CC BY
218using Simulation Models for Green Economy policy Making: A Comparative Assessment*
Dr. Andrea M. BASSI
Founder and CEO, KnowlEdge Srl, Italy; Extraordinary Associate Professor, Faculty of Economic and Management
Sciences, Stellenbosch University, Republic of South Africa
Abstract. In the run up to Rio+20 one area of research that has received considerable attention is the development of methodologies and tools for the identification of worrying economic, social and environmental trends, and the evaluation of potential interventions through the analysis of alternative future scenarios. In fact, the main contribution of a green economy approach has been identified as being the integration of sectoral interventions in a coherent, cross-sectoral framework of analysis. This study analyses the main methodologies and models currently available to carry out a green economy assessment and concludes that the selection of these tools should be done on a case-by-case basis, depending on specific country needs and priorities. In addition, different models could be combined to inform the various phases of the policy cycle, and maximize the effectiveness of green economy interventions.
Аннотация. В период до «Рио+20» (Конференции ООН по устойчивому развитию в Рио-де-Жанейро 2012 г.) значительное внимание уделялось исследованиям методик и инструментариев для выявления тревожных экономических, социальных и экологических тенденций и оценки возможных интервенций на основе анализа альтернативных сценариев будущего. Основным вкладом «зеленой экономики» как подхода является интеграция секторальных интервенций в единую межсекторальную систему анализа. Это исследование анализирует основные сегодняшние методики и модели для оценки «зеленой экономики» и делает вывод, что выбор этих инструментов должен проводиться на индивидуальной основе в зависимости от потребностей и приоритетов каждой страны. Кроме того, различные модели могут быть объединены, чтобы обеспечить поддержку политического цикла в различных его фазах и максимально повысить эффективность интервенций «зеленой экономики».
Key words: green economy, simulation models, quantitative methods, integrated analysis, sustainable development.
introduction
The misallocation of capital in the last two decades has contributed to the manifestation of several concurrent crises: climate, biodiversity, energy, food, water, as well as the global financial and economic crisis. In response to these systemic crises, UNEP has stressed the need for a shift to more sustainable and inclusive economic, social and environmental policies, which can enable the transition to a green economy. At the visionary level, UNEP (2011) considers the green economy as:
"An economy that results in improved human well-being and social equity, while significantly reducing environmental risks and ecological scarcities."
The report by the United Nations Environment Management Group's (EMG) (2011) points out that at the operational level, the green economy is seen as one whose
growth in income and employment is driven by investments (e.g. in human capital, social institutions, infrastructure) that:
• Reduce carbon emissions and pollution;
• Enhance energy and resource efficiency;
• Prevent the loss of biodiversity and ecosystem services.
In short, the green economy represents an attempt to guide countries through more action-oriented pathways to sustainable development. In this context, given that no single approach exists for sustainable development, policy makers need support through studies and analyses to help them better identify and understand upcoming challenges and opportunities, as well as to design, choose and implement policy interventions. The Agenda 21 (UNCED, 1992) also reflects the main goal of strategy and planning exercises on the green economy,
* Использование имитационных моделей принятия решений в области «зеленой экономики»: сравнительная оценка
that is, to inform and influence the policymaking cycle so as to effectively progress towards sustainable development. On the other hand, while there is a common and broadly shared goal, involving cross-sectoral amalgamation and integrated assessments, the vast majority of models available to governments are sectoral and cannot be easily coupled with each others. As a result of this disconnect (single and integrated goal vs. multiple disconnected tools), the selection and use of models and their effectiveness in informing decision makers ultimately highly depends on what needs to be measured and analyzed.
For the reasons outlined above, starting from the definition of the green economy (what needs to be measured and analyzed), this paper aims to provide (1) a framework for the review and selection of methodologies and models and (2) information on what tools are available to governments, and are currently being used to support the analysis of green economy strategies at the national and sectoral levels.
This study offers a critical review of the strengths and weaknesses of various methodologies, and of the adequacy of models to help countries to assess their economies and develop green economy strategies. On the other hand, this report does not aim at identifying the best approaches for the formulation and evaluation of green economy strategies, instead it provides key information for the ministries tasked with planning and implementing responsibilities to evaluate the adequacy of various models in meeting their specific needs.
1. criteria for review and evaluation
Various criteria are considered for the review and assessment of methodologies and models with a particular focus on the concerns and needs of developing countries. These criteria primarily focus on tangible dimensions to be reviewed and analyzed1. Further, the present study only assesses the methodologies and models that are most commonly used in developing countries for analyzing interventions for a green economy transition. Many2 studies provide more information on the breadth of models available, but the framework presented here does not intend to be exhaustive but rather to contribute
1 While it is acknowledged that the political dimension is crucial, there are limitations on the extent to which this can be captured by the use of models. As a result, although the process of model building and calibration is often affected by political dynamics, this criterion is not explicitly considered in the analysis presented in this report.
2 These include, among others: CE and SERI (2010), for a review of macroeconomic models and their approach to sustainability, IEEP et al. (2009) for a review of models used to project scenarios of biodiversity and ecosystem services, GEO-5 (UNEP, 2012) for a review of scenarios across sectors.
to a better understanding of how useful and adequate methodologies and models can be in supporting country-led green economy policy formulation and evaluation exercises.
The key criteria to assess the methodologies depend on the extent or nature of their contribution to various stages of the policymaking process. In addition, the complementarity of the methodological approaches is also considered, together with the inclusiveness (i.e., stakeholder involvement) of the process to implement them.
The following methodologies are considered: data frameworks and modeling methodologies that can be used to generate and analyze simulations of social, economic and environmental pathways or scenarios. Methodologies, or the underlying body of knowledge for the creation of different types of simulation models, can be "static" (data frameworks) or "dynamic" (modeling approaches). Both types are then used to create and simulate quantitative models.
— It is worth noting that modeling approaches may make use of data frameworks. Actually, data frameworks often represent the backbone of models, also depending on the flexibility and degree of customization offered by the modeling approach utilized. Data frameworks are "static", and can be used in two main ways: (1) in isolation, to investigate and understand the history and current state of system, and (2) embedded in simulation models, to generate simulations of future trends for all the indicators included in the framework selected. Data frameworks include:
• Indicators;
• Input-Output frameworks (I-O);
• Social Accounting Matrix (SAM); and
• Geographic Information Systems (GIS).
— Modeling approaches instead refer to the underlying mathematical theories and frameworks that can be used to create and simulate (or solve) quantitative simulation models. These methodologies could therefore be considered "dynamic" as they allow to generate future projections. Modeling methodologies include:
• Econometrics;
• Optimization; and
• System Dynamics (SD).
Concerning models, the criteria focus more explicitly on the definition of the green economy (which would vary depending on the national context), and the quantitative outputs required to effectively inform decision-making. As a result, the main criteria that were considered include the capability of models to represent the social, economic and environmental dimensions of the problems and opportunities analyzed, as well as their capability to carry out investment and policy analysis.
Further, as an additional layer of the analysis, models are assessed for their ease of customization and use. This
is relevant for specific country implementation, where data, time and financial resources may be scarce, and tradeoffs need to be addressed.
The following sectoral and macro models are considered:
• Input-Output (I-O) models (employment ILO, 2012a and 2012b; material flows IGES, 2010);
• System engineering models (models of an engineered system, e.g., energy supply) (energy planning Loulou et al., 2005; IAEA, 2001);
• Geographic Information System (GIS) and natural capital valuation models (e.g. WWF and Natural Capital Project, Van Paddenburg et al. 2012);
• Computable general Equilibrium (CGE) models (including those coupled with system engineering modules) (IFPRI model Lofgren et al., 2002; MAMS Logfren and Diaz-Bonilla, 2010); and
• System Dynamics (SD) models (Bassi, 2009, 2010, 2011; UNEP, 2013).
These models use different data frameworks (e.g., CGE models use the SAM) and modeling approaches (e.g., optimization in systems engineering models).
1.1. METHODOLOGIES
The growth in income and employment in the green economy is driven by investments. For these investments to be catalyzed and leveraged, public expenditure, policy reforms and regulation changes are needed. As a result, methodologies and models are needed that can support the policymaking process (see Figure 1), allowing to quantitatively project and evaluate trends (issue identification, stage 1), identify entry points for interventions and set targets (policy formulation, stage 2), assess ex-ante the potential impact across sectors and the effectiveness in solving stated problems (or exploiting opportunities) of selected interventions (policy assessment, stage 2), as well as monitor and evaluate the impact of the interventions chosen against a baseline scenario (policy monitoring and evaluation ex-post assessment / analysis, stage 5).
Various methodologies can be utilized to effectively support policy formulation and assessment (identification of problems, and then policy options that would have the desired impact, also of the magnitude desired, on the system) and evaluation (simulation of selected intervention options against real events). In this respect, it is worth mentioning that the methodologies presented in this report are most commonly used when the analysis is done ex ante, or before the actual implementation of the interventions (issue identification and agenda setting, and policy formulation and assessment), but they can also be used to carry out ex post (policy monitoring and evaluation) analysis:
• Ex-ante modeling methodologies can generate "what if" projections on scenarios with no action, and on the expected (and unexpected) impacts of proposed policy options on a variety of key indicators. In addition, various methodologies can assist in the cost-benefit and multi-criteria analysis, and subsequent prioritization of policy options.
• Ex-post modeling methodologies can support impact evaluation by improving the understanding of the relations existing among key variables in the system and by comparing the projected performance with initial conditions and historical data. This can be done by considering individual interventions or a policy package. Improvements to the model and updated projections allow decision-makers to refine targets and objectives, building on synergies and positive spillovers across sectors.
On top of their capability to support issue identification and agenda setting, policy formulation and assessment, and policy monitoring and evaluation, the methodologies are evaluated based on their complementarity with other approaches and their capability to involve a variety of stakeholders in model development and use.
Complementarity is important as it allows to strengthen the analysis and address some of the weaknesses of each methodology with inputs from others. Further, the simultaneous use of different methodologies supports the broader involvement of various stake -holders (technical and political) in policy formulation and evaluation. This latter aspect is particularly important in the context of the green economy. The goal being sustainable development, it is crucial that a green economy strategy is developed, and analyzed, for its impacts across sectors. The simultaneous evaluation of social, economic and environmental dimensions can only be carried to with the adoption of a multi-stakeholder approach in which projected impacts are evaluated, and if necessary, mitigating and/or complementary actions are designed and evaluated.
1.2. MODELS
The review and assessment of models use two main criteria: relevance to the concept and definition of the green economy, and ease of creation and use.3 More specifically, the former is assessed by evaluating the capability of models to:
3 A variety of additional, and more technical criteria could be proposed to assess models, such as the use of discrete or continuous simulation, and how they handle uncertainty. On the other hand, this report is seen as an introductory document, and a more detailed assessment of the strengths and weaknesses of simulation models in relation to specific policy analysis will be included in the upcoming Volume II of this study.
Policy Implementation
Decision Making
Decision making is based
on the results of the policy formulation stage, and should account for the forecasted impacts of policy implementation on the environment, the economy and overall well-being of the population.
figure 1. The Integrated Policymaking cycle, highlighting the three main stages supported by the use of quantitative methodologies
• Represent the social, economic and environmental dimensions of the problems and opportunities analyzed, also incorporating human, economic and natural capital in a single framework of analysis;
• Address climate change, a fundamental upcoming challenge, by forecasting impacts as well as analyzing mitigation and adaptation options; and
• Contribute to green economy investment and policy analysis.
The second set of criteria instead considers model creation and use from a developing country perspective, where data, time and financial resources may be scarce, and tradeoffs need to be addressed. In this case factors such as applicability to a country's context, transparency, implementation and maintenance time, and audience and IP support, take center stage.
2. review of methodologies
The review of methodologies starts with a brief introduction of their strengths and weaknesses to continue with a comparative analysis of their contribution to the policymaking process, respective complementarity with other approaches and accessibility, or multi-stakeholder participation, in the process of model creation.
2.1. DATA FRAMEWORKS 2.1.1. Indicators
An indicator is an instrument that provides an indication, generally used to describe and/or give an order of magni-
tude to a given condition. Indicators provide information on the historical and current state of a given system, and are particularly useful to highlight trends that can shed light on causal relations among the elements composing the system and in analyzing whether progress is made in reaching a given policy target.
When used in the context of policymaking, indicators are useful instruments to inform decision-making (UNEP, 2012). Using inventory data and/or surveys, indicators can be grouped in four main categories (1) indicators for issue identification and agenda setting; (2) indicators for policy formulation; (3) indicators for policy assessment and (4) indicators for policy monitoring and evaluation.
2.1.2. Input - Output
Input-Output (I-O) frameworks depict inter-industry relationships within an economy or across economies, estimating how output from one sector may become an input to another sector. Inputs and outputs can be measured in economic (e.g., the monetary value of trade) and physical terms (e.g., material flows and emissions, or employment). In a typical I-O matrix, columns would represent inputs to a sector, while rows would represent outputs from a given sector. This approach is frequently used to estimate impacts of investments and policies on the value chain of specific products and industries.
2.1.3. Social Accounting Matrix
A Social Accounting Matrix (SAM) is an accounting framework that captures the transactions and trans-
fers between the main actors in the economy. As a result, for any given year, the SAM provides information on the monetary flows that have taken place between, for instance, the government and households, ensuring that all inflows equal the sum of the outflows. The focus on households makes the SAM "social", and makes it an adequate backbone for Computable General Equilibrium (CGE) and other macroeconomic models to carry out analysis that spans across the whole economy.
2.1.4. Geographic Information System
A Geographic Information System (GIS) is a system designed to capture, store, manipulate, analyze, manage, and present all types of geographical data. In the simplest terms, GIS is the merging of cartography, statistical analysis, and computer science technology, and is used to analyze land use changes.
GIS applications use geographically disaggregated data presented in maps. Technically there is no restriction in the type of data that can be included in GIS tools, which often incorporate social, economic and environmental indicators. On the other hand, there could be a scaling problem when the coupling of spatially disaggregated data is not possible (e.g., when attempting to couple detailed local GIS information with economic data that may only be available at the national level).
2.2. MODELING APPROACHES
2.2.1. Econometrics
Econometrics measures the relation between two or more variables, running statistical analysis of historical data and finding correlation between specific selected variables. Econometric exercises include three stages — specification, estimation, and forecasting. The structure of the system is specified by a set of equations, describing both physical relations and behavior, and their strength is defined by estimating the correlation among variables (such as elasticities: coefficients relating changes in one variable to changes in another) using historical data. Forecasts are obtained by simulating changes in exogenous input parameters that are then used to calculate a number of variables forming the structure of the model (e.g., population and economic growth).
The most important limitations of econometrics are related to the assumptions characterizing the most commonly used economic theories: full rationality of human behavior, availability of perfect information and market equilibrium. When looking at the results produced by econometric models, issues arise with the validation of projections (that cannot backtrack historical data) and with the reliability of forecasts that are only
based on historical developments and on exogenous assumptions.
2.2.2. Optimization
The use of optimization in policymaking generates "a statement of the best way to accomplish some goal" (Sterman, 1988). Optimization leads to models that are normative, or prescriptive, and provide information on what to do to make the best of a given situation (the actual one). In order to optimize a given situation, these models use three main inputs: (1) the goals to be met (i.e., objective function, such as minimizing the cost of energy supply), (2) the areas of interventions and (3) the constraints to be satisfied.
Optimization is also used to estimate the impact of external shocks (e.g., policies), such as in the case of CGE models. Here optimization is primarily used to solve the mathematics underlying the model. The assumption is that agents are maximizing welfare (profits or consumption), and the model is solved by finding the price vector that optimizes overall welfare as a representation of how the economy might be thought of as functioning.
The challenges related to optimization models include the correct definition of an objective function, the extensive use of linearity, the limited representation of feedback and dynamics. Such models usually do not provide forecasts, but some of them, such as CGE models (Coady, 2006) as well as MARKAL (Fishbone et al., 1983; Loulou et al., 2004) and MESSAGE (IIASA, 2001, 2002) in the energy sector, provide snapshots of the optimum state of the system with specific time intervals. Such models use exogenous population and economic growth rates, among other exogenous variables.
2.2.2. System Dynamics
System Dynamics is a methodology used to create models that are descriptive, and focuses on the identification of causal relations influencing the creation and evolution of the issues being investigated. System Dynamics models are in fact most commonly used as "what if" tools that provide information on what would happen in case a policy is implemented at a specific point in time and within a specific context.
System Dynamics aims at understanding what the main drivers for the behavior of the system are. This implies identifying properties of real systems, such as feedback loops, nonlinearity and delays, via the selection and representation of causal relations existing within the system analyzed. Potential limitations of simulation models include the correct definition of system's boundaries and a realistic identification of the causal relations characterizing the functioning of systems be-
ing analyzed (e.g., relating to the use of causality rather than correlation).
2.3. COMPARATIVE ASSESSMENT
A comparative assessment of the methodologies analyzed in this study is provided in Table 1. This table does not aim at identifying what is the best methodology, but to review their main strengths and weaknesses, how they contribute to the policymaking process, as well as their complementarity and accessibility. The choice of the best methodology and model to use depends on a variety of additional criteria.
With regard to data frameworks, and concerning the policy process, while the use of indicators can support each phase, I-O and SAM can primarily support policy formulation and assessment, by testing the impact of policies. GIS tools instead can be used to identify problems (by observing trends), support policy formulation (by testing the extent to which a policy, often regulation, would impact land use, among others) as well as policy M&E (by monitoring the evolution of the system over time). Concerning complementarity, indicators, SAM and GIS could be relatively easily incorporated in other types of assessments (provided that data are coherently disaggregated), while the specificity of I-O tables (especially concerning employment and material flows) makes them particularly useful for detailed studies but of more difficult incorporation in other analyses. Regarding accessibility, indicators and GIS are likely to capture the interest of a larger set of stakeholders, mostly due to their cross-sectoral coverage.
With regard to modeling approaches, System Dynamics provides a degree of flexibility that makes it useful and relevant for all policymaking stages. While this does not mean that a single model may be relevant throughout the policy cycle, the methodology allows for the creation of a suite of models that can effectively inform decision makers. Further, econometrics can most effectively contribute to issue identification (by projecting trends based on historical observed behavior), and optimization is better suited for policy formulation and assessment (especially by setting targets and providing information on the best system setup to reach them). Concerning complementarity, elements of econometrics and optimization (especially if used in simulation mode, for solving the underlying mathematics of models) can be easily utilized in several models used for green economy assessments. System Dynamics facilitates the incorporation of knowledge in a single framework of analysis, and can also be coupled with other approaches (e.g., econometrics and optimization, and more increasingly GIS as well). Regarding accessibility, econometrics and optimization generally target a focused target audi-
ence, which would change depending on the scope of the analysis (e.g., energy, economic planning). The use of a systemic approach to develop System Dynamics models makes it instead better suited to broaden the range of stakeholders involved in the modeling process and planning. This is primarily due to the ease of incorporating cross-sectoral factors in the model (e.g., energy-economy-environment nexus).
Beyond Table 1, the study also presents the complementarity of various methodologies through the coupling of models. In fact, among others, System Dynamics models can be optimized and may use econometric inputs, and/or include I-O tables and spatially disaggregated data. CGE models run an optimization routine, use the SAM as their underlying economic accounting framework and can include I-O employment and material flow tables. Spatial models, using GIS data as foundation, can be used to run simulations (optimizing future trend and/or simulating "what if" scenarios).
3. review of models
The review of models focuses on the comparative assessment of their respective key contributions to a green economy assessment. More details on the main characteristics of these models are available in UNEP (2014).
A comparative assessment of the models is provided in Table 2, which shows how models include the different dimensions relevant to green economy. No model can capture all the facets of the green economy. However, CGEs coupled with sectoral biophysical models and System Dynamics could potentially satisfy most criteria if some information from the other models is available (e.g., InVEST, concerning natural capital).
More specifically, I-O models can provide a high level of sectoral disaggregation and generate results analyzed across the value chain of selected products and technologies, tracking employment, material and/or emission flows. Regional I-O models extend this analysis to trade among countries. These models can capture economic and human capital, sustainable consumption and production (SCP) and competitiveness, as well as support investment analysis.
Energy and other system engineering models specifically focus on one or two sectors and can track manufactured capital (even if expressed in physical terms, as built up capital), climate change mitigation options (e.g., in the case of energy) and potentially also climate change adaptation (e.g., in the case of water). These models can support both green economy investment and policy analysis (especially regulation).
GIS-based models (e.g., LCM) and InVEST, being spatially disaggregated and focusing on land use changes, specialize in natural capital and are able to capture eco-
Table 1. Review of methodologies for green economy assessments; contribution to the policy process, complementarity and stakeholder participation.
Methodology Main strengths in assessing the Green Economy Main weaknesses relative to the Green Economy Problem identification Policy formulation Policy assessment Policy M&E Complementarity Accessibility — participation
Static
Indicators Support the entire policy cycle, quantify trends. Require harmonization; primarily limited to (quantitatively) measurable variables. y y y y y y
Input-Output Represent value chain impacts, and ripple effects across sectors. Data intensive; material flows not generally available. y y
Social Accounting Matrix Estimates economic flows across the main economic actors. Covers exclusively monetary flows; lacks feedbacks. y y y
Geographic Information System Captures local trends, based on geographical maps; fully accounts for natural resources and ecosystem services. Data intensive; may miss economic dimensions; uneven data resolution may pose challenges. y y y y y
Dynamic (Projections)
Econometrics Entirely based on historical trends; quick implementation. Lacks the explicit representation of feedbacks and does not capture possible emerging dynamics. y y
Optimization Supports the estimation of targets, understanding key limits of the system. Provides and "end" with little insights on the "means". Not viable for highly dynamic and cross-sectoral systems. y y
System Dynamics Focuses on structure to drive behavior; horizontal sectoral representation; knowledge integrator (ad hoc). Highly reliant on knowledge available in other fields; relatively long implementation time for national models. y y y y y y
logical scarcities and environmental risks. These tools can also support the analysis of human well-being, with access to resources and vulnerability to climate change, being capable of analyzing impacts, mitigation (especially sinks, through land use) and adaptation options. Spatial models are generally better suited to analyze policy impacts (e.g., regulation), rather than green economy investments.
CGE models cover the economic sphere of sustainable development, accounting for manufactured capital, competitiveness and social equity (e.g., through the estimation of income distribution). Human capital can also be estimated, despite methodological constraints, regarding employment, skills, as well as salary and wages.
CGE models can effectively support both investment and (fiscal and monetary) policy analysis.
When coupled with system engineering models, CGEs can more effectively incorporate natural capital (primarily by representing natural resource stock and flows) and ecological scarcities. This allows a fuller estimation of competitiveness, also including SCP and the analysis of capital misallocation (now possible due to the cross-sectoral nature of the model, capable of estimating ecological scarcities). Further, by adding natural resources, the model would be able to analyze climate change mitigation and adaptation options, and make use of spatial information to potentially incorporate impacts as well.
Table 2. Review of models for green economy analysis; relevance to the green economy definition and assessment.
Representation of key pillars (and capitals) of sustainable development Analysis of Climate Change GE Intervention Analysis
Economic dimension Social dimension Environmental dimension
Model Scope of the analysis Economic capital Sustainable Consumption and Production Competitiveness Capital misallocation Human capital Human well being Social equity Natural capital Ecological scarcities Environmental risks CC impacts CC mitigation CC adaptation Investment analysis Policy Analysis
Input-Output (I-O) Macro, with high level of sectoral disaggregation, for monetary and physical flows ✓ ✓ ✓ ✓ ✓
Energy and other System Engineering models Sectoral analysis, with high level of detail ✓ ✓ ✓ ✓ ✓
Geographical Information System (GIS) and Invest Highly geographically disaggregated, with analysis ranging from local to national * ✓ ✓ ✓ ✓ ✓ ✓ ✓
Computable general Equilibrium (CGE) Macro, with sectoral disaggregation ✓ ✓ * ✓ ✓ ✓
CGE and System Engineering (energy and natural resources) Macro, with sectoral detail. ✓ ✓ ✓ ✓ * * ✓ * ✓ * ✓ ✓ ✓ ✓
System Dynamics (SD) models (e.g., T21) Macro, with the possibility to add sectoral detail with social, economic and environmental variables ✓ ✓ * ✓ ✓ * ✓ * ✓ * * ✓ ✓ ✓ ✓
The * indicates the possibility to include basic variables and to address the criteria more extensively with the availability of information generated by other models.
System Dynamic models, both sectoral and integrated, can endogenously represent economic, human and natural capital. The strength of the model and the level of detail of the analysis depend on the identification and understanding of the key drivers of the system, and on the availability of inputs from more detailed employment and natural capital assessments. By accounting for natural resource stocks and flows, ecological scarcities can be estimated, with resulting environmental risks and vulnerabilities (incorporated using results of an InVEST analysis, for instance). At the economic level, given the typical high level of aggregation of System Dynamic models, SCP could be simulated and analyzed from a macro perspective, tracking
consumption of the most relevant inputs to production (especially natural resources). Further, competiveness and capital misallocation would be endogenously estimated, providing insights on the key — past, present and future — drivers of economic growth. Concerning social dimensions, while social equity would be estimated through income distribution, the calculation of human well-being could use indicators from a variety of sectors, including environmental ones. As in the case of CGEs with system engineering modules, climate change impacts could be incorporated if science is available, and the model could simulate and support the evaluation of mitigation and adaptation options using cross-sectoral indicators (including direct, indirect
and induced impacts). Finally, System Dynamics models can be used to carry out both green economy investment and policy analysis.
This section reviewed some of the various criteria for choosing the best model to use, criteria which relate primarily to the problem to be analyzed, the stage of the policymaking process to influence and the constraints relating to timing, budget and human resources (e.g., local knowledge of modeling techniques and time availability).
4. exploiting the complementarity of existing methodologies and models for a green economy assessment
The green economy work carried out so far has focused on filling gaps in the policy process to fully incorporate the environmental dimension in national planning. With respect to simulation models, gaps have been identified in the inclusion of the social and environmental dimensions of development in modeling exercises related to national planning, and in the use of long-term projections to inform policy making.
As the field of the green economy evolves and new challenges arise, it is becoming more and more evident — and not surprisingly so — that no perfect model exists, and that certain contexts may require a suite of highly customized models to effectively inform pol-icymaking. Before model (s) can be selected, the goals and issues should be clearly identified and the state of the policymaking process analyzed. It may well turn out that more than a model is needed to solve a given problem or support a single policy process, especially if cross-sectoral linkages are important and both short and longer term analysis is needed. In addition, certain governments may have already embarked in modeling work and may want to make use of their existing models and knowledge to support ongoing planning efforts. They may at the same time need to develop other models ad hoc. For these reasons, complementarity will play a key role in determining the usefulness of quantitative assessments for policy formulation and evaluation at the country level.4
Table 3 highlights the key commonalities and inter-dependencies of various models, indicating how apparently different streams of work can complement each other. This table should be analyzed in conjunction with Table 2. While the specific technical characteristics and
4 Considerable amount of work is being done to improve the integration of social, economic and environmental factors in Integrated Assessment Models. While this type of work is beyond the scope of the report, more information can be found at the Integrated Assessment Modeling Consortium website (http://www.globalchange. umd.edu/iamc).
capabilities of each model are analyzed in Table 2, the following one refers to the analysis that could originate from aligning the use of different models.
The complementarity of models in creating a coherent green economy analysis is evident when considering their sectoral coverage, time horizon, and also when taking into account their support to the policymaking process (see Table 1). Some models are better suited to help decision makers in the issue identification phase (e.g., the analysis of historical trends with InVEST), while others are designed to shape policy formulation (e.g., with the identification of goals and targets, such as in the case of optimization). As mentioned above, an additional layer of complementarity is the time horizon of the analysis created with simulation models. Due to methodological characteristics (both strengths and weaknesses), some models are better suited for short-term analysis (e.g., I-O), while others are designed to address medium and longer-term trends (e.g., T21).
A case study is provided to highlight the potential to combine various models for a comprehensive green economy assessment. The example refers to a specific problem or policy objective, namely fossil fuel subsidy removal.
4.1. CASE STUDY: FOSSIL-FUEL SUBSIDY REMOVAL
Assessing the implications of the rationalization of fossil-fuel subsidies requires a cross-sectoral analysis, touching upon economic and social indicators, and also affecting energy (fossil fuel) consumption and production, and as a consequence impacting the environment (Bassi, 2012). As a result, several methods and tools are needed to carry out a comprehensive and solid analysis of the impacts of fossil-fuel subsidies removal.
Most of the efforts in designing the analytical framework and methodology for a study carried out by IISD for the Asian Development Bank (ADB)5 consisted in the identification of the synergies that can be created by using existing methods and tools. With the goal to adopt a framework that is of easy implementation and replication, but still solid and rigorous, the team has selected three main tools, all based on (and supported by) quantitative data. These are (1) the MARKAL energy model, (2) a Macro Economic CGE model, and (3) the SAM. Strong synergies can be created when using simultaneously the three tools, as they can (and should) all feed results to each other to allow for a comprehensive and solid assessment.
With respect to the analysis of the impacts of fossil-fuel subsidies it can be argued that:
5 The project is "TA-7834 REG: Assessment and Implications of Rationalizing Fossil-Fuel Subsidies", done in the context of IISD's Global Subsidies Initiative.
Table 3. Review of the complementarity of models in creating a green economy analysis.
Information Provided
Model Input-Output (I-O) Energy and other System Engineering models Geographical Information System (GIS) and InVEST Computable General Equilibrium (CGE models) System Dynamics (SD) (e.g.,T21)
Input-Output (I-O) * Projections, planned capacity expansion Spatial distribution of employment, material flow * Projections, economic growth across sectors * Projections, with feedbacks across sectors
Energy and other System Engineering models Energy flow, for value chain analysis Water availability for cooling, proximity of transport means for fuels (e.g., coal) GDP, for energy demand estimation Socio-economic impacts of energy choices, repercussions on energy demand
Information received Geographical Information System (GIS) and InVEST Employment in sectors affecting or impacted by the environment Emissions and fuel/water requirements from/for power generation Economic growth Socio-economic impacts of environmental trends/policies: direct, indirect and induced.
Computable General Equilibrium (CGE models) Employment and material flow (for extended CGEs) Energy price (production cost) and investment information Spatial information, natural resource stocks (for extended CGEs) Long-term feedback responses (e.g., rebound effect)
System Dynamics (SD) (e.g.,T21) Employment, material and energy flow Energy system structure, construction, O&M costs Spatial information, natural resource stocks, ecosystem services SAM structure
* Input-Output tables generally only provide information to other models. These data, used as inputs are then simulated using econometrics, optimization and System Dynamics.
• The SAM and MACRO are particularly useful in analyzing consumer subsidies from a macroeconomic perspective, with data (such as household surveys) being necessary to carry out a detailed assessment of the impact on household (e.g., considering income classes, regional differences among the population and other social factors).
• MARKAL and MACRO are needed instead to analyze producer subsidies, with the former emphasizing the biophysical dimensions of the energy sector, and the latter estimating the economic impacts of decisions on energy supply.
As mentioned above, the MARKAL, MACRO and the SAM can be very complementary, strengthening the analysis that otherwise would be carried out with each of them used independently. In fact, the basic accounting structure and much of the underlying data of CGE models are derived from a SAM, making them useful and easy to implement for the analysis of consumer subsidies; and MARKAL has been improved and expanded by linking it with CGE models, making them excellent tools to use in conjunction when analyzing producer subsidies.
The figure below shows how the three tools can be used in a synergetic manner to make use of the strengths of each of them and carry out a solid and easily replicable analysis of the implications of rationalizing fossil fuels.
The combined utilization of these three tools is necessary to generate coherent projections, and analyze them in the context of the rationalization of fossil-fuel subsidies. In this respect, several policy-related questions can be analyzed, starting from the various options for subsidy removal (e.g., what reduction, by when, and in which shape/form), and ending with the potential reallocation of avoided public expenditure (to which household groups, and with which policy intervention option). Additional analyses become relevant depending on the scenarios simulated, and these include the differences between short and long term impacts, as well as policy and system responses.
5. conclusions
While not aiming at identifying the best methodologies and models for the definition and analysis of green
Sectoral and geographically disaggregated impact analysis for households (e.g., savings)
Reallocation of funding. ' Economic flows across Distributional effects and
opportunities.
the key actors of the economy.
Energy sector analysis. Optimization of energy supply, at least cost.
MACRO
(CGE model)
Macroeconomic assessment. Economic impact of energy prices.
Producer subsidies
k Energy _
production costs
Energy (market) prices
Inflation
Consumer subsidies
Figure 2. Key elements of integration of MARKAL, MACRO and SAM for the analysis of the implications of rationalizing fossil
fuels.
MARKAL is used to generate information on energy production costs, taking into account producer subsidies. With energy costs and consumer subsidies, energy market prices can be estimated. Energy prices are then used to estimate energy demand (and possibly GDP and other macro-economic flows) in MACRO, and inflation. MACRO is used in conjunction with the SAM to estimate GDP and household income, as well as consumption, savings and investment. With data from the household survey it is possible to disaggregate impacts by income classes and location, and also estimate household energy costs. This is done by combining estimates on energy demand (potentially using driving needs - e.g., distance from the workplace - and household or housing size) and energy prices, obtained from the MARKAL and MACRO. Finally, income, originating from the SAM and the household survey will also be used to estimate household energy demand at the micro level, on top of using prices and specific needs.
economy strategies, this study provides key information for researchers and policy makers interested in green economy policy assessment to evaluate the adequacy of various models in meeting their specific and unique needs. As discussed in the paper, integrating and/or linking different modeling approaches and models is often required for the types of complex questions posed by a green economy assessment. In this respect, there are two critical factors to consider:
1. The peculiarities of the local context; and
2. The analysis carried out with models.
The former was already addressed in this report and includes, among others, data availability, knowledge and skills available within critical ministries and national research organizations. The latter refers to the fact that the potential to effectively inform the policymaking process is highly dependent on the type and breadth of the analysis carried out, as models may be misused or utilized below their potential. For this reason the evaluation of models emphasizes the need to ensure broad stakeholder involvement, the necessity to estimate impacts across sectors, for the short, medium and longer term, while considering direct, indirect and induced impacts of action and inaction.
references
Bassi, A.M. (2009). "An Integrated Approach to Support Energy Policy
Formulation and Evaluation". University of Bergen, Norway. Bassi, A.M. (2010). Evaluating the use of an integrated approach to support energy and climate policy formulation and evaluation". Energies. 2010; 3 (9):1604-1621. Bassi, A.M. (2011). A Context-Inclusive Approach to Support Energy Policy Formulation and Evaluation. Regional Environmental Change, Volume 11, Issues 2 (2011), Page 285-295. Bassi, A.M. (2012). "Understanding the role of energy subsidies using a systemic approach". Working Paper 5, Development Series. KnowlEdge Srl, Castellanza, Italy. Cambridge Econometrics (CE) and Sustainable Europe Research Institute (SERI) (2010). A Scoping Study on the Macroeconomic View of Sustainability. Cambridge, United Kingdom. Coady, D. (2006). Indirect Tax and Public Pricing Reforms. In A. Cou-douel, Analyzing the Distributional Impact of Reforms: A Practitioner's Guide to Pension, Health, Labor Markets, Public Sector Downsizing, Taxation, Decentralization and Macroeconomic Modeling (pp. 255-312). Washington D.C.: World Bank. Fishbone, L.G., G. Giesen, G. Goldstein, H.A. Hymmen, K.J. Stocks, H. Vos, D. Wilde, R. Zocher, C. Balzer, H. Abilock (1983). User's Guide for MARKAL. IEA Energy Technology Systems Analysis Programme. IEEP, Alterra, Ecologic, PBL and UNEP-WCMC (2009) Scenarios and models for exploring future trends of biodiversity and ecosystem services changes. Final report to the European Commission, DG Environment on Contract ENV.G.1/ETU/2008/0090r. Institute for
European Environmental Policy, Alterra Wageningen UR, Ecologic, Netherlands Environmental Assessment Agency, United Nations Environment Programme World Conservation Monitoring Centre.
Institute for Global Environmental Strategies (IGES) (2010). Carbon Emissions Embodied in International Trade. An assessment from the Asian perspective
International Atomic Energy Agency (2001). Wien Automatic System Planning (WASP) Package. A Computer Code for Power Generating System Expansion Planning. Version WASP-IV. User's Manual, Vienna: IAEA
International Institute for Applied Systems Analysis IIASA (2002). Achieving a Sustainable Energy System.
International Institute for Applied Systems Analysis IIASA. (2001). Model MESSAGE. Command Line User Manual, Version 0.18.
International Labour Organization (ILO) (2012a). Working Towards Sustainable Development: Opportunities for Decent Work and Social Inclusion in a Green Economy
International Labour Organization (ILO) (2012b), Proposals for the Statistical Definition and Measurement of Green Jobs, Geneva: ILO
Lofgren, H. and C. Diaz-Bonilla (2010). MAMS: An Economy-wide Model for Development Strategy Analysis. The World Bank. Washington DC, USA.
Lofgren, H., R.L. Harris, S. Robinson (2002). A Standard Computable General Equilibrium (CGE) Model in GAMS, International Food Policy Research Institute (IFPRI)
Loulou R, Remme U, Lehtila A, Goldstein G. (2005). Documentation for the TIMES model, Part II. Energy technology systems analysis programme (ETSAP).
Loulou, R., G. Goldstein, K. Noble (2004). Documentation for the MARKAL Family of Models. IEA Energy Technology Systems Analysis Programme.
Sterman J.D. (1988). A Skeptic's Guide to Computer Models. In: Barney GO, et al, editor. Managing a nation: the microcomputer software catalog. Boulder, CO: Westview Press. p. 209-229.
United Nations Conference on Environment and Development (UNCED). Agenda 21, as adopted by the Plenary in Rio de Janeiro, on 14 June 1992.
United Nations Environment Management Group (EMG) (2011). Working Towards a Balanced and Inclusive Green Economy, A United Nations System-wide Perspective.
United Nations Environment Programme (UNEP) (2011). Towards a Green Economy: Pathways to Sustainable Development and Poverty Eradication, Nairobi: UNEP.
United Nations Environment Programme (UNEP) (2012). Measuring Progress Towards an Inclusive Green Economy, Geneva: UNEP
United Nations Environment Programme (UNEP) (2013). Green Economy Scoping Study: South African Green Economy Modelling Report (SAGEM) — Focus on Natural Resource Management, Agriculture, Transport and Energy Sectors. Geneva, Switzerland.
United Nations Environment Programme (UNEP) (2014). Using Indicators for Green Economy Policymaking. Geneva, Switzerland.
Van Paddenburg, A., Bassi, A., Buter, E., Cosslett C. & Dean, A. (2012). Heart of Borneo: Investing in Nature for a Green Economy. WWF Heart of Borneo Global Initiative, Jakarta.
