Structural Changes in Chile’s Industries to Reduce Carbon Dioxide (CO₂) Emissions: An Emissions Multiplier Product Matrix Analysis (eMPM)

Abstract

Most countries in the world have agreed to reduce their emissions following the COP21 agreement in Paris, and as a result, each nation has presented suitable plans to do so. Chile is not an exception in this regard. This article examines the emissions of Chilean industries using the emission multiplier product matrix (eMPM), a cutting-edge method that estimates the pollution caused by inter-industrial activity in the country’s regions by integrating CO₂ emissions with multi-region input–output table (MRIO) databases and elasticities. This approach connects the major emissions-producing sectors to the regions where these emissions come from, thereby accounting for existing interregional linkages. The application of technology, along with adequate state regulation in compliance with Chile’s pledges, acquired following the COP25 call, will decide the level of improvement in emissions reduction.

1. Introduction

Economic growth and greenhouse gas emissions are closely related [1,2]. For many years, the predominant topic of conversation has been climate change [3]. It is common knowledge that each nation’s production matrix, and consequently, its growth strategies and ability to recover from crises like COVID-19, determine how much greenhouse gas (GHG) emissions are produced [4,5]. These strategies can be combined with ethical sustainable growth options like the bioeconomy [6] or green finance [7,8].

Leontief’s 1963 notion of efficiency states that their productive processes are comparable to those of industrialized economies [9]. The other economies are going to carry out this process in a similar way if the productive process of a given economic activity or industrial process is the most efficient. This will have an impact on input costs and usage as well as output emissions. Soza-Amigo and Aroca [10] argue that there are economies that, despite not being classified as developed, present the structures of developed economies, including Brazil, Chile, Mexico, and Turkey, using the economic structures of developed and underdeveloped countries registered in the OECD (a total of 116 input–output matrices), during three instances of time (from the mid-1990s to the mid-2000s (Mid90; Early00 and Mid00), and 37 economic activities.

Moreover, Soza-Amigo and Aroca [10] conclude that the developed economies demonstrated a maximum global similarity of 85.00%; 85.7% and 84.6%, respectively, over the three time periods, and the average similarity was 75.57% ((75.35 + 75.83 + 75.55)/3 = 75.57%). Furthermore, eighty percent of the activities in Chile exhibit a similarity of more than eighty percent when sector by sector comparisons are made with other industrialized economies. The economy of Austria was judged to be the most like Chile’s, with an average global similarity of 75.09 percent; Portugal came in second, with an average global similarity of 73.28%.

Assuming one understands the GDP of Chile, the CO₂ emissions for every nation (ideally Portugal and/or Austria), and their proportional GDP by activity, one can determine their CO₂ contribution [11,12]. We believe that this study makes an innovative contribution based on the EUROSTAT suggestion, using the Full International and Global Accounts for Research in IO (input-output) analysis (FIGARO).

FIGARO uses the contribution of emissions from final demand. This makes use of a reference economy very similar to that of Chile, instead of calculating fuel use. The nomenclature used in the article is detailed in Nomenclature.

Notably, in their ability to serve as the foundation for a comparative examination of the economic makeup of every region within a nation, the multiregional input–output tables (MRIO) provide significant insights on the composition of economies and interregional disaggregation within nations. In Chile, MRIO has been utilized to conduct a comparative study of biomass pellet social hotspots within the framework of a circular bioeconomy [13].

There are three objectives for this paper. The starting point is using the emission multiplier product matrix analysis (eMPM) to examine how changes in Chile’s productive sectors have affected the generation of emissions (as a by-product). This can be studied from a hierarchical point of view, as the MPM prioritizes the change in activities or economic sectors through a pyramid [14]. Secondly, the study of structural change, proposed by Soza-Amigo and Ramos [15], is reflected in the differences in elasticities between economic sectors. Thirdly, it emphasizes that many of these countries make significant contributions to climate change and aims to serve as a starting point for the implementation of the eMPM in developing nations.

The main contribution of the paper is to utilize the eMPM to analyze how the structural changes occurring in a developing country like Chile have affected the generation of emissions. Through analytical effort, the technique enables us to concentrate on the differences, similarities, and relationships between industries and regions. In order to analyze the influences [16], examine the interrelationships and intersectoral feedback [17], or uncover the economic repercussions of a sectoral change in the economy [10], the multiplier product matrix analysis (MPM) can also be combined with other approaches. A collection of input and output data has been used to demonstrate the concept; social and demographic accounting matrices, like Pyatt and Round [18], can be used with this technique. To assess industrial-level productivity in the United States in the presence of intersectoral links, Han and Sickles [19] used the MPM in a regional study.

The innovations that are presented in the paper are listed below. Firstly, keep in mind that it is built upon an expansion of Soza-Amigo and Ramos ‘s proposal [15], in which the notion of elasticity is first reconstructed within the context of an input–output system. Schintke and Stäglin first suggested this in 1988. The principle of adapting productive phases of the production function is reformulated in a way that maintains the desired percentage increase in each sector’s output. This can be seen as the first difference, since it includes the weighting that is dependent on the system’s intensity from the outset (typically, 1%, in line with the conventional definition of the term elasticity).

According to Soza-Amigo and Ramos, to adjust the contribution in terms of the weight of the activity (concentration of an activity), the scope of the effects (especially the indirect ones), or the adequate representation of each element in the response of the original formulation based on the inverse Leontief matrix, one no longer needs to resort to variables like production or final demand. This is made possible by the authors’ suggestion of using the MPM matrix—which Sonis and Hewings published in 2009—instead of the conventional Leontief matrix.

A second innovation in our paper is the integration of emissions per unit of production, in recognition of the fact that Ali et al. [20] propose to premultiply a pollution vector (e) by the MPM matrix (what they call, eMPM), thereby achieving, ordering, and quantifying the real impact of the emissions of an economic system that is based on a weighting according to its own intensity.

A third innovation combines the Soza-Amigo and Ramos formulation with the eMPM matrix of Ali et al. This serves two purposes. In this case, the formulation now determines how much each component of the production functions must change for a sector to increase its production by a specific percentage determined by the researcher. This includes the emissions that such a change would bring, allowing for an adequate assessment of the actual effects and changes in emissions in the system given a specific increase in production. First, there are elasticities that incorporate emissions. Second, if this is relevant and the full formulation provided by Soza-Amigo and Ramos is applied, the effects of the modifications can be determined in terms of the direct and indirect effects.

Chile was chosen for this research because it is a large, almost-linear country—nearly 4000 km long—whose regions can be geographically very distant from each other, which makes trade between regions more costly and more dependent on both air and land transportation. Based on the productive specialization of each region, Chile can be subdivided into four macroeconomic zones. Huge copper mines are in Norte Grande (Far North) and agriculture, fishing, and commerce in Norte Chico (Near North), while Santiago, the nation’s capital, is near to significant commerce, industrial, and agricultural activities in the Zona Central (Central Zone).

In addition to the previously stated characteristics, Chile’s economy is distinguished by its diversity as well as the stark contrasts between its regions (forests and deserts). It also exhibits significant regional variations in GDP per capita. For instance, the region known for extensive mining (Norte Grande) had a GDP per capita in 2014 of roughly 37,000 dollars, which is comparable to that of England; however, it is up to eight times higher than that of the nation’s poorest regions, which had a GDP per capita of roughly 4500 dollars, which would be equivalent to very underdeveloped economies. When presenting heterogeneous regions, there is an obligation in relation to public policies to be rigorous in compliance with environmental commitments. We believe that what is proposed will help to review how the less developed regions will be indirectly affected because of the activity and emission of the most productive and pollution-emitting regions.

The cities and regions that make up Chile’s southern region are more diverse. Valdivia is dedicated mainly to tourism (a city located in the northern part of Patagonia in R12), the industrial production of milk and beef is concentrated in Osorno (a city located in the northern part of Patagonia, south of Valdivia and corresponding to R13) and seafood, particularly salmon, in the Chiloe Archipelago (a set of islands located in R13). Lamb is produced in the city of Punta Arenas (a city located in the southern part of Patagonia and which corresponds to R15), which is in the southernmost region of Chile.

This study would fill the gap in the scholarly world in the identification of key sectors for the development of the territories analyzing CO₂ emissions, utilizing eMPM and the concept of interrelation between economic activities (elasticity). In addition to obtaining these findings, the study hopes to influence the national policy agenda aimed at fulfilling the SDG goals and the UN climate change plan.

Indeed, this article uses the eMPM to link the major industries that produce emissions with the geographic areas where those emissions originate. The use of technology in conjunction with appropriate governmental regulation via initiatives will enable Chile to fulfill its emissions reduction pledges prior to the road map for the energy policy named Energia2050 (Energia2050 available at https://www.energia2050.cl/ accessed on 13 July 2024) and the Chilean Climate Change Plan of Action 2017–2022 (Chilean Climate Change Plan of Action available at http://portal.mma.gob.cl/wp-content/uploads/2016/04/Anteproyecto-PANCC-2017-2022-FINAL-2016-04-18.pdf accessed on 12 May 2024) [21].

The layout of the article is as follows: Using MRIO in the context of Chile, Section 2 describes the connection between emissions and economic growth. The study’s data and methodology are presented in Section 3. The primary findings including the main structural changes, and the relationships between economic sectors and regions are listed in Section 4. Finally, Section 5 suggests future research directions and details constraints.

2. Background

2.1. Relationship between Emissions and Economic Growth Using MRIO

A number of contributions have been made to the study of the relationship between emissions and economic growth, including (1) the potential relationship between GDP and the Employment Multiplier [22]; (2) the study of the concentration or dispersion of the polluting effects of CO₂ [23]; (3) the determination of whether changes in CO₂ emissions are the result of technological advancements or changes in final demand [24]; (4) the proposal of development based on significant activities while also being environmentally aware, especially in developing economies [25]; (5) the establishment of CO₂ pollution from a new perspective, namely, through a structural similarity based on developed economies using their production functions and EUROSTAT criteria [26,27]; (6) the study of the relationship between emissions and economic growth using eMPM [20] and (7) the fact that to the best of our knowledge, the sensitivity of productive activities and the study of eMPM have not yet been used to link the placement of economic activities to pollution. By examining these subjects, we seek to contribute to the body of knowledge and public policies concerning CO₂ emissions and climate change.

San Miguel [28] evaluated the carbon footprint and employment creation linked to Internet access in six Peruvian geodemographic scenarios using MRIO. The findings demonstrated that most of the employment and carbon emissions are linked to more extensive information technology networks. Song [29] emphasized the significance of China’s energy use in the global production chain using the MRIO model.

There are various examples in the scholarly world regarding CO₂ emissions and structural change in Austria and Portugal. There is a link to research on Austria’s CO₂ emissions, where emissions are assessed at both the source and destination using data from the project database multi-region input–output table (MRIO) and Global Trade Analysis Project (GTAP) (Global Trade, Assistance, and Production: The GTAP 7 Data Base, Center for Global Trade Analysis, Purdue University: Available at: https://www.gtap.agecon.purdue.edu/databases/v7/v7_doco.asp accessed on 13 July 2024) [12]. The spatial pattern and underlying drivers of energy change as they relate to international trade between the Global North and South were examined by Yang utilizing GTAP [30].

In the case of Portugal, the economic ramifications of burning fossil fuels are considered [11]. It is evident that the suggested use of information for economies lacking it and the use of information itself are connected to each other, and encompass the boundaries of what the literature has acknowledged.

Miller and Blair [31] conducted a thorough investigation on the correlation between emissions and economic growth through the application of the MRIO model.

Multi-region input–output modeling is an economic technique that contains all the financial flows between pairs of regions and sectors. Its main power resides in accounting for all the purchases and sales of intermediate inputs. Some examples of MRIOT extensions are available in the literature as the National Footprint for Sweden [32] or for Europe [33] and Biocapacity Accounts data [34] that can be used to extend MRIO techniques from financial flows to resource flow estimation. In Chile, the usage of MRIO has been linked to research on the environmental responsibilities of individual regions, as well as the link between interregional trade and CO₂ emissions [35]. Ramos [36] examined the variations in CO₂ emissions throughout Chile’s regions.

It is possible to propose further subcategories for dividing the national Footprint data into more precise consumption and industry related components by following the primary resource flows between a country’s major economic sectors using the MRIO-based data. From this approach, studies have focused on countries that emit high GHG pollution, such as China [37], and in others such as Japan and Korea, where the phenomenon has been studied save for the domestic final demand (DFD) [38], an approach that is also repeated in several OECD countries [39,40], or in others such as Spain [41], and the same approach has been applied for several countries in Latin America [42].

A significant example is Brazil, which reveals the presence of a relationship between economic growth and GHG emissions [37]. However, in terms of public policy, some studies using the input–output matrix have examined how taxes affect Chile’s greenhouse gas emissions [43]. They have also determined how some important sectors, like agriculture, contribute to Chile’s greenhouse gas emissions [44].

However, considering any potential changes that the various productive sensitivities may bring about, it is necessary to consider the potential economic implications. Think about how much sector “i” and sector “j” should relate to each other in order to allow sector “i” to increase production by a certain percentage while maintaining environmental sustainability. If such an idea is assimilated to the concept of “structural characteristics of a developed economy” presented by Leontief in 1963, the latter would indicate greater development. Given that the productive framework becomes more profuse, interactive, and would contribute less CO₂ emissions, this condition and characteristic would aid the future development of other friendly activities.

There are not many critical economic activities at first, but there are a lot of structural adjustments and few links between the various industries. The collection of activities grows increasingly sensitive and related as economies age; however, this will change as structural changes become smaller and sensitive interrelationships rise [45,46]. Furthermore, the extractive industry’s sensitivities are moving toward the service sector as development advances, which are increasing the sensitivity and interrelationships in the activities [47]. In that regard, we would like to investigate whether this has anything to do with initiatives that reduce CO₂ emissions.

It should be noted that Chile does not have statistics on CO₂ emissions broken down by type of activity. The most recent data used, as reported by Ramos [36], is the division of energy consumption into seven groups (see National Energy Balance, Ministry of Energy; https://energia.gob.cl/pelp/balance-nacional-de-energia accessed on 12 May 2024)). Information from the Ministry of Environment regarding sectors like agriculture and its greenhouse gasses is added to these data (GHG; https://snichile.mma.gob.cl/documentos/ accessed on 12 May 2024); using these data, the estimation and distribution of CO₂ emissions for Chile, as well as its regions and sectors, has been made.

While there is a shortage of data concerning CO₂ contamination, the European Commission through EUROSTAT suggests to its members that as a mitigating methodology for its estimation, the use of formulas is decided upon, based on the combined contamination of Europe, which has been expressed in an MRIO matrix, plus the participation of the sector in total and that of the sector of the economy without related information. The sector’s contribution is then estimated in terms of both the economies and the total contribution; this is based on the distribution of final demand.

2.2. MPM and eMPM

Multiplier Product Matrix Analysis (MPM Analysis) is a tool that may be used to illustrate the changes in economy structures over time. This method can also be applied to interregional structural analysis, which compares two economies at the same time.

Several Hewings and Sonis studies for the US utilizing the MPM have shown that: 1. There are fewer interdependencies throughout national industries because of a general decline in total requirements multipliers, among other reasons, due to the outsourcing of activities and the dynamics of structural changes themselves [48,49]. 2. The decline in local pricing has been influenced by the increase in imports by industry interdependencies, since a greater portion of US production is dependent on foreign inputs rather than domestic output. 3. The US economy’s non-manufacturing sectors are growing at a faster rate and with more interconnected industries [50,51,52]. In more contemporary research such as Han and Sickles [19], they combine spatial analysis to estimate productivity at the industrial level in the presence of intersectoral links utilizing the MPM in the US.

The study that Haddad [53] have written about Latin America is worth mentioning since it analyzes how the economies of the region have changed structurally in response to developments in industrialized nations. The authors do this by creating two blocks: one containing the countries of Latin America and the other containing the rest of the world. Some other examples are presented in

Table 1. MPM and country examples.

Country	Method	Reference
Brazil	Interregional input–output tables for the year of 1992 for two Brazilian regions (Northeast and rest of the economy)	[54]
US	Sectoral dependencies are the consequences of indirect effects via the supply chain network of industries.	[19]
UK	IO tables and emission data for the period 1995–2009	[20]
Japan	The evaluation of economic structure represented in IO matrix	[55]
India	The MPM analysis is carried out for the years 1989–1990, 1993–1994 and 1998–1999.	[56]
Indonesia	The degree of structural change in the Indonesian economy from 1971 to 2008	[57]
Morocco	Social accounting matrix assessment in Morocco	[58]
Poland	Polish input–output tables from the period 1990–2000	[59]
South Africa	An analysis of economic growth and poverty using a social accounting matrix	[60]

.

The form of reduction in the extended IO pollution model that can be articulated must be taken into consideration to construct the eMPM; the methodology section will clarify this.

2.3. The Chilean Case

Chile is divided into different subregions and 12 economic sectors, including agriculture, forestry, and fishing (s01); mining (s02); manufacturing industry (s03); electricity, gas, water, and waste management (s04); construction (s05); trade, lodging, and dining (s06); transportation, communications, and information services (s07); financial intermediation (s08); real estate and housing services (s09); business services (s10); personal services (s11) and public administration (s12) (see in Figure 1).

Figure 1 illustrates the main sectors/regions that contribute to CO₂ emissions. The trend line is an average of efficiency by sector and region in Chile. Above the line it pollutes more than the average, and below the line it pollutes less than the average. Certain activities contribute significantly to the GDP of the territory, such as mining (s02) in Norte Grande, which has high CO₂ emissions in this territory. Sturla-Zerene [61] has conducted research in Chile on greenhouse gas emissions from the mining industry and how they affect sustainability in the context of copper mining. An interesting finding from this study is that improving the sea transport network for Chilean copper exports can contribute significantly to the emission reduction.

Other economic sectors, such as s08 to s12, do not contribute to the GDP of this territory. The s01 (agriculture, forestry, and fishing) sector, through the production of lamb meat, contributes enormously to the GDP in Patagonia.

A number of scholars in Chile have conducted extensive research on the relationship between emissions and GDP (data available in Appendix A). Mardones and Saavedra [62] examined the effects of emissions inside a particular administrative region. Mardones and Muñoz [43] analyzed environmental taxation in Chile to reduce greenhouse gas emissions using IO methods. O’Ryan [63] investigated the connection between Chile’s carbon footprint and the rise of renewable energy sources in the country’s electrical sector. Fuel consumption was employed by Ramos [36] to study each activity’s contribution to CO₂ emissions. The subject of the study related to which areas and activities cause the most pollution. This is the first record of its sort in Chile, and it is enhanced by a multiregional matrix (MRIO) created by Haddad [64] for 12 economic sectors, dated 2014.

As we already mentioned, many investigations have looked at the connection between economic activity and greenhouse gas emissions, namely carbon dioxide emissions. However, these studies have not covered Chile or some of its macro-territories (Norte Chico and Patagonia) in a thorough way.

3. Methodology

This section will present the methodological proposal that is used in this paper, which is based on multiregional input–output modeling. The goal of this proposal is to try to address some of the criticisms that have been highlighted in the literature regarding the use of an appropriate procedure when evaluating the impact of the productive CO₂ pollution [65]. In this concept, we use three approaches combined in a way that we think is novel and that advances the stated goal.

First, proper weighting is necessary to ensure that the values obtained are more in line with economic realities; particularly because it is crucial to emphasize the sector coefficients that are most significant relative to the remaining sectors and because appropriate weighting facilitates industry comparison. This will address a number of issues, including the concentration of activity and pollution, which is particularly important. Commodity production, which tends to concentrate output and distort its direct and indirect consequences on the economy, is a defining feature of those economies. With the implementation of this change, an attempt is made to approximate the potential impact of an activity more precisely. This is accomplished by using the Sonis and Hewings [66] matrix weighted by intensity method, or MPM, which authors like Ali [20] have suggested using in conjunction with a pollution vector to observe the hierarchical importance of activities in terms of pollution and production.

Let us examine an initial scenario where the following equation represents the whole output of an economy:

$$\mathit{x}={(\mathit{I}-\mathit{A})}^{-1}\mathit{y}=\mathit{L}\mathit{y}$$

(1)

where x is the total output, (I − A)⁻¹ is the well-known inverse Leontief matrix (with l_ij ∈ L obtained from the matrix A of technical coefficients: A = [a_ij]; with a_ij = x_ij/x_j; and, y, the final demand.

According to Ali [20], premultiplying matrix L by a pollution vector will incorporate the influence of pollution.

$$\mathit{B}=\widehat{\mathit{e}}\mathit{L}$$

(2)

where $\widehat{\mathit{e}}$ corresponds to the diagonalized vector of CO₂ pollution and B reflects the direct and indirect CO₂ multipliers, i.e., the total amount of CO₂ generated when each sector produces one unit of output.

Thus, and based on Equation (2), Ali et al. [20] propose establishing the following matrix, called “eMPM”:

$${\mathit{e}\mathit{M}\mathit{P}\mathit{M}}_{\mathit{B}}={\displaystyle \frac{1}{V_B}}{\mathit{B}}_{\mathit{i}.}{\mathit{B}}_{.\mathit{j}}={\displaystyle \frac{1}{V_B}}\left[\begin{array}{c}{\mathit{b}}_{1.}\\ ⋮\\ {\mathit{b}}_{\mathit{n}.}\end{array}\right]\left[\begin{array}{ccc}{\mathit{b}}_{.1}& \dots & {\mathit{b}}_{.\mathit{n}}\end{array}\right]; \hspace{1em}{\mathit{e}\mathit{M}\mathit{P}\mathit{M}}_{\mathit{B}}=\left\{m_{\mathit{i}\mathit{j}}\right\}$$

(3)

where V_B represents the global intensity of matrix $\mathit{B}$, defined as $V_B={\sum }_{i=1}^n{\sum }_{j=1}^n{b}_{ij}$.

In other words, the total of all the components of the matrix representing the immediate, direct, global indirect, and indirect effects—aside from the effects on the sector and its self-consumption—determines the global intensity of the impact, caused by both production and CO₂ emission.

Researchers, such as Lahr [67], Soza-Amigo [15,68], and Tarancon [69], contend that the Schintke and Stäglin [70] formula is an appropriate means of assessing sensitivity in economic activities, using the following expression:

$$w_{ij}(p)=a_{ij}\left({\displaystyle \frac{l_{ji}}{100}}p+l_{ii}{\displaystyle \frac{X_j}{x_i}}\right)$$

(4)

where $w_{ij}$ represents the degree of importance that the $a_{ij}$ coefficient achieves, if the total effects on the economy are considered.

Furthermore, Soza-Amigo and Ramos [15] also indicate that to adequately evaluate the different impacts, it is better to replace the Leonfief inverse matrix in that formulation with the MPM matrix proposed by Sonis and Hewings [66], which turns out to be similar to Equation (3) but without considering the vector “e” in “B”, that is, without including CO₂ emissions in the eMPM. In this way, they indicate that to analyze the importance of the a_mn coefficient, it is better to use the following formulation:

$$w_{mn}\left(p\right)=a_{mn}{\displaystyle \frac{1}{V}}\left(b_{n.}{b}_{.m}p+100b_{m.}{b}_{.m}{\displaystyle \frac{X_n}{x_m}}\right)=a_{mn}{b}_{.m}{\displaystyle \frac{1}{V}}\left(b_{n.}p+100b_{m.}{\displaystyle \frac{X_n}{x_m}}\right)$$

(5)

In other words, the term w_mn(p) allows us to quantify the importance of the a_mn coefficient, considering the total effects on the economy (matrix L), weighted by the normalized dispersion power index (b_.m [1/V]).

As is apparent, Ali [20] propose using the eMPM matrix to weight the CO₂ emissions in the economy appropriately. Similarly, other authors propose evaluating the impact of production between activities using an elasticity approach and the MPM matrix, but without weighting, since the goal was to assess the impact of production alone rather than its combined effect [15]. Consequently, we think that this approach can be integrated with CO₂ by suggesting the measurement of elasticity using an eMPM type matrix, which brings us to:

$$r_{ij}(p)={\displaystyle \frac{1}{w_{ij}}}={\displaystyle \frac{p}{a_{ij}\left[m_{ji}p+100m_{ii}\frac{X_j}{x_i}\right]}}$$

(6)

where $r_{ij}\left(p\right)$ indicates the maximum value, in percentage, that does not cause changes greater than the production percentage “$p$“ set for the condition established in Equation (1).

We consider that the suggested technique is a generalization and an improvement over the formulation of Schintke and Stäglin [70] as well as over the modifications suggested by for some authors [15,20]. This approach now enables us to obtain elasticities that are weighted according to both CO₂ emissions and the system’s intensity. This is one method in which weighing incorporates the consequences through their dispersion. We agree that the formulation that has been presented adequately captures the interactions between CO₂ emissions and the various productive stages in the economic system. By obtaining elasticities, one can distinguish the impact of a sectoral production increase of a given percentage on CO₂ emissions in addition to determining how much a productive stage should be modified.

In summary, based on what has been stated, three aspects can be answered, firstly, considering the effects of CO₂, we can question how much the relationship between sector j and a particular productive stage, an element of the production function, needs to change to achieve the 1% increase in production. This means that for sector j to raise its production by 1%, the relationship with sector i, including CO₂ emissions, must change by 5% if the resulting elasticity r_ij (p = 1) between sector j and productive stage i is 5; secondly, it can be assumed that the relationship between the two is unaffected since it has a high value, or low sensitivity, if the result is high, indicating that the increase in production and CO₂ pollution of sector j requires many changes through the interaction with “i”; and thirdly, on the contrary, if the value is low, it indicates that the 1% increase in “j” is highly sensitive to the change indicated by the relationship “i” with a low value; in this regard, the change is more likely because the modifications are small, and the production and pollution per unit produced will increase when this change exists.

A Boolean matrix is created once the findings are determined based on the suggested methodology and the pollution vector. A value of 1 denotes that the obtained elasticity is less than or equal to 15%. After this is acquired, the most sensitive sectors are identified by creating graphs that show how the regions are related to one another.

4. Data

Regarding the data source, how it was treated, and other works that have made use of it, let us first point out that the MRIO of Chile is used. It was compiled using data for 2014 and is the only one available to the best of our knowledge. Haddad [64] built the MRIO database, which comprises 12 economic activity branches and considers all 15 of the country’s regions. Similar uses of the matrix have been made in the past, such as by Ramos [36].

Regarding the pollution vector, we believe that by taking CO₂ consumption from the FIGARO data source and weighting it appropriately first, with respect to Chile, and then repeating the process for the regions of this country based on these results, the CO₂ consumption can be reached for each of the regions and their sectors. This is based on the works of [10,47], where it is argued that Chile and Turkey present specific productive structures and production functions—maintaining the proportion given for them—similar to those of developed economies like New Zealand, Australia, Austria, Finland, France, and Greece.

The FIGARO base was utilized to collect the emissions data, considering Austria’s CO₂ emissions in 2015. Since the original base was created for 65 different economic activity branches, they were divided into 12 groups to align with MRIO operations in Chile, such as (s01 (CPA_A01-CPA_A03); s02(CPA_B); s03(CPA_C10(12)-CPA_C33); s04 (CPA_D35-CPA_E37(39)); s05(CPA_F); s06(CPA_G45-CPA_G47(+)CPA_I); s07 (CPA_H49-CPA_H53(+)CPA_J58-CPA_J62_63); s08(CPA_K64-CPA_K66); s09(CPA_L68B(+) CPA_L68A); s10 (CPA_M69_70-CPA_N80(82)); s11(CPA_P85-CPA_U) y s12 (CPA_O84).

Determining the emission units per output for Austria (CO₂/VBP) was the next stage. The production of each activity in Chile was then multiplied by this relation, providing an information estimate of the nation’s sectoral emissions. A location criterion (region GDP/country GDP) was used to allocate Chile’s emissions among the regions. This produced the regional sectoral emissions, which, when added, resulted in Chile’s emissions overall.

Since works such as Alcántara and Padilla [71] and Duro and Padilla [72], which used emissions data from the International Agency of Energy to refer to 15 economies and the period 1971–2001, conclude that the GDP is the primary explanatory factor of the differences in CO₂ emissions, we believe that the CO₂ vector based on the information available in FIGARO is a good approximation. Consequently, using a weighting based on structural similarity is practical for situations when you do not have information, or you want to conduct a study from a different angle.

The geographical structure of Chile can be understood by examining Figure 2, left, which shows the country divided into four macrozones, each represented by a different color. It is pertinent to point out that the territories and areas are presented in the order that has been determined to carry out this study; nevertheless, this order may change based on the sort of study being conducted.

The first macrozone, known as Norte Grande, is made up of the regions of Arica and Parinacota (r01), Tarapacá (r02), and Antofagasta (r03). It is primarily responsible for the mining of copper and its derivatives in Chile; the second, known as Norte Chico (Atacama (r04) and Coquimbo (r05)), has less mining activity related to iron but is distinguished by its greater industrial and commercial development. In addition to its mining activity, the Central Zone comprises the regions of Valparaíso (r06), Metropolitan (where Santiago, the capital of Chile, is located; r07), Libertador General Bernardo O’Higgins (r08), Maule (r09), Biobío (r10), and La Araucanía (r11). The Metropolitan region is particularly significant for its industrial activity, and it also houses the various headquarters of numerous companies operating in the country. Patagonia, on the other hand, is defined by industrial development; in particular, the regions of Los Ríos (r12), Los Lagos (r13), Aysén del General Carlos Ibáñez del Campo (r14), De Magallanes and Chilean Antarctica (r15) are closely related to developments in the aquaculture industry.

5. Results

Economic growth and greenhouse gas emissions are two topics that are currently being discussed among citizens in most nations. The connection between these two current issues will be detailed in the study’s findings that will be presented below.

As we previously stated, for a particular year, data were used in matrix format. It was challenging to update these data because they relate to the cross-regional exchanges of goods, services, and input purchases and sales. Unfortunately, Chile does not currently have access to these data. In an effort to keep these data up to date, the Central Bank of Chile is working to enable and display this kind of data under the purview of “experimental statistics” (experimental statistics; review in Banco Central de Chile; https://si3.bcentral.cl/siete/ accessed on 13 July 2024).

Given the existence of data on extra-regional and interregional sales and purchases, this information is presented in aggregate form without specifying the sectors or destinations. Because of this, updating the matrix is difficult and outside the purview of this paper, which offers a novel method for assessing the impact of emissions. In any case, we anticipate having access to the data needed to update the matrix shortly, and this makes it easier to display more recent work in the future.

This study uses FIGARO. There are two justifications for using FIGARO data. First, according to an economic perspective, it relates to the relationship between efficiency in production processes and specific dynamics in their evolution. This has been examined by numerous writers [9,50,72,73,74,75,76].

Furthermore, the studies by Soza-Amigo and Aroca [10] and Soza-Amigo [47], which used input–output tables from economies in three different stages of development (the mid-1990s, early 2000s, and mid-2000s) with 116 tables each, have been used as references. These investigations showed that there is productive sensitivity and exchange interrelationships in Chile’s production functions. Moreover, Chile’s productive structure is comparable to that of several developed economies; these structures and productive processes are shared by economies like Greece, Australia, New Zealand, France, and Finland. Chile even outperformed Germany in terms of ratios; it also outperformed Canada, Poland, and Portugal somewhat and Belgium somewhat less.

Hence, in the absence of more detailed information, assuming that emissions are also similar is not a poor premise when considering production functions from a productive standpoint where similar production functions exist; that is, how things are done, or in the words of Leontief [9], a similarity in the technological recipe (that is, the same proportional use of inputs), especially given the lack of information regarding emissions for the case of Chile.

The available research is currently restricted to two areas: research on burning fossil fuels using equations that allow for the indirect determination of the fuels; in this sense, Ramos [36] evaluated how gas emissions affect Chile’s various regions using the same matrix as this study, utilizing formulations from the Spanish Inventory System of the year 2016 and the Chilean National Energy Commission’s report on fuel consumption by sector; additionally, the use of indirect information like that used here may lead to biased results, as the same production system does not always imply that it is contaminated to the same degree.

Additionally, the indirect links that are developed or the indirect requirements that are required for the production process may differ from the technology itself in terms of emission.

Figure 2 is generated by using the suggested formulation (Equation (6)), which is displayed on the right side of Figure 2. In this context, three points are highlighted: First, each of the previously determined macro-territories is linked to a certain hue of the nodes; second, there is mention of the 15 regions and the 12 economic sectors. Regarding this, it is observed that the region numbers are listed this time in a north-to-south order; that is, the numbers normally linked with the regions are not utilized; agriculture, forestry, and fishing (s01); mining (s02); manufacturing industry (s03); electricity, gas, water, and waste management (s04); construction (s05); trade, lodging, and dining (s06); transportation, communications, and information services (s07); financial intermediation (s08); real estate and housing services (s09); business services (s10); personal services (s11) and public administration (s12); third, the direction of the arrows will show whether the sector needs to change on its own to increase production (self-consumption) if they are on the same node. On the other hand, if the arrows start in activity “i” and end in activity “j”, it means that, in order for activity “j” to change its production by 1% in this instance, at least one of the 12 productive stages in sector “i” needs to change by a maximum of 15% in terms of pollution units as a result of the production that was undertaken.

Based on these results, three things can be seen; first, the high dependence on three regions in the central zone of the country (r06; r07 and r10) and, on four very specific activities (s01; s02; s04 and s05); second, the inter-industrial relationship according to geographical position; and third, the limited relationship that exists between the northern and central parts of the country with respect to Patagonia.

From the first perspective, it was noted that the nation’s primary activities are concentrated in the regions in question (r06; r07 and r10); additionally, the Valparaíso region (r06) is home to Chile’s main port. On the other hand, the activities are evidently concentrated in agriculture (s01), mining (s02), and the production of electricity, gas, and water (s04). In considering this, the country bases its activities and projects on what occurs in these regions and with those activities. Consequently, should other regions choose to increase production, irrespective of the sector in which this is desired, they will require the support of the aforementioned regions and sectors; that is, the input requirements and, as a result, the CO₂ pollution, which is related to and can be attributed to the demanding regions, will be impacted by their contributions and cease to pollute (r06s01; r07s01; r07s02; r07s04; r07s05 and r10s02).

Figure 2 shows that there are two regions and their economic activities within the territory located in the Norte Chico facilitate their development with small modifications (r04s12 and r05s02). In the first example, the public administration sector (s12) and the Atacama region (r04) are based on the results of r10s02 and r07s02. It is relevant to keep in mind that the population of Norte Chico is significantly smaller than that of Norte Grande, which depends on government activities. Furthermore, this type of territory based its growth on the same activities and functions that the Chilean government is in charge of; this also occurs in the extreme regions of southern Chile, such as Patagonia [77].

As a result, it is not surprising that its growth is dependent on the mining sector in other areas, given the continuous interactions between the State and the activities that are typically based in the Metropolitan area (the corporate offices). Regarding the Coquimbo region (r05) and its iron mining industry, the growth of the latter is contingent upon the investments made in the sector, which are likely to originate from the house matrix (r07) and the activities conducted in the Biobío region (r10).

Since the Central Zone of Chile is known to have the highest level of interactivity in the nation and is highly interconnected, any activity that wishes to increase its output by 1% will have an impact on the input requirements and pollution levels of the other activities located in this zone. As the primary activities that encourage greater coordination, this is noted for the sets of regions and activities, particularly for the Metropolitan region (r07). If this region wants to increase its production, it will impact the nation in terms of input requirements and through CO₂ pollution. Similar CO₂ emission patterns have been seen in other Latin American metropolitan areas, including Rio de Janeiro in Brazil [78], and Medellín in Colombia [79].

In parallel with the previously displayed results, it was noted that, for the applied filter, the economic sectors in Patagonia do not relate to the rest of the nation. This is likely because most of the activity in this region, including industrial activity, contributes very little when compared to the rest the nation.

If the nation, and because the changes needed are significantly larger than those that have been shown.

If the rest of Chile and Patagonia are compared, Figure 2 indicates that the latter is primarily dependent on a small number of industries, especially those related to sectors 11 (personal services) and s01 (agricultural, forestry, and fisheries) and s02 (mining). The latter relate to the Magallanes region of Chile, which is highly reliant on public administration due to its geographic isolation from the rest of the country. Additionally, this area is dependent on the extraction of gas, which is used for producing methanol. Consequently, it can be established that Patagonia is less economically integrated with the rest of the nation and more dependent on the state of the central region than the other regions already examined.

While the previously displayed results are contrasted with the conventional method—that is, the formulation of Schintke and Stäglin [70]—the latter without including the MPM matrix and weighting for contamination, Figure 3 shows the changes in elasticities that can be seen. First, we note the significance that r07s03 upholds; in this instance, however, it would suggest that in order for sector 03 of region 07 to grow by 1%, there must be little or equivalent adjustments of 15% made in at least one productive stage from those sectors that have an arrow pointing towards r07s03; Moreover, the observation is upheld on the interactions that occur in the Norte Grande between its own exchanges and the part this region plays in the growth of the rest of the nation.

Furthermore, the two approaches are clearly different from one another, as seen in Figure 2 and Figure 3. While the traditional approach involves Patagonia, the proposed approach excludes it, which makes sense given its industrial importance to the Chilean economy. In contrast, the traditional approach tends to concentrate on everything, while the proposed approach reveals relationships that were not previously observed.

Taking these findings into consideration, we may cite research that details several public policy initiatives Chile has implemented to lessen its climate change effects. One of the industries that contributes the most to greenhouse gas emissions is undergoing a significant shift, as mentioned by Simsek [80], referring to the energy sector in Chile, which has experienced significant change. In addition to committing to creating green policies to suggest a more sustainable energy system, Chile joined the Paris Agreement in 2017. With a current target of at least 70%, the nation was able to encourage the generation of renewable electricity without feed-in tariffs. In accordance with Mardones [81], a selection of public policy tools, including the carbon tax, the emissions trading system, and renewable energy subsidies, reduce greenhouse gas emissions in Chile by 40–57%. As noted by Jorquera-Copier [82], climate policies with high carbon prices or a mandate for all energy sources to be 100% renewable could be essential to Chile’s goal of having a completely renewable electricity generation system by 2050 that includes green hydrogen.

6. Conclusions and Upcoming Directions for Research

Ever since the Donella Meadows on Limits to Growth book study in the 1970s, climate change and how to address it, while considering industrial activity and limiting resources in a finite world, have dominated discourse [83,84]. Although it has not been explicitly mentioned in the book, it is relevant to mention some authors [85,86] who analyze the growing decoupling process between CO₂ emissions and GDP growth. To encourage green development, these measures must be coordinately incorporated into an new framework by coordinating public policy players across regional boundaries [83].

This article has three objectives, which were outlined in the beginning. Because of this, this paper has employed a novel methodology that involves using the eMPM analysis to investigate structural changes in economic sectors and their link with CO₂ emissions. Three themes emerge from this: first, the country’s high reliance on four highly specialized industries, including mining, agriculture, and the production of gas, electricity, and water; second, the inter-industrial relationship based on geographic location is emphasized; and third, the relatively small economic structure and exchange between the country’s northern and central regions and Patagonia.

In the second objective, the differences in elasticity between economic sectors and geographic regions represent the link between structural change and emissions. The final objective has to do with the fact that many developing nations contribute significantly to climate change. As such, this study can act as a springboard for the eMPM’s implementation in other developing nations and conduct comparative analyses that support public policy and the COP agenda.

Given that waste and waste management landfills are frequently managed by multiple territories in coordination, we believe that this eMPM methodology could also be extended to the study of waste. We propose a wMPM (waste MPM) using Waste Input–Output Analysis (WIO), initially proposed by Nakamura and Kondo [87], that would allow the relationships between territories and their waste generation to be studied. Doussoulin and Bittencourt [88,89] conducted a comparative analysis using the WIO to examine the building industries in Brazil and France.

A new perspective to consider would be the commutation of an essential sector like mining in Chile. In this regard, Aroca and Atienza [90] and Jamett Sasonov and Paredes Araya [91] have conducted research in Chile on the commutation flows between Antofagasta in the Norte Grande and the Central Zone. Ferrada [92] examines how employment is shifting in some areas of Chilean Patagonia in accordance with the direction of their nodal centers. The study concludes that the commutation is evident in low-skilled employment. The distances between cities and the cost of transportation influence this commutation. The investigation of CO₂ emissions from aircraft used to ferry mining labor from the Norte Grande to the Central Zone is one potential area for future research. Additional economic activities that take place between Patagonia and the Central Zone could be the subject of this investigation.

We consider that the methods that have already been discussed can be enhanced by other ones, like the hypothetical extraction methodology (HEM) to assess the opposite scenario [93,94]; for instance, what percentage would a given sector’s emissions drop? Does it reduce emissions all the way, and does this cut production by 1%? Alternatively, which industries would be most impacted by this reduction considering their apparent relationships and the emissions they would no longer produce? In this way, the amounts that are being suggested remaining unclear even though the proposal clarifies the situation.

Finally, future studies focused on the Chilean case should consider the multiplier impacts of employment and environmental respect, which are based on those essential activities in terms of forestry, mining, fishing, and other natural resource exploitation. In the case of Chile or for regions that rely on these activities, this has not been suggested.