1. Introduction
Involving southern countries in the fight against climate change, without hindering their development, is a major challenge for the international community. The Clean Development Mechanism (CDM), under the Kyoto Protocol, currently constitutes the primary tool for enabling northern countries (Annex I countries) to make environmentally friendly direct investments at low costs in developing countries (Non-Annex I countries),Footnote 1 through the carbon credit granted in proportion to the greenhouse gas reductions achieved. The CDM has enjoyed great success since its introduction in November 2004. As of April 2013, 6,600 CDM projects had been registered worldwide, for a total investment of approximately US$215.4 billion (UNFCCC, 2012; Fenhann, Reference Fenhann2013). However, despite extensive studies of the CDM in economic literature, mainly to assess the costs of climate policy, its impact on the development of host countries remains insufficiently explored.
The main objective of this study is to assess the impact of CDM on a host country, using the specific case of Mexico, where CDM has been a significant funding source since its ratification of the Kyoto Protocol in September 2002. At the start of 2013, 215 CDM projects had been registered or planned in Mexico, equivalent to nearly US$13.2 billion, making it the fifth largest recipient of such funds. Using the data from the CDM pipeline (Fenhann, Reference Fenhann2013), we numerically simulate the economic impacts of these foreign CDM investments on the Mexican economy, using a computable general equilibrium (CGE) model that includes environmental variables, which reveal the ecological consequences of these economic changes.
In the next section, we outline some limitations of extant CDM assessments in economic literature, before we detail our main hypotheses and present the CGE model for Mexico. Next, we describe our model-based scenario of the impact of CDM investments in Mexico and our quantitative results. Finally, we consider some alternative scenarios, to test the sensitivity of our results.
2. Assessing the CDM in economic literature
Since the establishment of the Kyoto Protocol, the goal of assessing the CDM has emerged progressively as a critical research area for economic literature, focusing on two complementary issues: how CDM achieves its environmental objective in terms of climate policy; and how it achieves its development objective in southern countries.
The analyses in the first group are more numerous, because there are strong incentives to analyze the effect of CDM on greenhouse gas emission reductions in order to quantify the revenues generated by each project in the carbon market. Most studies in this tradition either estimate marginal abatement cost curves or use microeconomic CGE models (for a review, see Böhringer et al., Reference Böhringer, Rutherford and Springmann2013). For example, Klepper and Peterson (Reference Klepper and Peterson2005) use a multi-region, multi-sector, recursive, dynamic CGE model of the world economy to compare the costs for investors stemming from the different flexibility mechanisms provided by the Kyoto Protocol. Jotzo and Michaelowa (Reference Jotzo and Michaelowa2002) and Michaelowa and Jotzo (Reference Michaelowa and Jotzo2005) apply a numerical partial equilibrium approach to assess how transaction costs and institutional rigidities influence the market size of a CDM. Finally, Anger et al. (Reference Anger, Böhringer and Moslener2007) use a large-scale CGE model of international trade and thereby quantify the macroeconomic impact of CDM, while accounting for the relative importance of investment barriers and risks and CDM regulations.
The second group of analyses assessing the CDM's potential impact on southern countries is smaller, perhaps because its contributions to development appear de facto less important than its environmental impacts. Development is just one objective required in the project design documents (PDD)Footnote 2 that each CDM must submit to the DNA of a host country. One group of econometric studies attempts to classify the types of CDM projects according to their technology transfer contributions (Haites et al., Reference Haites, Duan and Seres2006; Dechezleprêtre et al., Reference Dechezleprêtre, Glachant and Menière2008; Seres and Haites, Reference Seres and Haites2008). Another group focuses on more global development issues. Early studies in this tradition tended to adopt prospective quantitative approaches (Banuri and Gupta, Reference Banuri, Gupta and Manila2000; Mathy et al., Reference Mathy, Hourcade and De Gouvello2001), although most research features qualitative analyses with multiple criteria (Kolshus et al., Reference Kolshus, Vevatne, Torvanger and Aunan2001; Huq, Reference Huq2002; Begg et al., Reference Begg, Parkinson and Horst2003; Sutter, Reference Sutter2003; Anagnostopoulos et al., Reference Anagnostopoulos, Flamos, Kagiannas and Psarras2004; Olhoff et al., Reference Olhoff, Markandya, Halsnæs and Taylor2004; Olsen, Reference Olsen2007; UNFCCC, 2012). However, these heterogeneous studies are sometimes controversial (Olsen, Reference Olsen2007; Olsen and Fenhann, Reference Olsen and Fenhann2008; Boyd et al., Reference Boyd, Hultman and Roberts2009; Alexeew et al., Reference Alexeew, Bergset, Meyer, Petersen, Schneider and Unger2010; Bumpus and Cole, Reference Bumpus and Cole2010; Subbarao and Lloyd, Reference Subbarao and Lloyd2011. Their relevance appears limited by the arbitrary weight studies granted to the different development indicators and, in many cases, they depend on criteria specified by the DNA of each host country to select projects. Without ex post verification by independent studies, these selections usually reflect national priorities and conceptualizations of development or even are guided by competition with other DNAs that also are trying to attract CDM investments. Thus several studies call for an international standard for rating CDM projects, according to their contributions to various development dimensions (Cosbey et al., Reference Cosbey, Parry, Browne, Babu, Bhandari, Drexhage and Murphy2005; Schlup, Reference Schlup2005; Cosbey, Reference Cosbey2006; Sutter and Parreño, Reference Sutter and Parreño2007; Olsen and Fenhann, Reference Olsen and Fenhann2008).Footnote 3
Our study can be included in this second research stream. In an effort to assess the potential impact of CDM on Mexico's development, we adopt a quantitative approach and simulate numerically the impact of foreign CDM investments on the Mexican economy. With this perspective, we use a CGE model to reveal the various interdependencies that generate exogenous shocks on demand, supply, income and prices in the economy. Noting criticisms of the use of CGE models (e.g., Böhringer and Löschel, Reference Böhringer and Löschel2006) – especially their inability to estimate social and environmental dimensions of development accurately (Scrieciu, Reference Scrieciu2007) or assess the fundamental challenge of establishing a monetary valuation of natural resources – we extend the standard CGE analysis by including an environmental perspective that reveals the ecological consequences of economic changes, in accordance with the system of environmental accounting used in Mexico.
3. CGE model
The model, as detailed in the Appendix, encompasses eight main productive activities (agriculture, mining, electricity, energy, construction, industry, transport, services) and three domestic institutions (households, government and firms).
3.1. Main economic features
Because we adopt a developmental perspective, the internal economic logic of the model is fairly standard to CGE modelling for a developing country (see Dervis et al., Reference Dervis, De Melo and Robinson1982; Bourguignon et al., Reference Bourguignon, Branson and De Melo1989; Robinson, Reference Robinson, Chenery and Srinivasan1989). On the supply side, activities maximize profit by combining fixed capital with labour (equations (1)–(3)). Their products may be sold locally or to the rest of the world, given a constant elasticity transformation specification (equations (6)–(9)), and they are imperfectly substitutable with imported products, given an Armington (Reference Armington1969) specification (equations (10)–(13)). Primary incomes generated by the activities are distributed to the different agents, on the basis of their factor endowments and access to transfer and foreign incomes (equations (14)–(28)). On the demand side, households' consumption is a linear expenditure system function; nominal demand from the government is assumed to be exogenous (equations (29)–(31)). Intermediate consumption is driven by fixed technical coefficients in each production process. The volume of investment demand (including foreign CDM investments addressed to each activity,
$\overline{CDMQ_{i}}$
) is assumed to be exogenous (equations (33)–(34)).
Prices, nominal investments and row saving ensure the closure mechanisms for product markets, the capital market and the external accounts market, respectively (equations (35)–(38)). However, the closure rule for the labour market (equations (39)–(40)) deserves more attention, in that it reflects the dual nature of the Mexican economy, as is often assumed for developing economies. That is, the inefficient social insurance system in Mexico leads us to assume that unemployment might not be an option for workers excluded from activities. If the overall labour supply is fixed in volume and a real wage is assumed to be rigid, labour market balance can be achieved through the presence of an informal sector that serves as a shelter for ‘unemployed’ workers. This informal sector is composed of individual, non-capitalistic microenterprises (equations (4)–(5)), whose products are only sold in domestic markets for final household consumption (equation (30)) and whose value added is fully distributed to individual entrepreneurs (equation (16)).
3.2. Environmental features
Some additional equations link existing economic variables to some environmental variables, following the logic of the international System of Environmental and Economic Accounting (SEEA) used by the Mexican Environmental Accounts System.Footnote 4 Two environmental costs can be associated with the production of each (formal) activity. First, depletion costs represent the monetary valuation of the depreciation of natural resources used in the production process. For Mexico, these costs mainly refer to hydrocarbons, forest resources and underground water. Their valuation relies on a market value approach, which covers natural assets using the economic value indicated by current or potential market transactions (United Nations, Reference Nations2000).Footnote 5 Secondly, degradation costs represent monetary assessments of the cost of restoring environmental damages caused by each economic activity. For Mexico, these costs mainly involve the pollution of water, air and soil. Their valuation is more difficult and relies mainly on a maintenance cost valuation method (i.e., estimate the costs of restoring an environmental asset to its original or a tolerable level of degradation) and a contingent valuation method for environmental services.
In the model, each environmental cost is assumed to be proportional to the volume of production of each activity (equations (45)–(46)). Subtracted from the value added of each activity, these overall environmental costs indicate the ecological gross domestic product, or green GDP (Boyd, Reference Boyd2007), of the Mexican economy (equations (43)–(44)).
3.3. Model databases and initial equilibrium
Most of the parameters and the initial level of the economic variables in the model were calibrated from Mexico's 2004 Social Accounting Matrix (2004 SAM), as built by Barboza-Carrasco et al. (Reference Barboza-Carrasco, Vázquez-Alvarado and Matus-Gardea2009), adjusted to the needs and objectives of our analysis. First, we reduced the number of activities in the 2004 SAM to match those in the SEEA environmental accounting. Secondly, considering the nature of the CDM projects, we isolated a fossil energy sector (Fossil energy), to specify energies responsible for emissions of a greenhouse gas (gas, oil, coal), which are the main target of CDM projects in Mexico. Thirdly, because of the closure rule chosen for the labour market, we included the informal sector in the 2004 SAM, using data from Mexican employment surveys (INEGI, 2006), so that we could estimate the contributions of informal activities to employment and value added.Footnote 6 The parameters and initial levels for the environmental variables came from the Mexican Environmental Accounts System (INEGI, 2010). For 2005–2009, these costs accounted for an average of 8 per cent of Mexico's GDP: 6 per cent for degradation costs and 2 per cent for depletion costs (INEGI, 2010).
Table 1 describes the contribution of each activity to GDP, green GDP and environmental cost at the initial equilibrium. Nearly 60 per cent of the Mexican national value added is generated by Services activity, with 17 per cent by the Industry. Depletion costs come mainly from Agriculture and Mining, whereas the degradation costs are generated by Transport, Services and Agriculture activities. Globally, these environmental costs represent 7.3 per cent of Mexico's GDP.
Table 1. Economic and environmental contributions of each activity at initial equilibrium (share in %)
Source: Own calculations, based on 2004 SAM and INEGI, 2010.
4. Model-based scenario of the impact of CDM investments in Mexico
4.1. Scenario 1: Main hypotheses and empirical database
The main objective of our first model-based simulation is to evaluate the potential impact of the CDM projects registered for Mexico – that is, projects that have been validated since 2005 by the CDM Executive Board supervising the Kyoto Protocol's CDM. Data from the CDM pipeline (Fenhann, Reference Fenhann2013), supplemented by data from the UNFCCC (2012) about the type, location, amount and stage of completion of each project, confirm the importance of these investments (table 2). As of April 2013, 171 registered projects represented US$11,532.1 million (nearly 1.25 per cent of Mexico's GDP). Projects involving methane avoidance were the most numerous, but the data also indicate a preponderance of wind, hydro and transport projects, equivalent to 98.2 per cent of investments. These projects, and renewable energy projects more particularly, appear to offer a sustainable alternative to the highly polluting thermal generation of electricity, which currently accounts for nearly 71 per cent of installed electricity capacity in Mexico (SENER, 2010).
Table 2. Registered CDM projects in Mexico
Notes: This table refers to CDM projects registered between 2005 and 2013.
Source: CDM pipeline, April 2013; UNFCCC, (2012).
In our model-based scenario, these CDM investments should have three main impacts on the Mexican economy. First, on the demand side, they should generate additional demand for activities that produce the fixed capital goods required for their implementation. Therefore, in the CGE model, exogenous foreign CDM investments addressed to each activity (
$\overline{CDMQ_{i}}$
) constitute the first critical variable for this demand shock (equation (34)). Second, on the supply side, the CDM investments should generate a technological shock, by changing the nature of the production process for Electricity activity toward more environmentally friendly technology. Relying on the principle of additionality from the Kyoto Protocol,Footnote
7
we assume that the generation of ‘clean’ electricity replaces the use of fossil fuels and reduces the value of the technical coefficient parameter between Electricity and Fossil energy activities. Thus in the CGE model, the parameter ica
ener,elec
constitutes the second critical simulation variable. Ceteris paribus, a reduction of this parameter should generate an increase in the Electricity value added (equation (3)) and decreased demand in the Fossil energy sector (equation (32)). Thirdly, we introduce another environmental effect of CDM investments in the simulation. Mexican environmental accounting rules link the pollution costs of Electricity activity proportionally to its production level, given the initial technology process used. Because this technology process should have been modified by the previous technological effect, we assume that the environmental impact of Electricity activity also is affected. In this case, the parameters θp
elec
and θg
elec
constitute the third critical variable for our simulation (equations (45) and (46)). Their values should decrease with the same magnitude as that of the previous technological shock.
To determine the nature and magnitude of these shocks, we analyze the data in each PDD detailed in the CDM pipeline (Fenhann, Reference Fenhann2013), as well as data from the UNFCCC (2012). For demand shock, it appears that wind, hydro and transportFootnote 8 projects largely determine the allocation of CDM investments. For example, for a wind project, 64 per cent of investments go to the Industry branch (e.g., purchase of wind turbines), 24 per cent to the Construction branch (e.g., engineering and electrical installations), 9 per cent to the Services branch (e.g., legal fees, project engineering, control systems, financial and import costs), and 3 per cent to the Transport branch (e.g., transportation costs). For the technological and the environmental shocks, the CDM pipeline indicates that the additional production of electricity expected from registered projects would be nearly 8.1 per cent (4,142 MW) of 2004 total production (EIA, 2004).Footnote 9 Using these elements, we present the first simulated scenario from the CGE model in table 3.
Table 3. Scenario 1
Notes: a Million US$.
Source: Own calculations, based on CDM pipeline, April 2013.
4.2. Simulation results for Scenario 1
Table 4 shows the results of the simulation of this first scenario. As defined in the CGE model, the CDM investments would contribute significantly to Mexican economic growth (+0.50 per cent), mainly due to the demand effect generated by these additional investments for different activities. However, the overall environmental impact of CDM investments appears negative, such that green GDP grows at a slower pace than economic growth (0.45 per cent vs. 0.50 per cent). Economic growth is accompanied by parallel increases of depletion (+3.02 per cent) and degradation (+0.53 per cent) costs. This global effect would stem essentially from the contributions of Agriculture and Mining activities to overall depletion costs (nearly 98.3 per cent) and of Transport, Services and Agriculture activities to overall degradation costs (nearly 95.2 per cent). Finally, the sectors would increase their activity and thus the environmental degradation, which explains the increase (+0.78 per cent) in the share of GDP represented by environmental costs.
Table 4. Scenario 1: simulations of economic and environmental impacts of registered CDM investments in Mexico (percentage variation)
Source: Own calculations, using GAMS software.
With value added growth rates of 9.73 per cent, Electricity activity would be logically the one most affected by the shocks, considering both the nature of the CDM projects and the close links of this activity with other sectors. However, the supply effects of the reduced use of fossil fuels would lead to substantial decreases in the degradation costs generated by this activity, by −5.72 per cent, such that the share of its environmental costs as a proportion of total GDP would decrease by −14.08 per cent. For the other activities, the simulation shows that variations in sectoral production induced by CDM investments would modify their respective contributions to the environment. They would help slightly reduce the environmental impacts generated by Agriculture, Mining, Industry and Transportation. In contrast, Construction and Energy sectors would exert stronger environmental impacts.
5. Alternative scenarios and sensitivity analyses
The model-based Scenario 1 shows that even if CDM investments would contribute positively to economic growth, they would also have negative environmental impacts in Mexico. However, these results may be difficult to interpret in isolation, considering the circular causalities that drive our equilibrium general analysis. Moreover, the validity of the results also depends strongly on our scenario assumptions about the nature and the intensity of the potential effects of CDM projects in the CGE model. To provide points of comparison and gain further insights into the results, we thus change some key elements of Scenario 1. First, with the same empirical and theoretical methodology, we simulate alternative scenarios based on other CDM project portfolios, which should generate different magnitudes of shocks. Secondly, we simulate a hypothetical scenario, ‘business as usual’, for which investments in the power sector should not have any technological effect in terms of electricity generation. Thirdly, we test the sensitivity of the scenarios' results to various magnitudes of the technological effect included in the numerical simulations.
5.1. Scenarios 2 and 3: Changes to the CDM projects' portfolios
Scenario 1 reflected the potential impact of all registered CDM projects validated by Mexico's Executive Board since 2005. However, many other CDM projects remain awaiting validation; as of April 2013, 44 such projects in Mexico accounted for US$2,010.9 million, or nearly 0.22 per cent of Mexico's GDP, which could represent additional electricity production of 1,757.8 MW (+3.4 per cent of current capacity) (UNFCCC, 2012; Fenhann, Reference Fenhann2013). Scenario 2 integrates at-validation projects together with the registered ones and thus envisages broader demand and supply effects on the Mexican economy. In addition, noting their prominence in the CDM project portfolio, we sought to isolate the impact of renewable energy projects, biomass, wind and hydro; as of April 2013, these projects accounted for US$7,615.1 million (0.83 per cent of Mexico's GDP) and could potentially increase electricity production of 4,080 MW (+7.9 per cent of current capacity). Scenario 3 features only these renewable energy projects and thus envisages smaller effects on the Mexican economy. Table 5 describes the main assumptions of Scenarios 2 and 3; table 6 contains the simulation results obtained with the CGE model.
Table 5. Scenarios 2 and 3
Notes: a CDM projects registered between 2005–2013. b Million US$.
Source: Own calculations, based on CDM pipeline, April 2013.
Table 6. Scenarios 2 and 3: simulations of economic and environmental impacts of different CDM project portfolios in Mexico (percentage variation)
Source: Own calculations, using GAMS software.
As we might have predicted, the impact of CDM projects gets amplified when we include projects awaiting validation, along with registered projects. Because of the stronger demand effect of CDM investments (+US$2,010.9 million vs. Scenario 1), expected economic growth would increase to about 0.70 per cent (vs. 0.50 per cent in Scenario 1). The negative overall environmental impact would also appear stronger, such that green GDP grows at a slower pace than economic growth (0.65 per cent). As in Scenario 1, this environmental negative impact occurs because growth would be accompanied by parallel increases of depletion costs (3.96 per cent) and degradation costs (0.46 per cent). Ultimately, the share of GDP represented by these environmental costs would increase by 0.82 per cent for both scenarios. The supply effects on Electricity activity, such as reduced use of fossil fuels and the associated effects on degradation costs, would exert a substantial influence. Furthermore, Electricity activity appears most affected, relative to Scenario 1, such that its value added grows by 38.28 per cent (vs. 9.73 per cent in Scenario 1). Despite this increasing production, environmental costs would increase only slightly by +1.34 per cent (vs. −5.72 per cent in Scenario 1), and the share of environmental costs from the Electricity sector, as a proportion of total GDP, would decrease by −26.71 per cent (vs. −14.08 per cent in Scenario 1). This shift relates strongly to the nature of the projects awaiting validation, 88 per cent of which are dedicated to deriving new electric capacity (+1,758 MW) through biomass, hydro and wind projects.
These trends are also confirmed by the numerical simulation of Scenario 3, although the magnitude of change is smaller. With a smaller demand shock than Scenario 1 (−US$3,917.0 million), the impact on economic growth would appear positive (+0.36 per cent) but attenuated compared with the other scenarios. Green GDP would increase by 0.35 per cent. The depletion costs would increase by 1.92 per cent, and degradation costs would increase by 0.19 per cent; their total share of the GDP would increase to 0.37 per cent. Because of the nature of renewable energy projects, the technological shock is particularly strong (proportional to the demand shock) compared with the other scenarios. In this context, the contribution to GDP of Electricity activity would rise by 18.92 per cent and generate lower degradation costs of −1.01 per cent. Finally, we find a decrease in the share of environmental degradation in GDP due to the electric sector (−16.77 per cent).
5.2. Scenario 4: Business-as-usual scenario
Because the technological effect is a critical variable in our analysis, we also simulated a ‘business-as-usual’ scenario, in which we assume that investments in the power sector are ‘conventional’, that is, without any technological effect on electricity generation. To facilitate the comparison with Scenario 1 and ensure some level of consistency and equivalency across the scenarios in our general equilibrium framework, we assume that these conventional investments generate the same additional demand as in Scenario 1. Table 7 contains the simulation results for this Scenario 4.
Table 7. Scenario 4: Simulation of economic and environmental impacts of conventional investments in Mexico (percentage variation)
Source: Own calculations, using GAMS software.
Two notable main differences arise between Scenario 1 and this business-as-usual scenario. First, the technological shock appears to induce a growth effect that was not captured in the previous simulations. That is, CDM projects are likely to have stronger growth effects than conventional investments, in terms of both economic growth rate (0.50 per cent vs. 0.43 per cent) and Green GDP growth rate (0.45 per cent vs. 0.36 per cent). Secondly, the business-as-usual scenario reveals that the technological effect associated with the CDM investments mitigates the negative impact of the demand effect on the environment. If both types of investments would degrade the environmental performance of growth, the growth process generated in Scenario 1 would be of comparatively better environmental quality than the one obtained in the scenario with conventional investments. For example, the share of GDP represented by environmental costs would increase by 0.82 per cent in the business-as-usual scenario and 0.78 per cent in Scenario 1. Furthermore, CDM investments show a greater capacity to generate green growth than do more conventional investments. As figure 1 shows, when global economic growth increases by one point, green growth increases by 0.90 points in the case of the CDM scenario and 0.84 points in the case of a conventional investment scenario. This differentiated effect is particularly strong for the Electricity sector and the Fossil energy sector, the two areas most affected by the technological shock in terms of environmental costs.
Figure 1. Percentage change in green growth in response to a 1 per cent change in economic growth in Scenario 1 vs. Scenario 4
$\left(\displaystyle{{\Delta \hbox{Green \ GDP} \over \hbox{Green \ GDP}}{\Delta \hbox{GDP/GDP}}}\right)$
5.3. Technological shock sensitivity results
One of the main insights from Scenarios 1–4 is that the overall impact of CDM investments on the environment and the Mexican economy depends on the respective magnitude of the demand and technological shocks included. Demand shock always has a negative environmental impact, through the growth effect it generates. The impact of the technological shock appears more ambiguous; it generates a growth effect too, but also invokes positive environmental effects that can mitigate the latter. Yet in every scenario, the negative environmental impacts surpass the environmentally friendly technological shock.
To better understand the balance of these opposing forces, we test the sensitivity of the simulations' results to various magnitudes of the technological effect. We repeat Scenarios 1, 2 and 3 with different values for the technological and environmental shocks (Δica ener,elec and Δθ elec ), ranging from 0 per cent (no shocks) to 30 per cent (additional production of electricity expected from CDM projects would replace 30 per cent of initial production achieved using fossil fuels). We present the results of these sensitivity tests in figure 2.
Figure 2. Impacts of CDM investments on GDP and green GDP growth rates in Mexico for different technological shock levels
The sensitivity tests reveal first that, ceteris paribus, the growth effect induced by the technological shock increases with the intensity of this shock. In Scenario 1, without any technological change, expected economic growth due to CDM investments would be about 0.43 per cent (0.51 per cent in Scenario 2; 0.30 per cent in Scenario 3). With a maximum technological shock, these growth rates would increase to 0.90, 0.98 and 0.74 per cent, respectively, in the three scenarios. Secondly, a progressive technological shock generates a progressively positive impact on the environment. Thirdly, each sensitivity test reveals a breakpoint at which the friendly environmental effect of the technological shock starts to surpass the negative impact of the demand shock; that is, green GDP starts growing at a faster rate than the GDP. Scenario 3, with only renewable energy projects, achieves this breakpoint more quickly, by a technological shock of 11.5 per cent; this breakpoint occurs at 19 per cent for registered projects and 25 per cent for the registered and awaiting validation sample.
6. Conclusion
This study explores a model-based assessment of the impact of CDM investments on the Mexican economy since 2005, using a CGE model that also includes environmental topics. The numerical simulation of the impact of the projects already registered in Mexico shows that the CDM investments, as modelled in this study, partially meet their objectives. They would contribute positively to economic growth but also have negative environmental impacts. The simulations that include projects that are awaiting validation or focus solely on renewable energy projects confirm these trends; the numerical results reveal differences only in the magnitude of the impacts. These insights show that the overall environmental impact of CDM in Mexico depends on the nature of the sectors it affects and the intensity of the shocks it would generate. At a sectoral level, current CDM investments do not significantly reduce the costs of degradation and depletion of natural resources in Mexico, where negative environmental effects mainly stem from the mining, agriculture and services sectors, whereas these sectors generally attract few investments. Furthermore, as revealed by the business-as-usual scenario and sensitivity tests, a better balance between the environmentally harmful growth effects and the environmentally friendly technology effect could be achieved by projects that induce a stronger technological shock on electricity activity than have appeared previously, as the main target of projects in Mexico.
These results thus can guide the Mexican DNA in choosing which projects to support in order to improve the environmental quality of Mexican economic growth, although they would benefit from added nuance. For example, further research might pursue an even more integrated approach, combining simultaneously a genuine sectoral analysis of the production processes affected by CDM investments with global macroeconomic and environmental approaches.
Appendix: CGE model
Formal activities or products: i or j = Agriculture, Mining, Electricity, Fossil energy, Construction, Industry, Transport, Services Informal activities or products: inf = Informal sector Domestic institutions: H = Households; G = Government; F = Firms Rest of the World: Row
Production block
Formal production process
-
1.
$QX_{i}=A_{i}^{p}\cdot \lsqb \alpha_{i}^{p}\cdot{LD}_{i}^{-\mu_{i}^{p}} + \lpar 1-\alpha_{i}^{p}\rpar \cdot \overline{K_{i}}^{-\mu_{i}^{p}}\rsqb ^{-\displaystyle{{1}\over{\mu_{i}^{p}}}}$
-
2.
$LD_{i} =QX_{i} \cdot \left[\displaystyle{{\alpha_i^p \cdot PVA_{i}} \over {\lpar A_i^p\rpar^{{\mu}_i^p} \cdot w_i }}\right]^{\sigma_{i}}$
-
3.
$PVA_{i}=PX_{i}-\sum_{j}{PQ}_{j}\cdot{ica}_{j\comma i}-tx_{i}$
Informal production process
-
4. QD inf r = γinf·LD inf
-
5.
$PVA_{\rm inf}={PD}_{\rm inf}-\sum_{j}{PQ}_{j}\cdot{ica}_{j\comma \rm inf}$
Trade block
Exports and domestic sales (CET specification)
-
6.
$QX_i=A_i^t \cdot \lsqb \alpha_i^t \cdot QE_i^{-{\mu}_i^t} + \lpar 1-{\alpha}_i^t\rpar \cdot QD_i^{-{\mu}_i^t}\rsqb ^{-\displaystyle{{1} \over {{\mu}_i^t}}}$
-
7.
$\displaystyle{{QD_i}\over{QE_i}}=\left[\displaystyle{{\alpha_{i}^{t}} \over{\lpar 1-\alpha_{i}^{t}\rpar }}\right]^{\sigma_{i}^{t}}\cdot \left[\displaystyle{{PD_i} \over {{PE}_{i}}}\right]^{\sigma_{i}^{l}}$
-
8.
$PX_{i}=\displaystyle{{{PD}_{i}\cdot{QD}_{i}+{PE}_{i}\cdot{QE}_{i}}\over{{QX}_{i}}}$
-
9.
$PE_{i}=EXR \overline{{PWE}_{i}}$
Imports and domestic sales (Armington specification)
-
10.
$QQ_{i}=A_{i}^{m}\cdot \lsqb \alpha_{i}^{m}\cdot{QD}_{i}^{-\mu_{i}^{m}}+\lpar 1-\alpha_{i}^{m}\rpar \cdot{QM}_{i}^{-\mu_{i}^{m}}\rsqb ^{-\displaystyle{{1}\over{\mu_{i}^{m}}}}$
-
11.
$\displaystyle{{{QM}_{i}}\over{{QD}_{i}}}=\left[\displaystyle{{\alpha_{i}^{m}}\over{\lpar 1-\alpha_{i}^{m}\rpar }}\right]^{\sigma_{i}^{m}}\cdot \left[\displaystyle{{{PD}_{i}} \over {{PM}_{i}}}\right]^{\sigma_{i}^{m}}$
-
12.
$PQ_{i}=\displaystyle{{{PD}_{i}\cdot{QD}_{i}+{PM}_{i}\cdot{QM}_{i}} \over{\lpar 1-tq_{i}\rpar \cdot{QQ}_{i}}}$
-
13.
$PM_{i}=EXR \overline{{PWM}_{i}}$
Income and savings block
Factors
-
14.
$RL={RL}_{for} +{RL}_{\rm inf}+\overline{{RL}_{Row}}$
-
15.
$RL_{for}=\sum_{i}{LD}_{i}\cdot w_{i}$
-
16.
$RL_{\rm inf}={PVA}_{\rm inf}\cdot{QX}_{\rm inf}$
-
17.
$RK=\sum_{i} \lpar {PVA}_{i}\cdot{QX}_{i}-{LD}_{i}\cdot w_{i}\rpar +\overline{{RK}_{Row}}$
Households
-
18.
$YH=\lambda_{l}^{H}\cdot{RL}+\lambda_{k}^{H}\cdot{RK}+\overline{DivF}+ \overline{{Trsf}_{H}^{G}} +\overline{{Trsf}_{H}^{Row}}$
-
19. YDH = (1 − t y H )·YH
-
20. SH = mps·YDH
-
21. CFMH = YDH − SH
Firms
-
22.
$YF=\lambda_{k}^{F}\cdot{RK}+\overline{{Subv}_{F}^{G}}$
-
23.
$SF=\lpar 1-t_{y}^{F}\rpar \cdot{YF}-\overline{{DivF}}$
Administration
-
24.
$YG={TAXY}+{TAXS}+{TAXX}+\lambda_{k}^{G}\cdot{RK}+ \overline{{Trsf}_{G}^{Row}}$
-
25. TAXY = t y H ·YH + t y F ·YF
-
26.
$TAXS=\sum_{i}{tva}_{i}\cdot{PQ}_{i}\cdot{QQ}_{i}$
-
27.
$TAXX=\sum_{i}{tx}_{i}\cdot{PX}_{i}\cdot{QX}_{i}$
-
28.
$SG={YG}-\overline{{CFMG}}-\overline{{Trsf}_{H}^{G}}-\overline{{Subv}_{F}^{G}}$
Demand block
Final consumption
-
29.
$\eqalign{{PQ}_{i}\cdot{CFQH}_{i} &= C\hbox{min}_{i} \cdot{PQ}_{i}+{pmc}_{i} \cr & \quad \cdot\left[{CFMH}-\sum_{j\ne i} c\min_{j}\cdot{PQ}_{j}-{PD}_{\inf}\cdot{CFQH}_{\inf}\right]}$
-
30.
${PD}_{\inf}\cdot{CFQH}_{\inf}\, {=}\, C\min_{\inf}\cdot{PD}_{\inf}\, {+}\, {pmc}_{\inf}\cdot \left[{CFMH}\, {-}\, \sum_{i} c\min_{j}\cdot{PQ}_{j}\right]$
-
31.
$PQ_{i}\cdot{CFQG}_{i}=\varphi_{i}\cdot \overline{CFMG}$
Intermediate consumption
-
32.
$DIQ_{i}=\sum_{j}{ica}_{j\comma i}\cdot{QX}_{j}+{ica}_{i\comma \rm inf}\cdot{QD}_{\rm inf}$
Investment
-
33.
$IT=\sum_{i} \overline{{INVQ}_{i}}\cdot{PQ}_{i}+{CDM}$
-
34.
$CDM=\sum_{i} \overline{{CDMQ}_{i}}\cdot{PQ}_{i}$
Closure rules
Commodities
-
35.
$QQ_{i}={CFQH}_{i}+{CFQG}_{i}+{DIQ}_{i}+\overline{{INVQ}_{i}}+\overline{{CDMQ}_{i}}$
-
36. QD inf = CFQH inf
Foreign account
-
37.
$\eqalign{\sum_{i}{EXR}\cdot \overline{{PWM}_{i}}\cdot{QM}_{i} & =\sum_{i}{EXR}\cdot \overline{{PWE}_{i}}\cdot{QE}_{i} \cr & \quad +\overline{{RL}_{{Row}}}\, {+}\, \overline{{RK}_{{Row}}}\, {+}\, \overline{{Trsf}_{G}^{\, {Row}}} +\overline{{Trsf}_{H}^{\, {Row}}}+{Srow}}$
Capital
-
38. IT = SF + SH + SG + Srow
Labour market
-
39.
$\overline{{LDS}^{S}}=\sum_{i}{LD}_{i}+{LD}_{\rm inf}$
-
40.
$w_{i}=\overline{w}$
Price index and numeraire
-
41.
$Pindex = \sum_{i} \pi_{i}\cdot{PQ}_{i}$
-
42. EXR = 1
Environmental equations
-
43.
$GDP=\sum_{i}{PVA}_{i}\cdot{QX}_{i}$
-
44.
$GreenGDP={PIB}-\sum_{i}{DEGC}_{i}-\sum_{i}{DEPC}_{i}$
-
45. DEGC i = θg i ·QX i
-
46. DEPC i = θp i ·QX i
Simulation impacts on parameters
$$\theta_i = \overline{\theta_i^0} \quad \forall i\ne \hbox{{\'{e}lec}}$$
$$\theta_{{elec}} =\lpar 1-{cdm}_{{effect}}\rpar \cdot \overline{\theta_{{elec}}^0}$$
$$ica_{j\comma i}=\overline{ica\, 0_{j\comma i}} \quad \forall i\ne elec$$
$$ica_{{Ener}\comma {Elec}}=\lpar 1-{cdm}_{{effect}}\rpar \cdot \overline{{ica}\, 0_{{Ener}\comma {Elec}}}$$
Endogenous variables
- PVA i
-
Value-added activity price
- PVA inf
-
Informal value-added activity price
- PX i
-
Aggregate producer commodities price
- PD i
-
Domestic commodities price
- PD inf
-
Informal commodities price
- PE i
-
Export commodities price
- PM i
-
Import commodities price
- PQ i
-
Composite commodities price
- P index
-
Consumer price index
- EXR
-
Exchange rate
- QX i
-
Quantity of aggregate commodity i output
- QD i
-
Quantity of domestic supply of commodity i
- QD inf
-
Quantity of supply of informal commodity
- QE i
-
Quantity of commodity i exports
- QM i
-
Quantity of commodity i imports
- QQ i
-
Quantity of composite commodity i
- CFQH i
-
Quantity of consumption of composite commodity i by households
- CFQH inf
-
Quantity of consumption of informal commodity by households
- CFQG
-
Quantity of consumption of composite commodity i by government
- DIQ i
-
Quantity of intermediate demand for composite commodity i
- RL
-
Labour incomes
- RL for
-
Labour incomes from formal activities
- RL inf
-
Labour incomes from informal activities
- RK
-
Capital incomes
- YH
-
Households' income
- YDH
-
Households' disposable income
- SH
-
Households' savings
- CFMH
-
Households' consumption expenditures
- YF
-
Firms' income
- SF
-
Firms' savings
- YG
-
Government's income
- TAXY
-
Income tax
- TAXS
-
Sales tax
- SG
-
Government's savings
- Srow
-
Foreign savings
- CDM
-
Nominal CDM investment
- IT
-
Nominal investment
- LD i
-
Quantity of labour in activity i
- LD inf
-
Quantity of informal labour
- w i
-
Labour wage rate in activity i
- GDP
-
Gross domestic product
- GreenGDP
-
Green gross domestic product
- DEGC i
-
Degradation costs from activity i
- DEPC i
-
Depletion costs from activity i
Exogenous variables
-
$\overline{{CFG}}$
-
Government consumption expenditures
-
$\overline{{Trsj}_{H}^{G}}$
-
Transfers from government to households
-
$\overline{{Subv}_{F}^{G}}$
-
Transfers from government to firms
-
$\overline{{DivF}}$
-
Dividends from firms to households
-
$\overline{{Trsf}_{H}^{R}}$
-
Transfers from rest of the world to households
-
$\overline{{PWE}_{i}}$
-
Foreign export commodities price
-
$\overline{{PWM}_{i}}$
-
Foreign import price
-
$\overline{K_{i}}$
-
Quantity of capital factor in activity i
-
$\overline{w}$
-
labour wage rate
-
$\overline{{INVQ}_{i}}$
-
Quantity of investment demand for composite commodity i
-
$\overline{{CDMQ}_{i}}$
-
Quantity of CDM investment demand for composite commodity i
-
$\overline{{RL}_{R}}$
-
Labour income from rest of the world
-
$\overline{{RK}_{R}}$
-
Capital income from rest of the world
-
$\overline{{LDS}_{S}}$
-
Quantity supplied of labour factor
-
$\overline{{CQNMG}_{c}}$
-
Quantity of government consumption of non-marketed commodities
-
$\overline{{Subv}_{{Healthg}}^{h}}$
-
Price subvention of public health commodities
Parameters
- A i p
-
Production function efficiency parameter for commodity i
- α i p
-
Production function share parameter for commodity i
- μ i p
-
Production function exponent for commodity i
- σ i p
-
Production function substitution parameter for commodity i
- ica i,j
-
Quantity of intermediate input i per unit of product j
- A i m
-
Armington function shift parameter for commodity i
- α i m
-
Armington function share parameter for commodity i
- μ i m
-
Armington function exponent for commodity i
- σ i m
-
Armington function substitution parameter for commodity i
- A i t
-
CET domestic–export function shift parameter for commodity i
- α i t
-
CET domestic–export function share parameter for commodity i
- μ i t
-
CET domestic–export function exponent for commodity i
- σ i t
-
CET domestic–export function substitution parameter for commodity i
- tva i
-
Sales tax rate for commodity i
- tx i
-
Production tax rate for commodity i
- t y H
-
Direct tax rate on households
- t y F
-
Direct tax rate on firms
- c min i
-
Subsistence consumption of commodity i for households
- c mininf
-
Subsistence consumption of informal commodity for households
- pmc i
-
Marginal share of consumption of spending on commodity i for households
- pmc inf
-
Marginal share of consumption spending on informal commo-dity for households
- φ c
-
Share of commodity i in total government consumption
- mps
-
Marginal propensity to save for households
- π i
-
Weight of commodity i in the consumer price index
