The FEEM Sustainability Index (FEEM SI)
The FEEM SI is a composite index measuring progress in sustainable development. It considers 19 indicators pertaining to economic, social and environmental dimensions. The indicators computed in a Computable General Equilibrium (CGE) framework are projected over time and across different scenarios.
The FEEM SI methodology can be described in four different steps:
1. Indicator selection
Indicators selected for FEEM SI come from the literature on sustainability (Commission on Sustainable Development of the United Nations, EU Sustainable Development Strategy, and World Development Indicators from World Bank). The indicator set tries to consider the main dimensions of sustainability given the constraints imposed by the CGE framework, being aware about the evident trade-off between a detailed representation and the ex-ante analysis of sustainability. Table 1 reports the FEEM SI indicators considered in each sustainability dimension, and a brief description.
Table 1: List of FEEM SI indicators
DIMENSION / NAME / DESCRIPTIONEconomic / R&D / This indicator assumes a positive relationship between investment in R&D and growth, by maintaining that increased investment in R&D can bring more R&D output that will eventually lead to more innovation and increased productivity
Investment / Investment is one of the main drivers of economic sustainability, allowing for capital accumulation, which boosts economic growth. This indicator is weighted by considering the country specific capital stock.
GDP per capita / It is a measure of the per capita value of all market goods and services produced within a country. GDP p.c. is the typical indicator used to define the average well-being within a country.
Relative Trade Balance / The Relative Trade Balance measures the degree of a country’s exposure in the global commodities markets. It considers the net export value and weighs it with the country specific market openness (exports + imports). Relying, relatively, more upon exports is an indication of strong competitiveness.
Public Debt / Public Debt has an important role on the future perspective of a country’s economy. It depends on current government choices on expenditure and taxation, and on previously accumulated debt.
Social / Population Density / Population Density evaluates the population concentration in a specific country or macro-region (excluding uninhabitable areas). It represents the pressure on the available living space and resources for each individual.
Education / Expenditure in Education constitutes an investment in human capital. The role of education in improving future economic conditions and enhancing mobility as well as gender equality, which is supported by several studies.
Health / The generalised access to basic Health services is a major concern throughout the world. Monitoring the growth of expenditures in health by summing public and private expenditures allows measuring the degree of support on this issue.
Food Relevance / This indicator is used as a proxy for the poverty level. Furthermore, according to Engel’s law, the higher the proportion of national income spent on food, the lower the level of a country’s welfare.
Energy Imported / This is an indicator of energy security. The higher the Energy Dependence from abroad, the higher the risks deriving from changes in energy prices and political instability in energy-rich countries.
Energy Access / Access to Energy is important with reference to living conditions and future prospectives of well-being. This indicator considers the share of population having access to electricity. It allows capturing the intra country aspect of energy security, being more focused on distribution of energy resources than on availability at the country level.
Private Health / Monitoring the balance between public and private contribution to the health sector is essential for sustainability, because it determines the availability of a primary service to the whole society. The higher the share of Private Health expenditure, the lower the ability of poorer people to access to health care.
Environmental / GHG per capita / Greenhouse Gases are considered as described in the Annex I of the Kyoto Protocol. Emission per capita is a measure of the burden that the society imposes on climate and environment.
CO2
Intensity / This indicator is fundamental to monitor the improvement of the environmental performance of production and consumption activities, the latter playing a major role in the release of Carbon Dioxide into the atmosphere.
Energy Intensity / This indicator aims to assess the evolution of energy use efficiency.
Renewables / The gradual reduction of fossil fuel use is an important step towards security and sustainability of energy systems. The higher the share of green energy, the higher the environmental performance of the energy sectors.
Plants at risk / This indicator represents an alarm signal of the general worsening of habitats. It provides a comparable measure of endangered Plant species throughout the world, by considering the number of endangered species over the number of total known species present in that country.
Animals at risk / As in the previous indicator, it also represents an alarm signal of the general worsening of habitats. It is calculated in the same way, but focuses on animal biodiversity.
Water / Human pressure on water is an important indicator of resource pressure. It is estimated as water consumed in a country (for agriculture, industry and private uses) over the total renewable water resources available in that specific country.
2. ICES model and indicator computation
ICES model
The indicators are calculated using the outputs of the Intertemporal Computable Equilibrium System (ICES) model (Parrado and De Cian 2014; Eboli et al. 2010). ICES is a recursive-dynamic CGE model with World coverage based on the GTAP-E model (Burniaux and Truong 2002) and GTAP 7 database[1] (Narayanan and Walmsley 2008).
As in every model, there are several simplifications made in the CGE framework in order to represent the underlying phenomena in the most parsimonious way, and all results are conditioned by those assumptions. However, these simplifications allow modelling the most important elements of a complex environment in a controlled and coherent way. Nonetheless, it is of crucial importance to acknowledge the limitations of this approach when analysing its results. The main simplifying assumptions of our CGE framework are the existence of perfect competition that clears all markets along with a full employment of production factors. In addition, the future scenarios rely on external projections of key exogenous variables driving also its results. Despite these elements are considered limitations when comparing the modelling framework to the actual world, they provide the basis for a solid and consistent scenario analysis.
In the model, the economy of each country is characterised by n industries, a representative household and government. Industries minimize production costs and have nested Constant Elasticity of Substitution production functions that combine primary factors (natural resources, land, and labour), a capital&energy composite, and intermediates in order to generate the output. The ʺArmington assumptionʺ introduces some frictions on the substitutability of inputs imported from different countries.
A regional household in each region receives income, defined as the service value of national primary factors (natural resources, land, labour, capital). Capital and labour are perfectly mobile domestically but immobile internationally; instead land and natural resources are industry-specific. Regional income is used to finance three classes of expenditure: private household consumption, public consumption and savings (Cobb-Douglas specification), where the utility of private household consumption has a Constant Difference of Elasticities functional form.
A fictitious world bank collects savings from all regions and allocates investments in order to equalise the current rates of return.
Dynamics inside the ICES model are driven essentially by two sources: one endogenous and one exogenous. The first involves capital accumulation and foreign debt evolution governed by endogenous investment decisions. On the other hand, we make several exogenous assumptions concerning trends of population stock, labour stock, labour, land and total factor productivity over time in order to obtain a reference scenario in line with main economic indicators.
Our reference scenario focuses on the short run (2004‐2020) and it does not include any mitigation policy. It replicates the historical economic trend until 2009 using Eurostat (2010) and World Bank (2010) data for EU and non‐EU, respectively; for the period 2009-2020, we use a medium scenario (Medium Population ‐ Medium Growth ‐ Fast Convergence) developed within the FP7 RoSE project (Kriegler et al. 2013) and World Economic Outlook 2010 (IMF 2010) data for downscaling at country level.
The economic database is complemented with satellite databases on energy volumes (McDougall and Aguiar 2008), CO2 (Lee 2008) and non‐CO2 emissions (Rose and Lee 2008), which include nitrous oxide (N2O), methane (CH4), and three fluorinated gases (F‐gases). Both energy volumes and emission have an endogenous dynamic in the models and evolve the former according to energy sectors production and the latter proportionally to energy combustion process (CO2 emissions) and sector and household use of agricultural and energy commodities. CO2 trend are in line with information in IEA (2010) and fossil fuel prices changes according to Eurelectric (2010).
Extensions to the ICES model
In order to perform a sustainability analysis, we extended ICES to consider a more detailed sectoral aggregation and additional variables. We introduced 5 new sectors: Research and Development (R&D), Education, Private and Public Health, and Renewable Energy Sources (RES). All of them were split from the original GTAP 7 sectors according to the available international statistics.
For the R&D sector, we used the indicator “R&D expenditure as percentage of GDP” from the World Development Indicators - WDI (World Bank 2010) and the “share of R&D financed by Government, Firms, Foreign Investment and Other National” from the OECD Main Science and Technology Indicators (OECD 2010) for attributing R&D to the different economic agents.
A similar approach has been used for Education, Private Health and Public Health sectors. Data on overall expenditure on health and education have been obtained from the WDI database (World Bank 2010) and the data used for splitting the public and private health sector are from the World Health Organization (WHO 2010).
In order to regard separately the RES, namely wind, solar and hydro-electricity, they were split from the original electricity sector. The data collection refers to physical energy production in Mtoe (Million tons of oil equivalent) from different energy vectors and for each GTAP 7 country/region. The data source is Extended Energy Balances (both OECD and Non‐OECD countries) provided by the International Energy Agency (IEA). We complemented the production in physical terms with price information (OECD/IEA 2005; EC 2008; Ragwitz et al. 2007; GTZ 2009; IEA country profiles and REN21). The explicit consideration of the RES sector implied some modelling changes: the production function of electricity sector considers a new nest allowing the inter‐electricity substitution between RES and traditional fossil electricity.
The additional variables considered are water volumes, biodiversity at risk, energy access and inhabitable land.
The water volumes data come from the Food and Agriculture Organization (FAO’s Aquastat database) and account for the use of water in agriculture, industry, and by privates. We included also data on Total Renewable Water Resource (WTR) as a proxy of available water, and we considered this variable constant throughout time, according to Aquastat database. In the model, water use in agriculture, industry, and private agents has been linked respectively with demand of water services by agriculture, industry and households.
Biodiversity at risk is approximated using an index that quantifies the number of endangered species for both animals and plants over the overall number of species. Our data source is the World Conservation Union Red List of Threatened Species Database (IUCN). To obtain an endogenous dynamics of the indicator, we inversely linked the number of endangered species to CO2 concentration according to Thomas et al. (2004). The overall number of species is instead considered constant.
We used the indicator “share of population with access to electricity” from World Energy Outlook (IEA 2010) as a proxy of energy access. The indicator changes over time endogenously driven by the reduction of the gap between a country’s GDP per capita and the OECD average GDP per capita.
Data on inhabitable land comes from the GTAP 7 land use data base, developed using the Food and Agriculture Organization of the United Nations (FAO) 2004 data and FAO and IIASA methodology (2000).
Indicator computation
The ICES model computes indicators for each year and country/macro-region using simulation results. Table 2 outlines the FEEM SI list of indicators and the ICES variables used to project indicator values in the future.
Table 2: FEEM SI indicators and their dynamics
DIMENSION / NAME / FORMULA / INDICATOR DYNAMICSEconomic / R&D / R&D Expenditure / GDP (%) / R&D is a production sector in ICES SI model. R&D and GDP have both endogenous dynamics. Numerator and denominator are expressed at nominal values (2007US$).
Investment / Net Investment / Capital Stock (%) / Investment net of capital depreciation has an endogenous dynamics as well as Capital Stock which depends on the capital endowment and net investment of the previous year. Numerator and denominator are at nominal values (2007US$).
GDP per capita / GDP PPP / Population / GDP per capita considers real GDP PPP (2007US$), whose dynamic is endogenous and exogenous Population data.
Relative Trade Balance / Trade Balance / Market Openness / The Trade Balance (Export-Import) and the Market Openness (Exports + Imports) are both endogenous. Numerator and denominator are at nominal values (2007US$).
Public Debt / Government Debt / GDP (%) / Government Debt has an endogenous dynamics depending on yearly government budget deficit and stock of debt. GDP has an endogenous dynamics. Numerator and denominator are at nominal values (2007US$).
Social / Population Density / Population / Country Surface / Population dynamics is exogenous; Country Surface remains constant at 2000 levels.
Education / Education Exp. / GDP (%) / Education is a productive sector in ICES SI and has an endogenous dynamics as GDP. Numerator and denominator are at nominal values (2007US$).
Health / Total Health Exp. / GDP (%) / ICES SI considers two sector producing health services, one private and one public. Total Health Expenditure considers both sectors and has an endogenous dynamic as well as GDP. Numerator and denominator are at nominal values (2007US$).
Food Relevance / Food Cons. / Private Exp. (%) / Food Consumption coincides with household expenditure on agricultural and food products. Private Expenditure considers household total expenditure on commodities and services. Both variables are endogenous in the model. Numerator and denominator are at nominal values (2007US$).
Energy Imported / Energy Imported / Energy Cons. (%) / Energy Imported and Consumed are endogenous in the model and depends on agents’ behaviour in production and consumption. Numerator and denominator are expressed in Million Tons of Oil Equivalent.
Energy Access / Population with Access to Electricity / Total Population (%) / The share of Population with Access to Electricity has ana endogenous dynamics: it depends on the reduction of countries’ GDP per capita gap with the OECD average level, and on exogenous population dynamics.
Private Health / Private Health Exp. / Total Health Exp.(%) / Private and Total Health have an endogenous dynamics. Numerator and denominator are at nominal values (2007US$).
Environmental / GHG per capita / Kyoto GHGs Emissions / Population / GHGs emissions have and endogenous dynamics depending on energy combustion process (CO2 emissions) and sectors and household use of agricultural and energy commodities. Population moves exogenously. The indicator is expressed in tons of CO2 equivalent per capita.
CO2
Intensity / CO2 Emissions / Total Primary Energy Cons. / Both numerator and denominator have an endogenous dynamics linked to power sector production and input used. The indicator is expressed in tons of CO2 per Tons of Oil Equivalent.
Energy Intensity / Total Primary Energy Supply / GDP PPP / Both numerator and denominator have an endogenous dynamics linked to primary energy supply and the real GDP PPP. The indicator is expressed in Tons of Oil Equivalent over Million $ 2007US$.
Renewables / Renewable Cons. / Total Primary Energy Cons.(%) / Both numerator and denominator have an endogenous dynamics depending on the production of the sector Electricity from Renewables and the overall primary energy consumption.
Plants at risk / Endangered Species / Total Species (%) / The number of endangered species is endogenous and related to CO2 concentration
Animals at risk / Endangered Species / Total Species (%) / The number of endangered species is endogenous and related to CO2 concentration
Water / Water Use / Total Available Water (%) / Water use in agriculture, industry and by privates is endogenous and depends on the demand of water services by agriculture, industry and households. The available water remains constant across time.
3. Normalisation