Energy Policy 158 (2021) 112551 Contents lists available at ScienceDirect Energy Policy journal homepage: www.elsevier.com/locate/enpol Relevant sectors in CO2 emissions in Ecuador and implications for mitigation policies Edwin Buenaño a, Emilio Padilla b, *, Vicent Alcántara b a School of Physical and Mathematics Sciences, Pontificia Universidad Católica Del Ecuador, Quito, Ecuador b Department of Applied Economics, Universidad Autónoma de Barcelona, Campus de Bellaterra, 08193, Cerdanyola Del Vallès, Spain A R T I C L E I N F O A B S T R A C T Keywords: We analyse the relationship between the economic structure and CO2 emissions from fossil sources for Ecuador, a CO2 emissions small developing country that exports raw materials. We use an input–output method to identify the relevant Input–output analysis economic sectors in CO2 emissions. Sectoral emissions are decomposed into an own component (emitted directly Key sectors by its productive process) and an induced component (induced by its interrelation with other sectors). We use the Own and induced multipliers input–output table for 2013 and construct a highly disaggregated vector of CO2 emissions based on fossil energy consumption. The results show that 19 economic sectors (from a total of 71) are relevant in CO2 emissions, of which 8 are classified as key sectors: transportation, refined petroleum, crude oil, electricity, trade services, construction, public administration services and telecommunication services. Despite the last four sectors represent only 9.8% of direct CO2 emissions, they are indirectly responsible for 27.1% of total emissions due to their production chains. Our research orientates effective mitigation policy, as it makes possible to determine which sectors, with a high own component of emissions, require measures such as technological improvements and best practices, and which sectors, with a high induced component of emissions, require intersectoral policies, depending on their supply or demand linkages. 1. Introducción which in the case of Ecuador comes mainly from fossil sources. Although in recent years the production of energy from renewable sources, Ecuador has developed the concept of “good living” (“buen vivir”) as a particularly hydroelectric, has increased significantly (by 33% between roadmap for its growth and development (SENPLADES, 2013). The 2008 and 2018), oil remains, by far, the most important primary energy “good living” has its origin in the Andean concept called “Sumak Kaw­ source. In 2018, total primary energy production (including oil exports) say”", which served as inspiration for the drafting of the 2008 consti­ was 216 million barrels of oil equivalent (BOE), of which 87.5% was oil tution of Ecuador. The Sumak Kawsay is a multidimensional paradigm and only 7.8% was from renewable sources. that proposes new ways of life and has as its central axis the harmonious CO2 emissions are closely linked to the consumption of fossil fuels, development between human beings and nature, from a biocentric hence the importance that different institutions attach to energy policies coexistence perspective (Ramírez-Cendrero et al., 2017). However, to mitigate them. In order to contextualise and see in historical combining these development principles with economic practice can be perspective the importance of CO2 emissions linked to the consumption complex without adequate information and tools to harmonise and find of fossil energy sources on total greenhouse gas (GHG) emissions in an adequate balance between the economic, environmental and social Ecuador, we have drawn Fig. 1. dimensions. Both CO2 emissions and total GHG emissions experienced a sustained According to the Ministry of Energy and Non-Renewable Natural growth since 1999. The share of CO2 in total GHG emissions, measured Resources (MERNNR, 2018), between 2008 and 2018, Ecuador’s pop­ in terms of CO2 equivalent, has been increasing throughout the period ulation increased by 3.2 million people, its GDP by 33%, and per capita reaching 61.1% in 2014, almost two thirds of total GHGs.1 The CO2 energy by 15.3%. This led to greater energy demand and production, emissions from industrial processes, not related to energy consumption, * Corresponding author. E-mail addresses: evbuenano@puce.edu.ec (E. Buenaño), emilio.padilla@uab.es (E. Padilla), vicent.alcantara@uab.es (V. Alcántara). 1 The emissions of nitrous oxide (N2O) and methane (CH4) are related to agriculture and waste treatment. https://doi.org/10.1016/j.enpol.2021.112551 Received 9 April 2021; Received in revised form 19 July 2021; Accepted 26 August 2021 Available online 7 September 2021 0301-4215/© 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). E. Buenaño et al. Energy Policy 158 (2021) 112551 have remained more or less constant, representing only around 6% of Ecuador has limited official information on CO2 emissions at the sectoral total CO2 emissions.2 Thus, without ignoring the importance of the rest level. The most complete official source of information on GHG emis­ of GHG, given its weight on total emissions, the particular analysis of sions and environmental policy is the Third National Communication of CO2 emitted from the combustion of fossil fuels is of great relevance in Ecuador to the United Nations Framework Convention on Climate the Ecuadorian case. Change (Ministry of the Environment of Ecuador, 2017), which contains The concern about the increase in energy consumption and CO2 estimates of CO2 emissions for the year 2012. However, this information emissions in Ecuador and the policies implemented in recent years has only provides a breakdown into four sectors according to the informa­ attracted the attention of several authors. Robalino-López et al. (2014) tion from the National Energy Balance.3 Therefore, although analysed how changes in the energy matrix and in GDP would affect the policy-makers and researchers have analysed sectoral energy con­ country’s CO2 emissions and the importance of energy efficiency and sumption and CO2 emissions, these analyses have not considered a reducing the share of fossil energy. Arroyo and Miguel (2019) and Castro broader sector breakdown, which would make it possible to highlight et al. (2019) projected the energy demand of the country until 2030, its other sectors that may be relevant in terms of CO2 emissions. Nor have implication in terms of CO2 emissions, and the effect of certain energy they taken into account the sectoral interrelationships that exist in the policies at the sectoral level. Cevallos (2016) analysed the transport economy, which would make it possible to show the total weight (direct sector with respect to its energy consumption, efficiency problems and and indirect) that the different economic activities (or sectors) have in Fig. 1. CO2 emissions and their link with total GHG emissions and non-energy emissions in Ecuador (1990–2016). (Source: prepared by the authors based on World Resources Institute (2019)) its impact on CO2 emissions. Martínez et al. (2017) analysed the CO2 emissions and the economic impacts of the energy efficiency program called the National Efficient Kitchen Program (NECP) in Ecuador. 3 The most up-to-date information regarding CO2 emissions from fossil Nevertheless, a major problem for the research on the issue is that sources, published in the National Energy Balance, is of 2019; however, the breakdowns are only available for the following large groups: Energy in­ dustries, Manufacturing and construction industries, Transportation, and Other 2 The CO2 emissions related to industrial processes refer almost entirely to the sectors (Commercial/Institutional, Residential and Agriculture/Forestry/ manufacture of cement. fishing). 2 E. Buenaño et al. Energy Policy 158 (2021) 112551 CO2 emissions. That is why, for a small developing country that exports production and y is the vector of final demand. x can be expressed as raw materials, such as Ecuador, an analysis of this type is relevant, since (I − A)− 1 y, being (I − A)− 1 = B the Leontief inverse matrix.4 The it would shed light on the most efficient way to establish mitigation methodology to determine the key sectors in the emission of CO2 measures proposed by both policy-makers and researchers. described below follows the proposal by Alcántara (2007). We define The objective of our work is to analyse the links between the pro­ cnx1 as a vector of relative CO2 emission coefficients, the result of ductive structure and CO2 emissions in Ecuador. For this reason, we dividing the emissions of each economic activity (in tons of CO2) by the have omitted the study of the emissions directly linked to the activities sectoral production (in millions of dollars), such that c x = E, where E is ′ of families (heating, cooling and lighting). According to IEA (2019), a scalar that measures total CO2 emissions. Replacing x with its equiv­ residential emissions in 2017 amounted to 3930.6 kt of CO2, 11.5% of alent of the Leontief model, we have ednx1 = ̂ c (I − A)− 1 y, where ednx1 is total emissions. Given the nature of consumer activities, the analysis of the vector of sectoral direct emission. We can now obtain the demand these emissions should be approached from another perspective. We multipliers by expressing the final demand of a particular sector as a propose the use of the environmentally extended input–output (IO) proportion of the total final demand (Hazari, 1970). From the final de­ approach. This tool provides information that allows us to understand ∑ mand vector, a distribution vector ̃y is defined, such that ̃ y = 1. In this the existing relationships between productive structures and the effects i that their interrelation generate on the environment. way, the backward linkage is redefined as: The initial IO model of Leontief (1936) has been developed and used μy 1xn = c (I − A)− 1 ̂ (1) ′ ′ in the study of multiple issues related to economics. Regarding the y ̃ determination of “key sectors”, the first proposals were those of Ras­ mussen (1956). The information contained in an IO matrix allows the where, μy is a vector of weighted emission multipliers, in response to an ′ detection of the key sectors of an economy based on the set of sectoral expansion of demand. interdependencies. In general, the methodology of key sectors can be Regarding the supply multipliers, Jones (1976) proposed to use the applied to any vector of sectoral coefficients. In the environmental field, Ghosh model (1958) to measure forward linkages.5 Following the Ghosh Alcántara (1995) analysed the key sectors in SOx emissions for Catalonia model, x = v (I − D)− 1 , where (I − D)− 1 = G is the Ghosh inverse ma­ ′ ′ (Spain), Lenzen (1998) investigated GHGs and primary energy con­ trix, D is the matrix of distribution coefficients, x is the vector of pro­ sumption through an analysis of Australian domestic production and duction and v the vector of primary inputs used for production. Relating imports, and Duarte et al. (2002) applied this method to the use of water it to the emission vector, we have ed1xn = v [I − D]− 1 ̂c , where ed1xn is the ′ in the Spanish Economy. vector of sectoral direct emission obtained from a supply perspective. As The method applied in this work follows the line of studies such as in the demand case, we define a new distribution vector of primary in­ Alcántara (2007), who related the structure of the Spanish economy to puts, ̃ ∑ v, such that ̃ v = 1. The supply multiplier6 is then redefined as: CO2 emissions, using supply and demand multipliers, determining the i key sectors in the emissions, Alcántara and Padilla (2006), who identi­ fied “key” productive sectors analysing the impact that an increase in the μvnx1 = ̂ v (I − D)− 1 c ̃ (2) value added of the different productive sectors would have on total CO2 emissions and identifying the productive sectors responsible for the in­ where μvnx1 is a vector of weighted emission multipliers in response to an crease in emissions when income increases, and Imori and Guilhoto expansion of primary inputs. ′ ′ (2007), who used this methodology to analyse the key sectors in CO2 The mean of the multipliers is given by μ = ny = nv , where n is the μ u μ u emissions in Brazil. Alcántara et al. (2010) broaden the analysis of key number of sectors considered and u a unit vector. The comparison of the sectors through a decomposition of the supply and demand multipliers multipliers obtained in (1) and (2) with the mean allows us a first sec­ into “own” and “pure” components (with “pure” component referring to toral classification, defined in the upper part of Table 1, which is the what we call “induced” component in this paper), applied to the analysis conventional typology in the analysis of key sectors. To calibrate of electricity consumption in Spain, in order to make recommendations whether the j-th activity affects all activities uniformly or not, we use the of energy policy. In the same line, Jodar (2011) determined the key coefficient of variation following the proposal of Rasmusen (1956). sectors of the Swiss economy and decomposed forward and backward Thus, for the demand case, the coefficient would be: linkages into induced and own components. Piaggio (2013) and Piaggio √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ( )2 et al. (2014) also identified the key sectors in GHG emissions for 1 ∑n 1 ∑n a − a Uruguay and applied multipliers decomposition. Regarding the Ecua­ n− 1 i=1 ij n i=1 ij dorian case, although the IO approach has been used in several in­ CVyj = 1 ∑n (3) i=1 aij vestigations for the determination of key sectors (see Gachet, 2005; n Artola, 2009; Fernández, 2007; Aguilar, 2014), it has not been applied to For the supply side, the influence on the i-th industry would be environmental issues. In this paper we will show the sectoral in­ calculated as follows: terdependencies between the different sectors, in order to form a basic framework for the design of CO2 mitigation policies. The study is structured as follows. Section 2 develops the method­ ology used. Section 3 makes a brief description of the data. Section 4 presents the main findings and their discussion. Finally, policy impli­ 4 The matrices are represented by uppercase letters, while the vectors are cations and conclusions are presented in Section 5. represented by lowercase letters. Both vectors and matrices are indicated with bold. The symbol “^” represents the diagonalisation of a vector. 5 Several authors argue that the Leontief inverse matrix is not adequate to 2. Methodology for the analysis of key sectors measure (direct or total) forward linkages (Dietzenbacher and Van der Linden, 1997; Miller and Blair, 2009). The discussion of this point is extensive and can We start from the Leontief model Ax + y = x, where A is the matrix of be reviewed in Lenzen (2003). As noted by Oosterhaven (1988), the Leontief coefficients or technical requirements, x is the vector of total domestic and Ghosh models can be used together only as descriptive tools for forward linkage analysis and key sectors, since it is essential for impact analysis that matrix A be constant. 6 The terms backward linkage, demand multipliers or pull effect, will be used interchangeably throughout the text; as well as the terms forward linkage, supply multipliers or push effect. 3 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table 1 3. Data description Classification of sectors. Key sector μv,i/ > 1 μv,i/ < 1 For the development of this research, the IO tables published by the μ μ classification Central Bank of Ecuador (BCE) were used for 2013. The official publi­ / μy,j > 1 Key sectors Demand drivers cation is a 71 × 71 square matrix, disaggregated into 16 activities /μ μy,j < 1 Supply drivers Rest of sectors belonging to the primary sector, 36 to the manufacturing sector and 18 μ to the service sector. Regarding CO2 emissions, given the lack of data Backward linkage μinduced > μinduced μinduced < μinduced classification yj yj yj yj disaggregated at that level, we made our own estimate to allow a more μown yj > μown yj Relevant sectors in induced Relevant sectors only in own in-depth look at sectoral interrelationships. We used the methodology of and own backward linkage backward linkage Alcántara and Roca (1995) and Alcántara et al. (2008) for the estimation components component of the vector of CO2 emissions. This required to estimate a vector of μown < μown Relevant sectors only in Rest of sectors yj yj induced backward linkage primary energy consumption disaggregated in the 71 activities of the IO component matrix. We used data on energy balances published by the IEA (2016), Forward linkage μinduced vi > μinduced vi μinduced vi < μinduced vi data on energy balances 2013 published by the Coordinating Ministry of classification Strategic Sectors of Ecuador (MICSE), conversion factors from IPCC μown > μown Relevant sectors in induced Relevant sectors only in own vi vi (2008), final energy consumption data disaggregated in supply and use and own forward linkage forward linkage component components tables (SUT) and information of the IO matrix 2013 provided by BCE μown vi < μown vi Relevant sectors only in Rest of sectors (2016). induced forward linkage The procedure used was to relate the information published in the component SUT for 2013 (disaggregated to 277 products) with the information Source: Adapted from Jodar (2011). coming from the energy balance published by the Coordinating Ministry of Strategic Sectors of Ecuador (MICSE). The goal was to build a total √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ( )2 utilization vector of KBOEs for each energy source, for the productive ∑n ∑n 1 n− 1 j=1 d ij − 1 n j=1 d ij sector (excluding residential use and exports). This energy use vector CVvi = ∑n (4) shows the total primary and secondary energy used by the productive sector, including the oil refining process. However, some energy sources 1 n j=1 dij in the use table had to be added so that they were comparable with the Moreover, distinguishing between the pollution generated by an energy balance categories.7 activity due to its own production process and the pollution induced in The data to build the emission matrix are the result of linearly other activities is very useful for the design of mitigation policies. distributing the transformations described in the previous paragraph to Alcántara et al. (2010) propose decomposing the backward and forward the total KBOEs per energy source among the different industries. This linkages into two induced and own components. We use expressions (1) distribution is based on their participation in the total intermediate and (2) for decomposing the supply and demand weighted multipliers, consumption of each energy source, assuming that prices are the same separating the external and internal sectoral links, respectively. between different industries, which, given the existing policies on fuels, In the case of demand, the decomposition of backward linkages into turns out to be a valid assumption. The results of the estimation, their own and induced components is given by: together with other relevant variables of each activity, are presented in μown yj yj = cj Bjj̃ (5) Table 2. This estimate allows us to have the direct CO2 emissions for 2013, ∑ μinduced = ci Bij̃yj (6) disaggregated for the 71 activities of the IO matrix, which opens the yj i∕ =j door to carry out more in-depth analyses of the emissions from fossil sources. Similarly, in the case of supply, the decomposition of forward link­ ages into their own and induced components is given by: 4. Main results and discussion μown vi vi = ci Gii ̃ (7) The total (direct and indirect) emissions from the different sectors ∑ μinduced = vi cj Gij ̃ (8) were calculated for the 71 economic activities of the IO matrix in 2013 vi j∕ =i (Table A.1 in the Annex). A large part of the direct emissions of the different productive sectors ultimately serve to facilitate the production In an activity with a high own component, the measures adopted of other sectors (Alcántara and Padilla, 2009). A clear example is the should focus on technological improvements or better practices, since activity (58) “Transport and storage services”, which represents 21.9% they would be effective in reducing pollution (Alcántara et al., 2010; of direct emissions of the productive system, while it only represents Piaggio et al., 2014). In activities with a high induced forward compo­ 11.9% of total (direct and indirect) emissions from a final demand nent, policies should be implemented to reduce the emissions associated with the activities where the production of these is destined, and policies can also be designed to help to reduce the pollution process of the input that such activities require. Furthermore, when more than one activity is 7 The national accounts heading “Other fuels and petroleum oils (including involved in the process, specific policies are not enough, but intersec­ lubricants)” includes “airplane gasoline (Jet Fuel/Turbo)” and “other non- toral policies must be used. Finally, activities with a high induced energy petroleum products”. For this reason, all consumption outside the backward component demand products from directly polluting activ­ transport industry is eliminated and it is assumed that the total energy con­ ities. Therefore, measures that reduce their demand, as well as techno­ sumption of “Other fuels and petroleum oils (including lubricants)” is concen­ logical measures or best practices that reduce their demand for polluting trated in the transport industry. Subsequently, the “Gasolines” and “Naphtha inputs, can be effective. and natural gasoline” were added to the previous items in the use table in such Table 1 shows the classification of sectors according to their a way that it is comparable to the “Petroleum products” heading of the energy relevance. balance. The uses of “Other fuels and petroleum oils (including lubricants)” outside the transport industry are eliminated and it is assumed that the use in the transport industry corresponds to “Kerosene and Turbo”. Finally, the items “Electric power” and “Transmission and distribution services” are added. 4 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table 2 Estimation of CO2 emissions by economic activity for Ecuador for 2013, total output, value added and final demand. Cod. Activity Direct Emission % Total output Direct emission intensity Value Added Final Demand (Kt CO2) (million US$) (TCO2/million US$) (million US$) (million US$) 58 Transport and storage services 9189 21.87% 8505 1080.4 4337 3495 38 Refined petroleum oils and other products 7399 17.61% 3830 1931.9 439 1759 51 Electricity 6773 16.12% 2769 2445.5 671 1006 12 Crude oil and natural gas 2200 5.24% 14,424 152.5 10,014 13,356 44 Cement, articles of concrete and stone 1870 4.45% 1409 1326.8 613 − 19 64 Services provided to companies and 1275 3.03% 8648 147.4 6639 564 production 54 Trade services 1261 3.00% 14,903 84.6 9763 8004 53 Construction and construction works 1231 2.93% 17,829 69.0 10,151 16,843 65 Administrative services of the government and 856 2.04% 7790 109.9 5849 7790 for the community in general 60 Telecommunications, transmission and 762 1.81% 3750 203.2 1925 2937 information services 40 Other chemical products 691 1.64% 1937 356.9 941 1049 10 Fish and other aquatic products (except 610 1.45% 685 890.2 460 237 shrimp) 9 Live or fresh shrimp and shrimp larvae 563 1.34% 1130 498.3 515 52 57 Restaurant services 517 1.23% 2754 187.8 1518 2432 7 Live animals and animal products 504 1.20% 2415 208.8 505 964 36 Products of treated wood, cork and other 443 1.06% 1514 292.9 543 53 materials 70 Association services, leisure, cultural and 416 0.99% 2484 167.6 1351 2069 sports 43 Glass, ceramics and refractories 397 0.94% 401 989.8 216 58 5 Oilseeds and industrialised products 395 0.94% 1700 232.5 1247 142 16 Meat, meat products and by-products 242 0.57% 2329 103.7 638 1862 63 Real estate services 232 0.55% 6189 37.5 4300 4994 47 Machinery, equipment and electrical 193 0.46% 1267 152.2 476 863 appliances 46 Processed metal products 190 0.45% 1373 138.1 249 1020 56 Accommodation services 188 0.45% 542 347.1 291 274 17 Elaborated shrimp 187 0.44% 1995 93.7 458 1937 69 Non-market social and health services 183 0.44% 3093 59.1 1979 3093 1 Banana, coffee and cocoa 175 0.42% 2833 61.7 1839 2475 42 Plastic products 164 0.39% 871 188.3 178 159 37 Pulp, paper and cardboard, editorial products 158 0.38% 1551 101.7 562 499 and others 6 Services related to agriculture 142 0.34% 351 406.0 266 0 19 Prepared and preserved fish and other aquatic 142 0.34% 1468 97.0 816 1398 species 13 Services related to oil and natural gas 136 0.32% 1875 72.3 1397 0 50 Other manufactured products 130 0.31% 880 147.2 482 141 33 Threads, spinning, weaving and confection 129 0.31% 835 154.7 344 383 25 Sugar, brown sugar and molasses 129 0.31% 445 288.5 114 300 61 Financial intermediation services 127 0.30% 3455 36.9 1989 1204 4 Tubers, Vegetables, melons and fruits 118 0.28% 932 126.8 721 836 15 No metallic minerals 117 0.28% 208 564.9 117 − 1 45 Common metals 117 0.28% 1380 85.1 406 667 29 Various food products 107 0.25% 780 137.0 153 741 14 Metallic minerals 106 0.25% 261 407.2 188 49 67 Public education services 101 0.24% 3717 27.1 3442 3717 18 Fish and other processed aquatic products 87 0.21% 685 127.2 201 465 30 Alcoholic beverages 81 0.19% 978 83.0 483 898 48 Transportation equipment 77 0.18% 1094 69.9 211 925 3 Flowers and buds 76 0.18% 1213 62.4 803 1205 21 Processed dairy products 67 0.16% 1233 54.2 323 1113 2 Cereals 60 0.14% 932 64.2 752 70 52 Water, sanitation and gas services (oil exc) 56 0.13% 557 100.6 413 180 20 Crude and refined oils 55 0.13% 1230 44.7 403 854 23 Bakery products 50 0.12% 894 55.9 375 832 49 Furniture 49 0.12% 1156 42.4 349 1088 8 Forestry products 46 0.11% 1245 36.9 1062 376 31 Non-alcoholic beverage 45 0.11% 710 62.9 279 669 28 Processed coffee products 42 0.10% 270 156.7 110 251 59 Postal and courier services 39 0.09% 319 122.9 156 210 24 Noodles, macaroni and other similar 39 0.09% 100 389.4 26 84 farinaceous products 11 Aquaculture products (except shrimp) 39 0.09% 217 178.6 129 125 34 Clothing 35 0.08% 903 39.1 409 763 26 Elaborated cocoa, chocolate and confectionery 31 0.07% 546 57.4 4 316 products 39 Basic chemical products, fertilizers and 30 0.07% 418 71.6 158 32 primary plastics (continued on next page) 5 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table 2 (continued ) Cod. Activity Direct Emission % Total output Direct emission intensity Value Added Final Demand (Kt CO2) (million US$) (TCO2/million US$) (million US$) (million US$) 22 Grain mill products 29 0.07% 1308 22.5 245 974 35 Leather, leather products and footwear 25 0.06% 518 47.7 222 398 68 Social and health services 22 0.05% 1408 15.9 896 1407 66 Private education services 20 0.05% 1798 10.9 1178 1778 55 Repair and maintenance services of motor 19 0.04% 1148 16.4 1001 460 vehicles and motorcycles 41 Rubber products 14 0.03% 202 66.8 98 149 27 Animal food 12 0.03% 896 13.1 168 288 62 Insurance services and pension funds 11 0.03% 1183 8.9 549 662 32 Elaborate tobacco 0 0.00% 93 4.9 43 93 71 Domestic service - 0.00% 382 0.0 382 382 TOTAL 42,019 100.0% 161,145 90,534 106,446 Source: prepared by the authors based on IEA (2016) energy balances, BCE (2016) data and IPCC (2008) guidelines. perspective. In contrast, the activity (54) “Trade services” represents 3% multipliers close to 1 and coefficients of variation higher than the of direct emissions, but its weight rises to 5.8% of total (direct and in­ average, in most cases. This means that, faced with an increase in de­ direct) emissions from a final demand perspective. This is because other mand, the impact is small and expands heterogeneously to the rest of the activities have to increase their production to meet the demand for in­ activities (with the exception of activities 57 “Restaurant services” and puts required by this activity. 70 “Association services, leisure, culture and sports”, which have low coefficients of variation and, therefore, the impact is more homogeneous). 4.1. Key sectors Finally, six activities have higher than average multipliers, but only on the supply side (forward linkages). These activities have weighted For the analysis of key sectors, we used the weighted multipliers that multipliers close to unity (with the exception of activity (64) “Services would generate an expansion of the demand or of the primary inputs, provided to companies and production”) and coefficients of variation calculated on the basis of the expressions (1) and (2), respectively. lower than the average (with the exception of activities 44 “Cement” and Additionally, their coefficients of variation were calculated, based on 13 “Services related to oil and natural gas”). This means that, in the face expressions (3) and (4). Table 3 presents the classification of the activ­ of an increase in productive inputs, the impact will be small but ho­ ities that were relevant according to the methodology used (following mogeneous to the rest of the activities, while an activity such as 44 the classification criteria described in first three rows of Table 1). generates a high and also very heterogeneous impact; that is, affecting Table 3 contains 19 of the 71 activities of the IO matrix of Ecuador. differently to other activities.8 These 19 activities are the most relevant in terms of CO2 emissions and are divided into: 8 key sectors (relevant from demand and supply), 5 sectors that are relevant only from demand and 6 sectors that are rele­ 4.2. Own and induced components of the relevant activities in the vant only from supply. The eight key sectors are responsible for 70.6% of emission of CO2 direct emissions and 67.1% of total emissions (see Table A1 in the Annex). They are key sectors because they have the highest demand To obtain a more exact measure of the backward and forward (weighted backward linkages) and supply (weighted forward linkages) dependence of each sector with the rest of the economy, this section multipliers. In other words, these activities would be the ones that presents the decomposition of the backward and forward linkages into generate the greatest push and pull effects on the rest of the economy in their induced and own components for the 19 relevant sectors, based on the generation of CO2 emissions. Regarding their coefficients of varia­ equations (5)–(8) (Table 4). This decomposition is useful because miti­ tion, they are below the average in most cases. This means that the gation policies require to understand sectoral linkages since they would impacts are distributed evenly to the rest of the activities. However, not work in the same way in activities that have a high own component, there are differences within this group of key sectors. but a low induced component, or vice versa. Some examples will help the reader understand the classification. The activity (38) “Refined petroleum oils and other products” is key because both weighted multipliers of supply and demand are high. On 8 To verify if similar results are obtained in the determination of forward the demand side, the coefficient of variation, being higher than the linkages using the Leontief inverse instead of the Ghosh inverse, the calcula­ average, is telling us that this impact (that is 6.14 times the average) tions presented in Table A.3 were made. The result was that none of the key expands heterogeneously to the rest of the activities. The coefficient of sectors are affected by this calculation. There are two activities that are clas­ variation of the supply indicates that this impact will spread more ho­ sified as influential from the supply side, regardless of the methodology used; mogeneously to the rest of the activities. The activity (60) “Telecom­ and, obviously, the influential activities on the demand side do not suffer any munications” is key because its weighted demand and supply multipliers alteration. Therefore, 10 of the 14 activities classified as relevant do not un­ dergo any change in their classification and only four would change their ty­ are greater than unity and, therefore, its total impacts are greater than pology depending on the calculation method used. The relevant activities from the average. In the case of the weighted supply multipliers, the coeffi­ the supply-side that do change are: (13) “Services related to oil and natural cient of variation is indicating that the impact spreads homogeneously to gas”, (5) “Oilseeds and industrialised products” and (61) “Financial interme­ the rest of the activities; while on the demand side, the impact expands diation services” are classified as relevant from the supply-side only when the in a more heterogeneous way. Ghosh inverse is used. In contrast, the activity (10) “Fish and other aquatic Five sectors have higher than average multipliers but only on the products (except shrimp)” is classified as relevant only when the Leontief in­ demand side (backward linkages). These activities have weighted verse is used. 6 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table 3 Classification in key sectors and activities that are relevant from a demand- or supply-side perspective. Cod. Activity Weighted Backward Linkages Weighted Forward Linkages Classification / / μy,j CVyj μy,i CVvj μ μ 58 Transport and storage services 8.42 3.04 12.65 4.30 key sector 38 Refined petroleum oils and other products 6.14 4.65 11.04 3.41 key sector 51 Electricity 7.19 3.03 5.76 3.14 key sector 12 Crude oil and natural gas 6.63 3.42 6.18 3.80 key sector 54 Trade services 4.12 3.42 4.83 5.19 key sector 53 Construction and construction works 10.00 2.51 1.55 5.10 key sector 65 Administrative services of the government and for the community in general 2.96 2.11 1.15 - key sector 60 Telecommunications, transmission and information services 2.16 3.75 1.09 3.59 key sector 17 Elaborated shrimp 1.83 3.94 relevant from demand 63 Real estate services 1.60 3.41 relevant from demand 57 Restaurant services 1.49 1.81 relevant from demand 16 Meat, meat products and by-products 1.45 3.49 relevant from demand 70 Association services, leisure, cultural and sports 1.23 2.40 relevant from demand 64 Services provided to companies and production 6.76 3.30 relevant from supply 44 Cement, articles of concrete and stone 1.88 8.18 relevant from supply 40 Other chemical products 1.13 3.83 relevant from supply 13 Services related to oil and natural gas 1.09 8.06 relevant from supply 5 Oilseeds and industrialised products 1.07 2.93 relevant from supply 61 Financial intermediation services 1.00 4.28 relevant from supply Averages 1.00 3.24 1.00 4.32 Source: prepared by the authors based on Table A.2 information (in the Annex) from BCE (2016) data. Table 4 Own and induced components of the relevant activities in the emission of CO2 (normalized). Weighted Backward Linkages Weighted Forward Linkages Classification (Normalized) (Normalized) Cod. Activity μinduced yj /μ μown yj /μ μinduced vi /μ μown vi /μ 58 Transport and storage services 2.97 13.61 4.63 17.11 key activity 51 Electricity 1.76 12.36 0.66 8.60 key activity 38 Refined petroleum oils and other products 0.66 11.35 8.35 12.54 key activity 12 Crude oil and natural gas 6.52 6.75 9.76 4.18 key activity 53 Construction and construction works 16.47 3.84 0.54 2.10 key activity 65 Administrative services of the government and for the community in general 3.10 2.82 0.00 1.78 key activity 54 Trade services 6.05 2.28 9.07 2.47 key activity 60 Telecommunications, transmission and information services 2.25 2.09 0.74 1.29 key activity 63 Real estate services 2.60 0.64 relevant from demand 57 Restaurant services 1.47 1.51 relevant from demand 70 Association services, leisure, cultural and sports 1.27 1.18 relevant from demand 16 Meat, meat products and by-products 2.28 0.67 relevant from demand 17 Elaborated shrimp 3.12 0.60 relevant from demand 64 Services provided to companies and production 13.57 2.97 relevant from supply 44 Cement, articles of concrete and stone 0.29 2.76 relevant from supply 40 Other chemical products 0.59 1.43 relevant from supply 13 Services related to oil and natural gas 2.54 0.29 relevant from supply 5 Oilseeds and industrialised products 1.39 0.88 relevant from supply 61 Financial intermediation services 2.40 0.22 relevant from supply Averages 1.00 1.00 1.00 1.00 Source: prepared by the authors based on Table A.4 (in the Annex) with BCE (2016) data. As mentioned above, the own component of the weighted backward example, the decomposition of the weighted backward linkages of the linkages shows how changes in the final demand of activity j affect the activity (58) “Transport and storage services” indicates that a variation CO2 emissions of the activity itself; while, the induced component of the in the demand for this activity generates an induced effect on emissions weighted backward linkages shows how changes in the final demand of that is 2.97 times higher than the average for the economy; while in the sector j cause CO2 emissions from other activities. On the other hand, the sector itself this increase is 13.61 times higher than the average. On the own component of the weighted forward linkages shows how the vari­ side of the weighted forward linkages, it is observed that a variation in ations in the production of sector i affect the CO2 emissions of the sector i the productive inputs of the activity (58) generates an increase in the itself, while the induced component of the weighted forward linkages emissions of other sectors that is 4.63 higher than the average; while, in shows how the changes in production of sector i causes CO2 emissions the sector itself, this increase is 17.11 times higher than the average. The from other activities. sectors are classified following the criteria described in the last six rows Table 4 contains the decomposition of the backward and forward of Table 1. linkages into their induced and own components, for the 19 relevant The first three activities, (58) “Transportation and storage services”, activities (the data were normalized to have the unit as average). For (51) “Electricity” and (38) “Refined petroleum oils and other products”, 7 E. Buenaño et al. Energy Policy 158 (2021) 112551 are responsible for 55.6% of direct emissions of CO2, but only 30.6% if variation coefficients, the component is more homogeneous on the side the direct and indirect effects are measured (see Table A1 in the Annex). of its forward linkages. This suggests that an adequate mitigation policy We can understand this observing the relevance of the own components requires influencing the final demand and production of other sectors (in of the backward and forward linkages in these sectors, both in absolute particular transport, financial services and professional services) and not magnitude and in relative terms, compared to the induced components. the demand of the activity itself. However, Table 3 indicates that this induced effect of lower magnitude The last two activities classified as key in CO2 emissions are (65) is not negligible since it is distributed homogeneously among various “Administrative services of the government and for the community in activities. An implication of these findings, in terms of mitigation policy, general” and (60) “Telecommunications, transmission and information can be clearly seen in the activity (51). Since the own effect is much services”. They have similar characteristics on the demand side, since higher than the induced effect, the mitigation policy must be directed both show slightly higher induced components than their own compo­ towards the sector itself, for example, through energy efficiency and the nent. Therefore, a combination of direct policies, such as technological generation of electricity with renewable sources. improvements, is required, but also on those activities that provide their The analysis for activities (58) “Transportation and storage services” inputs. With regard to the supply perspective, the effect of these activ­ and (38) “Refined petroleum oils and other products” is similar, since ities is concentrated in the own component. Therefore, a direct their own components are more important than their induced compo­ improvement policy on these activities will be important; in particular in nents. However, an important nuance is that the induced component of the telecommunications service since its coefficient of variation in­ the activity (38) is not minor, which means that an increase in this ac­ dicates that the effect will be homogeneous. In the case of the admin­ tivity will cause other sectors to pollute. istrative services of the government, as it does not provide inputs to A characteristic of these 3 sectors is that all of them are directly other activities, measures to reduce its emissions should focus exclu­ related to the existing energy subsidies in Ecuador for transport, elec­ sively in the activity itself. tricity and gasoline. The induced and own effects of these sectors show The activities that are relevant only from the demand side are: (63) that, an elimination of subsidies (causing an increase in prices and a fall “Real estate services”, (57) “Restaurant services”, (70) “Association in demand), would involve important reductions in emissions in these services; leisure; cultural and sports”, (16) “Meat, meat products and by- activities, but the drag effect they may have on other sectors would be products” and (17) “Elaborated shrimp”. All of these are characterised much smaller. by having close to average multipliers. Likewise, their induced back­ A different case is the activity (12) “Crude oil and natural gas”, which ward linkage component is, in general, greater than their own compo­ has lower demand and supply multipliers, but whose induced and own nent. This means that they are activities that pollute indirectly, so they components are more symmetrical. In this case, a variation in its de­ also require multisectoral policies. mand and in the primary inputs of the sector itself, but also of the other Finally, various activities are relevant from the supply-side sectors, has a significant effect on its CO2 emissions. This is particularly perspective: (64) “Services provided to companies and production”, the case of its backward linkages, where the coefficient of variation in (44) “Cement, articles of concrete and stone”, (40) “Other chemical Table 3 indicates that its induced effect is distributed more homoge­ products”, (13) “Services related to oil and natural gas”, (5) “Oilseeds neously than the average of the activities. On the supply side, the own and industrialised products” and (61) “Financial intermediation ser­ effect is lower than the induced effect and it is also distributed in a vices”. Sector (64) stands out, since it presents the strongest forward heterogeneous way. This is due to providing more than 92% of its pro­ linkage of the whole group (the other relevant sectors from the supply- duction to the oil refining activity. Given these characteristics, inter­ side have multipliers close to the average). The induced forward linkage vention policies cannot be approached in the same way as in the of sector (64) is higher than its own component; therefore, it is also an previous cases, since in addition to direct measures of technological activity that is indirectly induced to emit CO2 because its sales are improvement within the activity, indirect measures are also necessary destined to highly polluting activities, such as transportation, cement, on the sectors that supply and demand inputs from this activity. Thus, electricity, or refined petroleum, among others. Therefore, the policies the policies for activity (12) “Crude oil and natural gas” and activities focused on the sectors where its production is destined would allow a (58) “Transportation and storage services”, (51) “Electricity” and (38) reduction of the emissions of this activity. Another activity with similar “Refined petroleum oils and other products” must be articulated to have characteristics is (61) “Financial intermediation services”. Finally, the the desired effect. activities (44) “Cement, articles of concrete and stone” and (40) “Other On the other hand, the activity (53) “Construction and construction chemical products”, are activities with a high own component, so direct works” has different characteristics from the previous ones, since its policies of best practices and technological improvement should be relevance in emissions is produced indirectly. That is why it has the implemented. Finally, the activities (13) “Services related to oil and highest induced component on the demand side and a much lower (but natural gas” and (5) “Oilseeds and industrialised products” require not negligible) own component. It also has a low coefficient of variation. mixed measures, which mitigate their own emissions and those of ac­ This means that, in the face of a variation in its demand, an increase in tivities to which they provide inputs. emissions from the other sectors is generated that exceed 15 times the We relate our findings with the environmental policy guidelines average and are distributed homogeneously. On the supply side, it has a established in the Third National Communication of Ecuador to the much lower magnitude and its effect is concentrated in the own United Nations Framework Convention on Climate Change, which, as component. These characteristics are explained because it has a low indicated above, constitutes the most complete document with official intensity of direct emission, only 69 t CO2 per million dollars of total information for Ecuador in terms of GHG emissions and environmental production, but it is combined with a high participation in total pro­ policies. Table 5 presents the classification of activities based on the duction (11%) and in final demand (16%) (see Table 2). importance of the induced and own components, the mitigation mea­ As for the activity (54) “Trade services”, it has low own component sures established in national policy and the need for adequate envi­ and high induced component, both in its backward and forward link­ ronmental policy depending on the characteristics of the activity. ages; therefore, its importance in terms of emissions is indirect. This is Source: prepared by the authors based on Third National Commu­ observed in Table 2, since activity (54) is not intensive in CO2 emissions nication of Ecuador on Climate Change (Ministry of the Environment of (84.6 t CO2 per million dollars of total production) but it has an Ecuador, 2017). important participation in the economy of 8% of the final demand and 10.8% of the value added. Furthermore, it should be noted that its 5. Conclusions and policy implications induced forward component is greater than its induced backward component (9.07 and 6.05, respectively, see Table 4); and, observing its Ecuador is a country with limited information on environmental 8 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table 5 Weighted linkages decomposition, mitigation actions and necessary policies. Strong effect Cod. Activity Mitigation actions and projects (Third National Necessary policy Communication of Ecuador on Climate Change, 2017) Own backward or 58 Transport and storage services Energy efficiency in transportation: replacement old Direct sectoral measures like technological forward vehicles (public and commercial) improvement and best practices to reduce component Mass use of bicycle resource use or emissions, Quito Metro, Cuenca Tram 51 Electricity Hydroelectric development, renewable energy Energy efficiency: public sector (luminaires); private (cooking, lamps); industry (ISO standards) Optimization of Electricity Generation and Energy Efficiency in the interconnected oil system: efficient use of non-renewable natural resources 38 Refined petroleum oils and other Biofuel products 12 Crude oil and natural gas 57 Restaurant services 65 Administrative services of the Improvement of energy efficiency in public lighting by government and for the community in replacing technology general 60 Telecommunications, transmission and information services 44 Cement, articles of concrete and stone Destruction of Ozone Depleting Substances (ODS) in cement kilns Productive improvement of smelting plants 40 Other chemical products Induced forward 64 Services provided to companies and Inter-sectoral policies component production 12 Crude oil and natural gas 54 Trade services 13 Services related to oil and natural gas 5 Oilseeds and industrialised products 61 Financial intermediation services Induced backward 53 Construction and construction works Technological improvement or better practices component 12 Crude oil and natural gas are effective if they reduce intermediate 54 Trade services demand to directly polluting sectors 65 Administrative services of the government and for the community in general 60 Telecommunications, transmission and information services 70 Association services, leisure, cultural and sports 16 Meat, meat products and by-products 63 Real estate services 17 Elaborated shrimp 57 Restaurant services pollution disaggregated data. This explains why policy guidelines and increasing this percentage becomes difficult since not all the installed previous empirical research only focus on activities traditionally related capacity comes into operation and problems begin to arise in several of to pollution. An exercise of greater sectoral disaggregation (71 sectors) these hydroelectric plants. Given the importance of fossil sources in the makes visible the relevance in CO2 emissions of other economic activ­ generation of electricity, the construction of natural gas combined cycle ities. This disaggregation and input–output analysis allow us to identify plants should be promoted to generate cleaner energy. Furthermore, a set of relevant sectors in CO2 emissions in Ecuador. The contribution of efficiency polices and reducing losses in energy transformation and this paper is to make this relevance evident by linking environmental distribution are also a priority. pollution with the economic structure of the country and achieving an Two other sectors that are important for their high induced effect are integrative vision of the relationship between the environment and the refined petroleum oils and transportation. Direct measures on these economy. In addition, according to the nature of their backward and sectors are also relevant, especially those based on technological im­ forward linkages, sectoral emissions are decomposed into an own provements. In this regard, there have been two failed projects in the component (emitted directly by its productive process) and an induced country: the Pacific refinery, whose construction was halted after cor­ component (induced by its interrelation with other sectors) to help the ruption scandals; and the Esmeraldas refinery, which, despite the in­ correct design of mitigation policies. vestment made, continues to produce fuels with high sulphur content One of the key sectors in CO2 emissions is electricity, which requires and low octane. However, it is striking that, despite the importance of a direct mitigation policies (given its high own component). Policy- measure like this, the multi-criteria analysis for prioritisation of actions makers should focus on measures that promote clean energy produc­ from The Third National Communication of Ecuador to the United Na­ tion and energy efficiency. The change of the energy matrix has certainly tions Framework Convention on Climate Change places improving the been one of the most important policies in Ecuador. Between 2008 and quality of fuels at the Esmeraldas refinery as the last priority. This is so 2018, the construction of several hydroelectric plants allowed the gen­ due to data availability, project development, and project benefits and eration of clean energy to increase by approximately 33%. In 2018, barriers; elements that policy-makers could change if they recognised hydroelectric energy represented 44%, while energy from fossil sources the importance of this measure. As for the transport sector, studies such represented 48% of electricity generation (MERNNR, 2018). However, as those by Castro et al. (2019) and Cevallos (2016) have shown the 9 E. Buenaño et al. Energy Policy 158 (2021) 112551 importance of the sector regarding energy consumption and the rele­ There are various sectors, such as trade services, administrative vance of passenger and cargo transport, as well as the increase in the services of the government and for the community in general, tele­ fleet of light vehicles. Therefore, some measures could be the increase in communications, transmission and information services, which are key both the private and public electric vehicle fleet, which have been raised sectors in CO2 emissions that are not identified by policy-makers and through incentives for the replacement of old cars and the elimination of researchers because they are not direct polluters. However, they have tariffs for the import of electric cars, as well as larger policies such as the strong links with sectors that pollute directly, such as transportation, Quito subway and Cuenca tram projects. electricity, crude oil, or refined petroleum oils, among others. This type The policy of subsidies for energy consumption in transportation, of interrelationship requires the design of multisectoral policies that are fuels and electricity is significant in the Ecuadorian economy, which not analysed by the current actions and projects of public policy, but that makes it difficult to adopt policies focused on increasing prices in these are necessary to have the desired mitigation impact. key sectors. Therefore, policy-makers should look at the characteristics Finally, we find a group of activities that also pollute indirectly and of the sectors for policy design, since it is important to identify whether a are not monitored by academics or policy makers. However, their sector pollutes directly in its production process or indirectly and mitigation potential is more conservative, as they are relevant only from whether it is responsible for the pollution generated by other sectors. a supply or demand perspective. In the case of relevant activities only The extraction of crude oil and natural gas is other key activity. This from a demand perspective (such as real estate services, restaurant sector is characterised by being one of the main sources of government services, meat products or elaborated shrimp), policies should focus on financing and having an important weight in the economy, around 10% the activities that are suppliers of their inputs, as the increase in demand of GDP (BCE, 2018). Moreover, it has losses of around 2300 KBOEs in will pull polluting sectors to increase their emissions. For example, the recent years, which represents 1% of energy production (MERNNR, government can review tariff and tax policies that have privileged 2018). This activity requires direct and indirect policies (given the certain activities, such as agriculture and livestock, which have magnitude of its induced and own components). In the case of direct polluting technologies and provide inputs to activities that are relevant policies, it is important to reduce the burning of gas generated in the from the demand-side. In contrast, direct policies would not have the production process, which involves significant CO2 emissions and eco­ desired effect (unless they had a direct impact on their final demand), nomic waste. The various alternatives to improve this process go since neither their direct emissions nor their own backward linkages are through careful analysis and evaluation of the plants to know the flow significant. The activities that are relevant only from the supply-side and composition of the gases and the possibility of reusing them. Other perspective, such as services provided to companies and production or direct measures are the recovery of energy by means of recovery tur­ services related to oil and natural gas, are highly related to sectors such bines or turbo expanders and efficiency improvements in steam as an as transportation, electricity, and oil extraction. The policies focused on energy source. These policies, together with those indicated above for the sectors where their production is destined would allow a reduction the electricity, transportation and oil refining sectors constitute the mix of the emissions of these activities. In the case of financial intermedia­ of direct and indirect policies for this activity. However, despite their tion services, as pointed out by Piaggio et al. (2014), an efficient policy relevance, no mitigation policies were identified in the Third National would be the creation of credit access incentives, as well as preferential Communication of Ecuador to the United Nations Framework Conven­ interest rates aimed at financing non-polluting activities. tion on Climate Change (see Table 5), nor in the energy efficiency studies reviewed for the preparation of this work. CRediT authorship contribution statement The construction activity is also a key sector, but it is approached tangentially by mitigation plans because the emphasis is placed on the Edwin Buenaño: Conceptualization, Methodology, Formal analysis, cement, articles of concrete and stone sector. This is an error because the Investigation, Writing – original draft, Writing – review & editing. construction sector has stronger impacts than the cement sector due to Emilio Padilla: Conceptualization, Methodology, Formal analysis, its multisectoral linkages. Therefore, variations in the demand of this Investigation, Writing – review & editing. Vicent Alcántara: Concep­ activity generate impacts, in CO2 emissions, which are equivalent to 10 tualization, Methodology, Formal analysis, Investigation, Writing – re­ times the average of the rest of the economy. In this case, mitigation view & editing. policies based on technological improvements would not be useful, because this activity is not directly responsible for emissions. What is Declaration of competing interest required is to intervene to reduce the final demand. However, this is not very viable given its importance in the economy. Then, the emphasis The authors declare that they have no known competing financial should be on the sectors from which construction demands inputs, so interests or personal relationships that could have appeared to influence that they adopt better technologies, as well as on the substitution of the work reported in this paper. materials with low emissions. The activity has important links with sectors that supply raw materials, such as forestry, and with Acknowledgements manufacturing, such as cement or wood, but it also demands a large part of inputs from service activities such as trade, professional services or We acknowledge support from project RTI2018-095484-B-I00 transportation. (Spanish Ministry of Science and Innovation and ERDF). Annexes Table A.1 Direct and total CO2 emissions by economic activity. Cod Activity Direct emission % Total emission % 58 Transport and storage services 9189 21.9% 4986 11.9% 38 Refined petroleum oils and other products 7399 17.6% 3632 8.6% 51 Electricity 6773 16.1% 4257 10.1% 12 Crude oil and natural gas 2200 5.2% 3926 9.3% 44 Cement, articles of concrete and stone 1870 4.5% − 36 − 0.1% (continued on next page) 10 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table A.1 (continued ) Cod Activity Direct emission % Total emission % 64 Services provided to companies and production 1275 3.0% 148 0.4% 54 Trade services 1261 3.0% 2437 5.8% 53 Construction and construction works 1231 2.9% 5916 14.1% 65 Administrative services of the government and for the community in general 856 2.0% 1752 4.2% 60 Telecommunications, transmission and information services 762 1.8% 1280 3.0% 40 Other chemical products 691 1.6% 492 1.2% 10 Fish and other aquatic products (except shrimp) 610 1.5% 244 0.6% 9 Live or fresh shrimp and shrimp larvae 563 1.3% 38 0.1% 57 Restaurant services 517 1.2% 881 2.1% 7 Live animals and animal products 504 1.2% 525 1.2% 36 Products of treated wood, cork and other materials 443 1.1% 26 0.1% 70 Association services, leisure, cultural and sports 416 1.0% 725 1.7% 43 Glass, ceramics and refractories 397 0.9% 81 0.2% 5 Oilseeds and industrialised products 395 0.9% 51 0.1% 16 Meat, meat products and by-products 242 0.6% 861 2.0% 63 Real estate services 232 0.6% 946 2.3% 47 Machinery, equipment and electrical appliances 193 0.5% 290 0.7% 46 Processed metal products 190 0.5% 284 0.7% 56 Accommodation services 188 0.4% 157 0.4% 17 Elaborated shrimp 187 0.4% 1081 2.6% 69 Non-market social and health services 183 0.4% 584 1.4% 1 Banana, coffee and cocoa 175 0.4% 519 1.2% 42 Plastic products 164 0.4% 44 0.1% 37 Pulp, paper and cardboard, editorial products and others 158 0.4% 145 0.3% 6 Services related to agriculture 142 0.3% 0 0.0% 19 Prepared and preserved fish and other aquatic species 142 0.3% 520 1.2% 13 Services related to oil and natural gas 136 0.3% 0 0.0% 50 Other manufactured products 130 0.3% 44 0.1% 33 Threads, spinning, weaving and confection 129 0.3% 136 0.3% 25 Sugar, brown sugar and molasses 129 0.3% 192 0.5% 61 Financial intermediation services 127 0.3% 211 0.5% 4 Tubers, Vegetables, melons and fruits 118 0.3% 211 0.5% 15 No metallic minerals 117 0.3% − 1 0.0% 45 Common metals 117 0.3% 194 0.5% 29 Various food products 107 0.3% 359 0.9% 14 Metallic minerals 106 0.3% 30 0.1% 67 Public education services 101 0.2% 259 0.6% 18 Fish and other processed aquatic products 87 0.2% 280 0.7% 30 Alcoholic beverages 81 0.2% 245 0.6% 48 Transportation equipment 77 0.2% 146 0.3% 3 Flowers and buds 76 0.2% 296 0.7% 21 Processed dairy products 67 0.2% 411 1.0% 2 Cereals 60 0.1% 17 0.0% 52 Water, sanitation and gas services (oil exc) 56 0.1% 41 0.1% 20 Crude and refined oils 55 0.1% 211 0.5% 23 Bakery products 50 0.1% 186 0.4% 49 Furniture 49 0.1% 348 0.8% 8 Forestry products 46 0.1% 47 0.1% 31 Non-alcoholic beverage 45 0.1% 202 0.5% 28 Processed coffee products 42 0.1% 84 0.2% 59 Postal and courier services 39 0.1% 60 0.1% 24 Noodles, macaroni and other similar farinaceous products 39 0.1% 54 0.1% 11 Aquaculture products (except shrimp) 39 0.1% 44 0.1% 34 Clothing 35 0.1% 146 0.3% 26 Elaborated cocoa, chocolate and confectionery products 31 0.1% 100 0.2% 39 Basic chemical products, fertilizers and primary plastics 30 0.1% 6 0.0% 22 Grain mill products 29 0.1% 201 0.5% 35 Leather, leather products and footwear 25 0.1% 86 0.2% 68 Social and health services 22 0.1% 111 0.3% 66 Private education services 20 0.0% 122 0.3% 55 Repair and maintenance services of motor vehicles and motorcycles 19 0.0% 21 0.1% 41 Rubber products 14 0.0% 23 0.1% 27 Animal food 12 0.0% 37 0.1% 62 Insurance services and pension funds 11 0.0% 48 0.1% 32 Elaborate tobacco 0 0.0% 15 0.0% 71 Domestic service 0 0.0% 0 0.0% Total 42,019 100% 42,019 100% Source: prepared by the authors based on the estimation of CO2 emissions (Table A.1) and IO Matrix 2013 (BCE, 2016). 11 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table A.2 Classification of economic activities in key activities, and relevant activities from the supply or demand perspective. Cod. Activity Backward Linkages Forward Linkages Classification / y / μy,j CVj μy,i CVvj μ μ 1 Banana, coffee and cocoa 0.9 2.8 0.3 3.0 rest 2 Cereals 0.0 2.8 0.2 3.6 rest 3 Flowers and buds 0.5 3.0 0.1 3.2 rest 4 Tubers, Vegetables, melons and fruits 0.4 3.6 0.2 4.2 rest 5 Oilseeds and industrialised products 0.1 3.0 1.1 2.9 relevant from supply 6 Services related to agriculture 0.0 - 0.3 3.6 rest 7 Live animals and animal products 0.9 3.0 0.3 4.5 rest 8 Forestry products 0.1 4.2 0.4 7.8 rest 9 Live or fresh shrimp and shrimp larvae 0.1 2.9 0.7 6.9 rest 10 Fish and other aquatic products (except shrimp) 0.4 2.7 0.8 5.3 rest 11 Aquaculture products (except shrimp) 0.1 3.8 0.1 3.7 rest 12 Crude oil and natural gas 6.6 3.4 6.2 3.8 key sector 13 Services related to oil and natural gas 0.0 - 1.1 8.1 relevant from supply 14 Metallic minerals 0.1 2.3 0.2 3.9 rest 15 No metallic minerals 0.0 3.0 0.2 8.0 rest 16 Meat, meat products and by-products 1.5 3.5 0.2 5.4 relevant from demand 17 Elaborated shrimp 1.8 3.9 0.1 5.4 relevant from demand 18 Fish and other processed aquatic products 0.5 3.0 0.1 4.0 rest 19 Prepared and preserved fish and other aquatic species 0.9 3.5 0.2 4.3 rest 20 Crude and refined oils 0.4 3.5 0.1 2.4 rest 21 Processed dairy products 0.7 3.1 0.0 3.8 rest 22 Grain mill products 0.3 4.8 0.0 3.3 rest 23 Bakery products 0.3 2.8 0.1 2.9 rest 24 Noodles, macaroni and other similar farinaceous products 0.1 3.1 0.0 4.0 rest 25 Sugar, brown sugar and molasses 0.3 5.0 0.1 2.3 rest 26 Elaborated cocoa, chocolate and confectionery products 0.2 3.0 0.0 3.8 rest 27 Animal food 0.1 2.9 0.3 3.3 rest 28 Processed coffee products 0.1 4.8 0.0 3.3 rest 29 Various food products 0.6 2.1 0.1 2.1 rest 30 Alcoholic beverages 0.4 4.7 0.1 3.2 rest 31 Non-alcoholic beverage 0.3 2.8 0.1 4.6 rest 32 Elaborate tobacco 0.0 4.1 0.0 - rest 33 Threads, spinning, weaving and confection 0.2 3.0 0.2 2.6 rest 34 Clothing 0.2 4.1 0.1 5.6 rest 35 Leather, leather products and footwear 0.1 2.6 0.0 4.4 rest 36 Products of treated wood, cork and other materials 0.0 3.9 0.4 7.7 rest 37 Pulp, paper and cardboard, editorial products and others 0.2 2.9 0.8 3.1 rest 38 Refined petroleum oils and other products 6.1 4.7 11.0 3.4 key sector 39 Basic chemical products, fertilizers and primary plastics 0.0 2.9 0.2 3.6 rest 40 Other chemical products 0.8 2.9 1.1 3.8 relevant from supply 41 Rubber products 0.0 2.8 0.0 3.6 rest 42 Plastic products 0.1 3.2 0.5 4.0 rest 43 Glass, ceramics and refractories 0.1 2.7 0.4 8.1 rest 44 Cement, articles of concrete and stone − 0.1 2.5 1.9 8.2 relevant from supply 45 Common metals 0.3 3.6 0.3 4.5 rest 46 Processed metal products 0.5 3.3 0.3 6.6 rest 47 Machinery, equipment and electrical appliances 0.5 3.1 0.3 2.9 rest 48 Transportation equipment 0.2 3.0 0.2 4.0 rest 49 Furniture 0.6 3.6 0.0 4.6 rest 50 Other manufactured products 0.1 3.3 0.6 4.0 rest 51 Electricity 7.2 3.0 5.8 3.1 key sector 52 Water, sanitation and gas services (oil exc) 0.1 3.3 0.3 4.1 rest 53 Construction and construction works 10.0 2.5 1.5 5.1 key sector 54 Trade services 4.1 3.4 4.8 5.2 key sector 55 Repair and maintenance services of motor vehicles and motorcycles 0.0 3.5 0.7 3.9 rest 56 Accommodation services 0.3 2.0 0.3 3.9 rest 57 Restaurant services 1.5 1.8 0.7 3.4 relevant from demand 58 Transport and storage services 8.4 3.0 12.6 4.3 key sector 59 Postal and courier services 0.1 5.9 0.1 4.4 rest 60 Telecommunications, transmission and information services 2.2 3.7 1.1 3.6 key sector 61 Financial intermediation services 0.4 3.9 1.0 4.3 relevant from supply 62 Insurance services and pension funds 0.1 4.5 0.3 3.1 rest 63 Real estate services 1.6 3.4 0.8 3.9 relevant from demand 64 Services provided to companies and production 0.2 2.3 6.8 3.3 relevant from supply 65 Administrative services of the government and for the community in general 3.0 2.1 1.1 - key sector 66 Private education services 0.2 2.5 0.0 3.0 rest 67 Public education services 0.4 2.3 0.2 - rest 68 Social and health services 0.2 2.8 0.0 7.6 rest 69 Non-market social and health services 1.0 3.0 0.2 - rest 70 Association services, leisure, cultural and sports 1.2 2.4 0.6 3.6 relevant from demand 71 Domestic service 0.0 - 0.0 - rest Source: prepared by the authors based on the estimate of CO2 emissions (Table A.2) and IO Matrix 2013 (BCE, 2016). 12 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table A.3 Comparison of type of classification of sectors according to the Ghosh and Leontief matrixes to calculate the forward linkage. Forward Linkages Leontief Ghosh / Cod. Activity μj μj >1 μi μ/i >1 Result μ μ 1 Banana, coffee and cocoa 1.6 0.3 0.0 1.4 0.3 0.0 not change 2 Cereals 0.6 0.1 0.0 1.1 0.2 0.0 not change 3 Flowers and buds 0.7 0.1 0.0 0.5 0.1 0.0 not change 4 Tubers, Vegetables, melons and fruits 1.1 0.2 0.0 1.0 0.2 0.0 not change 5 Oilseeds and industrialised products 3.7 0.7 0.0 5.9 1.1 1.0 FL Ghosh 6 Services related to agriculture 1.3 0.2 0.0 1.4 0.3 0.0 not change 7 Live animals and animal products 4.7 0.9 0.0 1.9 0.3 0.0 not change 8 Forestry products 0.4 0.1 0.0 2.1 0.4 0.0 not change 9 Live or fresh shrimp and shrimp larvae 5.3 1.0 0.0 3.9 0.7 0.0 not change 10 Fish and other aquatic products (except shrimp) 5.7 1.0 1.0 4.6 0.8 0.0 FL Leontief 11 Aquaculture products (except shrimp) 0.4 0.1 0.0 0.4 0.1 0.0 not change 12 Crude oil and natural gas 20.7 3.7 1.0 34.3 6.2 1.0 not change 13 Services related to oil and natural gas 1.3 0.2 0.0 6.1 1.1 1.0 FL Ghosh 14 Metallic minerals 1.0 0.2 0.0 1.1 0.2 0.0 not change 15 No metallic minerals 1.1 0.2 0.0 1.3 0.2 0.0 not change 16 Meat, meat products and by-products 2.3 0.4 0.0 0.9 0.2 0.0 not change 17 Elaborated shrimp 1.8 0.3 0.0 0.4 0.1 0.0 not change 18 Fish and other processed aquatic products 0.8 0.1 0.0 0.4 0.1 0.0 not change 19 Prepared and preserved fish and other aquatic species 1.3 0.2 0.0 1.0 0.2 0.0 not change 20 Crude and refined oils 0.5 0.1 0.0 0.4 0.1 0.0 not change 21 Processed dairy products 0.6 0.1 0.0 0.3 0.0 0.0 not change 22 Grain mill products 0.3 0.0 0.0 0.2 0.0 0.0 not change 23 Bakery products 0.5 0.1 0.0 0.3 0.1 0.0 not change 24 Noodles, macaroni and other similar farinaceous products 0.4 0.1 0.0 0.1 0.0 0.0 not change 25 Sugar, brown sugar and molasses 1.2 0.2 0.0 0.4 0.1 0.0 not change 26 Elaborated cocoa, chocolate and confectionery products 0.3 0.1 0.0 0.1 0.0 0.0 not change 27 Animal food 0.1 0.0 0.0 1.5 0.3 0.0 not change 28 Processed coffee products 0.4 0.1 0.0 0.2 0.0 0.0 not change 29 Various food products 1.0 0.2 0.0 0.3 0.1 0.0 not change 30 Alcoholic beverages 0.8 0.1 0.0 0.5 0.1 0.0 not change 31 Non-alcoholic beverage 0.4 0.1 0.0 0.3 0.1 0.0 not change 32 Elaborate tobacco 0.0 0.0 0.0 0.0 0.0 0.0 not change 33 Threads, spinning, weaving and confection 1.2 0.2 0.0 1.2 0.2 0.0 not change 34 Clothing 0.3 0.1 0.0 0.3 0.1 0.0 not change 35 Leather, leather products and footwear 0.2 0.0 0.0 0.2 0.0 0.0 not change 36 Products of treated wood, cork and other materials 4.2 0.7 0.0 2.3 0.4 0.0 not change 37 Pulp, paper and cardboard, editorial products and others 1.5 0.3 0.0 4.6 0.8 0.0 not change 38 Refined petroleum oils and other products 69.5 12.5 1.0 61.4 11.0 1.0 not change 39 Basic chemical products, fertilizers and primary plastics 0.3 0.1 0.0 1.0 0.2 0.0 not change 40 Other chemical products 6.5 1.2 1.0 6.3 1.1 1.0 not change 41 Rubber products 0.1 0.0 0.0 0.3 0.0 0.0 not change 42 Plastic products 1.5 0.3 0.0 2.8 0.5 0.0 not change 43 Glass, ceramics and refractories 3.7 0.7 0.0 2.5 0.4 0.0 not change 44 Cement, articles of concrete and stone 17.6 3.2 1.0 10.4 1.9 1.0 not change 45 Common metals 1.1 0.2 0.0 1.8 0.3 0.0 not change 46 Processed metal products 1.8 0.3 0.0 1.6 0.3 0.0 not change 47 Machinery, equipment and electrical appliances 1.8 0.3 0.0 1.8 0.3 0.0 not change 48 Transportation equipment 0.7 0.1 0.0 1.0 0.2 0.0 not change 49 Furniture 0.5 0.1 0.0 0.2 0.0 0.0 not change 50 Other manufactured products 1.2 0.2 0.0 3.5 0.6 0.0 not change 51 Electricity 63.6 11.4 1.0 32.0 5.8 1.0 not change 52 Water, sanitation and gas services (oil exc) 0.5 0.1 0.0 1.9 0.3 0.0 not change 53 Construction and construction works 11.6 2.1 1.0 8.6 1.5 1.0 not change 54 Trade services 11.8 2.1 1.0 26.9 4.8 1.0 not change 55 Repair and maintenance services of motor vehicles and motorcycles 0.2 0.0 0.0 3.9 0.7 0.0 not change 56 Accommodation services 1.8 0.3 0.0 1.6 0.3 0.0 not change 57 Restaurant services 4.9 0.9 0.0 3.8 0.7 0.0 not change 58 Transport and storage services 86.3 15.5 1.0 70.3 12.6 1.0 not change 59 Postal and courier services 0.4 0.1 0.0 0.4 0.1 0.0 not change 60 Telecommunications, transmission and information services 7.2 1.3 1.0 6.1 1.1 1.0 not change 61 Financial intermediation services 1.2 0.2 0.0 5.6 1.0 1.0 FL Ghosh 62 Insurance services and pension funds 0.1 0.0 0.0 1.5 0.3 0.0 not change 63 Real estate services 2.2 0.4 0.0 4.6 0.8 0.0 not change 64 Services provided to companies and production 12.0 2.2 1.0 37.6 6.8 1.0 not change 65 Administrative services of the government and for the community in general 8.0 1.4 1.0 6.4 1.1 1.0 not change 66 Private education services 0.2 0.0 0.0 0.2 0.0 0.0 not change 67 Public education services 0.9 0.2 0.0 0.9 0.2 0.0 not change 68 Social and health services 0.2 0.0 0.0 0.2 0.0 0.0 not change 69 Non-market social and health services 1.7 0.3 0.0 1.3 0.2 0.0 not change 70 Association services, leisure, cultural and sports 3.9 0.7 0.0 3.6 0.6 0.0 not change (continued on next page) 13 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table A.3 (continued ) Forward Linkages Leontief Ghosh / Cod. Activity μj μj >1 μi μ/i >1 Result μ μ 71 Domestic service 0.0 0.0 0.0 0.0 0.0 0.0 not change Average 5.6 5.6 Source: computed by the authors with the CO2 emission estimation (Table A.2) and the IO Matrix 2013 (BCE, 2016). Table A.4 Pure and own components of backward and forward linkages. Backward Linkage Forward Linkages Cod. Activity μj Own induced μi own induced 1 Banana, coffee and cocoa 4.9 1.5 3.4 1.4 0.5 1.2 2 Cereals 0.2 0.0 0.1 1.1 0.0 0.0 3 Flowers and buds 2.8 0.7 2.1 0.5 0.2 0.8 4 Tubers, Vegetables, melons and fruits 2.0 1.0 1.0 1.0 0.3 0.4 5 Oilseeds and industrialised products 0.5 0.3 0.2 5.9 0.1 0.1 6 Services related to agriculture 0.0 0.0 0.0 1.4 0.0 0.0 7 Live animals and animal products 4.9 1.9 3.0 1.9 0.7 1.1 8 Forestry products 0.4 0.1 0.3 2.1 0.0 0.1 9 Live or fresh shrimp and shrimp larvae 0.4 0.3 0.1 3.9 0.1 0.0 10 Fish and other aquatic products (except shrimp) 2.3 2.0 0.3 4.6 0.7 0.1 11 Aquaculture products (except shrimp) 0.4 0.2 0.2 0.4 0.1 0.1 12 Crude oil and natural gas 36.9 19.2 17.7 34.3 6.7 6.5 13 Services related to oil and natural gas 0.0 0.0 0.0 6.1 0.0 0.0 14 Metallic minerals 0.3 0.2 0.1 1.1 0.1 0.0 15 No metallic minerals 0.0 0.0 0.0 1.3 0.0 0.0 16 Meat, meat products and by-products 8.1 1.9 6.2 0.9 0.7 2.3 17 Elaborated shrimp 10.2 1.7 8.4 0.4 0.6 3.1 18 Fish and other processed aquatic products 2.6 0.6 2.0 0.4 0.2 0.8 19 Prepared and preserved fish and other aquatic species 4.9 1.3 3.6 1.0 0.4 1.3 20 Crude and refined oils 2.0 0.4 1.6 0.4 0.2 0.6 21 Processed dairy products 3.9 0.6 3.3 0.3 0.2 1.2 22 Grain mill products 1.9 0.2 1.7 0.2 0.1 0.6 23 Bakery products 1.8 0.4 1.3 0.3 0.2 0.5 24 Noodles, macaroni and other similar farinaceous products 0.5 0.3 0.2 0.1 0.1 0.1 25 Sugar, brown sugar and molasses 1.8 0.8 1.0 0.4 0.3 0.4 26 Elaborated cocoa, chocolate and confectionery products 0.9 0.3 0.7 0.1 0.1 0.2 27 Animal food 0.3 0.0 0.3 1.5 0.0 0.1 28 Processed coffee products 0.8 0.4 0.4 0.2 0.1 0.1 29 Various food products 3.4 1.0 2.4 0.3 0.3 0.9 30 Alcoholic beverages 2.3 0.7 1.6 0.5 0.3 0.6 31 Non-alcoholic beverage 1.9 0.4 1.5 0.3 0.1 0.6 32 Elaborate tobacco 0.1 0.0 0.1 0.0 0.0 0.0 33 Threads, spinning, weaving and confection 1.3 0.6 0.6 1.2 0.2 0.2 34 Clothing 1.4 0.3 1.1 0.3 0.1 0.4 35 Leather, leather products and footwear 0.8 0.2 0.6 0.2 0.1 0.2 36 Products of treated wood, cork and other materials 0.2 0.2 0.1 2.3 0.1 0.0 37 Pulp, paper and cardboard, editorial products and others 1.4 0.6 0.8 4.6 0.2 0.3 38 Refined petroleum oils and other products 34.1 32.3 1.8 61.4 11.3 0.7 39 Basic chemical products, fertilizers and primary plastics 0.1 0.0 0.0 1.0 0.0 0.0 40 Other chemical products 4.6 3.6 1.0 6.3 1.3 0.4 41 Rubber products 0.2 0.1 0.1 0.3 0.0 0.0 42 Plastic products 0.4 0.3 0.1 2.8 0.1 0.0 43 Glass, ceramics and refractories 0.8 0.5 0.2 2.5 0.2 0.1 44 Cement, articles of concrete and stone − 0.3 − 0.3 − 0.1 10.4 − 0.1 0.0 45 Common metals 1.8 0.6 1.2 1.8 0.2 0.5 46 Processed metal products 2.7 1.3 1.3 1.6 0.5 0.5 47 Machinery, equipment and electrical appliances 2.7 1.4 1.4 1.8 0.5 0.5 48 Transportation equipment 1.4 0.7 0.7 1.0 0.2 0.3 49 Furniture 3.3 0.4 2.8 0.2 0.2 1.0 50 Other manufactured products 0.4 0.2 0.2 3.5 0.1 0.1 51 Electricity 40.0 35.2 4.8 32.0 12.4 1.8 52 Water, sanitation and gas services (oil exc) 0.4 0.2 0.2 1.9 0.1 0.1 53 Construction and construction works 55.6 10.9 44.6 8.6 3.8 16.5 54 Trade services 22.9 6.5 16.4 26.9 2.3 6.1 55 Repair and maintenance services of motor vehicles and motorcycles 0.2 0.1 0.1 3.9 0.0 0.0 56 Accommodation services 1.5 0.9 0.6 1.6 0.3 0.2 57 Restaurant services 8.3 4.3 4.0 3.8 1.5 1.5 58 Transport and storage services 46.8 38.8 8.1 70.3 13.6 3.0 59 Postal and courier services 0.6 0.2 0.3 0.4 0.1 0.1 60 Telecommunications, transmission and information services 12.0 5.9 6.1 6.1 2.1 2.2 (continued on next page) 14 E. Buenaño et al. Energy Policy 158 (2021) 112551 Table A.4 (continued ) Backward Linkage Forward Linkages Cod. Activity μj Own induced μi own induced 61 Financial intermediation services 2.0 0.5 1.5 5.6 0.2 0.6 62 Insurance services and pension funds 0.5 0.1 0.4 1.5 0.0 0.1 63 Real estate services 8.9 1.8 7.1 4.6 0.6 2.6 64 Services provided to companies and production 1.4 0.9 0.5 37.6 0.3 0.2 65 Administrative services of the government and for the community in general 16.5 8.0 8.4 6.4 2.8 3.1 66 Private education services 1.1 0.2 1.0 0.2 0.1 0.4 67 Public education services 2.4 0.9 1.5 0.9 0.3 0.5 68 Social and health services 1.0 0.2 0.8 0.2 0.1 0.3 69 Non-market social and health services 5.5 1.7 3.8 1.3 0.6 1.4 70 Association services, leisure, cultural and sports 6.8 3.4 3.5 3.6 1.2 1.3 71 Domestic service 0.0 0.0 0.0 0.0 0.0 0.0 Source: prepared by the authors based on the CO2 emissions estimation (Table 2) and the IO Matrix 2013 (BCE, 2016). References Imori, D., Guilhoto, J.J.M., 2010. Estrutura produtiva brasileira e emissão de CO2. XLV Congresso da Sociedade Brasileira de Economía, Administração e Sociología Rural. Editora Senac, São Paulo, pp. 205–233. Alcántara, V., Roca, J., 1995. Energy and CO2 emissions in Spain. Methodology of International Energy Agency (IEA), 2016. Energy Balances of OECD-Countries, analysis and some results for 1980–90. Energy Econ. 17, 221–230. Documentation for beyond 2020 Files. International Energy Agency, Paris. Alcántara, V., 2007. Análisis input-output y emisiones de CO2 en España: un primer International Energy Agency (IEA), 2019. CO2 Emissions from Fuel Combustion. OCDE/ análisis para la determinación de sectores clave en la emisión. Working Paper 07.08, IEA, Paris, France. Department of Applied Economics, Univ. Autónoma de Barcelona. IPCC, 2008. In: Eggleston, H.S., Miwa, K., Srivastava, N., Tanabe, K. (Eds.), 2006 IPCC Alcántara, V., Del Río, P., Hernández, F., 2010. Structural analysis of electricity Guidelines for National Greenhouse Gas Inventories – A Primer, Prepared by the consumption by productive sectors. The Spanish case. Energy 35, 2088–2098. National Greenhouse Gas Inventories Programme. Published: IGES, Japan. Alcántara, V., Padilla, E., 2006. An input–output analysis for the “key” sectors in CO2 Jodar, E., 2011. Key sectors in greenhouse gases emissions in Switzerland: an emissions from a production perspective: an application to the Spanish economy. In: input–output approach. In: Master Dissertation, Master in Applied Economics. Working Paper 06.01. Department of Applied Economics at Univ. Autónoma de Universidad Autónoma de Barcelona. Barcelona. Jones, L., 1976. The measurement of Hirschman linkages. Q. J. Econ. 90, 323–333. Alcántara, V., Padilla, E., 2009. Input–output subsystems and pollution: an application to Lenzen, M., 1998. Primary Energy and greenhouse gases embodied in Australian final the service sector and CO2 emissions in Spain. Ecol. Econ. 68, 905–914. consumption: an input-output analysis. Energy Pol. 26, 495–506. Alcántara, V., Padilla, E., Roca, J., 2008. De los consumos finales de energía a los Lenzen, M., 2003. Environmentally important paths, linkages and key sectors in the requerimientos de energía primaria y las emisiones de CO2. In: Aproximación a Australian economy. Struct. Change Econ. Dynam. 14, 1–34. partir de los balances de energía. Aplicación a Cataluña, 1990–2005, vol. 67. Leontief, W., 1936. Quantitative input-output relations in the economic system of the Ekonomiaz, pp. 302–337. United States. Rev. Econ. Stat. 18, 105–125. Aguilar, L., 2014. Interdependencias sectoriales, año 2007: Un análisis insumo producto Martínez, J., Martí-Herrero, J., Villacís, S., Riofrio, A., Vaca, D., 2017. Analysis of energy, para el Ecuador, Proyecto de titulación previo a la obtención del título de Ingeniera CO2 emissions and economy of the technological migration for clean cooking in en Ciencias Económicas y Financieras, Facultad de Ciencias. Escuela Politécnica Ecuador. Energy Pol. 107, 182–187. Nacional. Miller, R., Blair, P., 2009. Input–Output Analysis: Foundations and Extensions, second Arroyo, F., Miguel, L., 2019. The trends of the energy intensity and CO2 emissions related ed. Cambridge University Press. to final energy consumption in Ecuador: scenarios of national and worldwide Mernnr, 2018. Balance Energético Nacional 2018. Quito, Ecuador. strategies. Sustainability 12 (1), 20. Ministerio del Ambiente del Ecuador, 2017. Tercera Comunicación Nacional del Ecuador Artola, V., 2009. Evolución del consumo de energía primaria en el Ecuador entre los años sobre Cambio Climático. Quito, Ecuador. 2000 y 2006: un análisis de productos clave y descomposición estructural a través Oosterhaven, J., 1988. On the plausibility of the supply-driven input–output model. del modelo insumo–producto. Department of Applied Economics, Univ. Autónoma J. Reg. Sci. 28, 203–217. de Barcelona. Master dissertation. Piaggio, M., 2013. From Global to Local behind the Relationship between the Economy Banco Central de Ecuador (Bce), 2016. Matrices Insumo Producto. Dirección Nacional de and the Environment, PhD Dissertation. Department of Applied Economics, Univ. Síntesis Macroeconómica - Cuentas Nacionales. Autónoma de Barcelona. Castro, P., Vidoza, J., Gallo, W., 2019. Analysis and projection of energy consumption in Piaggio, M., Alcántara, V., Padilla, E., 2014. Greenhouse gas emissions and economic Ecuador: energy efficiency policies in the transportation sector. Energy Pol. 134, structure in Uruguay. Econ. Syst. Res. 26, 155–176. 110948. Ramírez-Cendrero, J., García, S., Santillán, A., 2017. Sumak kawsay in Ecuador: the role Cevallos, J., 2016. Estimating road transport fuel consumption in Ecuador. Energy Pol. of communitarian economy and the experience of the rural communities in Sarayaku 92, 359–368. (Ecuadorian Amazonia). J. Rural Stud. 53, 111–121. Dietzenbacher, E., Linden, J., 1997. Sectoral and spatial linkages in the EC production Rasmussen, N.P., 1956. Studies in Inter-sectorial Relation. North-Holland Publishing structure. J. Reg. Sci. 37, 235–257. Company. Duarte, R., Sánchez-Chóliz, J., Bielsa, J., 2002. Water Use in the Spanish Economy: an Robalino-López, A., Mena-Nieto, A., García-Ramos, J.E., 2014. System dynamics input–output approach. Ecol. Econ. 43, 71–85. modeling for renewable energy and CO2 emissions: a case study of Ecuador. Energy Fernández, N., 2007. Análisis input-output: identificación de los encadenamientos for Sustainable Development 20, 11–20. productivos y los sectores claves de la economía ecuatoriana para el año 2007. Tesis SENPLADES, 2013. Plan Nacional del Buen Vivir 2013–2017. Secretaría Nacional de para obtener el título de maestría en economía con mención en gestión de pymes. Planificación y Desarrollo. FLACSO Ecuador. World Resources Institute, 2019. “Country Greenhouse Gas Emissions”, CAIT Climate Gachet, I., 2005. Efectos multiplicadores y encadenamientos productivos: análisis input- Data Explorer. World Resources Institute, Washington, DC. Available online at. output de la Economía ecuatoriana. Cuestiones Económicas 21, 97–133. http://cait.wri.org. July 15th, 2020. Ghosh, A., 1958. Input–Output approach in an allocation system. Economica 25, 58–64. Hazari, B., 1970. Empirical identification of key sectors in the Indian economy. Rev. Econ. Stat. 52, 301–305. 15