02 1.93 2.07 1.98 1.99 1.91
B1 1.81 1.60 1.51 1.70 1.54 1.55 1.57 1.58 1.56 1.52 1.63 1.62 1.62 1.63 1.63 1.63 1.62 1.61 1.55 1.62 1.61 1.61 1.56 1.56 1.58 1.63 1.63 1.63 1.56 1.62 1.71 1.70 1.67
B2 1.24 1.24 1.10 1.24 1.20 1.24 1.25 1.25 1.24 1.19 1.25 1.26 1.25 1.28 1.25 1.26 1.24 1.24 1.20 1.28 1.27 1.27 1.24 1.23 1.26 1.27 1.24 1.30 1.24 1.29 1.36 1.38 1.34
Again as an example, considering the present Italian cereal productivity (5.4 t/ha) and a CC coefficient value of 0.94 for the scenario A1FI, this country’s cereal yield in 2050 will be 5.4 t/ha × 0.94 = 5.08 t/ha as a result of the future climatic variation. However, a combination of the two impact factors, i.e. (CC + T) will have a much higher effect (1.81) on the relative changes in crop productivity for the same scenario (corresponding to a predicted yield of 5.4 t/ha x 1.81 = 9.78 t/ha. Furthermore, with respect to biodiversity impacts, the coefficient (Bio) was calculated on the basis of an econometric exercise that isolated the marginal impact of biodiversity as modelled by equation 1. We created an ad hoc database for the analysis on wheat yields, covering 19 countries over the period 1974 and 2000, see a sample in Table 20, Appendix. Moreover, information regarding wheat yield, grassland and cropland areas, total fertilizers used, and total tractors is derived from FAO statistics whereas information about temperature and precipitation is derived from the Tyndall database. The regression model results are summarized in Table 4. We can see that the model is statistically significant (P
