Mining tailings dumps and socio-territorial inequalities in Chile: an exploratory study

ABSTRACT

The aim of this study is to territorially link the distribution of mining tailings and multi-dimensional poverty at a local level. For this purpose, a database was constructed with governmental information regarding municipalities that present tailings, and subsequently, identified their levels of multidimensional poverty. The results show: (i) the presence of tailings throughout the national territory, although their greater concentration is found in the country's center-north; (ii) that the majority of tailings at the national level is located in socially and economically disadvantaged municipalities; and (iii) that only 4 municipalities concentrate 321 of the 757 tailings, which represent 42,4% of the national total. This information allows us to conclude that the task of surveying and remediating mining tailings is a national governance challenge and that in order to address it, the context and the social injustice it causes must be considered.

KEYWORDS:

• Social sciences
• extractivism
• environmental sociology
• governance
• environmental injustice

View correction statement:

Correction

1. Introduction

In the space of three decades, Chile has achieved a high per capita income, reduced poverty, and exhibited relative political stability (Ffrench-Davis, 2014), all of which has allowed it to accede as member number 31 of the select group of OECD countries in 2010. This international image of ‘as a country that is an example of transition to development’ is underpinned by a growth model based on the limitless extraction of natural resources (Maillet et al., 2021; Urbina et al., 2021) and a social concentration of the fruits of economic growth, which has generated socioeconomical inequality (Candia et al., 2021; Kennedy & Murray, 2012; Mieres, 2020; PNUD, 2017) as well as social, political, and territorial tensions (Donoso & Von Bulow, 2016; Ffrench-Davis, 2014).

An important part of Chile´s economical growth has been based on mining, a productive activity that has been carried out in the country since Colonial times, and which has expanded throughout the nation’s history, achieving its maximum peak with saltpeter in the nineteenth Century and, presently, with copper mining. South American models of growth of similar characteristics have been referred to as ‘extractive’, highlighting their dependence on the international markets and their focus on intensified exploitation-exportation of commodities (Gudynas & Acosta, 2011; Svampa, 2019). A situation that, while generating economic growth, has also given rise to social-environmental unrest (Allain, 2019; Bebbington, 2010; Irarrazaval, 2022; Maillet et al., 2021; Smart, 2017) and the erosion of institutions by private interests (Leiva, 2019; Madariaga et al., 2021).

In this text, we understand extractive mining in Chile as a socio-territorial ‘paradox’. On one hand, it is the main producer of copper in the world, generating important contributions to the national economy and fiscal revenues (Soza-Amigo et al., 2021; Sturla et al., 2018), which has allowed the post-dictatorship progressive administrations to finance social programs and public infrastructure that have contributed to the reduction of poverty indicators. In Latin America, this is referred to as the consensus of the commodities, or neo-extractivism (Svampa, 2019). On the other hand, mining has generated significant impacts on the ecosystems, affecting the flora and fauna, and community's health located in the vicinity of extractive infrastructures. For instance, Yohannessen et al. (2015) recorded the presence of heavy metals and toxic elements in schools in Chañaral, as by-products of tailings dumped in its bay, a situation that increases the probability of developing severe illnesses such as cancer.

The mining production process is extremely complex and may generate socio-territorial impacts at different moments (Agboola et al., 2020). In this case, we pay attention to mining tailing deposits, infrastructures containing residues from the chemical process in which minerals of greater and lesser economic interest are separated (Edraki et al., 2014). Thus, the deposits contain minerals that are not economically interesting for commercialization and that may be toxic. The relation existing between mining production and its residues must be taken into consideration (Campos-Medina et al., 2021; Edraki et al., 2014; Kossoff et al., 2014), in the case of Chile, as the main world exporter of copper, it officially registers 757 tailings, being the 3rd country with the largest number of these type of infrastructures and failures in the planet (Islam & Murakami, 2021). In this line, it is estimated that, by 2026, 915,000,000 tons of tailings material will have to be managed annually (Fundación Chile, 2019).

1.1. Relevance and objectives of the research work

Having sufficient evidence about the deposits of tailings in Chile is relevant due to the risks they generate in the socioecological systems. It has even been documented that, after being abandoned, tailings continue to be significant sources of contamination by heavy metals (Angelovicova et al., 2014). In parallel, research has shown that tailings materials have affected watercourses and soils with chemical contamination, which results in an impact to the flora and fauna (Armienta et al., 2004; Corcho Alvarado et al., 2014; Fernández-Caliani et al., 2009; Hettler et al., 1997), besides having had an impact on the cardiovascular and respiratory health of the population (Kan et al., 2021; Mpode & Fantke, 2017; Yohannessen et al., 2015). To this we must add the complex international situation of the last 20 years, about which there is some scientific consensus in terms that the number of failures – and of catastrophic failures – in deposits is growing (Islam & Murakami, 2021; Owen et al., 2020) and, additionally, few incidents generated a large impact (Hatje, 2017), as for instance, Mount Polley in Canada and Samarco and Córrego de Feijao in Brazil.

In 2014 in the British Columbia region of Canada, the Mount Polley Tailings Dam collapsed. Government and independent investigations showed that construction was carried out on unsuitable ground, and that after its completion in 1997, the owner Imperial Metals had added height to the dam annually to increase its capacity, which gave way (Government of British Columbia, n.d..; Lamberti, 2015; Schoenberger, 2016; Zabolotnii et al., 2022). As a result of this failure, approximately 17 million cubic meters of water and 8 million cubic meters of Tailings were spilled, i.e. a total of 25 million cubic meters of contaminants (heavy metals). This severely affected the environment and fauna, especially Polley and Quesnel Lakes and Rios (Byrne et al., 2016; Byrne et al., 2018; Cuervo et al., 2017; McMahen & Hughes, 2016), where most of the material arrived which also forced to cut the drinking water supply of more than 300 families in the region of British Columbia (Lamberti, 2015).

In 2015 in the state of Minas Gerais in Brazil, two tailings dams at the Samarco mine, owned by the Brazilian company Vale and Anglo-Australian BHP, collapsed. Unfortunately, the cause of the disaster has not been clarified to date; however, its effects were evident: after the dams gave way, approximately 55 million cubic meters of sludge and mining waste were released and flowed more than 600 km down the Doce River. Probes of the material concluded that there were minerals, such as Iron, that exceeded by more than 1,000% what was allowed (Menezes, 2015) by international standard, there is scientific consensus regarding the environmental contamination that was generated (Biscotto et al., 2020; Burritt & Christ, 2018; De Oliveira et al., 2017; Fernandes et al., 2020; Galindo Da Fonseca & Galindo Da Fonseca, 2016; Hatje et al., 2017; Queiroz et al., 2018). The government recognizes 12 people dead, despite locals reporting 120 people dead and 200 missing (Lamberti, 2015).

In 2019 another mining industry tragedy struck Brazil, a tailings dam collapsed at the Córrego de Feijao mine located in the municipality of Brumadinho, also in the region of Minas Gerais. The tailings was owned by the same company Vale (De Almeida et al., 2019). After the dam gave way, 13 million cubic meters of tailings spilled, which advanced at a speed of 80 kilometers per hour. This wiped out practically the entire downstream ecosystem, forests, agricultural areas and a significant number of villages (Parente et al., 2021; Silva et al., 2020; Teramoto et al., 2021; Thompson et al., 2020), which has led to attempts to identify more than 400 biological samples of dead people (Moreira Araújo et al., 2022). Darlington et al. (2019) developed a visually interesting chronology and impact map.

On another note, recent studies on international literature have identified that the development of research from within the social sciences and the Global South are a challenge that has been sparsely addressed at an international level (Campos-Medina et al., 2022; Ojeda-Pereira & Campos-Medina, 2021). This work is groundbreaking and a significant contribution to the study field of mining tailings by combining sociologic and spatial analysis. The article inquires about the territorial link between the number of tailings deposits and the level of multidimensional poverty. By doing so, we seek to reconstruct the geography of tailings, observing the socioeconomic inequalities linked to territorial characteristics.

In accordance with other investigations that have explored the link between socioeconomic conditions and extractive activities in Chile (Hofflinger et al., 2021), this study contributes to the socio-spatial understanding of tailings deposits in Chile. This study is innovative and relevant for the international literature on tailings deposits, where there are no publications that address the geographies constructed as a result of the location of these infrastructures.

Our argument poses that tailings deposits, as contaminating and unsafe infrastructures, do not affect the national territory in the same way. Thus, mining in Chile not only concentrates the economic lead in advantaged municipalities, but also concentrates the ‘risks and impacts’ of its activities in disadvantaged municipalities. After this introduction, the methodology and results are discussed, followed by a conclusion highlighting the contributions of the analysis are presented.

2. Methodology

Methodologically, this investigation had a quantitative approach, georeferencing the information with GIS tools in order to elaborate charts and cartographic representations of the municipal averages of the multidimensional poverty index (in Spanish: Índice de pobreza multidimensional), and the location of tailings deposits.

2.1. Databases

The databases consulted are two. Firstly, to provide an account of the distribution of tailings deposits at a national level, we used the Tailings Deposits Cadastre developed by Chile’s National Geology and Mining Service, updated 2020 (in Spanish: Catastro de Depósitos de Relaves by Servicio Nacional de Geología y Minería). Secondly, in order to observe the level of multidimensional poverty at a municipal level in Chile, we used the Municipal Report from the Chilean Congress Library, where the data obtained from the 2017 Socioeconomic Characterization Survey are presented (last version) dependent from Chile’s Ministry of Social Development (in Spanish: Encuesta de Caracterización Socioeconómica del Ministerio de Desarrollo Social).¹

In order to integrate both variables, we constructed a database selecting the municipalities where one or more tailings deposits are present (available free of charge). So, from the 346 municipalities at a national level, we worked with 68, which represent a fifth of the national total, considering the following variables: (i) name of municipality (ii) number of deposits, (iii) recoding of the number of deposits, (iv) level of multidimensional poverty, (v) recoding of level of multidimensional poverty to dichotomic category variable regarding the national average of multidimensional poverty.

2.2. Data processing in software

From this database, the following model of information processing was elaborated, represented in Figure 1.

Figure 1. Data processing diagram. Source: authors’ own.

Read the detailed description of this figure

Flowchart describes the data processed through GIS in order to represent the socioeconomic inequality related to tailings deposits in Chile. Two input databases are used; multidimensional poverty index by communes and mining deposits cadastre obtained from the National Geology and Mining Service, updated to 2020. From the latter, we proceed to select only the communes that have mining tailings deposits. This selection of data is crossed with the multidimensional poverty data by commune, creating a new attribute. According to these results, the communes are grouped considering the reference line (national average of the poverty index). Then the groupings are categorized according to the concentration of tailings deposits. Finally, the output data will be the following classifications: (i) 1 to 2 deposits low concentration (ii) 3 to 9 deposits medium concentration (iii) 10 to 44 deposits high concentration (iv) 45 to 123 deposits very high concentration. The results are mapped for subsequent comparative analysis between groups above and below the national average of the multidimensional poverty index used as a baseline.

For the database processing, subsequent analysis and cartographic preparation, ArcMap 10.8 software was used, authenticated by means of a license for students associated with the University of Chile (in Spanish: Universidad de Chile). The inclusion of the value contained in the Multidimensional Poverty Index variable was done for each particular commune by selecting and modifying attributes, on the other hand, the crossing of the databases of the Cadastre of tailings deposits with the database of the Communal Political Administrative Division was done using the ‘Spatial Join’ tool to determine the number of tailings deposits (Points) associated with each polygon.

3. Results

The results of this research are presented in two sections. The first one addresses descriptive information of the relevant variables (Figure 2, Tables 1 and 2); in the second section we focus on representing the link between the variable Average Municipal Multidimensional Poverty Index (under or above the national average) and the number of deposits per municipality (Table 3, and Maps 1 and 2).²

Figure 2. Scatter diagram. Source: authors’ own.

A purple horizontal line is drawn on the Y-axis at 20.7, a yellow line is drawn at 25.5, which establishes the average multidimensional poverty index of the communes with at least one tailings deposit, a red line at 28.07 represents the average of map 2, and a green line at 17.27 represents the average of map 1.

Table 1. Municipalities and concentration of tailings deposits.

	No. of municipalities	% In respect to the total no. of municipalities with tailings deposits	No. of tailings	% Respect to the total no. of tailings deposits
Low concentration	27	39.7%	35	4.6%
Medium concentration	21	30.9%	103	13.6%
High concentration	16	23.5%	298	39.4%
Very high concentration	4	5.9%	321	42.4%
Total	68	100%	757	100%

Source: authors’ own.

Table 2. Category in relation to multidimensional poverty average.

Category	Number of municipalities	%
Above national average	49	72.1%
Below national average	19	27.9%
Total	68	100%

Source: authors’ own.

Table 3. Municipalities, multidimensional poverty and tailings deposits (frequency in relation to concentration of tailings deposits).

	Low concentration	Medium concentration	High concentration	Very high concentration	Total
Above national average	77.78% (21)	66.67% (14)	62.50% (10)	100.00% (4)	72% (49)
Below national average	22.22% (6)	33.33% (7)	37.50% (6)	0.00% (0)	28% (19)
Total	100% (27)	100% (21)	100% (16)	100% (4)	100% 68

Source: authors’ own. Note: values approximated to the hundreds.

3.1. Concentration of tailings deposits

Table 1 shows that 29.4% of the municipalities selected for the study concentrate 81.8% of the tailings deposits. In other words, 20 municipalities accommodate 619 of the 757 tailings. Along the same line, 5.9% of the municipalities show a very high concentration, which are: Andacollo, La Higuera, Copiapó and Illapel, host 42.4% of the country’s tailings (321 of the 757 deposits).

The national average of the Multidimensional Poverty Index is 20.7, and we use this figure as a parameter to distinguish between those municipalities that are situated above the national average (with higher poverty levels, disadvantaged municipalities) and under it (with lower levels of poverty, advantaged municipalities). Table 2 shows that of the 68 municipalities with presence of tailings deposits in Chile, about 72% show a level of multidimensional poverty above the national average. With this information it is possible to argue that the majority of tailings deposits are located in municipalities with high levels of multidimensional poverty, or, less advantaged.

3.2. Territorialization of inequality

Figure 2 and Table 3 present information on the crossing of both variables. Figure 2 shows the distribution of municipalities according to the number of tailings and the multidimensional poverty index. In general, most of the municipalities that have tailings are above the multidimensional poverty average, i.e. they are disadvantaged. It is also observed that municipalities with the highest number of Depósitos are all above the multidimensional poverty average.

Table 3 shows that municipalities that are above the national average in the Multidimensional Poverty Index (the least advantaged) exceed those situated below the national average in the Multidimensional Poverty Index in all levels of concentration. The most striking percentage difference is that of very high concentration. Although in absolute terms it refers to 4 municipalities, it is no coincidence that 100% of them have averages of multidimensional poverty above the national average. Indeed, it becomes evident that there is inequality in the territorialization of these mining waste infrastructures in Chile.

The cartographic observation is developed in two stages. Firstly, by showing the concentration of tailings in municipalities that present levels of multidimensional poverty below the average (advantaged municipalities) in Map 1. Secondly, by analyzing the concentration of tailings patterns in municipalities with levels of multidimensional poverty above the average (disadvantaged municipalities) in Map 2.

Map 1 allows us to observe that from the Antofagasta Region (north of the country) to the Valparaiso Region (center of the country) there are municipalities that we can consider in general terms as ‘advantaged’, presenting high concentrations of tailings. Such are the cases of the municipalities of Sierra Gorda, Antofagasta, Diego de Almagro, La Serena, Coquimbo and Petorca. On the contrary, in municipalities located in mountainous areas in the Valparaiso Region (Aconcagua Province), and the Metropolitan area, as well as in the corridor running between The Andes and the Coastal mountain ranges in the O’Higgins Region, we find a lower concentration of tailings in terms of quantity. A relevant case to highlight is the municipalities of Coyhaique and Chile Chico in the Chilean Patagonia, where abandoned tailings that are the result or a by-product of the area's mining past history can be found. In a different level of analysis for the case of ‘advantaged’ municipalities, we find as many cases of municipalities with ‘high concentration’ as with ‘low concentration’ (6 cases).

Map 2 shows the territorial distribution of mining tailings in the municipalities we have described as ‘disadvantaged’. The first thing to note is that in these cases we can find the category ‘Very high Concentration’ of tailings, and that these are situated in the Atacama and Coquimbo regions. We should bear in mind that, in this case, we are talking about number of tailings and not about their size. In territorial terms, the Coquimbo Region (389 deposits), and the municipality of Copiapó in the Atacama Region (88 deposits) function as an axis where most of the tailings are located, their presence decreasing towards the north and the south.

A closer look into the Coquimbo region shows that all of its municipalities present at least one tailings deposit. Similarly, 13 municipalities in the region are above the multidimensional poverty average, and at least 11 of them host at least 3 tailings deposits. However, the municipalities with ‘Very High Concentration’ in this region are: Andacollo, La Higuera and Illapel, and each one of them reflects the existence of large-scale mining in the transverse valleys of the region, historically characterized by the development of agriculture and livestock. The cases of low concentration of tailings in the Maule and Aysén regions are traces of a historical mining activity which is not longer active, but show ongoing prospections for its reactivation.

4. Conclusions: environmental injustice and governance of recognizing the difference

In this article we have explored the socio-territorial link between multidimensional poverty and the location of tailings deposits in Chile. For this purpose, the level of multidimensional poverty has been recategorized according to the national average, and the municipal density of tailings location has been described. Thus, shed light on the inequality phenomenon regarding the location of these infrastructures. In other words, the tailings deposits represent a hazard that does not affect the country’s 346 municipalities equitably. Tailing deposits are situated in only 68 municipalities, where most of them are situated above the national multidimensional poverty average (that is to say, they are on the disadvantaged spectrum of the index). Now, analyzing in greater detail: there are 20 municipalities that present a high concentration of tailings, and 4 that show very high concentrations. In the latter, approximately 40% of the existing deposits in the country are located, and all of them are situated above the national average of multidimensional poverty.

During the last two decades, the environmental justice paradigm has gained particular strength (Taylor, 2000). Pellow (2020) argues that the environmental justice and injustice approach describes and highlights inequality and the disproportionality that exists in the distribution of environmental and health damages in contexts of natural resource extraction. Generally, it is the marginalized populations that are left out of the decision-making process and have to bear the brunt of the damages. This phenomenon was explored by Taylor (2000) in the early 2000s by social movements and Martinez-Alier (2003) called it environmentalism of the poor. There is an interesting research agenda on environmental justice and injustice, which has explored: socio-environmental conflicts (Martinez-Alier, 2001), the international construction of a movement for environmental justice (Martinez-Alier et al., 2016), the environmental justice content agenda and its links with other ideas (Martinez-Alier, 2012), the relationships between the local and the global in environmental inequalities (Carmin et al., 2011), development in a specific context such as Latin America (Carruthers, 2008), among others. The results of this research contribute significantly to studies on tailings deposits, where research works from social sciences, political sciences and human geography are scarce. In this sense, linking the results with a focus on environmental justice and injustice allows us to open a discussion on the governance of tailings deposits, and our research work contributes with a first approach to the socio-territorial inequalities of the ‘risks and impacts’ produced by the mining industry. The difference that exists between municipalities is an environmental injustice, and it is precisely in the face of injustice that one of the management alternatives proposed by sociology is recognition (Fraser & Honneth, 2006; Honneth, 1997).

The article seeks to contribute to the debate and public policies in Chile, and make its results available to decision-makers. In this sense, a governance that recognizes this socio-territorial inequality aims at validating that tailings do not affect the totality of the country in a similar way and that the affected territories counts on fewer social, economic, and technical resources to face this problem. In the same vein, this governance recognizes the need to consider the uniqueness of these territories and the need for greater local sensitivity to define the measures for intervention. When we talk about the uniqueness of territories, we must contextualize ourselves to a country like Chile. When we make a geographical comparison between Chile and Europe, we observe that the length of Chile would extend from the south of Portugal to the north of Sweden. First of all, this length means that Chile has a great variety of ecosystems from North to South, and from East to West. In the North we find the most arid desert in the world, the Atacama Desert (in Spanish: Desierto de Atacama), and in the South we find the most important fresh water reserve on the planet in Patagonia (in Spanish: Patagonia). From east to west there is an extensive coastline, which is followed by the Coastal Mountain Range (in Spanish: Cordillera de la Costa) and then by the Andes Mountain Range (in Spanish: Cordillera de los Ándes). Secondly, diversity also has a historical and cultural dimension, with respect to access to services, transportation, quality of political functioning, presence of the state, and others.

Regarding the limitations of this study, it is necessary to advance research on multidimensional poverty at the municipal level beyond the construction of a distinction of two groups based on the national average. In this way, it will be possible to observe more clearly the socioeconomic conditions in which most of the country's municipalities find themselves. A second limitation that we recognize is that, although we refer to the National Tailings Deposits Cadastre, which is the most reliable database that exists and is prepared by the authority, there is no absolute certainty that all tailings infrastructure in Chile has been registered. The database is incomplete, for example, in the definition of the owners of the deposits, the date of opening or the tonnage they have. At the same time, there is no clarity regarding the mineral content of each deposit; only the main minerals that are commercialized in the operations linked to these deposits are described. A geological team is needed to take soil samples at the deposits and determine what heavy metals are present in each tailings deposit, which would allow estimating the magnitude of the threat. There is also no information regarding the spatial area covered by each deposit, which is essential for risk assessment. It is urgent to use other methodologies in order to map their locations, for instance with satellite data (Sarker et al., 2022) or, also, through citizen mechanisms that can be subsequently verified by State teams and entered into the national database. Thus, the quality and quantity of reliable state data available to date is considered as an important limitation to research, and also as a challenge for public policies based on evidence.

Software

Spatial information analysis, geoprocessing and cartographic design were carried out with ArcGIS Desktop 10.8.2 software.

Acknowledgements

Iván Ojeda-Pereira thanks the support of the Center for Integrated Disaster Risk Management ANID/FONDAP/1511001 (CIGIDEN). We thank the Vicerrectoría de Investigación y Desarrollo Tecnológico of the Universidad Católica del Norte for their support for the translation and open access publication of this paper. Fernando Campos-Medina is grateful for the support of the FONDECYT Regular Nº1231353 project ‘Energy Transition and Critical Minerals, Tensions for divergent socio-ecological futures’, funded by the Agencia Nacional de Investigación y Desarrollo (ANID). We also thank the peer reviewers of this article; any mistakes are our sole responsibility.

Data availability statement (DAS)

The database ‘Tailings deposits and multidimensional poverty in Chile’ that we built in this research, is available in open access so that researchers can replicate the study or use the data: https://doi.org/10.5281/zenodo.7044851

 Google Scholar

.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article was originally published with errors, which have now been corrected in the online version. Please see Correction (https://doi.org/10.1080/17445647.2024.2357006).

Additional information

Funding

This work was supported by Fondo de Financiamiento de Centros de Investigación en Áreas Prioritarias CIGIDEN/ANID/FONDAP Nº1511001; Vicerrectoría de Investigación y Desarrollo Tecnológico Universidad Católica del Norte, Fondos para traducción y Open Access y Fondo Nacional de Desarrollo Científico Tecnológico FONDECYT Regular Nº1231353.

Notes

1 The multidimensional poverty index emerges from an international debate on how to understand poverty, it is argued that a broader approach to the strictly economic is necessary, thus, the multidimensional poverty index is a way of measuring poverty that considers monetary (e.g. income) and non-monetary (e.g. quality of life, health, education) elements (Tsui, 2002; Alkire et al., 2020; Santos & Villatoro, 2018). Alkire et al. (2020) integrate the health, education and standard of living dimensions to estimate multidimensional poverty indices for different countries and compare on a global scale. For the Chilean case, the 2017 Socioeconomic Characterization Survey considered the dimensions Education, Health, Labor and Social Security and Housing to estimate a multidimensional poverty index at the regional level (see technical document of the Social Observatory of the Government of Chile, 2018).

2 During the analysis, the terms favored and disadvantaged are used. We have selected this terminology considering that a parameter has been established and the municipalities are separated into two sets (depending on that parameter). We did not want to use concepts such as more or less poor municipalities in order not to fall into stigmatization or generalizations.

Appendix