UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mineworkers' quality of life in remote communities : a multiple case study in the Brazilian Amazon Costa, Silvana Dunham da 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2008_spring_costa_silvana.pdf [ 6.92MB ]
Metadata
JSON: 24-1.0066394.json
JSON-LD: 24-1.0066394-ld.json
RDF/XML (Pretty): 24-1.0066394-rdf.xml
RDF/JSON: 24-1.0066394-rdf.json
Turtle: 24-1.0066394-turtle.txt
N-Triples: 24-1.0066394-rdf-ntriples.txt
Original Record: 24-1.0066394-source.json
Full Text
24-1.0066394-fulltext.txt
Citation
24-1.0066394.ris

Full Text

MINEWORKERS’ QUALITY OF LIFE IN REMOTE COMMUNITES: A MULTIPLE CASE STUDY IN THE BRAZILIAN AMAZON  by Silvana Dunham da Costa B.A., Universidade Federal da Bahia, 1996 M.Env.Des., University of Calgary, 2000 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Mining Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (VANCOUVER) April 2008  © Silvana Dunham da Costa, 2008  ABSTRACT The mining industry has long played a significant role in regional development in remote regions throughout the world. For the last two decades, the industry has faced high expectations regarding sustainable development and corporate social responsibility, particularly in remote and environmentally sensitive areas. Mining community models and mineworkers’ accommodation strategies in remote locations have varied greatly, yet there has been little documented reflection on the various models’ performance or on their implications for the quality of life (QOL) of mineworkers and their families and for the pre-existing local communities. This multidisciplinary case study research used a subjective quality of life approach to investigate the levels of satisfaction with QOL and specific aspects of QOL domains in three communities: the company town, the gate development community and the integrated community. The triangulation of data from qualitative and quantitative methods was used to examine the major QOL factors that should be taken into account by mining companies, local governments and policy makers when planning for mine development in remote areas. Findings of this research suggest that differences exist between the mineworkers’ levels of satisfaction with specific QOL aspects and how QOL predictors are defined in distinct mining community models. Even though the case studies represent three clearly different models of mining communities, in general, mineworkers in all three communities seem to be only moderately satisfied with their quality of life. It is also suggested that employees living in two almost opposite models—the company town and the gate development community—seem to have similar levels of satisfaction with overall quality of life, suggesting that the investment in infrastructure and services limited to the boundaries of the company town is not reflected in a generally improved perception of overall quality of life in this community. Findings also support the argument for an environmental and social impact assessment process for new mines in remote areas. This process should include a full and integrated consideration of the economic, environmental and social impacts of the workforce migration to remote areas and the consequent intensification of the already rapid urbanization of environmentally sensitive areas such as the Brazilian Amazon.  ii  TABLE OF CONTENTS ABSTRACT................................................................................................................................... ii LIST OF TABLES........................................................................................................................ vii LIST OF FUGURES...................................................................................................................... x GLOSSARY ................................................................................................................................. xi ACKNOWLEDGEMENTS ........................................................................................................... xii DEDICATION..............................................................................................................................xiii CHAPTER 1: INTRODUCTION....................................................................................................1 1.1. 1.2. 1.3. 1.4. 1.5. 1.6. 1.7. 1.8.  Problem Statement.........................................................................................................1 Rationale ........................................................................................................................4 Objectives and Research Questions ..............................................................................5 Approach: Subjective Quality of Life .............................................................................6 Applied Contribution of Research...................................................................................7 Academic Contribution of Research...............................................................................9 Limitations of Research................................................................................................10 Thesis Structure ...........................................................................................................11  CHAPTER 2: CASE STUDIES ...................................................................................................14 2.1. 2.2. 2.3. 2.4.  2.5.  2.6.  The Brazilian Mining Context .......................................................................................14 The Development of the Carajás Mining Province .......................................................16 The Mining Company: Vale ..........................................................................................17 Núcleo Urbano de Carajás: The Closed Company Town ............................................21 2.4.1. Development .................................................................................................22 2.4.2. Urban Infrastructure.......................................................................................23 2.4.3. Population......................................................................................................24 2.4.4. Education.......................................................................................................25 2.4.5. Health ............................................................................................................25 2.4.6. Culture, Recreation and Entertainment .........................................................26 Parauapebas: The Gate Development Community......................................................27 2.5.1. Development .................................................................................................28 2.5.2. Population......................................................................................................29 2.5.3. Urban Infrastructure.......................................................................................31 2.5.4. Economy........................................................................................................32 2.5.5. Human Development.....................................................................................34 2.5.6. Education.......................................................................................................35 2.5.7. Health ............................................................................................................35 2.5.8. Culture, Recreation and Entertainment .........................................................36 Canaã dos Carajás: The Integrated Community ..........................................................37 2.6.1. Development .................................................................................................39 2.6.2. Population......................................................................................................40  iii  2.7.  2.6.3. Urban Infrastructure.......................................................................................43 2.6.4. Economy........................................................................................................44 2.6.5. Human Development.....................................................................................47 2.6.6. Education.......................................................................................................47 2.6.7. Health ............................................................................................................47 2.6.8. Culture, Recreation and Entertainment .........................................................48 Reflecting on the Case Study Communities.................................................................49  CHAPTER 3: APPROACH TO RESEARCH..............................................................................51 3.1. 3.2. 3.3. 3.4.  3.5.  Quality of Life ...............................................................................................................51 Measuring Quality of Life..............................................................................................52 3.2.1. Internal versus External Measurements of Quality of Life .............................53 Subjective Wellbeing ....................................................................................................55 Measuring Subjective Wellbeing ..................................................................................58 3.4.1. Health-Focused versus Holistic Measurements ............................................58 3.4.2. Self-Reported Levels of Satisfaction .............................................................61 3.4.3. Reflective versus Formative Indicators..........................................................61 3.4.4. Quality of Life Domains .................................................................................65 Reflecting on the Quality of Life Literature ...................................................................68  CHAPTER 4: METHODOLOGICAL STRATEGY ......................................................................69 4.1  4.2. 4.3.  4.4. 4.5  Methods Overview........................................................................................................70 4.1.1. Literature Review...........................................................................................71 4.1.2. Participant Observation .................................................................................72 4.1.3. Interviews ......................................................................................................73 4.1.4. Survey Questionnaire ....................................................................................75 Phase I: Preliminary Research.....................................................................................76 Phase II: Quality of Life Model .....................................................................................76 4.3.1. Phase II: Methods..........................................................................................78 4.3.2. Quality of Life Domains and Indicators..........................................................80 4.3.3. Quality of Life Survey Questionnaire Development.......................................84 Phase III: Interviews and Questionnaire.......................................................................89 4.4.1. Interviews ......................................................................................................90 4.4.2. Administration of the Quality of Life Survey Questionnaire ...........................99 Analytical Strategy .....................................................................................................105 4.5.1. Quantitative Analysis ...................................................................................105 4.5.2. Qualitative Analysis .....................................................................................115  CHAPTER 5: RESULTS...........................................................................................................117 5.1.  Quantitative Analysis Results.....................................................................................117 5.1.1. Establishing Reliability.................................................................................117 5.1.2. Calculating Mean Scores.............................................................................119 5.1.3. Comparing Means .......................................................................................121 5.1.4. Comparing Means: Average Quality of Life (AQOL) ...................................123 5.1.5. Comparing Means: Economic Domain ........................................................123 5.1.6. Comparing Means: Work Domain ...............................................................124  iv  5.2.  5.1.7. Comparing Means: Environment Domain....................................................126 5.1.8. Comparing Means: Social Domain ..............................................................129 5.1.9. Comparing Means: Health Domain .............................................................131 5.1.10. Defining QOL Satisfaction Predictors ..........................................................132 5.1.11. Defining LAW Satisfaction Predictors..........................................................135 5.1.12. Subjective Wellbeing Index .........................................................................137 Qualitative Analysis Results .......................................................................................139 5.2.1. Interview Results: Defining Quality of Life ...................................................139 5.2.2. Núcleo Urbano de Carajás ..........................................................................141 5.2.3. Canaã dos Carajás......................................................................................150 5.2.4. Parauapebas ...............................................................................................156  CHAPTER 6: DISCUSSION .....................................................................................................163 6.1.  6.2.  6.3.  6.4. 6.5.  6.6.  Addressing Research Question 1: Main Factors That Appear to Impact Perceived Quality of Life .............................................................................................................165 6.1.1. Predictors of Satisfaction with Quality of Life ..............................................166 6.1.2. Predictors of Satisfaction with Life as Whole (LAW) ...................................174 Addressing Research Question 2: Comparing Satisfaction Levels in the Núcleo Urbano, Canaã dos Carajás and Parauapebas .........................................................176 6.2.1. Comparing Perceptions of Environment Domain ........................................178 6.2.2. Comparing Perceptions of the Work Domain ..............................................180 6.2.3. Comparing Perceptions of the Health Domain ............................................185 Addressing Research Question 3: Aspects that Most Impact Satisfaction with Quality of Life .........................................................................................................................187 6.3.1. Criteria for Most and Least Satisfying Aspects of Quality of Life.................188 6.3.2. Núcleo Urbano de Carajás ..........................................................................189 6.3.3. Canaã dos Carajás......................................................................................195 6.3.4. Parauapebas ...............................................................................................200 Addressing Research Question 4: Respondents’ Characteristics and Satisfaction with Quality of Life .............................................................................................................206 Opportunities to Improve Quality of Life in Remote Mining Communities ..................213 6.5.1. Mining Companies.......................................................................................214 6.5.2. Local Governments .....................................................................................230 6.5.3. Opportunities for Policy Development .........................................................234 Other Considerations .................................................................................................236 6.6.1. Reflecting on Concepts of Quality of Life and Life as Whole.......................236 6.6.2. Defining Quality of Life Domains .................................................................236 6.6.3. Reflection on the QOL Questionnaire and the SWI Pilot Test.....................237  CHAPTER 7: CONCLUSION ...................................................................................................238 7.1. 7.2.  Highlights of Quality of Life Findings ..........................................................................238 7.1.1. Satisfaction with Quality of Life and Quality of Life Domains ......................238 7.1.2. Factors Affecting Satisfaction with Quality of Life........................................242 Highlights of Methodological Findings........................................................................245 7.2.1. Quality of Life Domains and Indicators........................................................245 7.2.2. Quality of Life Questionnaire and Subjective Wellbeing Index Pilot Test ....246  v  CHAPTER 8: CLAIM OF ORIGINAL CONTRIBUTION ...........................................................247 8.1. 8.2.  Original Contribution to Theory and Research ...........................................................247 Original Contribution to Professional Practice and Policy ..........................................248  CHAPTER 9: RECOMMENDATIONS FOR FUTURE RESEARCH.........................................249 9.1. 9.2.  Future Methodological Research ...............................................................................249 Future Research on Quality of Life in Remote Mining Communities..........................249 9.2.1. Investigating the Relationship between Quality of Life, Productivity and Occupational Health and Safety ..................................................................249 9.2.2. Comparative Study: Brazilian and Canadian Mining Communities .............250  REFERENCES .........................................................................................................................251 APPENDIX A ............................................................................................................................270 APPENDIX B ............................................................................................................................271 APPENDIX C ............................................................................................................................292 APPENDIX D ............................................................................................................................298  vi  LIST OF TABLES CHAPTER 2 Table 2.1: Yutaka Takeda Hospital, 2004 Data ..........................................................................25 Table 2.2: Indicators of Growth and Urbanization, Pará and Parauapebas................................29 Table 2.3: Population Gender Distribution, Parauapebas...........................................................30 Table 2.4: Age Distribution, Pará and Parauapebas (2007) .......................................................30 Table 2.5: Household Access to Selected Goods and Services, Parauapebas .........................32 Table 2.6: Poverty Indicators, Pará and Parauapebas (2000)....................................................33 Table 2.7: Income Indicators, Parauapebas (2000 Data) ...........................................................34 Table 2.8: Human Development Index (HDI) Indicators in the Carajás Region..........................35 Table 2.9: Health Indicators, Parauapebas 2006........................................................................36 Table 2.10: Indicators of Growth and Urbanization, Pará and Canaã dos Carajás ....................41 Table 2.11: Urban and Rural Populations, Pará, Parauapebas and Canaã dos Carajás (2006) 42 Table 2.12: Population Gender Distribution, Canaã dos Carajás ...............................................42 Table 2.13: Age Distribution, Pará and Canaã dos Carajás (2007)............................................43 Table 2.14: Permanent Housing Units and Selected Goods and Services, Canaã dos Carajás (2000)..................................................................................................................................44 Table 2.15: Economic Indicators, Canaã dos Carajás (2000) ....................................................45 Table 2.16: Poverty Indicators, Pará and Canaã dos Carajás (2000) ........................................46 Table 2.17: Health Care Indicators, Canaã dos Carajás and Parauapebas (2003)....................48 Table 2.18: Case Studies............................................................................................................50 CHAPTER 3 Table 3.1: Recent Quality of Life Measurement Instruments ......................................................59 Table 3.2: Examples of SWB Measurements Using Reflective Approach ..................................63 Table 3.3: SWB Measurements Using Formative Approach ......................................................64 Table 3.4: Domains in Smith’s Criteria (1973) ............................................................................66 Table 3.5: Quality of Life Domains in Selected Instruments .......................................................67 CHAPTER 4 Table 4.1: Phase II Data Collection ............................................................................................79 Table 4.2: Subjective Wellbeing Index and Personal Wellbeing Index .......................................86 Table 4.3: Questionnaire Items: QOL Domains and Indicators ..................................................88  vii  Table 4.4: Questionnaire Items: Demographic Information ........................................................89 Table 4.5: Interview Population Sample .....................................................................................92 Table 4.6: Interview Population Sample: Núcleo Urbano de Carajás .........................................95 Table 4.7: Interview Population Sample: Canaã dos Carajás.....................................................96 Table 4.8: Sampled Population for Interviews: Parauapebas Residents ....................................99 Table 4.9: Frequencies of Demographic Variables...................................................................102 Table 4.10: Quality of Life Questionnaire Statistical Analysis ...................................................110 Table 4.11: Variables of Interest for QOL Domains ..................................................................111 CHAPTER 5 Table 5.1: Domain Reliability Tests – Cronbach Alphas...........................................................118 Table 5.2: Indicators’ Mean Scores ..........................................................................................119 Table 5.3: Mean Scores............................................................................................................122 Table 5.4: Statistical Tests for AQOL........................................................................................123 Table 5.5: Satisfaction with Economic Domain by Education Level .........................................124 Table 5.6: Satisfaction with Work Domain by Community of Residence ..................................125 Table 5.7: Satisfaction with Work Domain by Job Class...........................................................125 Table 5.8: Satisfaction with Work Domain by Length of Employment ......................................126 Table 5.9: Satisfaction with Environment Domain by Community of Residence ......................127 Table 5.10: Satisfaction with Environment Domain by Housing Status ....................................128 Table 5.11: Satisfaction with Environment Domain by Job Class .............................................128 Table 5.12: Satisfaction with Environment Domain by Age Groups .........................................129 Table 5.13: Satisfaction with Environment Domain by Number of Dependent Children ...........129 Table 5.14: Satisfaction with Social Domain by Educational Level...........................................130 Table 5.15: Satisfaction with Social Domain by Number of Dependent Children .....................130 Table 5.16: Satisfaction with Health Domain by Community of Residence ..............................131 Table 5.17: Satisfaction with Health Domain by Educational Level ..........................................132 Table 5.18: Model Coefficients: Regression against QOL, all communities .............................132 Table 5.19: QOL Model Coefficients per Community ...............................................................134 Table 5.20: LAW Model Coefficients per Community ...............................................................136 Table 5.21: QOL Model Coefficients for SWI............................................................................138 Table 5.22: Pearson Correlations for SWI and Average QOL ..................................................139 Table 5.23: Major Concepts – How Do You Define Quality of Life? .........................................140 Table 5.24: Best Things about Living in the Núcleo Urbano de Carajás ..................................142 Table 5.25: Worst Things about Living in the Núcleo Urbano de Carajás ................................143 viii  Table 5.26: Areas of Priority: Núcleo Urbano de Carajás .........................................................144 Table 5.27: Best Things about Living in Canaã dos Carajás ....................................................151 Table 5.28: Worst Things about Living in Canaã dos Carajás ..................................................152 Table 5.29: Areas of Priority: Canaã dos Carajás.....................................................................153 Table 5.30: Best Things about Living in Parauapebas .............................................................157 Table 5.31: Worst Things about Living in Parauapebas ...........................................................158 Table 5.32: Areas of Priority: Parauapebas ..............................................................................159 CHAPTER 6 Table 6.1: Variation of Means by Community of Residence .....................................................178 Table 6.2: Environment Indicators by Community of Residence ..............................................179 Table 6.3: Satisfaction with Number of hours you work per week by Community of Residence ..........................................................................................................................................181 Table 6.4: Satisfaction with Opportunities to Grow Professionally by Community of Residence ..........................................................................................................................................182 Table 6.5: Satisfaction with “Ability to Make Decisions on Your Own” at Work by Community of Residence .........................................................................................................................184 Table 6.6: Satisfaction with Relationships at Work by Community of Residence .....................184 Table 6.7: Health Indicators by Community of Residence ........................................................186 Table 6.8: Most Satisfying Aspects of the Núcleo Urbano de Carajás .....................................189 Table 6.9: Least Satisfying Aspects of the Núcleo Urbano de Carajás ....................................192 Table 6.10: Most Satisfying Aspects of Canaã dos Carajás .....................................................195 Table 6.11: Least Satisfying Aspects of Canaã dos Carajás ....................................................197 Table 6.12: Most Satisfying Aspects of Parauapebas ..............................................................200 Table 6.13: Least Satisfying Aspects of Parauapebas .............................................................202 Table 6.14: Educational Level – QOL Domains ........................................................................209 APPENDIX A Table A: Vale Northern System Mines Profile (Production year 2006) .....................................270 APPENDIX B Table B: Population Frame: Northern System Workforce.........................................................271  ix  LIST OF FIGURES CHAPTER 1 Figure 1.1: Thesis Roadmap.......................................................................................................13 CHAPTER 2 Figure 2.1: Location Maps ..........................................................................................................18 Figure 2.2: Operations in Vale’s Northern System, in Relation to Case Study Communities .....19 Figure 2.3: Logistics Infrastructure for Vale’s Northern System..................................................20 Figure 2.4: The Núcleo Urbano de Carajás: Aereal View and Typical Housing Units ................22 Figure 2.5: The Núcleo Urbano: Banks and Recreation Centre .................................................24 Figure 2.6: Kuarup Brazilian Culture Centre and Yutaka Takeda Hospital, Núcleo Urbano de Carajás................................................................................................................................26 Figure 2.7: Parauapebas Streetscapes ......................................................................................27 Figure 2.8: Parauapebas: Open Sewage....................................................................................37 Figure 2.9: Canaã dos Carajás: Vale-Built Homes and an Existing Wood-Frame Home ...........38 Figure 2.10: Population Growth in Canaã dos Carajás...............................................................41 Figure 2.11: Canaã dos Carajás City Centre ..............................................................................45 Figure 2.12: Canaã dos Carajás GDP (1997–2004)...................................................................46 Figure 2.13: Canaã dos Carajás. Casa da Cultura .....................................................................48 CHAPTER 4 Figure 4.2: A Preliminary Quality of Life Conceptual Model .......................................................77 CHAPTER 6 Figure 6.1: Structure of Chapter 6 ............................................................................................164 Figure 6.2: Approach to Answering Research Question 1........................................................165 Figure 6.3: QOL Domains Affecting Perceptions of Quality of Life in the Population Sample ..168 Figure 6.4: QOL Domains Affecting Perceptions of Quality of Life in the Núcleo Urbano de Carajás..............................................................................................................................168 Figure 6.5: QOL Domains Affecting Perceptions of Quality of life in Parauapebas ..................168 Figure 6.6: QOL Domains Affecting Perceptions of Quality of Life in Canaã dos Carajás........168 Figure 6.7: Comparing Communities. .......................................................................................177 Figure 6.8: Population Samples: Education Levels ..................................................................182 Figure 6.9: Approach to Defining Aspects Most Affecting QOL Satisfaction ............................187  x  GLOSSARY AQOL: Average quality of life CVRD: Companhia Vale do Rio Doce (Vale) DIFN: Diretoria de Ferrosos (Ferrous Metals Division) DIOC: Diretoria de Operações de Não-ferrosos de Carajás (Non-ferrous Metals Division) HDI: Human Development Index LAW: Life as a whole IBAMA: Institute de Meio Ambiente e Recursos Naturais Renováveis do Brasil (Brazilian Institute for Renewable Natural Resources) PWI: Personal Wellbeing Index QOL: Quality of life QWL: Quality of work life SWB: subjective wellbeing SWI: Subjective Wellbeing Index WHO: World Health Organization WHOQOL: World Health Organization Quality of Life  xi  ACKNOWLEDGEMENTS I offer my enduring gratitude to the faculty, staff and my fellow students at UBC, particularly my Sustainability Working Group and Bridge Program colleagues, who have inspired and continue to inspire me to persevere in my work in this exciting field. It was a both a challenge and a rewarding experience to complete this program in the Mining Engineering Department. I feel it is important to recognize the professors and the leadership of this Department, particularly Dr. Marcello Veiga and Dr. Malcolm Scoble, who brought together students from different disciplines to conduct multidisciplinary and holistic mining research. The Sustainability Working Group is a key piece of this groundbreaking initiative. I owe particular thanks to co-supervisor Dr. Malcolm Scoble and my supervisory committee, Dr. Kay Teschke and Dr. Mary Louise McAllister. They endured the exciting yet challenging task of working in a diverse team. Special thanks to my supervisor Dr. Marcello Veiga for the unrelenting support and the incredible efforts to implement this research in our home country. I also thank Vale for their support of this research, and all the staff who helped me in the field, participated in the interviews and filled out the questionnaire. Wholehearted thanks go to Kaliane Guimarães, who more than once, as a true Brazilian (and Northeastern) sister, received me with open arms and heart in the fascinating community of Carajás. My sincere thanks also go to Doug White, who got me on the path of quality of life studies, and to Dr. Bruno Zumbo and Dr. Robert Cummins for the invaluable contributions to the development of the methodological strategy of this research. Thank you to Sara Bose for the excellent work editing this manuscript, and Danilo Sena for the being so generous with your time to make some of the figures and maps in this document. Finally, I must thank my husband, Ian, for his endless support—and editorial comments—and for putting up with the nuisance of sharing his life with a professional student for so many years. I love you and could not have made it without you.  xii  DEDICATION  To Ian  xiii  CHAPTER 1: INTRODUCTION 1.1.  Problem Statement  The mining industry has long played a significant role in regional development in rural and remote regions throughout the world. More recently, the industry has been challenged with social expectations regarding sustainable development and corporate social responsibility, particularly in remote locations. Mining community models used by the industry in remote locations have varied greatly throughout history, from paternalistic company towns to fly-in-fly-out remote mine camps (Storey 2001; Costa, 2004). Over time, there has been little documented reflection on and analysis of the various models’ performance or the implications they have for the quality of life of mineworkers and families living in these communities. In Brazil, this is also reflected in a general lack of coordinated and defined policy on frontier development (Godfrey, 1990; Hecth and Cockburn, 1990; Palheta da Silva 1998). In the last century, significant mining activity has occurred in the world’s frontier regions, such as the Canadian North and the Brazilian Amazon, and has often resulted in the prosperity of communities and corporations alike. However, economic prosperity and development often occurred with little formalized consideration for the wellbeing of the people and communities that fuel and provide critical support to the mining industry. The International Institute for Environment and Development’s Mining, Minerals and Sustainable Development (MMSD) Project’s South American Regional Report concluded that mining is necessary for the strategic development of countries such as Brazil, Peru, Chile and Bolivia. However, the report also points out that there is a clear need to reconsider where and how mining development takes place. Specifically, MMSD’s South American Regional Report indicates that one of the priorities for the mining industry in Latin America should be making a more direct and equalitarian contribution to local development (Mining Minerals and Sustainable Development, 2000). In many parts of the world, the resource industry’s lack of consideration for social and economic local development has resulted in communities with low urban standards, challenging living and working conditions and considerable psychological stress (Burvill, 1975; Riffel, 1975; Beach, 1999; British Columbia & Yukon Territory Building and Construction Trades Council, 2004; White, 2004; Costa et al., 2005; Sibbel et al., 2006).  1  Individuals and families in remote mining communities constitute a particular sociological group living under unique geographical and socio-cultural circumstances. Isolation from relatives and friends and the typically limited resources and opportunities for family members of mineworkers are some of the distinct disadvantages of these communities (Riffel, 1975; Sharma and Rees, 2007). Challenges such as harsh environmental conditions, feelings of isolation and inadequate social stimuli are significant for employees and families in remote mining communities (Riffel, 1975; Storey, 2001) and have resulted in cases of depression and incipient neuroses (Burvill, 1975), as well as high rates of divorce among mineworkers in such communities (Sharma and Rees, 2007). Other challenges identified in the resource community literature include inadequate housing, services and facilities, the predominance of single men and the absence of female companionship, inadequate educational and medical services, and few opportunities for employment for married women (The Institute of Local Government, 1953; Riffel, 1975; Neitzert et al., 1999; Storey, 2001; Sharma and Rees, 2007). Traditionally, human wellbeing has not been a critical consideration in decision making during mine lifecycle planning. Mineworkers’ accommodations (e.g., company towns, community-based housing, permanent and temporary mine camps and mineworker hostels) tend to be secondary planning concerns, subordinate to operations, and historically have not received much consideration 1 (Riffel, 1975; Bray, 1991; Costa, 2004). Decisions and investigations on workforce accommodations are mainly based on economic performance measures (Skaburskis, 1980; Bottge, 1986;) and, to a certain extent, on considerations such as the supply and local availability of skilled workers and training needs (Costa and Scoble, 2006). Decisions regarding the settlement mode chosen are typically made by the company, with little or no involvement of the workers and people who eventually will live or spend considerable amounts of time in these communities or camps (Canada Employment and Immigration Advisory Council, 1987). This trend is clearly observed in the Brazilian Amazon (Almeida, 1986; Perz, 2000; Godfrey, 1997; Hecth, 1990; da Silva, 1998; Barros et al., 2002). Even though mineral development is consistently accompanied by economic development and wealth creation, pre-existing local communities affected by mining development often do not  1  In Canada, a surge of research and reflection on the resource community seems to have started in the 1950s, upon the recognition of the severe shortcoming of this model of frontier development (e.g., Bancroft, 1975; Robinson, 1962; The Institute for Local Government, 1953). The recognition of the need to make life in resource communities more appealing was reflected in the development of towns with high levels of consideration for physical standards, such as Leaf Rapids in Manitoba in the 1970s and Tumbler Ridge in British Columbia in the 1980s (Costa, 2004).  2  participate and share in economic prosperity brought about by mineral exploitation in a meaningful way. Thus, they have experienced only marginal positive impacts on their wellbeing and quality of life (Weitzer, 2006). These communities have, however, often lived with the negative social and environmental impacts of such operations (Cleghorn, 1999; McAllister and Wismer, 1999; Veiga et al., 2001; Remy and McMahon, 2002; Dreyer and Myers, 2004) and shaped by powerful forces of metropolitan-based corporations (Bowles, 1982). Decisions regarding mineworkers’ settlements and accommodations, and the interface between the workforce and the existing local communities, are critical to the sustainable development of remote regions (Costa and Scoble, 2006) and reflect a company’s commitment to corporate social responsibility. Particularly in the developing world, the socio-economic characteristics of remote locations often are such that large industrial development projects attract large numbers of migrants to these communities. Most of these migrants come from poverty-stricken areas or neighboring countries, and they often have no relevant training. Small and remote communities worldwide are often not prepared—physically, socially or culturally—for such rapid and dramatic growth and change. This phenomenon is apparent in the Brazilian context (de Castro et al., 1995) and is also evident in many other areas throughout the world (de Castro et al., 1995; McMahon and Remy, 2001; Weber-Fahr, 2002; Weitzer, 2006). As stated by Veiga et al. (2001), mining companies’ attention to the surrounding social environment has historically been focused on reducing conflicts and complying with legal requirements rather than on the local communities’ long-term sustainability. The quality of life—that is, the level of wellbeing of residents and the suitability of the physical environment—in remote mining communities around the world significantly impacts the industry’s ability to attract and retain high-quality employees in these areas (Costa and Scoble, 2006). Satisfactory quality of life is also expected to promote higher productivity and efficiency and better health and safety performance of operations (Bancroft, 1975; Riffel, 1975; Costa et al., 2005). The rapid growth of mining development, along with the global shortage of skills and training 2 (Canadian Mining Industry Human Resources Council, 2005), makes it increasingly necessary to  2  According to the Canadian Mining Industry Human Resources Council, the mining industry is facing a human resources crisis, the effects of which are both global and long term, significantly impacting the ability of mining companies worldwide to attract and retain skilled workers (Canadian Mining Industry Human Resources Council, 2005).  3  identify the ideal conditions for employee wellbeing and investigate ways to improve the quality of life of miners and their families in remote locations. It is imperative that mining companies and governments recognize that their decisions impact not only the industry’s ability to fulfill mining workforce needs, but also the efficiency and health of the workforce and their families, as well as the sustainability of the local existing communities. It is critical that companies and government agencies involved in evaluating and permitting large-scale mines consider, from the earliest stages of mine planning, the human and social implications of mine development, particularly in remote locations. A significant change in practice is needed to favor the adoption of a new approach to planning the development of resource-based communities, an approach that focuses on human wellbeing.  1.2.  Rationale  The Núcleo Urbano de Carajás, Parauapebas and Canaã dos Carajás, located in the Brazilian Amazon, presented an opportunity for a close examination of the quality of life in remote mining communities. These communities were considered suitable for this research because: ƒ  they represent 3 different mining community models;  ƒ  they support operations of a single company; and  ƒ  they are located in relatively close proximity to one another.  Different mining community models offer distinct implications for the quality of life of mineworkers and their local communities. This research borrows its definition of mining community from Bowles (1982): “a community is the base camp or general headquarters where individuals engaged in work meet (as best they can) their daily needs for shelter, food, and rest so that they may work again another day.” The purpose of this particular definition is to emphasize the diversity of settlements in which individuals working in mining may live. In the context of this research, 3 mining community models were examined: the company town, the gate development 3 and the integrated community. Respectively, these models represent three distinct approaches to mineworkers’ accommodations: the paternalistic approach of a company-controlled community (in which all housing and facilities are owned and operated  3  Gate development refers to an informal settlement located adjacent to the entrance gate of the mine property. This term makes no reference to “gated communities”, which denotes a community with high level of security, usually accessible thorugh a gate.  4  directly or indirectly by the company); the “hands-off” approach (in which the company avoids intervention and allows for the local government and market forces to dictate the quality of the social and physical environment); and the collaborative approach, resulting in an integrated community development (in which the company works in partnership with the local government to provide for adequate development of the community). The 3 communities hence provided a unique opportunity to investigate the implications of these distinct approaches to mining communities and compare their impacts on the quality of life of mine employees in remote locations. The three case study communities are associated with three large-scale mining projects: Carajás Iron-ore Mining Complex, Azul Manganese and Sossego Copper Mine. These three operations are owned and operated by a single mining company, Vale 4—the largest Brazilian mining company and the second-largest diversified mining company in the world (Vale, 2007b). The single ownership is considered an advantage to this research as it is reflected in similar work and corporate cultures in the three communities and workforces. The fact that the 3 communities are located in a similar geographical location implies that environmental and political contexts are largely similar to one another, and it provided for short travel times between communities and thus more efficient use of limited research funds and time available.  1.3.  Objectives and Research Questions  The objectives of this research are: 1. To identify and prioritize factors that can impact mineworkers’ quality of life in remote mining communities; 2. To identify opportunities to improve mineworkers’ quality of life in the communities investigated; 3. To develop a quality of life framework that can be integrated into life-cycle planning of new operations in remote areas worldwide; and  4  In 2007, the Companhia Vale do Rio Doce (CVRD) announced its global brand unification and as of November 29th 2007 it is identified as Vale (Vale, 2007).  5  4. To develop a quality of life monitoring tool that can be used in remote mining communities. To reach the above objectives, this research examined mineworkers’ perceptions of quality of life in 3 case study communities to answer the following research questions: 1. What appear to be the main factors that impact the quality of life of mineworkers in the Núcleo Urbano de Carajás, Parauapebas and Canaã dos Carajás? 2. How differently do mineworkers living in these 3 communities perceive their quality of life? 3. What appear to be the most and least satisfying aspects of the quality of life in these communities? 4. How do the mineworkers’ characteristics appear to impact their perceptions of their own quality of life?  1.4.  Approach: Subjective Quality of Life  Subjective quality of life, or subjective wellbeing, is the theoretical approach for this research. The rationale for this approach is based on the limitations of externally measured quality of life evaluations and the need for deeper and more holistic measurements of quality of life as perceived and experienced internally by individuals. Quality of life researchers have defended this conceptual approach since the 1970s (e.g., Dalkey 1972; Andrews and Withey, 1976). The relevance of subjective quality of life also lies in the evidence that the more people become satisfied with their lives, the more they contribute positively to society. It has been widely argued that “happy” people have identity integrity, ego strength, mental maturity, inner control, social ability, activity and perceptual openness. Happiness and satisfaction lead people to become actively involved in society; positive perceptions of wellbeing free people to be creative and foster better social relationships; enjoyment of life fosters activity, strengthens social bonds, and preserves health (Veenhoven, 1991). Thus, to promote social and societal development, we need to enhance people’s perceptions of their quality of life—in other words, their subjective wellbeing (Sirgy, 2002). The Subjective Wellbeing approach considers a human-focused framework that can be used to support decision making regarding workforce accommodations and settlements that support natural resources development in remote areas. This framework takes into consideration the  6  wellbeing of workers, their families and their relationships with the local communities, and can be applied in other investigations of remote mining communities throughout the world. Subjective Wellbeing has been the approach taken by several quality of life studies, and several subjective wellbeing instruments have been created (e.g., Diener et al, 1985; Cummins, 1996; The International Well Being Group, 2005). A review of the literature and close examination of these instruments was critical when defining the conceptual model for this research, particularly regarding the definition of the quality of life domains and indicators adopted in this research.  1.5.  Applied Contribution of Research  Large-scale mining activity has long been an engine of development and employment of countries rich in natural resources, such as Canada (Robinson, 1962; Cranstone, 2002), Australia (Barnett, 1979) and Brazil (Koshiba and Pereira, 2006). However, the mining sector has also had to deal with negative public perception of its activities. Mining is an industry that has been widely perceived as a producer of a great deal of pollution. It has also been viewed as dangerous and negligent in protecting the health and safety of its workers and communities close to mining operations (Hilson, 2000; Richards, 2002; Schiavi, 2005). The early 1900s saw substantial efforts towards advancing mineworker health and safety (Hilson, 2000; Richards, 2002; Schiavi, 2005; National Institute for Occupational Safety and Health, 2008). Even though performance varies significantly from country to country, the mining industry of the 21st century has made some significant progress in health and safety performance (Hilson, 2002; National Institute for Occupational Safety and Health, 2008). The next revolutionary industry transformation occurred in the 1980s in response to growing environmental management and protection requirements. As a direct result of new regulations, intricate environmental and waste management systems were created. In addition, reclamation and mine closure procedures were taken into greater consideration. Today, there is evidence that the mining industry around the world is continuing to take steps to reduce its environmental footprint and ensure that land can be used in a productive and safe manner after mine closure (Hilson, 2002; Hodge, 2004). Since the early 1990s, the mining industry has faced the challenge of understanding and earning a “social license” to operate (Joyce and Thompson, 2000; Costa and Scoble, 2005; Kogel, 2006). The social license to operate concept implies the consideration of societal expectations that go  7  beyond workforce health and safety and environmental management. A social license to operate includes corporate social responsibility requirements, such as extensive community consultation, shared decision making and more equitable distribution of the costs and benefits of mine development (Joyce and Thompson, 2000; Kogel, 2006; Business for Social Responsibility, 2007). In this context, there is a need for companies to significantly change their approach to planning and developing mines in remote locations. There is also a need for changes in how government policy addresses mine development in remote areas. This research is timely with regards to the evolution of the global mining scenario. Mine development is experiencing a significant boom worldwide, particularly in countries where operations in remote locations are or will become increasingly common, such as Canada, Brazil and Australia. Mining companies and governments need to define appropriate approaches to development in remote areas whilst struggling to address the challenges of recruitment and retention of mining workforces. Combined with the typically high turnover rates in remote operations (Bancroft, 1975; Riffel, 1975; Bowles, 1982), the current skills shortage represents a human resources crisis for remote mining operations. A negative perception of the quality of life in remote mining communities compromises the industry’s ability to attract and retain high-quality employees in remote areas (Bancroft, 1975; Riffel, 1975) and, arguably, the industry’s ability to acquire and maintain a social license to operate. This research examined, through the perspective of mineworkers, the major quality of life factors that should be taken into account by both mining companies and policy makers when planning for mine development in remote areas. It thus provides a human-focused framework that can be used to support decision making in planning and implementing workforce accommodation strategies in remote areas, and a tool to monitor the performance of these strategies. Both the evaluation framework and tool include quality of life domains and indicators that could be applied to a range of settlements and accommodations options. Locally, this research also makes important contributions specific to the case study communities. It identifies factors that most impact the quality of life of mineworkers in each of the case study communities and examines the perspective of mineworkers to identify key areas of concern. Findings from this research are expected to provide for better-informed mining community development strategies to improve quality of life in these communities. The discussion of results provides some clear directions and identifies opportunities for specific interventions locally. It is also expected that findings will inform Brazilian policy makers and local governments that will  8  benefit from the identification of opportunities to address challenges related to quality of life in remote resource communities in the Amazon. With specific regards to the context of the 3 case study communities, Vale needs to address the challenges of recruitment and accommodations in the Brazilian Amazon as the company prepares for the future development of up to 5 more mines 5 and one hydrometallurgical plant in the same geographical region (Vale, 2007b). The most recently established community, Canaã dos Carajás, is expected to serve as the major centre for the development of these upcoming mines. It is intended to serve as a locus for the housing, education and health care for the mining workforce for several decades to come. Moreover, since Vale has recently expanded into a major global mining company through the purchase of Inco, a major Canadian mining company, it needs to address similar challenges at varying levels and in very different geographical and cultural contexts, such as Canada, the Pacific island of New Caledonia, the United Kingdom, the United States and countries throughout Asia.  1.6.  Academic Contribution of Research  Findings from this research complement the body of knowledge on resource communities and quality of life. This research contributes to the existing body of quality of life (QOL) literature by providing an evaluation of QOL in 3 case studies of resource communities in remote locations that were examined through a subjective wellbeing lens. This research intends to demonstrate how the subjective wellbeing approach can be integrated into lifecycle planning of new natural resource operations in remote areas. The research also intends to identify opportunities for positive intervention in existing resource communities. According to a recent publication on the past, present and future of the QOL movement (Sirgy et al., 2006), one of the critical needs in QOL research is to improve the applications of its findings. Few interventions based on research findings have taken place because policy makers tend to rely on objective social indicators. Significant tasks that remain to be completed in subjective quality of life research is a systematic approach to the assessment, the translation of the findings  5  Vermelho (Ni) expected start-up2008; Alvo 118 (Cu) expected start-up 2008, Cristalino (Cu/Au) expected start-up 2010, Salobo I (Cu/Au) expected start-up 2009, Salobo II expected start-up 2011, and Alemão (Cu/Au) expected start-up year 2011. A hydrometallurgical copper plant is also expected to process concentrate from Salobo II and Alemão (Vale, 2007b).  9  into practical applications and the development of an index of subjective quality of life indicators to compliment the existing social indicators already in use by policy makers (Sirgy et al., 2002). This research used a qualitative case study approach. Findings are not expected to provide definitive answers or solutions to broad quality of life theoretical inquiry. Findings are, however, expected to contribute to an improved theoretical understanding of the quality of life and the subjective wellbeing concepts. Specifically, findings are expected to provide insights on quality of life domains and indicators and the distinction between related constructs widely used in the applied and theoretical research: quality of life and life satisfaction. The quality of life domains and indicators stemming from this research are expected to contribute to existing quality of life theory as well as provide new hypotheses to be tested in future research. A new quality of life measurement tool is also a contribution to the field. This research also adds to the understanding of the implications of 3 distinct models and approaches to the development and management of resource communities to human wellbeing in remote locations worldwide. Results from this research also provide hypotheses for further research and investigation of the factors and circumstances that are most relevant to the improvement of quality of life in remote resource communities in Brazil and elsewhere. This study also provided evidence of some demographic factors that influence perceived quality of life in remote mining communities. Of broader application are the relationships between satisfaction with quality of life and age, gender, educational level, marital status and number of children. This research is a multidisciplinary case study undertaking which includes the triangulation of different sources of evidence. Lessons learned from its unique methodological strategy and specific methods are also expected to be of value for researchers intending to take similar approaches to future research.  1.7.  Limitations of Research  This research includes qualitative and quantitative methods. A comprehensive discussion of the methodological strategy, including the limitations of specific methods is presented in Chapter 4. One of the limitations of this research relates to the study population. Given that only current employees were research subjects, the framing does not include ex-employees or ex-residents who have left their jobs or moved away for reasons related to the quality of life in these  10  communities. Moreover, the researcher examined quality of life aspects in the 3 communities mostly through the lenses of the Vale employees, largely excluding individuals working for other companies and the unemployed population. Time constraints and limited resources to complete the research are also considered limitations. Given a limited time frame to complete Phase III of the research, which included cross-sectional data collection through questionnaire and qualitative interviews, both methods were used concurrently. The implication is that unexpected findings or new hypothesis that could be developed from questionnaire or interview results in Phase III could not be further explored in Phase III. Further investigation of hypotheses emerging from qualitative or quantitative results of this research is recommended as future research.  1.8.  Thesis Structure  This thesis is organized in 9 chapters. Figure 1.1 is a summary diagram that presents a roadmap to this thesis and is provided to guide the reader through this document. Chapter 2 aims to discuss the context in which the research was conducted. It includes a description of the historical development of the Carajás Mining Province as a whole, as well as the 3 individual case study communities—Núcleo Urbano de Carajás, Parauapebas and Canaã dos Carajás. This chapter also includes a summary of each community’s socio-economic and physical characteristics. The development of these 3 communities is directly associated with the development of the mining sector in the southeast of the State of Pará and of Vale’s Northern System mines, so general information about mining development in the region and the company is also included. Chapter 3 discusses the specific elements of the approach to this research, including a discussion of the subjective wellbeing approach. This chapter also includes a summary of a literature review on quality of life and subjective wellbeing, and discusses the most relevant and current quality of life measurement approaches, including tools that are particularly relevant to the methodological strategy of this research. Chapter 4 explains in detail the methodological strategy adopted in this research, including qualitative and quantitative methods, and discusses the specific choices of research methods. This discussion includes the details of the 3 research phases: preliminary research  11  (comprehensive literature review and quality of life domain development), quality of life model development (validation of domains and development of indicators and a questionnaire tool) and data collection (intensive data collection through interviews and a quality of life survey questionnaire). Chapter 5 presents the results from both the quantitative and qualitative analyses of the data collected for the research. The quantitative analysis refers to the data from the quality of life questionnaire, which was examined through statistical analysis. The qualitative analysis included the analysis of qualitative interviews with mineworkers and other residents in the 3 communities and participant observation notes. Chapter 6 presents a discussion of the results achieved with the qualitative and quantitative analysis of the data collected for this research. It also shows the triangulation of these results with participant observation notes taken during 3 fieldwork periods. In addition, this chapter includes a discussion of the opportunities for intervention identified with this research. This discussion is aimed at mining companies operating in remote areas, local governments, and policy makers involved in environmental and social impact assessments of mining operations in remote areas. Chapter 7 is the conclusion of this thesis and provides a concise summary of the main findings of this research. It includes an overview of how mineworkers living in the 3 case studies perceive quality of life. Chapter 8 discusses the original contributions of this research to scientific knowledge and professional practice. It includes a discussion on the local significance as well as the scope of other implications of its findings. Chapter 9 presents a discussion on opportunities for future research.  12  Figure 1.1: Thesis Roadmap  13  CHAPTER 2: CASE STUDIES This chapter discusses the context in which this thesis research was conducted, providing an introduction to the historical development of the Carajás Mining Province as well as the 3 case study communities. Since the case studies are directly associated with the Vale mining operations, an overview of Vale and its significance in the Brazilian mining context is also included in this chapter. The information in this chapter was compiled through a review of literature on the development of the southern State of Pará, particularly with regards to the Carajás Mining Province. Information in this chapter also reflects a comprehensive data-collection effort focused on statistical information available mostly through government-generated documents, such as census information and some Ministry-specific research and documentation. This data collection presented some challenges because of the limitations of the data. Specifically, because the Núcleo Urbano de Carajás is not an emancipated municipality, but part of the City of Parauapebas, statistical information specific to the Núcleo Urbano is fairly scarce. All data regarding population characteristics and services available in the community was supplied directly by Vale or through interviews with Vale personnel. Moreover, all census data obtained for Parauapebas include data from the Núcleo Urbano de Carajás. Discrepancies also exist between data available for Parauapebas and Canaã dos Carajás. Until 1993, the village of Canaã dos Carajás was part of the Parauapebas municipality; therefore, census data available until 1993 aggregates Parauapebas and Canaã’s information. Official demographic data for Canaã dos Carajás is only available starting at the 1996 census.  2.1.  The Brazilian Mining Context  The great diversity of Brazil’s landscapes and geological formations, partly attributed to the size of its territory—8.5 million km²—is one of the reasons why it has one the world’s largest mineral potentials (Departamento Nacional de Produção Mineral, 2006). Much like Canada, the mining industry in Brazil is responsible for significant development in remote regions and contributes considerably to the national economy. As in other parts of the world, the Brazilian minerals industry is experiencing substantial recent growth. From 2002 to 14  2005, the growth of the value of the Brazilian mineral production was of over 212%, from US$ 11.528.553 in 2002 to US$36.005.948 in 2005 (Departamento Nacional de Produção Mineral, 2006). In spite of the large number of mining companies operating in Brazil, the sector is quite concentrated in most relevant segments, which has high levels of integration between the mines and the mineral processing industry (Departamento Nacional de Produção Mineral, 2006). In fact, an interesting feature of the Brazilian mining industry is the dominant presence of one large diversified mining company: Vale. This company has played a significant role in the regional development of its 2 major operational centers: the Southern System (including the states of Goiás, Minas Gerais, Espírito Santo and Rio de Janeiro) and the Northern System (including the states of Pará, Maranhão, Sergipe and Bahia), where Vale has large operations and a very complex and comprehensive logistics scheme and infrastructure. In the Northern System, Vale operations include the largest iron-ore mining complex in the world. The Carajás Iron-ore Mining Complex, the Azul Manganese Mine, and the Sossego Copper Mine are all located in the south of the State of Pará. Appendix A includes a summary table with basic information, productivity and health and safety indicators for the 3 operations. The Carajás Iron-ore Mining Complex began operating in 1984. The Azul Manganese Mine has been operating since 1985 (Vale, 2007b) and is not expected to operate for more than 30 years (Pinto, 2007). Both operations are located in the Carajás Mountain range, where the Floresta Nacional de Carajás (Carajás National Park) was established in 1988, within the boundaries of the Parauapebas municipality (Coelho et al., 2002). The Sossego Copper Mine is the most recent mine in the Vale Northern System. It is also Vale’s first copper project, and it is located about 75 kilometers from the Carajás National Park, in the City of Canaã dos Carajás. The expected mine life of the Sossego Mine is about 17 (Vale, 2005). Canaã dos Carajás is also where 3 of Vale’s advanced exploration projects (copper, nickel and gold) are located (Vale, 2004). In developing the comprehensive Northern System, Vale has taken different approaches to establishing the communities that support their operations in what used to be a vastly unexplored and undeveloped Amazon forest in the south of Pará. These communities were established in a context of extreme remoteness, with a lack of infrastructure and tremendous land disputes that  15  made the State of Pará notorious for violent land and access conflicts involving Aboriginal groups, small farming settlers, large property owners, timber and other forest resources harvesters, cattle ranchers and artisanal miners (Marín, 2003; Monteiro, 2003; Coelho et al. 2002).  2.2.  The Development of the Carajás Mining Province  In the late 1960s, the world’s largest mineral province was found in the Brazilian Amazon, in the south of the State of Pará. The detection of the various mineral deposits in the area followed intense oil exploration, manganese, tin and bauxite discoveries. American multinationals Union Carbide and US Steel were exploring for manganese in the Amazon rainforest and discovered what came to be one of the largest iron-ore deposits in the world. On July 31, 1967, Brazilian geologist Breno Augusto dos Santos, who was working for Companhia Meridional de Minerações (CMM) (a subsidiary of US Steel) was part of a team of geologists who landed in a clearing at the top of the Carajás Mountain range, where the ground had a visible overlay of iron-ore (Chadwick, 2005; Santos, 2006). Shortly after, the Brazilian government granted mineral rights to the then state-owned Companhia Vale do Rio Doce (CVRD). In 1970, CMM and CVRD became associated and started the exploitation of iron-ore, manganese and gold in the traditional territory of the Xikrins do Cateté Aboriginal peoples (Prefeitura Municipal de Parauapebas, 2005). The discovery of the Mineral Province of Carajás was a milestone for the mining industry in Brazil. The estimated reserves in the Mineral Province of Carajás at the time of discovery were of 18 billion tonnes of iron-ore, 1 billion tonnes of copper, 40 million tonnes of bauxite, 60 million tonnes of manganese and 124 million tonnes of nickel (Chaves, 2004). This massive mineral discovery started an exploration rush in the Central-West and Northeast regions of the Amazon which was supported by Australian, Peruvian and American technical expertise. By the 1970s, several federal organizations had been established to facilitate and regulate exploration and mining activity in Brazil while a multitude of national and international companies were implementing very aggressive exploration programs throughout the country. The Carajás Iron-ore Project was a critical part of a large-scale federal development program that included mineral development as well as timber harvesting, agriculture, cattle ranching and the development of extensive infrastructure and logistics in a large area of the eastern Amazon, accounting for nearly 10% of Brazil’s total land mass (Chaves, 2004). This program was 16  designed in the 1980s to develop and integrate the eastern Brazilian Amazon (Coelho et al., 2002) and was planned and partly implemented by the military government of Brazil. The Carajás Iron-ore Project is one of the largest mining operations in the world. It is an integrated mail-railroad-port system, built and operated by Vale (Chadwick, 2005). It includes 5 open pits, and according to Chadwick (2005), it is expected to be in operation for over 4 centuries. The Carajás Iron-ore Project is situated in the heart of the Amazon rainforest—in the North region of Brazil—and is powered by the hydroelectric dam in Tucururí (Chaves, 2004). The Project is located between the Tocantins and Xingu rivers, at an average altitude of 650 metres above sea level (Chadwick, 2005). Figure 2.1 includes location maps that illustrate the geographical context of the case study communities.  2.3.  The Mining Company: Vale  Vale originated in 1942 as the Companhia Vale do Rio Doce, a state-owned resource development company (Kury, 1992). As of early 2008, it is the largest mining company in Latin America and the dominant force in the Brazilian minerals industry. Throughout its 60-plus years of existence, Vale has expanded its operations in Brazil from the Southeast to the Northeast and Northern regions, always facing significant challenges regarding socio-economic and environmental impacts of its operations to the over 500 communities in Brazil where it operates (Coelho et al., 2002; Instituto Ethos, 2006). Vale is present in 32 countries on 5 continents and operates in 13 of Brazil’s 26 states (Vale, 2007a, 2007b). On October 24, 2006, Vale obtained control of Canadian nickel producer Inco Limited, the second-largest producer and the number one holder of nickel reserves in the world. On January 4, 2007, Inco Limited became a wholly-owned subsidiary of Vale (Vale, 2007b).  17  SOUTH AMERICA  BRAZIL  SOUTHEAST PARÁ DETAIL: CARAJÁS AREA  Figure 2.1: Location Maps Adapted from Governo do Estado do Pará (2007)  18  Vale is the largest producer of iron-ore and pellets (mostly used for steel production) and one of the largest producers of nickel. Vale also produces copper, manganese, bauxite, alumina, aluminum, coal, potassium and kaolin, products extensively used by the global industrial sector (Vale, 2007a). In the past, the company’s Brazilian operations have produced significant amounts of gold.  Figure 2.2: Operations in Vale’s Northern System, in Relation to Case Study Communities Source: Adapted from Ribeiro, E. (2006). (Not to scale.)  To support its activities, Vale created a complex integrated logistics system and became the greatest provider of logistic services in Brazil. The company operates over 9,000 kilometers of railroad and 10 proper port terminals. Vale is also one of the country’s largest energy producers, with varying ownership of 6 hydroelectric plants. Along with its subsidiaries and joint ventures, Vale is the single biggest energy consumer in Brazil (Vale, 2007a). Figure 2.2 and Figure 2.3 illustrate the extent of Vale’s operations in the Northern System.  19  Figure 2.3: Logistics Infrastructure for Vale’s Northern System Source: Adapted from Ribeiro, E. (2006). (Not to scale.)  Vale is one of the wealthiest and largest Brazilian companies. At the time of the Carajás mineral discovery, Vale was a state-owned company, operating mostly in the Southeast of Brazil. It was privatized in the late 1990s. Vale’s total market value in 2006 was approximately US $170 billion. According to recent corporate information, the company’s third quarter gross income for 2007 was more than US $32 billion (Vale, 2007a). Vale holds more than 30% of the iron-ore transoceanic market share. The company is the biggest private investor in Brazil, responsible for around 14% of all Brazilian foreign trade (Instituto Ethos, 2006). At the end of 2006, Vale had almost 44,000 direct employees, along with 12,000 at Vale-Inco. Even though little occupational health and safety information from Vale is publicly available, the company’s 2006 Annual Report highlights continuous performance improvement. In 2003, the 20  company reported 15.1 accidents not causing days off work (per million hours worked). In 2006, this number was reduced to 8.8. The frequency of accidents causing days off work declined from 4.7 in 2003 to 1.9 per million hours worked in 2006. From 2005 to 2006, there was a reduction of 34% in the number of accidents causing days off work (Vale, 2006).  2.4.  Núcleo Urbano de Carajás: The Closed Company Town  The region of the Carajás has been explored by Vale since 1968, but 1979 marked the effective start of the development of the Carajás Iron-ore Project in the south of Pará. The Project became the primary goal of the corporate strategy (Vale, 2007a). In order to accommodate workers in the remote and isolated location, Vale built a mining camp in close proximity to the project site and began construction of its comprehensive logistical complex, including roads, rail and port facilities. The Núcleo Urbano de Carajás (“Núcleo”) is a result of the evolution of this original mining camp. The current Núcleo was built in 1986 and occupies an area of 308 hectares. The community provides all the conditions of comfort and security to accommodate a population of up to 7,000 inhabitants in 1,320 housing units. The Núcleo Urbano can be accessed by road (about 15 km from Parauapebas), or by plane. The airport is located within a national park, about 20 minutes by car from the Núcleo Urbano. The Carajás Airport is served by 2 airlines, with 5 daily weekday flights to a total of 14 destinations. The main destinations are Belém, Belo Horizonte, Brasília and Manaus. Each of the flights includes at least one stop in small towns in Pará, Tocantins and Minas Gerais 6. The Núcleo Urbano de Carajás is therefore a closed company town—access is controlled and mostly limited to residents and mine employees. Residence in the community is controlled by Vale’s local operations and is mostly a privilege and a benefit offered to higher-ranking employees. The Núcleo case study will allow for the description of a closed, controlled community established as a company town in the early 1980s, when Vale was a state-owned company and government policies supported the development of the eastern Amazon with little or no discussion of the social, economic and cultural implications of development and  6  In December 2007, other destinations were: Altamira, Araguaína, Parintins, Porto Trombetas, Santarém, Tucuruí, Ourilândia do Norte, São Félix do Xingu, Uberaba, and Uberlândia. The costs of one-way flights varied from R$281.00 (Araguaina) to R$859.00 (Uberaba) each way (CAD $140.00 to CAD $430.00).  21  urbanization. The development approach here was paternalistic: the company provided all the facilities and services and imposed a code of conduct. This case also allows for the investigation of the rationale for the use of such a model, and of the ways Vale has attempted to facilitate quality of life (addressed in this context as part of the employee retention plan).  2.4.1. Development The Núcleo Urbano de Carajás was developed in the early 1980s for Vale employees who came from diverse parts of Brazil to work for the company, generally in the Carajás Iron-ore Mining Complex, but also for the Azul Manganese Mine and Igarapé Bahia Gold Mine 7. The community has undergone virtually no changes in terms of physical layout throughout the years. It remains a small community (5,500 persons) (Correa, 2006) where employees and families live in housing provided by the company. Most services and infrastructure are also provided free of charge 8, except for transportation, telephone and Internet access (Cota, 2006).  Figure 2.4: The Núcleo Urbano de Carajás: Aerial View and Typical Housing Units © Silvana D. Costa. Photos taken in November 2006  Following construction of the Carajás Iron-ore Mine and the commencement of operations at the top of the Carajás Mountain Range, the Floresta Nacional dos Carajás (Carajás National Park) was established. The community and the large iron-ore reserves are located within this park, which was created in 1988 to “promote scientific research, sustainable exploitation of natural  7  In 2006, Igarapé Bahia was under reclamation and closure procedures. For all homes, there is a limit to how much water and electricity consumption Vale will cover, but residents are free to go above that quota and pay the service fees for the extra usage (Cota, 2006).  8  22  resources, the study and conservation of biodiversity and bring social benefits to the local communities through the rational exploitation of its resources as well as tourism” (Instituto Brasileiro do Meio Ambiente e Recursos Naturais Renováveis, 1998). Vale entered in an agreement with the Government of Brazil to assist the Instituto Brasileiro do Meio Ambiente e Recursos Naturais Renováveis (Brazilian Institute for Renewable Natural Resources—abbreviated as IBAMA), the federal environmental agency, in protecting and maintaining Carajás National Park. As a result, a large protective fence and several gates were erected around the Park, and have since also served as a means to control access to the Núcleo. Vale and IBAMA are responsible for the Park’s 1.2 million hectares of forests. Of the Carajás Iron-ore Project’s 412,000 hectares of land, only 1.6% is disturbed by the open pit mines (iron ore, manganese and, previously, gold), the mineral processing plants and the Núcleo itself. A new mining camp, the current-day Núcleo, was built very close to the Carajás Iron-ore Project site to satisfy the needs (i.e., housing, health, education, food and recreation) of the mineworkers and their families. The original mining community was decommissioned when construction of the current community was completed in the mid-1980s, and the land on which the original camp was situated was then mined. The original mining community comprised wood houses and facilities and was mostly accessed by airplane. Employees worked on a fly-in-fly-out schedule (Coelho, 2006). The Núcleo is known as the “city in the forest” because of its location on the top of a mountain in the Serra dos Carajás (Carajás Mountain Range), inside the Carajás National Park.  2.4.2. Urban Infrastructure Development of the Núcleo was planned by Vale staff and included carefully designed houses, town squares and small parks, as well as water and power plants to service the community. The community is carefully maintained, and includes regular street cleaning, garbage collection, recycling and security services. It is also serviced by regular bus and passenger van transportation to and from Parauapebas. Commercial venues in the Núcleo include grocery and general stores, 3 banks and a few bars and restaurants. Other public facilities in the Núcleo are the Zoobotânico—the Carajás Zoo,  23  which includes a botanical garden—a cinema/theatre and a recreation centre, the Docenorte Clube.  Figure 2.5: The Núcleo Urbano: Banks and Recreation Centre © Silvana D. Costa. Photos taken in November 2006  2.4.3. Population According to Kury (1992), by 1989, over 7,000 people, including Vale employees, contractors and their families, lived in the Núcleo. Current Vale employees living in the Núcleo today work at either the Carajás Iron-ore Mining Complex or the Azul Manganese Mine. In 2006, the Núcleo was home to about 5,500 people, including Vale employees and their families (Jatay, 2006). The population also includes some key staff not directly employed by the mining company, such as the school principal, Yutaka Takeda hospital staff and Docenorte recreation centre program managers and staff, among others. Demographically, the Núcleo is quite different from a regular community in Pará. Residents work directly or indirectly for Vale, or are employees’ family members. The population is therefore made up of mostly working-age individuals (i.e., ranging from about 18 to 65 years old) and their children. Even though there are no demographic data available for the Núcleo (and resident turnover seems to be quite high), one can assume that this community has virtually no elderly permanent residents.  24  2.4.4. Education The main educational facility in the Núcleo Urbano de Carajás is the community school (K–12). This school is a branch of the renowned Pitágoras School, which has branches in 5 other Brazilian states. The education offered by the Núcleo’s community school is considered to be of a high quality, and this is recognized in evaluation reports prepared by the Pitágoras School Corporation, which annually compares performance amongst its various branches in Brazil. In August 2005, there were a total of 1,303 students attending all branches of the Pitágoras School: 130 students in the pre-school, 927 in the elementary school and 246 in the high school. In addition, the School had 106 staff. The Núcleo also has a small day-care centre (available for employees’ children free of charge) and a private English language school for children and adults (available for a fee).  2.4.5. Health Information regarding the health of the population of the Núcleo Urbano de Carajás is not readily available through regular monitoring surveys implemented by health agencies. Municipal-level surveys aggregate data from the Núcleo with data from Parauapebas. However, some information from the private hospital (Yutaka Takeda) located in the community is available (Table 2.1). Table 2.1: Yutaka Takeda Hospital, 2004 Data Incidences and Procedures  Total  Tests  Monthly Average  Visits to Doctors  20,700  Endoscopies  31  Occupational Health–Related Visits*  75,000  Newborn Exams  11  Overnight Visits  1,000  Laboratory Exams  Emergency Room Visits  13,600  X-Rays  992  Ultrasounds  160  Electrocardiograms  373  Surgeries  700  10,434  *  Relates to the workers’ health; most of these visits involved evaluation of fitness to work and deal with work hazards and requirements. This data does not include accidents. Source: (Jatay, 2006). Note: Numbers are rounded.  25  The Yutaka Takeda Hospital has been in operation in the Núcleo since 1986. The hospital includes 50 beds, 12 examination offices and operating rooms for surgeries of average levels of complexity. In 2004, the hospital was the workplace of 13 doctors and 145 other employees. The following doctor specialties are available at the hospital: anesthesia, cardiology, general surgery, family medicine, gynecology, obstetrics, occupational health, orthopedy and pediatry. On a scheduled basis (once every month or 2), the following specialties are also available: neurology, otorrinolaringology and psychiatry (Jatay, 2006). The hospital also includes a medical laboratory and a blood bank, as well as services of radiology, endoscopy, ultrasound, electrocardiograms, phonoaudiology, nutrition, physiotherapy, blood pressure monitoring and Holter 24h heart monitoring tests.  Figure 2.6: Kuarup Brazilian Culture Centre and Yutaka Takeda Hospital, Núcleo Urbano de Carajás © Silvana D. Costa. Photos taken in November 2006  2.4.6. Culture, Recreation and Entertainment The Núcleo was designed to provide for all the needs of Vale employees and their families. Therefore, some cultural, recreational and entertainment options are available in the Núcleo, such as a zoo and botanical garden complex, a cinema/theatre, a large recreation centre and a few bars and restaurants. The Núcleo also includes a small facility called the Kuarup Brazilian Culture Centre, which is dedicated to teaching traditional Brazilian dance.  26  2.5.  Parauapebas: The Gate Development Community  A direct result of the Carajás mineral discoveries made at the end of the 1960s, Parauapebas originated from a small village built by Vale adjacent to the Carajás National Park gate (Secretaria Executiva de Estado de Planejamento, 2007b). When the Park boundaries were set, this area was largely untouched and used mostly for the harvest of castanha (Brazil nuts). Almost simultaneously, the region experienced a rapid increase in clear-cut timber extraction and the beginning of the cattle raising activity by large capitalist enterprises that take advantage of federal government policies and fiscal incentives for Northern development (Martins, 1984). Parauapebas is located in the south of Pará, 547 km from Belém, the capital of the State. The city is situated in the Parauapebas River valley and shares borders with the cities of Marabá (north), Curionópolis (east), Canaã dos Carajás and Água Azul do Norte (south), and São Félix do Xingu (west) (Prefeitura Municipal de Parauapebas, 2005; Instituto Brasileiro de Geografia e Estatística, 2007). Parauapebas is located in a relatively remote area. The city can be accessed by highway (Pará 275), aircraft (Carajás Airport) and train (company-controlled railroad). The airport is located at the top of the Carajás Mountain Range, within Carajás National Park, and mostly serves the workforce transportation needs.  Figure 2.7: Parauapebas Streetscapes © Silvana D. Costa. Photos taken in November 2006  27  The community of Parauapebas is different from the 2 other communities in this study because here Vale has made only marginal efforts to facilitate the quality of life of resident employees. The city has developed adjacent to the Carajás National Park gate and access to the mining operations and the Núcleo Urbano de Carajás. Parauapebas is also important because with continuous expansion of the Carajás iron-ore mine and the need to increase the mine workforce, Vale needs to find workforce accommodation solutions. There is very limited opportunity to increase the population density of the Núcleo, and Parauapebas has become an undesirable place for higher-ranking or well-educated employees, who usually aspire to higher standards of living conditions and quality of life. A community that grew spontaneously without addressing environmental or health concerns, the Parauapebas case allows for the examination of the implications of the “hands-off” approach to mining community development and planning for the quality of life of resident mineworkers.  2.5.1. Development In the early 1980s, Vale built a small village—the Vila de Parauapebas—to accommodate contractors, municipal services and the workforce for the development of the Pará-Maranhão railway project (Prefeitura Municipal de Parauapebas, 2005). The original Parauapebas village was designed for 5,000 people and was built in about 1.5 years. It included treated water, sewer treatment, electricity, a hospital, a police station, an administration building and a school. News of this development attracted large numbers of migrants to the area, and by the mid-1980s, the population had already reached 20,000 (Prefeitura Municipal de Parauapebas, 2005). The village of Parauapebas experienced rapid and uncontrolled growth as a consequence of high expectations of employment and business development in the region (Prefeitura Municipal de Parauapebas 2005). According to the municipal government, between 1981 and 2004, the population of Parauapebas increased tenfold (Prefeitura Municipal de Parauapebas, 2005). Today, migrants continue to flock to Parauapebas from across the country, but they mostly come from the North and Northeastern states, where high unemployment rates and environmental challenges have created large poverty pockets. A municipality since 1988, Parauapebas has a highly transient population and very low economic, social and environmental standards, with several neighborhoods being established through squatter settlements. In 2007, it was estimated  28  that more than 100,000 people lived in this community (Instituto Brasileiro de Geografia e Estatística 2007).  2.5.2. Population Between 2000 and 2007, Parauapebas’ annual growth rate was 4.71%—almost double the annual growth rate of the State of Pará as a whole (2.28% from 2000–2007) (Instituto Brasileiro de Geografia e Estatística, 2007), and almost triple the annual growth rate of Brazil (1.67% from 2001–2005) (Ministério da Saúde, 2007). Urbanization of Parauapebas has also occurred at a phenomenal rate—82.8% in 2000 (Table 2.2). According to the 1991 census carried out by the Instituto Brasileiro de Geografia e Estatística (the Brazilian Institute of Geography and Statistics—abbreviated as IBGE), the total Parauapebas population was 53,335 inhabitants distributed over 17,653.76 km² (including Canaã of the Carajás, which was considered part of Parauapebas until 1994). This averages to 3.02 inhabitants per square kilometre. In the 2000 census, Parauapebas’ population size soared to 71,568, and density increased to over 10 inhabitants per square kilometre—this even though Canaã dos Carajás was no longer included in the Parauapebas figures. The 2006 census accounted for 95,225 inhabitants, and a population density of over 13 inhabitants per square kilometre (Instituto Brasileiro de Geografia e Estatística, 2007). The April 2007 municipal population count was 133,261 (Instituto Brasileiro de Geografia e Estatística, 2007) 9. Table 2.2: Indicators of Growth and Urbanization, Pará and Parauapebas Annual Growth, 2000–2007 (%)  Urbanization Rate*, 2000 (%)  Pará  2.28  66.55  Parauapebas  4.71  82.80  Jurisdiction  *Proportion of urban population in relation to total population. Source: Instituto Brasileiro de Geografia e Estatística (IBGE) 2007  The majority of migrants looking for opportunities in Parauapebas arrive by road or train. The Agência Nacional de Transportes Terrestres (National Agency of Terrestrial Transportation)  9  However, methodological changes in the data collection have resulted in the population of the Contestado village (about 30,000 people) being counted as part of Parauapebas. This village had traditionally been considered part of the municipality of Marabá (Santos, 2007). But even if the Contestado population is not included,, Parauapebas still have a population is of over 100,000 people.  29  estimates that between 500 and 600 people come to the city each month (by bus or train), while only about 200 people leave Parauapebas monthly (Agência Nacional de Transportes Terrestres, 2006). According to IBGE, over 34% of the Parauapebas population is originally from the State of Maranhão—one of the poorest states in Brazil, where the Vale train makes its final stop to deliver mineral products at the Itaqui Port—while about 40% were born in the State of Pará. According to the most recent census (2006), the population of Parauapebas is rather balanced in relation to gender. The male-female ratio has steadily declined throughout the years (Table 2.3), reflecting a population diversification since its origins as a mining village with a predominantly male population. Table 2.3: Population Gender Distribution, Parauapebas Population  1991  1996  2000  2006  Male  27,982  33,001  36,269  48,258  Female  25,353  30,562  35,299  46,967  Gender Ratio*  110.37  107.98  102.91  102.74  *Number of men per 100 women. 2006 gender ratio calculated by author. Source: Secretaria Executiva de Estado de Planejamento, Orçamento e Finanças (SEPOF)/Diretoria de Estudos, Pesquisas e Informacoes Socio-economicas (DIEPI)/Gerência de base de Dados Estatísticos do Estado (GEDE) 2007. Data source: IBGE.  Parauapebas is a city of young people. More than 40% of the population is between 20 and 49 years of age, and over 66.5% is between 10 and 49 years of age. The community has a smaller percentage of older (i.e., over 50 years of age) people and higher percentages of people under 10 years of age than the State of Pará as a whole (Table 2.4). Table 2.4: Age Distribution, Pará and Parauapebas (2007)  Jurisdiction Pará Parauapebas  Less Than 1 – 9 Years %* %* 1 Year Old Old  10 – 19 Years Old  %*  20 – 49 Years Old  %*  50 Years Old or More  %*  177,596 2.45 1,629,035 22.47 1,743,795 24.06 2,894,304 39.93 804,430 11.10 2,697  2.73  23,346  23.63  23,424  23.71  42,352  42.87  6,977  7.06  *Percentage of total population. Source: Banco de dados do Sistema Único de Saúde (DATASUS)/IBGE 2007  30  2.5.3. Urban Infrastructure According to IBGE (2007), in 2000 there were 26,000 housing units in Parauapebas. In the urban areas, 92% of the streets had lighting, and 70% were paved. Of the total urban housing units, 70.83% were being used by owners or owners’ relatives and 20.14% by renters, 7.66% were company-provided housing and 1.33% were going through an acquisition process. The urban services provided in Parauapebas have increased throughout the years (Table 2.5). However, these services tend to be distributed in the more established neighborhoods of Parauapebas, in the Núcleo Urbano and closer to the downtown area and Vale’s gate. The urban areas that have experienced the most growth are situated along the railway, where the infrastructure is seriously lacking (Santos, 2006). There are only 5 small banks (all of which are branches of large banking institutions) in Parauapebas (excluding the Núcleo Urbano) (Prefeitura Municipal de Parauapebas, 2005). This number is extremely low relative to the size of the city’s population, and this in turn is a source of considerable disappointment and frustration for residents, particularly on paydays, when the demand for banking services is significantly higher than usual. There are 4 mobile phone operators —TIM, Vivo, Oi and Amazônia Cellular—4 TV stations, and 1 official FM radio station in Parauapebas. According to the National Agency of Telecommunications (Anatel), in 2006 there were about 45,000 registered private cellular phones in Parauapebas, which means 1 cellular phone for every 2 inhabitants—almost 2 phones per housing unit (Santos, 2006). Cable Internet is available in a limited area of the downtown core of Parauapebas, although Internet provision has been expanding in the last 5 years. In April 2007, Parauapebas celebrated the opening of the first computer lab in a public school; this lab houses 20 computers connected to the Internet and services 700 students of the Josias Leão School (Secretaria de Educação de Parauapebas, 2007).  31  Table 2.5: Household Access to Selected Goods and Services, Parauapebas Household Goods and Services  1991  2000  Housing Units Total 10,962  %  Housing Units Total 16,433  %  Garbage Collection  5,459  49.80  13,598  82,75  Electricity  6,881  62.13  15,714  95.63  Phone line  452  4.12  2,429  14.78  Radio  6,429  58.65  9,435  57.42  Television  5,073  46,28  13,161  80.09  Computer  -  -  958  5.83  1,080  9.29  2,192  13.34  Private Vehicle  Source: Secretaria Executiva de Estado de Planejamento, Orçamento e Finanças (SEPOF)/Diretoria de Estudos, Pesquisas e Informacoes Socio-economicas (DIEPI)/Gerencia de base de Dados Estatísticos do Estado (GEDE) 2007. Data source: IBGE. .  2.5.4. Economy Parauapebas’ economy is heavily dependent on large-scale mining (Prefeitura de Parauapebas, 2005; da Silva, 1998). Most of the infrastructure of the city, including roads, was developed with funds generated by taxes collected from the mining operations in the Carajás Mountain Range. Except for forestry farming and cattle ranching, most industries in Parauapebas are directly or indirectly related to the mining sector. In 2006, mining activity (ore extraction and transformation) was responsible for 70% of Parauapebas’ Gross Domestic Product (GDP), followed by the services sector (25%), and by the farming sector (5%). Before its emancipation in 1998, Parauapebas was considered part of the municipality of Marabá and thus did not have direct access to economic benefits (royalties) from the mining development in the Carajás Mountain Range. In 2006, Parauapebas was the 6th richest municipality in Pará, after Belém, Barcarena, Tucuruí, Ananindeua and Marabá, and it has the 7th highest income per capita in the state (Instituto Brasileiro de Geografia e Estatística, 2007). From 1991 to 1995, mining activity contributed to 23%–28% of Parauapebas’ municipal revenue (da Silva, 1998; Barreto, 2001). In Brazil, mining royalties are shared by the municipal (65%), state (23%) and federal (12%) governments (da Silva, 1998). As an example of the magnitude of the mining royalties Parauapebas earns every year, in 2000, the municipal government received  32  US $12,051,604 from iron-ore mining alone (Barreto, 2001). Because the mineral production in the Carajás Iron-ore complex has steadily increased since 2002, it is certain that mining royalty revenues have also increased significantly. Despite the economic growth and vibrancy of the city, the percentage of people living under the poverty line in Parauapebas is alarmingly high: 43.55% of the total population lives below the poverty line 10, and the poverty rate of the rural population is almost 50% (Table 2.6). When compared to the State of Pará as a whole, however, these rates indicate that Parauapebas is, in general, less poor than the general Pará population. The difference is most noticeable on the rural population poverty data. Table 2.6: Poverty Indicators, Pará and Parauapebas (2000)  Pará Parauapebas  Population Below the Poverty Line (1)  Population  Jurisdiction  Population Below the Poverty Line (%)  Total  Urban  Rural  Total  Urban  Rural  Total  Urban  Rural  6,195,974  4,122,105  2,073,869  3,246,882  1,784,787  1,462,095  52.40  43.30  70.50  71,571  59,261  12,310  31,166  25,017  6,149  43.55  42.21  49.95  (1) Poverty Line: Proportion of people who live in a family with a monthly income of less than one-half of the minimum wage per capita. Source: IBGE/SEPOF (2006)  Incomes in Parauapebas (Table 2.7) reflect the scale of the mining activity in Carajás and its direct and indirect impacts on employment and wages. Even though Vale has tended in the last few years to reduce its workforce and increase its use of contractors, these jobs still remain in the region and often with comparable wages. The mineral activity has facilitated the increased the income per capita in Parauapebas and the city’s contribution to state revenues. Excluding Belém, the state capital, Parauapebas was the city that most contributed to Pará’s revenues in 1995 (da Silva, 1998).  10  The term “poverty line” is used here as it is defined by IBGE: The proportion of people who live in a family with a monthly income of less than one-half of the minimum wage per capita. In 2000, the minimum wage in Brazil was US $84.00 per month (considering 1 Brazilian Real = 0.56 American Dollar), or US $2.8 a day.  33  Table 2.7: Income Indicators, Parauapebas (2000 Data)  % income from work  % income from government transfers  % of population with > 50% of income from government transfers  297.23  -  -  -  Pará  168.59  68.93  12.22  10.38  Parauapebas  221.48  73.39  6.73  5.02  Income per capita (monthly in Brazilian Reais)  Brazil  Jurisdictions  Source: United Nations Development Program (UNDP).New Atlas of Human Development in Brazil (2000)  There is clearly income inequality in Parauapebas, which has intensified throughout the years: the Gini Index 11 for Parauapebas in 1991 was 0.58, and it increased by 15% to 0.67 in 2000.  2.5.5. Human Development According to the most recent United Nations (UN) Human Development Index (HDI) 12 (based on 2000 census data), Parauapebas’ HDI grew 12.8% from 0.657 in 1991 to 0.74 in 2000, putting the city in the 11th position among the 143 cities in the State of Pará and in 2,137th position (out of a total of 5,507 cities evaluated) in all of Brazil 13 (Santos, 2006). Table 2.8 shows the Parauapebas HDI in relation to other jurisdictions.  11  The Gini coefficient is a measure of inequality of income distribution or inequality of wealth distribution. It is defined as a ratio with values between 0 and 1: 0 corresponds to perfect equality and 1 corresponds to perfect inequality (where one person has all the income, while everyone else has zero income). A low Gini coefficient indicates more equal income or wealth distribution, while a high Gini coefficient indicates more unequal distribution (United Nations Development Programme, 2000). 12  Launched in 1990, the Human Development Index (HDI) is the measure of life expectancy, literacy, education, and standard of living for countries worldwide. The HDI measures the average achievements in three basic dimensions of human development: long and healthy life, as measured by life expectancy at birth, knowledge and education, as measured by the adult literacy rate and the combined primary, secondary, and tertiary gross enrollment ratio (with one-third weighting) and decent standard of living, as measured by the log of gross domestic product per capita at purchasing power parity in US$. In the 2005, the country with highest HDI was Iceland (0.968). Canada was third in the world, with an HDI of 0.961(United Nations Development Programme, 2000; United Nations Development Programme, 2007).  34  Table 2.8: Human Development Index (HDI) Indicators in the Carajás Region 1991 Municipality  2000  HDI  Education  Income  Life Expectancy  HDI  Education  Income  Life Expectancy  -  -  -  -  0.699  0.792  0.627  0.679  Eldorado dos Carajás  0.542  0.472  0.556  0.598  0.663  0.725  0.551  0.712  Parauapebas  0.657  0.712  0.661  0.598  0.741  0.844  0.674  0.704  Canaã dos Carajás  Source: United Nations Development Program (UNDP). New Atlas of Human Development in Brazil (2000). 1991 data is not available for Canaã dos Carajás.  2.5.6. Education According to a 2006 school census carried out by the Brazilian Ministry of Education, Parauapebas has about 45,000 students (kindergarten to high school); almost 35,000 (78%) of these students were enrolled in municipal schools, while 22% were attending private schools. About 1,500 individuals were registered as university students in approximately 30 courses offered by 10 different (public and private) educational institutions in Parauapebas (Ministério da Educação, 2006). The 2006 illiteracy rate (of people 15 years of age or older) of Brazil’s North region was 12.7%, and that of the government of the State of Pará was 14.1%. At 12.23% (about 8,000 people), Parauapebas’ 2006 illiteracy rate was lower than the region’s average, but better than the state average, and was an improvement over the 2000 illiteracy rate, which was 16.3% (Instituto Brasileiro de Geografia e Estatística, 2007).  2.5.7. Health Thousands of people from neighboring cities come to Parauapebas in search of health care. Because Parauapebas is the largest city between Marabá (north) and Redenção (south), it has more than half of the medical facilities in the North region (Santos, 2006). In 2005, Parauapebas had a total of 33 health care facilities, of which 19 were public and 14 were private facilities (Instituto Brasileiro de Geografia e Estatística, 2007).  35  In 2003, Parauapebas had 131 hospital beds, but only 28 of these were located in public hospitals. According to the State of Pará, this number had not changed since 1999 (Secretaria Executiva de Estado de Planejamento, 2007b). The 2003 average number of beds—1.61 per 1,000 people (Secretaria Executiva de Estado de Planejamento, 2007b)—is extremely low if compared to the World Health Organization (WHO) minimum standard of 2.5 beds per 1,000 people 14 (World Health Organization, 2007). The situation is in fact worse because these beds are serving not only Parauapebas’ population, but also people living in neighboring towns with fewer health care facilities, such as Curionópolis and Canaã dos Carajás. The health care facilities available are under-resourced regarding medical staff and equipment and are clearly not sufficient for the growing population of Parauapebas. Signs of the poor health care in this city are the poor performance on indicators of health displayed in Table 2.9, which includes data for Pará. Table 2.9: Health Indicators, Parauapebas 2006 Jurisdiction  Infant Mortality Rate*  General Mortality Rate*  Pará  22.22  3.87  Parauapebas  30.06  4.60  Source: DATASUS/IBGE 2007 * Mortality rate expressed as number of deaths per 1,000 inhabitants per year  Considering that its economic performance in general is better than the state average, the health care challenges in Parauapebas can be seen as something of a paradox.  2.5.8. Culture, Recreation and Entertainment The cultural, recreational and entertainment options in Parauapebas are fairly limited. The only theatre in Parauapebas is actually located in the Núcleo Urbano de Carajás. This theatre is accessed made through the Vale/Carajás National Park gate with the purchase of a movie ticket. The Zoobotânico Park (zoo and botanical garden) is also located in the Núcleo. Parauapebas recreation centers—City Park Club, Juvenil Club and Docenorte Club—are all privately owned. The best-equipped of these is the Docenorte recreation centre, which is located at the Núcleo Urbano de Carajás.  14  In Brazil, the rate was 2.6 in 2002. Canada’s rate was 3.6 (World Health Organization, 2007)  36  Figure 2.8: Parauapebas: Open Sewage. © Silvana D. Costa. Photo taken in November 2006  One of most traditional recreation centers, the Asfep Club, was located next to the Carajás National Park gate (also Vale’s security access gate) but recently closed down to make way for a private company parking lot. Public spaces for leisure and recreation include only a few sports fields spread throughout the city: the Rosenão Stadium, the Poliesportivo Gymnasium, the Centro de Desenvolvimento Cultural (CDC) (cultural centre), and 8 small municipal squares. Sports events in Parauapebas are limited to school competitions and inter-municipal school games. Very little cultural activity occurs in Parauapebas. The organization that most supports cultural programs in Parauapebas is the Vale do Rio Doce Foundation. Other organizations that promote cultural events are the KUARUP Dance Group (Brazilian dance classes located in the Núcleo Urbano) and folklore dance groups Palmares I and II (Noel, 2006; Santos, 2006).  2.6.  Canaã dos Carajás: The Integrated Community  The third case study community, Canaã dos Carajás (Canaã) has its origins in the 1982 federal government resettlement program that brought migrant families, mostly from the States of Maranhão, Tocantins and Goiás to develop small scale agriculture in southern Pará (Lima,  37  2004). The Carajás Settlement Project was established to ease the violent land conflicts in the Araguaia–Tocantins region (Coelho et al., 2002; Martins, 1984). The youngest Vale mine, the Sossego Copper Mine, has been in operation since 2004. During this time, Canaã has experienced rapid growth and both physical and social transformation. Similar to Parauapebas, Canaã is a boomtown that receives migrants from across the country, most of whom are looking for employment in the nearby mines. From a small village of just over 1,500 families in the early 1980s, Canaã has grown to be home to more than 20,000 people in 2007 (Instituto Brasileiro de Geografia e Estatística, 2007). With advanced copper and nickel projects in the area and requirements from the Brazilian government associated with their development permits, Vale has engaged in a large-scale investment program in Canaã dos Carajás. In partnership with the municipal government, the company has invested in basic infrastructure, health, educational and cultural facilities and programs in Canaã. The company has also built houses for employees in 4 different neighborhoods, spatially dispersed throughout the city (Pavan, 2005). Thus Canaã can be designated the “integrated community”.  Figure 2.9: Canaã dos Carajás: Vale-Built Homes and an Existing Wood-Frame Home © Silvana D. Costa. Photos taken in August 2006  The case of Canaã dos Carajás allows for the investigation of Vale’s revised approach to mineworkers’ settlements and accommodations about 25 years after the establishment of the closed-company town of the Núcleo Urbano de Carajás and the hands-off approach to the development of Parauapebas. This case provides the opportunity to illustrate a contemporary 38  approach, under contemporary environmental legislation and mine permitting process, which includes a more formalized relationship with the local municipal government. This case study also allows for the investigation of the implications of these factors on the quality of life of employees who reside in Canaã dos Carajás. Canaã dos Carajás is located 760 km from Belém, the capital of the State of Pará. Canaã is in a fairly remote area, located 120 km from Marabá and 75 km from Parauapebas. The city shares borders with Parauapebas (north and west), Curionópolis and Sapucaia (east), Água Azul do Norte (south), and São Félix do Xingu (west) (Secretaria Executiva de Estado de Planejamento, 2007a). Canaã dos Carajás includes a town centre as well as 7 rural villages: Mozartinópolis, Planalto, Ouro Verde, Serra Dourada, Bom Jesus, Feitosa and Nova Esperança (Lima, 2004). Most of the rural villages are accessible only by dirt roads that often flood during the rainy season. Vale has built a highway to facilitate the development of the Sossego Copper Mine, and this highway provides the only road access to and from the city. The closest airports are located in the Núcleo Urbano de Carajás and in Marabá. Bus service is only available in Parauapebas, Núcleo Urbano de Carajás or Marabá. Train service is only available in Parauapebas.  2.6.1. Development Canaã dos Carajás originated from an agricultural settlement set by a federal government program. The Executive Group of Lands of the Araguaia and Tocantins (GETAT), established the Carajás Settlement Project in the Southeastern region of Pará, in 1982 (Lessa, 2007). The GETAT’s purpose was to ease the land conflicts in the Araguaia –Tocantins region, mainly in the highly disputed area known as Bico do Papagaio (“Parrot’s Beak”) (Martins, 1984). In 3 years, GETAT settled 1,551 families in the area that was known as a Regional Development Centre (CEDERE) in Canaã dos Carajás (Lessa, 2007). These families were mostly from the States of Maranhão, Tocantins and Goiás, and were settled in this area to develop small scale agriculture activity (Lima, 2004). However, by 1985, only 816 families had received formal ownership of their lands. Economic activity was limited to agriculture, mainly the growth and harvest of corn, beans and rice. Also in  39  1985, the settlement activities were halted and GETAT was disbanded for political reasons. In October of 1994, the CEDERE was separated from Parauapebas and became a municipality— Canaã of the Carajás. The city’s name was chosen through popular vote. The name “Canaã” has biblical origins and refers to the "Promised Land" (Lessa, 2007). In 2002, Vale started work on the Sossego Copper project, located on Sossego Mountain, a few kilometers away from Canaã’s town centre. A large migratory process was prompted by the company’s announcement of expectations to employ around 3,000 people. This population growth was accompanied by a growth in Canaã’s goods and services sector and a critical need to improve existing public services and urban infrastructure.  2.6.2. Population Between 2000 and 2007, Canaã’s population more than doubled (Figure 2.1). According to the 1996 census carried out by the Instituto Brasileiro de Geografia e Estatística (IBGE), Canaã was home to 11,139 inhabitants distributed over 3,161,50 km², which averages to 3.52 inhabitants per square kilometre. The 2000 census indicated a decline in land area and population, probably reflecting a downturn in the agriculture sector and migration to other parts of the state, such as neighboring Parauapebas, where the economy was booming due to mining activity. Since the establishment of Vale’s mineral development projects in 2002 and the opening of the Sossego Copper Mine in 2004, the population of Canaã has soared—the 2006 census counted 13,870 inhabitants, and the population density had increased to 4.41 inhabitants per square kilometre (Instituto Brasileiro de Geografia e Estatística, 2007). As of April 2007, the current municipal population count sits at 23,707, which would imply on a population density of 7.53 inhabitants per square kilometre (Instituto Brasileiro de Geografia e Estatística, 2007).  40  Canaã dos Carajas: Population Growth 1996-2007 25,000  23,700  Population  20,000 15,000  13,870  11,139 10,922  10,000 5,000  07  06  20  05  20  04  20  03  20  02  20  01  20  00  20  99  20  98  19  97  19  19  19  Years  96  0  Figure 2.10: Population Growth in Canaã dos Carajás Note: 1996, 2000 and 2006 figures are based on census data, while the 2007 population estimate is from a municipal government count. Source: IBGE, Population Count 2007 and Population Estimates 2007  Table 2.10 exhibits the annual growth rate for Canaã, and the State of Pará as a whole. At 3.94%, Canaã’s annual growth rate from 2000–2007 was smaller than Parauapebas’ growth rate of 4.71 for the same period (Table 2.2), but still quite higher than that of the State of Pará. Urbanization in Canaã (Table 2.10) has not been as intense as it has been in Parauapebas (Table 2.2). However, it is expected that Canaã’s urbanization rate will show a significant increase in the 2006 census results, reflecting the significant population growth. Table 2.10: Indicators of Growth and Urbanization, Pará and Canaã dos Carajás Annual Growth, 2000–2007 (%)  Urbanization Rate, 2000 (%)  Pará  2.28  66.55  Canaã dos Carajás  3.94  35.93  Jurisdiction  Source: IBGE 2007  In 2006, more than 60% of Canaã’s population lived in rural areas (Table 2.11). This was almost double than the percentage for the State of Pará as a whole, and almost 4 times the percentage in Parauapebas. 41  Table 2.11: Urban and Rural Populations, Pará, Parauapebas and Canaã dos Carajás (2006) Jurisdiction Pará Parauapebas Canaã dos Carajás  Urban Population  Urban % of Population  Rural Population  Rural % of Population  4,811,305 81,805 5,143  63.37 82.80 35.93  2,437,855 16,991 9,171  33.63 17.20 64.07  Source: DATASUS/IBGE 2007  Similar to Parauapebas (Table 2.3), the male population in Canaã is slightly larger than the female population (Table 2.12). Statistical data shows, however, that unlike Parauapebas, this balance changed significantly between 1996 and 2000 and has since remained relatively stable. The recent increase in the proportion of males in Canaã is certainly a reflection of the mineral development boom brought about by the Sossego Copper project. Such upturns in mineral development activity in an area generally attract a larger percentage of males for its workforce.  Table 2.12: Population Gender Distribution, Canaã dos Carajás Population  1996  2000  2006  Male  5,783  5,871  7,694  Female  5,356  5,051  6,620  Gender Ratio*  107.97  116.23  116.22  *Number of men per 100 women; 2006 gender ratio calculated by author. Source: Secretaria Executiva de Estado de Planejamento, Orcamento e Financas (SEPOF)/Diretoria de Estudos, Pesquisas e Informacoes Socio-economicas (DIEPI)/ Gerencia de base de Dados Estatísticos do Estado (GEDE) 2007. Data source: IBGE.  In terms of its population’s age distribution, Canaã dos Carajás is more similar to the State of Pará as a whole (Table 2.13) than it is to Parauapebas (Table 2.4). Canaã is a youthful community, with more than 65% of its population aged 10–49, and over 40% of its population aged 20–49.  42  Table 2.13: Age Distribution, Pará and Canaã dos Carajás (2007)  Jurisdiction Pará Canaã dos Carajás  Less than 1 year old 177,596 309  1–9 % % years old  10–19 years old  %  20–49 years old  %  50 years old or more  %  2.45 1,629,035 22.47 1,743,795 24.06 2,894,304 39.93 804,430 11.10 2.16  3,145  21.97  3,506  24.49  5,863  40.96  1491  10.42  Source: DATASUS/IBGE 2007  2.6.3. Urban Infrastructure According to IBGE, in 2000 there were 2.521 housing units in Canaã dos Carajás, not including the 100 housing units built for the Sossego Copper Mine workforce (Secretaria Executiva de Estado de Planejamento, 2007a). It is evident that with the development of the Sossego Copper Mine the urban infrastructure and services available in Canaã have improved; however, census data is only available for the year 2000 (Table 2.14). It is also clear that most improvements to Canaã’s infrastructure were accomplished in the neighborhoods where the Vale housing units have been built. Even though these have been distributed around the city in different neighborhoods, infrastructure improvements by 2006 had been implemented only in these particular areas and at the city’s main point of access, where the Sossego Mine Camp and some of the company offices are located. For example, before the Sossego Mine was developed, there were no paved roads in Canaã. In 2006 there were several roads paved with asphalt, including the main access road to city and the streets that lead to company housing units. Table 2.14 depicts the deficient infrastructure and services in Canaã in 2000. Unfortunately, data for 2006 was not available at the time that research for this thesis was completed. However, it is notable that a significant problem of access to electricity and garbage collection existed in Canaã before the Sossego Mine began operating. It is expected that these numbers, as well as access to consumer goods such as radio and televisions, have improved since 2000.  43  Table 2.14: Permanent Housing Units and Selected Goods and Services, Canaã dos Carajás (2000) Household Goods and Services  Housing Units*  (%)**  628  24.91  Electricity  1,569  62.24  Radio  1,425  56.53  Television  1,130  44.82  Computer  44  1.75  Private Vehicle  286  11.34  Garbage Collection  *Out of 2,521 total housing units. **Percentage of housing units which use service/own household good. Source: IBGE/SEPOF 2007  Communications systems in Canaã are fairly limited. Internet access is only available by radio, and there are very few establishments where Internet access is available for a fee. In 2001, the first 547 phone lines—for both residential and commercial uses—were installed (Secretaria Executiva de Estado de Planejamento, 2007a). Mobile phone services are only available through 2 service companies. There is only one post office and one banking institution in the city (not including a bank branch at the Sossego Mine).  2.6.4. Economy Since the 1980s, the economic activity in Canaã has diversified slowly to include cattle ranching, timber extraction and goods and services. Because of the disbandment of GETAT, cattle ranching (mostly for meat production) had developed as the main economic activity before mineral exploration and mining were established with the Sossego Project in the late 1990s. Table 2.15 summarizes data on the economic activity in Canaã. Even though the data in Table 2.15 does not depict exceptional economic performance from 1999–2000, the area has since experienced a significant boom. For example, the amount of taxes received annually has grown incredibly since 1997—from about one million Reais to eleven million Reais in 2006 15 (Secretaria Executiva de Estado de Planejamento, 2007a).  15  Total taxes, including mining royalties.  44  Figure 2.11: Canaã dos Carajás City Centre © Silvana D. Costa. Photos taken in November 2006  Table 2.15: Economic Indicators, Canaã dos Carajás (2000) Indicators  2000 Numbers  Population aged 10 years or older  8,286  Economically active population (EAP)*  4,031  Working population  3,824  Activity rate (percentage of EAP that are working)  48.65%  Inactivity rate (percentage of unemployment in relation to EAP)  5.14%  Income per capita  $167,46* monthly  Income from work  66.19%  Income from government transfers  6.16%  Population with > 50% of income from government transfers  4.53%  Source: IBGE/SEPOF 2007 * Reais (Brazilian Currency). In August 2006, 1 R$ ~ CAD$ 0.50  From 1997–2001, the gross domestic product (GDP) value in Canaã remained roughly the same. With mining development, the GDP value started to grow, and in 2004, when the Sossego Mine opened, the GDP value in Canaã dos Carajás spiked significantly (Figure 2.2.). In 2004, Canaã’s GDP per capita was 21,480.04, the third largest in the State of Pará (Secretaria Executiva de Estado de Planejamento, 2007a). 45  Canaã dos Carajas Gross Domestic Product (GDP) 1997–2004 Value in R$ 1,000  300,000 250,000 200,000 150,000 100,000 50,000 0 1997 1998 1999 2000 2001 2002 2003 2004 Years  Figure 2.12: Canaã dos Carajás GDP (1997–2004) Source: (Secretaria Executiva de Estado de Planejamento, 2007a)  Even though the percentage of people living in poverty in Canaã is smaller than that of the State of Pará, in 2000, almost half of the population in Canaã lived under the poverty line. The situation was even worse in the rural areas, where more the half of the population were living below the poverty line (Table 2.16). Poverty in the urban areas of Canaã dos Carajás in 2000 (Table 2.16) was more acute than in the urban areas of Parauapebas, where the percentage of people living under the poverty line was higher in the rural areas (Table 2.6). Moreover, the inequality in Canaã dos Carajás has in fact increased, since the Gini Inequality Index for Canaã increased from 0.52 to 0.62—a 19.2% increase in 9 years. Table 2.16: Poverty Indicators, Pará and Canaã dos Carajás (2000)  Total Pará Canaã dos Carajás  Population Below the Poverty Line (1)  Population  Jurisdiction  Urban  Rural  6,195,974 4,122,105 2,073,869  10,923  3,925  6,998  Total 3,246,882  5,303  Urban  Rural  Population Below the Poverty Line (%) Total Urban Rural  1,784,787 1,462,095 52.40  1,591  3,712  48.55  43.30  70.50  40.54  53.0  (1) Poverty line: Proportion of people that live in a family with monthly income of less than a ½ minimum wage per capita (US $2.8 dollars a day). Source: IBGE/SEPOF 2007  46  2.6.5. Human Development In 2000, the Human development Index (HDI) for Canaã dos Carajás was 0.699, well below the HDI for the State of Pará (0.723) and for Brazil (0.796).  2.6.6. Education A 2006 school census carried out by the Brazilian Ministry of Education shows that the Canaã dos Carajás student population (kindergarten to elementary school) has more than doubled since 2000, from 3,504 to 7,630 students. In 2006, 6,804 (89%) students were registered in municipal or state schools, while 826 (11%) were attending private schools. However, there are no municipal or state high schools in Canaã dos Carajás. In 2000, the literacy rate in Canaã dos Carajás was 81.5%, lower than Parauapebas (83.7%), the state average (83.2%) and Brazil’s average (86.37%) (United Nations Development Programme, 2000).  2.6.7. Health In 2005, there were only 9 health care facilities in Canaã dos Carajás, of which 6 were public and 3 were private (Instituto Brasileiro de Geografia e Estatística, 2007). The first and only municipal hospital in Canaã opened in the early 2000s (Secretaria Executiva de Estado de Planejamento, 2007a), and until 2006 was still the only one in town. A new private hospital, built by Vale, was under construction in 2006, and was designed to include some publicly accessible hospital beds (Pavan, 2005). In 2003, Canaã dos Carajás still had only 26 hospital beds in the municipal hospital. The 2003 average number of beds—2.14 per 1,000 people (Table 2.17)—and the number of health units per 1,000 people were better than Parauapebas but still below the World Health Organization (WHO) minimum standard of 2.5 beds per 1,000 people (World Health Organization, 2007). The situation in 2006 was particularly dramatic since the population had increased to 13,870 without any increase in capacity at local hospitals, which implies on a rate of 1.87, well below the WHO standard of 2.5 beds per 1,000 inhabitants.  47  Table 2.17: Health Care Indicators, Canaã dos Carajás and Parauapebas (2003) Cities  Units/10,000 Persons  Beds/1,000 Persons  Canaã dos Carajás  3.5  2.14  Parauapebas  3.4  1.61  Source: DATASUS/IBGE 2007  2.6.8. Culture, Recreation and Entertainment Vale built the only cultural facility available in Canaã, the Casa da Cultura, through the Itakyra Association 16 (independent but heavily funded by Vale). This centre includes an open-air theatre, arts and crafts workshops and a small library. In 2006, a company-owned recreation centre was being developed next to the Sossego Camp, at the entrance of the city. When completed, this centre will include sports fields, a gym and swimming pools. This facility will be a private enterprise available to the general public for a membership fee.  Figure 2.13: Canaã dos Carajás. Casa da Cultura © Silvana D. Costa. Photo taken in November 2006  Entertainment options are also few and far between in Canaã. The first public municipal square 17 was only recently built by Vale, and there are no public sports fields or arenas (except for the  16  The Itakyra Association is a not-for-profit organization, a partnership of the Vale Foundation and the Canaã dos Carajás municipal government. The Foundation manages investments in education, health and leisure/culture (Ethos Institute, 2006).  48  occasional open lot where soccer is played). Internet access is quite limited in private homes and available at only a few Internet cafés.  2.7.  Reflecting on the Case Study Communities  The above descriptions of the 3 case study communities reveal a clear distinction among them. Table 2.18 summarizes each of the 3 case study communities in relation to its mining community model and approach, associated mining operation and the parameters discussed in this chapter. Information in this chapter (highlighted in Table 2.18) reveals that the Núcleo Urbano de Carajás scenario is quite different from that of Parauapebas or Canaã dos Carajás. On-site observations support the obvious conclusion that the Núcleo provides better and more comfortable physical living conditions, services and resources for its residents. Despite the clear differences between the models and approaches to mining community development used in Canaã and Parauapebas, and despite the limited data available for Canaã, it seems that Canaã is facing challenges that are somewhat similar to those faced by Parauapebas since its inception in the early 1980s. With the new integrated approach taken in Canaã dos Carajás, Vale is clearly more involved with the physical and social development of the town. Moreover, local government and social organizations are operating under a new paradigm, one of sustainable development and environmental management. In Parauapebas, however, mining activity does not seem to be contributing to sustainable development in the region, but it is providing for improved quality of life for thousands of people. As observed in the historical development of Parauapebas, the most significant challenge for a booming small mining community is to be prepared to quickly adapt to the rapid and intense growth that accompanies increased industry activity. In Canaã, this adaptation does not seem to be happening at the pace necessary. Statistical measures presented in this chapter might indicate dramatic positive changes during this early phase of development but can be deceiving because of the limited ability to adequately account for the existing population.  17  In Brazil, city or neighbourhood plazas or squares are culturally important and historically have been the public space where residents meet and greet and where major civic or community events take place.  49  Table 2.18: Case Studies Human Development Index (health, income and education)  Culture, Recreation and Entertainme nt  Associated Operations  Community Model  Approach  Núcleo (built by the company in the 1980s).  ƒ Carajás Iron-ore Complex Azul Manganes e Mine.  ƒ A typical company town ƒ The entire infrastructure is owned by the company  Paternalistic— company controls physical and social environment  ƒ Fairly constant at 5,500 ƒ Mostly mineworkers and their families  ƒ Built/maintained by company ƒ High quality ƒ Relatively good health and education services and facilities  Activities revolve around the mines and supporting commercial services  Data included in Parauapebas’ numbers  Limited facilities and options, but certainly best in the south of Pará  Parauapebas (started as supporting village for the mining operations in the early 1980s, emancipated municipality in 1998).  ƒ Carajás Iron-ore Complex, ƒ Azul Manganese Mine, and ƒ Sossego Copper Mine.  ƒ Gate development ƒ Chaotic development which has started at the gate to the mine /park boundaries  Company built first facilities and infrastructure but has since taken a “hands off’ approach, with little intervention on social or physical environment  ƒ Population boom since its beginnings ƒ From 53,335 in 1988 to 133,261 in 2007  ƒ Clearly inadequate for population size ƒ Housing shortages ƒ Inadequate health and education services and facilities  Severely dependent on mining. Some services, agriculture and ranching  Grew 12.8% from 0.657 in 1991 to 0.74 in 2000, putting the city on the 11th position among the 143 cities in the state  Limited facilities and options  Canaã (started as a farming settlement in 1982, emancipated municipality in 1994)  ƒ Sossego Copper Mine.  ƒ Integrated community ƒ Company is extending infrastructure and building facilities and homes throughout the city (not in one specific neighborhood)  Collaborative, integrated development (extension) of independent existing community: company partnerships with foundations and government  ƒ Population boom since the beginning of the Sossego Project ƒ From 11,139 in 1995 to 23,000 in 2006  ƒ In development phase. ƒ Most isolated community case study ƒ Housing shortages ƒ Inadequate health and education services and facilities  Increasingly dependent on mining. Some family farms and large scale ranching  In 2000, the Human development Index (HDI) for Canaã dos Carajás was 0.699, below the HDI for the state (0.723) and Brazil (0.796)  Limited facilities and options  Case Study  Population (2006)  Urban Infrastructure, Education and Health Facilities  Economy  50  CHAPTER 3: APPROACH TO RESEARCH This chapter discusses the approach to this research. It includes a summary of a literature review on Quality of Life and Subjective Wellbeing, and introduces and discusses the most relevant and most current QOL measurement studies, including tools that are particularly relevant to the methodology used in this research.  3.1.  Quality of Life  The term Quality of Life (QOL) does not have an unambiguous and universally accepted definition; it has been used inconsistently and with diverse connotations (Cummins et al., 2002). Often, the term has been closely associated with the concept of wellbeing (Andrews and Withey, 1976; Diener, 1984; Glatzer, 1991; Eckermann, 2000; Noronha, 2001), happiness (Glatzer, 1991; Musschenga, 1997; Steel and Ones, 2002) and life satisfaction (Vittersø et al., 2002). For some authors, the term overlaps but is not synonymous with other terms, such as level of living, ways of life, and life satisfaction (Andrews and Withey, 1976). In the western world, quality of life is a terminal societal value. Different from instrumental values, terminal values are life goals that are not means to an end, but end-states such as happiness, a comfortable life, peace, equality, freedom and wisdom (Sirgy, 2002). Instrumental values are those that need to be achieved if the terminal goals are to be experienced. Examples of instrumental values are hard work, respect for others, co-operation, loyalty, commitment and respect for the environment. Quality of life, as a terminal value, is often captured by commonlanguage terms such as “happiness”, “a meaningful life” and “a comfortable life” (Sirgy, 2002). As a societal terminal value, governments and civil society organizations strive to attain a better quality of life for all. Recently, with the popularization of the notion of corporate social responsibility, corporations have also felt compelled to contribute, through their business, to the overall quality of life of the communities where they operate. One of the key distinctions that have been made is the difference between the quality of life concept in health-related and non-health related fields. In the health-related fields, QOL has often been defined in a one-dimensional way (i.e., in terms of the absence of illness and symptoms of the disease under consideration, or health status) (Power et al., 2002), but 51  exceptions exist, such as the World Health Organization (WHO) quality of life definition, which includes the culture and context which influence an individual's perception of health (American Thoracic Society, 2002). More recently, there has been wide agreement that quality of life is a multidimensional concept (e.g., Coffman and Adamek, 1999; Bowling and Gabriel, 2004) and needs to be understood in the context of culture and value systems in which individuals live and in relation to their goals, expectations, standards and concerns (Cummins et al., 2002; Power et al., 2002). Quality of life is here understood as the external conditions of one’s life—that is, what is available to individuals at a specific location and at a specific time of their lives. When making cognitive evaluations of one’s quality of life, immediate and current conditions are examined.  3.2.  Measuring Quality of Life  Quality of life (QOL) studies emerged during the social indicators movement of the late 1960s and early 1970s (Andrews and Withey, 1976), largely in response to dissatisfaction with traditional economic methods of measuring wellbeing. It was recognized that economic prosperity was often accompanied by degradation of the environment, increased poverty and inequality, and other social problems among some population groups (Beckley and Burkosky, 1999). Therefore economic measures such as growth in Gross Domestic Product (GDP) are often no longer considered enough for evaluating a nation’s, region’s or city’s quality of life (Australian Centre on Quality of Life, 2005). Historically, QOL measurement studies have been used for a number of purposes, including descriptive reporting of the state of a society or community; for the analysis of social change; for setting goals and priorities; and for developing a system of social accounts (Eyles 1994). According to Beckley and Burkosky (1999), there are 3 general types of quality of life studies: a) those conducted on a national or cross-national level to measure and compare social progress or “development” of nations; b) those that focus on the quality of life in local communities; and c) those that focus on the more subjective, individual view of quality of life.  52  The majority of the national, cross-national and community quality of life studies have trusted external judgments to define and measure quality of life.  3.2.1. Internal versus External Measurements of Quality of Life The lack of clarity of the quality of life (QOL) concept and its inconsistent use is reflected in the variety of ways researchers have attempted to measure QOL. A critical discussion in QOL measurement, closely related to the QOL definition discussion, involves the objective versus the subjective way to define and measure QOL (Veenhoven, 2006). Externally measured indicators are usually referred to as “objective”, although this work—including measures such as Human Development Index (HDI) 18, the Index of Social Progress (ISP) 19 and the Gross National Product (GDP)—inevitably requires subjective judgments by the analyst on the meaning and content of wellbeing (Sirgy et al., 2006). The term “objective measures” therefore refers to the degree a life meets explicit standards of what is “good” as assessed externally by an impartial outsider. The term “subjective measures” involves self-appraisals based on implicit, internally defined criteria (Veenhoven, 2006). Traditionally, QOL studies have almost exclusively relied on external measures (e.g., the environment) and the external circumstances (based on outsiders’ judgments) of an individual’s life, such as pollution, traffic congestion, access to health care, quality of housing and aesthetic surroundings. External judgments might be perceived to be more manageable by decision makers when compared to attitudes or feelings (e.g., feelings of love or belonging), which can only be measured internally and at an individual level. However, it has been widely argued that subjective measures are necessary to properly evaluate a society, and that they add substantially to the external measures that are often favored by policy makers (Diener and Suh, 1997), given that external circumstances “form only a limited aspect of the sum of satisfactions that make life worthwhile” (Dalkey, 1972, p. 9 ).  18  The United Nations Development Project HDI has reported on the development and progress of the world’s nations since 1990, combining indicators of life expectancy, educational attainment, and income (United Nations Development Programme, 2004). 19  The ISP is a framework for measuring national and international social development and examines the sectors of education, status of women, demography, political participation, cultural diversity, and welfare effort (Estes, 1998; Beckley and Burkosky, 1999).  53  In the early 1970s, Dalkey (1972) proposed that a fundamental question is whether external measures “constitute a major share of an individual’s wellbeing, or whether they are dominated by factors such as a sense of achievement, love and affection, perceived freedom and so on” (Dalkey, 1972, p.9). Dalkey defended a deeper and more holistic evaluation of quality of life as perceived and experienced by individuals. This proposition was supported by Bateson (1972), who strongly felt that quality of life is not adequately defined only by physical variables, and suggested that what human beings care about the most is not “episodes or things as such, but the patterns and setting of their personal relationships —how they stand in love, belonging, hate, respect, responsibility, dependency, trust, and other abstract but nonetheless real relations” (Bateson 1972 in Andrews and Withey, 1976, p. 5). Diener and Suh (1997) point out that in the last decades, researchers have suggested different approaches to measuring quality of life, such as through social indicators (e.g., health and levels of crime) (Sirgy and Cornwell, 2002; Requena, 2003), subjective wellbeing measures (e.g., individuals’ perceptions of their lives and societies) (Proshansky and Fabian, 1986; Sirgy, 1998; Zumbo and Michalos, 2000; Vittersø, 2004), and economic indices (Organization for Economic Co-operation and Development, 1971; Szalai and Andrews, 1980). According to Diener and Suh, these alternative measures appraise 3 philosophical approaches to wellbeing. These are based, respectively, on normative ideals, on subjective experiences, and on individuals’ abilities to select goods and services respectively. Diener and Suh argue that social indicators and subjective wellbeing measures are crucial to evaluating a society, and significantly add to the widespread economic indicators that are currently favored by policy makers and analysts. More recently, a significant number of studies involving QOL measurement efforts have focused on the subjective characteristics (Cummins et al., 2002). The rationale for the focus on subjectivity is twofold: a) objective and subjective measurement have been observed to be poorly related (Cummins et al., 2002; Rojas, 2006); and b) clarity and consistency within each measure tends to be lost when both objective and subjective measures are combined (Power et al., 2002). Even though there is an overload of external measures that relate to quality of life and wellbeing, the same is not true for subjective measures. Although interest in subjective wellbeing has grown over time and resulted in significant academic investigations in the last few  54  decades, there are only a few subjective quality of life measures that have been proved to be rigorous, comprehensive and systematic (Cummins et al., 2002). It is therefore the premise of this research that it is worthwhile to measure quality of life using rigorous, comprehensive and systematic subjective measurements that reflect how people feel about their quality of life and the material conditions under which they live. Even though one cannot deny the importance of externally measurable factors such as income and access to health services and drinking water, perceptions can often matter more than realities, and how people relate to one another can matter more than their material circumstances and assets. Based on this assumption, the literature reviewed for this research includes a relatively recent and innovative body of knowledge built on subjective measures of quality of life often described in the literature as “Subjective Wellbeing”.  3.3.  Subjective Wellbeing  The approach to this research, including the definition of the methodology and analytical framework are based on the theory of subjective wellbeing (SWB). SWB has a relatively short history in Psychology. Starting in the 1960s, the research on SWB within the field of Psychology slowly gained momentum. In the 1970s, large-scale national QOL surveys, which included assessments of SWB, were conducted in North America (e.g., Andrews and Withey, 1976; Campbell et al., 1976). The 1990s witnessed a significant expansion in SWB research (Sirgy et al., 2006). Subjective wellbeing—commonly referred to as “subjective quality of life” or ”perceived quality of life”—has become an increasingly popular research subject worldwide, and clear trends regarding this concept have emerged from the literature (Bramston, 2002). The literature demonstrates an evolution of the knowledge in the quality of life field resulting from the consideration of the implications of affect and cognition in measuring subjective quality of life, achieved through continuous research and theory development. As it has been widely acknowledged since Campbell et al. (1976), SWB includes both affect and cognition aspects (Cummins et al., 2002). The definition of cognitive versus affective is very important in quality of life research, and many researchers have made this differentiation. While  55  cognitive aspects involve conscious, judgmental processes that are dependent on a comparison of one’s circumstances with what is thought to be an “appropriate standard” (Diener et al., 1985), affective aspects involve immediate emotional reactions. For example, while happiness and sadness are affective concepts (positive affect and negative affect, respectively), satisfaction or dissatisfaction are cognitive concepts (Sirgy, 2002). The definition of SWB that has emerged over time is that of a broad, multi-faceted construct, which includes the affective and cognitive components that are gradually identified by early research. Diener et al. (1984, p. 277) offered the definition of SWB as “a broad category of phenomena that includes people’s emotional responses, domain satisfactions, and global judgments of life satisfaction”. According to Cummins et al. (2003), the single most important subjective wellbeing feature is that it is always positive—it is normal for individuals to feel good about themselves. This assertion is based on the Homeostasis 20 theory of subjective wellbeing, which suggests that each person has a “set-point” for subjective wellbeing that is internally maintained and defended. This internally regulated set-point is believed to be genetically determined (International Wellbeing Group, 2004). It has been established through extensive research that the normal level of individual set-point variation about 60–90% and that the provision of personal resources (e.g., money, material belongings, personal relationships) does not normally increase the set-point on a long-term basis, although such resources can strengthen personal defenses against negative experiences (International Wellbeing Group, 2004). For individuals suffering from homeostatic defeat (when homeostasis fails and no longer controls the SWB level), the provision of additional resources such as wealth and personal relationships may allow the individual to regain control of his/her wellbeing, causing it to rise until the set-point is achieved. Low or reduced levels of personal resources (e.g., poverty, low income, problems with personal relationships or the absence of a partner) can weaken or defeat subjective wellbeing homeostasis. When personal challenges such as stress, sadness or pain exceed personal resources, homeostasis is defeated, and subjective wellbeing decreases to levels below its normal 60–90% range (International Wellbeing Group, 2004).  20  In Psychology, homeostasis is a state of psychological equilibrium obtained when tension or a drive has been reduced or eliminated (Random House, 2006).  56  The second most important feature of subjective wellbeing is its stability, which has been demonstrated by studies using 2 types of data: the mean scores from population surveys and data obtained from individuals. Studies conducted by Cummins (1995) first demonstrated the strength of the SWB stability and consequently the predictability of population mean scores 21. When individual measures of satisfaction with SWB are used, the standard deviation is much larger, but is also very consistent. Psychologists—particularly social, personality, and developmental psychologists—and quality of life researchers often refer to the subjective aspects of quality of life as happiness, life satisfaction, subjective wellbeing, and perceived quality of life (Sirgy, 2002). It is critical, however, to pause here to discuss the differentiation between happiness, life satisfaction and subjective quality of life. All 3 terms are often used interchangeably with subjective wellbeing, and imply a subjective and internally defined concept. However, the bulk of research and literature on wellbeing suggests that these terms are not necessarily synonymous. Even though “happiness” and “life satisfaction” are quite often used interchangeably in the literature and frequently defined as subjective wellbeing, “subjective quality of life” is here conceptualized as peoples’ satisfaction with their quality of life—their perception of the external conditions of their lives; of what is available to them at a specific location and at a specific time of life. One’s satisfaction with one’s quality of life is thus conceptually different from one’s satisfaction with one’s life. When making cognitive evaluations of one’s quality of life, immediate and current conditions are examined, while one’s life is usually evaluated considering past events and accomplishments, current conditions, and often, future possibilities and opportunities. Perceived quality of life or subjective quality of life therefore is one the factors that impact one’s overall life satisfaction, but certainly not the only factor (Sirgy, 2002).  21  The first of these studies combined data from population surveys performed in different Western countries and were highly diverse, having been conducted by different researchers, using different scales of measurement, at different times over the decades 1970 to 1990. The disparate results were combined by converting all data to the standardized 0 – 100 range (Cummins, 1995).  57  3.4. Measuring Subjective Wellbeing 3.4.1. Health-Focused versus Holistic Measurements According to Cummins et al. (2002), a great number of studies of subjective wellbeing have generated over 600 measurement instruments, many of which were developed for specific population subgroups. These instruments have been largely designed for relatively disadvantaged groups, such as people with a medical condition, low income or a congenital condition (Cummins et al., 2002, Michalos, 2004). Consequently, some instruments have a deficit orientation, such that a high score indicates a relative lack of disability or condition, rather than quality of life (Cummins et al., 2002). Examples of these instruments are abundant in the medical literature examined, and are usually described as “Health-related Quality of Life” instruments. Depending on the scale chosen, the measures involve factors such as disease symptoms, patients’ perceptions of their health, functional status and standard of care, and are often combined in one single scale. As a result, these instruments are unsuitable for use with the healthy population. A well-established, reliable health-focused quality of life instrument is the SF-36 Health Survey (Jordan-Marsh, 2002; Roberts and Meletiche, 2001), which includes 36 health-related questions. The SF-36 is a multipurpose measure as opposed to instruments that target a specific group defined by age, disease, or treatment group (Ware, 2007). This research examined key instruments for self-reported subjective quality of life that can be used with the general population, and are therefore defined here as “holistic” 22 (Cummins et al., 2002). Table 2.1 includes details on some of self-reported subjective quality of life instruments, including health-focused and holistic tools.  22  Most holistic scales devised for use with the general population cannot be used with all population subgroups, such as people with cognitive impairment and children. These are important limitations since it means that the quality of life experienced by minority groups cannot be norm-referenced back to the general population (Cummins et al., 2002).  58  Table 3.1: Recent Quality of Life Measurement Instruments Instrument EuroQOL (EQ5D), (1990)  SF-36 Health Scale, (1988)  WHOQOL (World Health Organization Quality of Life), (1995)  WHOQOL –100, (1998) (short form of WHOQOL)  Domains mobility self-care usual activities pain or discomfort anxiety or depression Physical Health: ƒ general health ƒ bodily pain ƒ physical functioning ƒ role-physical ƒ Mental Health: ƒ social functioning ƒ vitality ƒ role-physical ƒ role-emotional ƒ physical ƒ psychological ƒ level of independence ƒ social relationships ƒ environment ƒ spirituality/religion/ personal beliefs ƒ ƒ ƒ ƒ  Same as above.  # of Items 15  Scale  Test Locations  Comments  Visual analogue scale of satisfaction (similar to a thermometer), from 0 to 100  England, The Netherlands and Sweden.  Focuses on health-related quality of life.  36  Multiple-choice scales as well as yes-no questions  United States of America, Sweden, United Kingdom and other 9 countries.  Most SF-36 items have their roots in instruments that have been in use since the 1970s and 1980s.  236  Likert scale 1–5 of satisfaction  Pilot study carried out in 15 field centres in Africa, Asia, Europe and Australia.  Focuses on health-related quality of life, but includes other areas besides physical, mental and social, such as environment and spiritual/religious aspects.  100  Likert scale 1–5 of satisfaction  Analysis carried out in 15 field centres in Africa, Asia, Europe and Australia.  The WHOQOL –100 is a refinement of the WHOQOL; a shortened scale.  59  Table 3.1: Recent Quality of Life Measurement Instruments (CONTINUED) Instrument  Domains  # of Items  Scale  Test Locations  Comments  WHOQOLBREF, (1998) (short form of WHOQOL – 100)  ƒ ƒ ƒ ƒ  physical psychological social relationships environment  24  Likert scale 1–5 of satisfaction  Analysis carried out in 15 field centres in Africa, Asia, Europe and Australia.  It is a further shortened form of the WHOQOL-100, with summary scores for 4 domains rather than detailed scores at the facet (sub-domain level).  Comprehensive Quality of Life Scale (ComQOL), (1997)  ƒ ƒ ƒ ƒ ƒ ƒ ƒ  material wellbeing health productivity relationships safety community emotional wellbeing  7  Several non-scale questions, scales of importance (5-point) and satisfaction (7-point)  Australia.  Following investigation of validity and reliability of the ComQol instrument, changes were made in domain determination, metrics, and the response scale format, resulting in the International Wellbeing Index (IWI).  Personal Wellbeing Index (PWI), (2001)  ƒ ƒ ƒ ƒ ƒ  standard of living health achievement in life safety personal relationships future security community connectedness  7  Likert 0–10 scale of satisfaction  Australia.  The PWI is applied along with the National Wellbeing Index (NWI) to form the International Wellbeing Index (IWI).  life expectations living conditions life accomplishments  5  Likert 1–7 scale of agreement  41 countries in Africa, Latin America, North America, Asia and Europe.  This tool measures satisfaction with “life”, which is conceptually different from satisfaction with quality of life.  ƒ ƒ Satisfaction with Life Scale (SWLS), (1985)  ƒ ƒ ƒ  60  3.4.2. Self-Reported Levels of Satisfaction Subjective wellbeing refers to wellbeing as declared by an individual; it is a self-reported measure based on a person’s answer to either one question or several questions about their wellbeing (Rojas, 2006). An important concept that has been widely accepted is that the cognitive component of subjective quality of life involves some form of internal comparison process (with past conditions, other people, etc), and as a consequence, it is generally recognized that the cognitive component of subjective quality of life can be effectively measured through questions of “satisfaction” (Cummins, 2002b). Levels of satisfaction can be measured through qualitative methods such as interviews or focus groups. Generally, however, studies of subjective quality of life have developed instruments for the measurement of self-reported satisfaction levels using psychometrics—the branch of Psychology that deals with the design, administration, and interpretation of quantitative tests for the measurement of psychological variables. Psychometric tools are quantitative measurement tools mostly used for the measurement of psychological variables such as intelligence, aptitude, and personality traits. These tools are widely used in Psychology, education, and the social sciences in general (The American Heritage Dictionary of English Language, n.d.). Psychometric scale development includes comprehensive processes of design and testing of questions and response scales for cognitive problems 23, and the testing of reliability and validity.  3.4.3. Reflective versus Formative Indicators A relevant observation for the development of psychometric instruments to measure quality of life is the distinction between reflective and formative quality of life indicators and quality of life domains. Reflective indicators are measures designed to capture the concept of quality of life directly, while formative indicators capture the concept of quality of life indirectly through other concepts  23  Cognitive problems are those with comprehension or understanding of the questions, which implied that some respondents might not in fact understand what the concept means or different interpretations, were possible. If cognitive problems exist, it is difficult to guarantee that all respondents answered the question proposed or their interpretation of the question.  61  believed to play a significant role in the formation of this concept. Reflective indicators of subjective wellbeing, such as “Generally speaking, are you a happy person?", aim at capturing the behavioral phenomenon in a holistic and global manner (Sirgy, 2002). Examples of how reflective indicators are used in Subjective Wellbeing (SWB) measurements are the Measure of Life Satisfaction (Spreitzer and Snyder, 1974) and the Satisfaction with Life Scale (Diener et al, 1985). In these instruments, reflective questions are asked or statements are made and respondents rate satisfaction or happiness with an item or agreement with the statement. Table 3.2 presents examples of some of such questions and statements. As described by Sirgy (2002), formative indicators aim at capturing subjective wellbeing by tapping its “causes”. For example, the Quality of Life Questionnaire (Greenley et al., 1997) is supported by the theory that overall life satisfaction can be understood by measuring and aggregating levels of satisfaction with 7 life domains: financial life, leisure life, social life, health and health care. Other examples of how formative indicators are used in SWB measurements are the Quality of Life Index survey (Ferrans and Powers, 1985), the Quality of Life Inventory (Frisch, 1993), the ComQol (Cummins et al. 1994), and the Personal Wellbeing Index (Cummins, 2002b). In these instruments, formative questions are asked or statements are made and respondents rate their satisfaction with an item or agreement with the statement. While reflective indicators are holistic, formative indicators are diagnostic—they provide an understanding of the causes of the levels of satisfaction. These causes are therefore critical to the definition and operationalization of what is being measured (Sirgy, 2002). Instruments using a formative approach tend to include several items, often called quality of life domains or dimensions. Table 3.3 exhibits a few instruments identified by Sirgy (2002) in a comprehensive review of quality of life instruments. The Australian Centre on Quality of Life (ACQOL) 24 database of quality of life instruments, which includes over 600 instruments, was also used to complement information in Table 3.3.  24  Located at the Deakin University, Melbourne, Australia (http://acqol.deakin.edu.au/instruments/instrument.php)  62  Table 3.2: Examples of SWB Measurements Using Reflective Approach Instruments  Questions  Scale  Measure of Life Satisfaction (for the elderly) (Spreitzer and Snyder, 1974)  “Taking things all together, how would you say things are these days—would you say that you are very happy, pretty happy, or not too happy?  One (1) single item, scores from 1–3: 1) very happy 2) pretty happy 3) not too happy  Delighted –Terrible (D-T) Scale (Andrew and Withey, 1976)  “How do you feel about your life as a whole?  Coded scale: delighted, pleased, mostly satisfied, about equally satisfied and dissatisfied, mostly dissatisfied, unhappy and terrible.  Happy Person (OHP) Scale (Veenhoven, 1974)  "Generally speaking are you a happy person?"  Open line scale, later coded in 7 categories, ranging from “very unhappy” (1) to “very happy” (7).  Satisfaction with Life one Leads (O-SLL) (Part of Eurobarometer surveys, in all EU member states, since 1973)  "On the whole how satisfied are you with the life you lead?"  Scale of satisfaction: 4) very satisfied 3) fairly satisfied 2) not very satisfied 1) not at all satisfied  Satisfaction with Life Scale (Diener et al., 1985)  “Below are 5 statements with which you may agree or disagree. Using the 1–7 scale below, indicate your agreement with each item by placing the appropriate number in the line preceding that item. Please be open and honest in your responding.”  7-point Likert scale of agreement: 7) Strongly agree 6) Agree, 5) Slightly agree, 4) Neither agree nor disagree, 3) Slightly disagree, 2) Disagree, and 1) Strongly disagree  Items: ƒ In most ways my life is close to ideal ƒ The conditions of my life are excellent ƒ I am satisfied with my life ƒ So far I have gotten the important things I want in life ƒ If I could live my life over, I would change almost nothing  This research adopted a combination of reflective and formative indicators as part of its methodology—more specifically, in the form of a quality of life questionnaire. Further discussion on the questionnaire design in presented in Chapter 4.  63  Table 3.3: SWB Measurements Using Formative Approach Instruments  Domains  Scale  The Quality-of-life Index (Ferrans and Powers, 1985)  14 in total: ƒ Your relationship with your spouse ƒ Your friends ƒ Your standard of living ƒ Your ability to meet nonfinancial family responsibilities ƒ Your usefulness to others ƒ Amount of non-job stress or worries in your life ƒ Your financial independence ƒ Your leisure time activities ƒ Your achievements and personal goals ƒ Your happiness in general ƒ Your health ƒ Size of the city in which you live ƒ Your religious life ƒ Your family’s happiness  Consists of 2 sections of 32 items each. One section measures satisfaction, and the other measures importance. 6-point scale of satisfaction, from “very dissatisfied" to “very satisfied”.  The Quality-of-life Inventory (Frisch, 1993)  16 in total: ƒ Health ƒ Self-esteem ƒ Goals and values ƒ Money ƒ Work ƒ Play ƒ Learning ƒ Creativity ƒ Helping ƒ Love ƒ Friends ƒ Children ƒ Relatives ƒ Home ƒ Neighborhood ƒ Community  3-point rating scale for importance and 6-point rating scale for satisfaction. Overall life satisfaction is calculated as the sum of satisfactions with the 16 life domains.  The needs hierarchy measure of life satisfaction (Sirgy et al., 1995)  4 in total (based on Maslow’s need-hierarchy theory): ƒ Survival needs ƒ Social needs ƒ Ego needs ƒ Self-actualization needs  “How much is there now?” ”How much should there be?” Answer scales from 1 to 7. The overall score is calculated by taking the absolute difference score between the 2 questions.  64  Table 3.3: SWB Measurements Using Formative Approach (CONTINUED) Instruments  Domains  Scale  The Quality of Life Questionnaire (Greenley et al., 1997)  7 in total: ƒ Living situation ƒ Finances ƒ Leisure ƒ Family ƒ Social life ƒ Health ƒ Access to medical care  24 items on 7-point scales.  Comprehensive Quality of Life Scale (ComQol), (Cummins, 1993)  7 in total: ƒ Material wellbeing ƒ Health ƒ Productivity ƒ Intimacy ƒ Safety ƒ Place in community ƒ Emotional wellbeing  Several non-scale questions, scales of importance (5-point) and satisfaction (7-point).  Personal Wellbeing Index (International Wellbeing Group, 2004)  7 in total: ƒ Standard of living ƒ Health ƒ Achievement in life ƒ Safety ƒ Personal relationships ƒ Future security ƒ Feeling part of the community  Likert 0–10 point scale of satisfaction: “How satisfied are you with…?” 0 = extremely dissatisfied 5 = not satisfied or dissatisfied 10 = extremely satisfied  3.4.4. Quality of Life Domains A widely accepted concept is that life satisfaction can be divided into a number of domains representing areas of life experience, and that satisfaction with each domain, in aggregate, reflects overall life satisfaction (Campbell et al., 1976; Diener, 1994). The notion of general quality of life being defined by domains has also been broadly employed and validated in quality of life studies in Canada and abroad (Michalos, 1985; Cummins, 1996; Michalos, 1997; Hagerty et al., 2001; Sirgy et al., 2001; Bernhard et al., 2004; Hsieh, 2004). Quality of life (QOL) domains can usually be determined through a logical process of “deconstructing” the general goal of quality of life. In subjective quality of life research, these domains would represent areas of life experience, and satisfaction with each domain, in aggregate, reflects overall life satisfaction (Campbell et al., 1976; Diener, 1994). Examples of domains used in subjective quality of life studies are presented in Table 3.1 and Table 3.3.  65  Advanced investigations of domains can be accomplished by further breaking those domains down into measurable elements—which have been defined as facets (e.g., WHOQOL 1995) and/or indicators (e.g., Smith Criteria 1973). An early example of a study that used objective measures of quality of life, QOL domains and indicators can be found in Smith (1973). Using an urban social geography perspective, Smith identified 6 major domains of QOL: economic status, environment, health, education, social disorganization, and participation and equality (Table 3.4). Smith further subdivided these domains into more detailed aspects and finally developed 48 QOL indicators. In a comprehensive literature review, Beesley and Russwurm (1989) draw from a broad variety of sources to identify a number of domains common to the majority of QOL studies. The literature that they referenced included the United Nations’ “level of living”, Drewnoski’s development of “living index”, Smith’s “criteria”, Harvey’s “list of needs”, Hagerstrand’s notion on “liveability”, and Shulman’s study on quality of life medium-sized Canadian cities. The criteria/domains Beesley and Russwurm identified included education, leisure, health, employment, transportation, social environment, security, physical environment, and social opportunity/participation. Table 3.4: Domains in Smith’s Criteria (1973) Economic Status ƒ ƒ ƒ  Income Employment Welfare  Environment ƒ ƒ ƒ ƒ  Housing Sewage Air pollution Open space  Health ƒ ƒ  General mortality Chronic diseases  Education ƒ  Duration  Social Disorganization ƒ ƒ ƒ ƒ ƒ  Personal pathologies Family breakdown Overcrowding Public order and safety Delinquency  Participation and Equality ƒ ƒ  Democratic participation Equality  Source: Knox (1987 p. 142)  However, the manner in which QOL domains are characterized has generated a variety of opinions. This problem has recently become less controversial, and several authors currently agree on the character of some central domains of quality of life, such as health, economic wealth, and social relationships. The consideration of these domains is important in this literature review because the approach for the methodology of this research is to examine these  66  domains from a subjective perspective. Therefore, domains examined and considered in the methodology and conceptual model for this study include those used in existing objective and subjective approaches. Table 3.5 illustrates common quality of life domains adopted by recent instruments with formative measures of subjective quality of life domains. Table 3.5: Quality of Life Domains in Selected Instruments  Satisfaction with Life Scale  WHOQOL- BREF  X  X  X X  X  X  X  Achievements in Life, SelfEsteem  Love  Safety  State of the Environment  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  The Quality of Life Questionnaire The Quality of Life Index  Standard of Living/ Living Conditions  Social Relationships  Productivity/Work  X X  Personal Wellbeing Index  Health  X  Comprehensive Quality of Life Scale  The Quality-of-life Inventory  Material Economic or Financial Situation, Future Security  Community Issues/ Social Conditions  Happiness, Emotional Wellbeing, Psychological Wellbeing  Domains  Life Expectations  Instruments  X X X  X  X  X X  X X  X  X  67  3.5. Reflecting on the Quality of Life Literature The review of the literature on quality of life, subjective wellbeing and measurement instruments provides a strong methodological and theoretical basis for this research. It seems sensible to define domains and indicators of quality of life and to use self-appraisal instruments to measure satisfaction. Moreover, the literature supports the multi-methods approach to the research design in order to support findings from quantitative analysis of instruments’ data. The review of the literature also provides this research with tools to identify and validate quality of life domains, and indicators that can be used equally in evaluating each of the 3 communities focused on in this research. Further discussion of the development of quality of life domains and indicators for this research is presented in the next chapter.  68  CHAPTER 4: METHODOLOGICAL STRATEGY This chapter discusses the methodological strategy chosen for this research. This research takes a case study approach and includes a combination of qualitative and quantitative sources of evidence, as well as the triangulation of findings. Research was implemented in three phases: a preliminary research phase, a model and questionnaire development phase, and a comprehensive data collection phase (Figure 4.1). Phases II and III included fieldwork in the 3 case study communities.  Figure 4.1: Research Phases and Methods  69  4.1  Methods Overview  This research project is an exploratory multiple case study. The case study is one of several approaches to conducting social science research 25. Yin described the case study as “an empirical inquiry that investigates a contemporary phenomenon within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident” (Yin, 1994). Since this research deliberately intends to cover contextual conditions—because these are believed to be pertinent to the phenomenon investigated—the case study strategy was deemed ideal. Case study inquiry benefits from existing theoretical propositions to guide data collection and analysis. This method copes with a large number of variables of interests by relying on multiple sources of evidence (Jick, 1983; Yin, 1994). Therefore, case study data need to converge in triangulation 26 (Yin, 1994). Denzin (1978, p. 291) described triangulation as “the combination of methodologies in the study of the same phenomenon”. The use of triangulation as a means to achieve convergence of findings is widely used in research 27. The cross validation achieved when two or more distinct methods are found to be congruent and yield comparable data enhances the validity of results (Jick, 1983). However, triangulation can achieve more than convergent validation: It can capture a more holistic and contextual description of the phenomena investigated. The triangulation of multiple measures may also illuminate elements of the context by eliciting data and suggesting conclusions that single methods might have neglected (Jick, 1983). Scientists have widely advocated the viability and importance of the integration of quantitative and qualitative methods (Jick, 1983). By using complementary methods, triangulation claims to exploit their strengths and mitigate their limitations (Rohner, 1977; Jick, 1983). The combination  25  Other approaches are experiments, surveys, and the analysis of archival information (Yin, 1994). The triangulation metaphor has its origins in navigation and military strategies that use multiple reference points to locate the exact position of an object. Multiple viewpoints are expected to provide for greater accuracy. This same theory can be applied to researchers looking to improve the accuracy of their findings (Jick, 1983). 27 Another conceptualization of triangulation is described by Denzin as the “within method”, which uses multiple techniques within one method to collect and interpret data. An example is a survey that includes different scales or questions that address the same construct (Denzin, 1978). 26  70  of qualitative fieldwork and survey methods is one of the most prevalent attempts to use triangulation. As sources of evidence, this research considers and triangulates data from the following: ƒ  the existing literature on quality of life and subjective wellbeing and resource communities;  ƒ  context-relevant literature particularly related to the geographical location of the case studies, and the 3 case study communities and mining company associated with them;  ƒ  data collected from interviews and focus groups;  ƒ  data collected through a quality of life survey; and  ƒ  participant observation.  4.1.1. Literature Review The literature review for this research included two main themes: quality of life and quality of life measurements (mostly in psychology and economic literature) as well as single-industry, resource communities, and mining communities’ literature. This review included scientific journal articles, books as well as grey literature and was completed using the following sources: ƒ  University of British Columbia Library (books, journals, and grey literature)  ƒ  University of British Columbia Library on-line journal database (urban, community and regional planning, mining engineering, psychology, economics, epidemiology and sociology databases)  ƒ  Government of British Columbia. Ministry of Community Services Bibliography on Local Government in British Columbia (single industry towns)  ƒ  World Wide Web (municipal, state and federal government sites, relevant blogs and news media)  ƒ  International Society for Quality of Life (ISQOL) publications database  ƒ  Australian Centre on Quality of Life (ACQOL) publications database  ƒ  Questia On-line Library (www.questia.com)  ƒ  HighBeam Research On-line Library (www.hghbean.com)  71  Other sources include Vale website and corporate materials, as well as online and print news associated with the case study communities and operations. Data collected through personal communication (interviews) were also used as reference throughout the thesis.  4.1.2. Participant Observation Case studies may take a variety of forms, most of which include participant observation (Yin, 1994). The main goal of participant observation is to generate “practical and theoretical truths about human life grounded in the realities of daily existence” (Jorgensen, 1989, p. 13). This approach is appropriate for studying processes, relationships between events, human relationships and organizations, patterns, and the immediate socio-cultural contexts in which human life takes place (Jorgensen, 1989). Participant observation may take place in a continuum that ranges from complete immersion in the setting as a full participant to spectator with a complete separation from the setting (Shipman, 1997; Patton, 2002). Participation can also be overt or covert (Jorgensen, 1989). For this research, the observer assumed overt yet varied roles. During the first fieldwork period (25 days), the researcher was an intern with the mining company. For the second fieldwork period (15 days) the researcher assumed a visitor role. During the third fieldwork period (30 days), the researcher assumed an academic researcher role. According to Patton (2002), direct, personal contact with and observation of a setting have clear advantages. Firstly, through observation, the researcher is better able to understand and capture the context of a phenomenon that occurs and how people interact with each other, which is essential to a holistic perspective. Secondly, firsthand experience with a setting allows the researcher to be open and discovery-oriented because while in the field, the observer has less of a need to rely on prior documented or oral conceptualizations of the setting. The third strength of observational fieldwork is the opportunity to discover things that do not come up in interviews because no one else has ever paid attention to them. Participant observers may perceive routines that may be taken for granted and/or remain unnoticed by others in the setting (Patton, 2002). “Questions regarding concept validity in participant observation revolve around whether or not the investigator has gained direct access to the insider’s world of meaning and action”  72  (Jorgensen, 1989 p. 36). Participant observation methodology requires the investigator to collect multiple forms of evidence regarding key concepts. It is therefore important to describe conflicts and disagreements over the meanings of basic concepts, and note differences among insiders. Thus actual use of the concept in the field while interacting with participants provides a powerful test of the validity of concepts (Jorgensen, 1989). Since participant observation is rarely performed as a measurement, the conventional notion of reliability is unsuitable. Observation is very concerned with dependable and trustworthy findings, and is thus very much interrelated with validity. Validity and reliability of participant observation are therefore mostly checked through the use of several forms of evidence; the extent to which the investigator had direct access to the subject’s world; and extensive discussion of procedures of data collection and analysis (Jorgensen, 1989). Limitations of the participant observation method are closely associated with the concept of researcher bias—the effect of the individual’s personal, professional and political views. Bias is, therefore, “part of any research, from design to implementation” (Shipman, 1997, p. 19). Since human beings construct their own knowledge of the world around them, a detached position for neutral observation is unattainable (Shipman, 1997).  4.1.3. Interviews Interviewing refers to the range of strategies for formally asking questions. The qualitative research interview seeks to describe the meanings of central themes in the experience of the subjects. The main task in interviewing is therefore understanding the meaning of what the participants say (Kvale, 1996; Strauss and Corbin, 1998). Since the researcher works directly with the respondent in interviews, these are a far more personal form of research than questionnaires (Kvale, 1996). Interviews are particularly useful for getting the story behind a participant’s experiences. Individual interviews are the best approach when the purpose of the data collection is to have an individual, open-ended discussion on a range of issues and to obtain in-depth information on an individual basis about perceptions and concerns (Kvale, 1996). Individual interviews are particularly useful when: ƒ  the topic is too sensitive for a group discussion; 73  ƒ  the target population differs in age, ethnicity, culture, or social background from investigators;  ƒ  participants may have a low reading ability; and/or  ƒ  it is difficult to gather people in groups due to conflicting schedules and/or geographic location.  In general, individual interviews can be classified as unstructured or structured interviews. In unstructured interviews, the flow of conversation is more open and natural. This allows respondents to express their views about a topic, including good and bad points, without being prejudiced by the interviewer’s own beliefs. It also allows participants to discuss points that are of importance to them and points that may not have occurred to the interviewer. Pre-determined questions are asked, but the interview remains as open and adaptable as possible to the interviewee’s nature and priorities. An unstructured interview is conducted in a question-andresponse format and may be free-flowing. However, it becomes structured in the sense that the interviewer has a purpose and requires skill to establish a relationship and ask well-structured questions to generate a conversational flow in which the participant offers information (factual, opinion, subjective and objective) about the matter under investigation (Kvale, 1996). Unstructured interviews were used in Phase II of this research. In structured interviews, the interviewer decides what to ask and in what order. As a result, structured interviews can only get participants' perceptions of the issues the interviewer believes to be important (Institute of Education Technology, 1996). Structured interviews can include standardized, open-ended questions, when the same open-ended questions are asked to all interviewees (this approach facilitates faster interviews that can be more easily analyzed and compared); or closed, fixed-response questioning, when all interviewees are asked the same questions and asked to choose answers from among the same set of alternatives. Phase III of this research includes structured interviews with open-ended questions. Interviewing requires a skilled researcher to make clear the meaning of questions, establish rapport with the respondent, and probe responses without leading the respondent to give certain answers (Strauss and Corbin, 1998). In the case of this research, the interviewer provided clarity regarding the purpose of the research and the particular focus of the investigation. In ensuring information received was properly heard and understood, listening, clarifying and reflective summarizing were important steps performed by the interviewer. To reduce the  74  uncertainty associated with the subjectivity of interview procedures, questions were tested on site. For the structured interviews in Phase III, the researcher closely followed a pre-defined interview protocol to reduce the interviewer bias (Groves et al., 2004).  4.1.4. Survey Questionnaire A survey can be described as a systematic method for gathering information from a sample of people for the purposes of constructing quantitative descriptors of the larger population of which the respondents are members. Survey questionnaires are popular research tools because they collect data from large numbers of participants quickly and systematically (Groves et al., 2004). In sample survey questionnaires, standardized measurement and sampling procedures enhance the reliability of observation, facilitate the replication of studies and allow for statistical analysis of data and generalizations for larger populations (McClintock et al., 1983). Questionnaire design is a meticulous and comprehensive process that implies careful question design to avoid cognitive problems and ensure accuracy of measurements (Shipman, 1997; Groves et al., 2004). Along with the definition of the population sampling approach, a critical issue in questionnaire design is the management of numerous potential errors 28 (Groves et al., 2004). Questionnaires are also evaluated for construct validity and reliability. In the case of this research, the questionnaire design also drew from existing validated psychometric quality of life and wellbeing instruments. Some limitations of survey questionnaires are associated with the way questions are written by the researcher and understood by respondents (Patton, 2002). At both steps, there are chances for error. Another challenge with questionnaire design is that the researcher can never be absolutely certain that cognitive problems do not exist, or of the reasons why some questions are left unanswered. To mitigate these limitations, the questions and response scales were emulated from previously tested quality of life questionnaires. Questionnaires were also pretested in the field.  28  Survey errors might include measurement errors, processing errors, coverage errors, sampling errors, nonresponse errors and adjustment errors (Groves et al., 2004).  75  A specific challenge of using questionnaires for measuring levels of satisfaction with quality of life in cross-sectional research designs is that participants respond only once. This is challenging because external transitory events might impact how individuals rate levels of satisfaction with domains or indicators. Longitudinal designs—when measurements are taken on multiple occasions—minimize the effects of the momentary contextual conditions (Sirgy et al., 2006). Given the time and resource limitations of this research, the methodology was strengthened by a multi-method strategy that included interviews and participant observation.  4.2.  Phase I: Preliminary Research  This phase consisted of a comprehensive review of the literature and methodology and the first attempt to define a conceptual model for the study. The broad literature review conducted in Phase I is summarized in the first chapters of this thesis. The literature on quality of life (QOL) measurement instruments provided information on which domain indicators have been tested in cross-cultural studies and proved to conform to validity and reliability expectations. Details on the major instruments and domains, and literature on quality of life and wellbeing studies also contributed to the initial definition of QOL domains for this research project (Tables 3.1, 3.2, 3.3 and 3.4). The intention of this research was, similar to several studies found in the literature (e.g., Michalos, 1985; Cummins, 1996; Michalos, 1997; Hagerty et al., 2001; Sirgy et al., 2001; Bernhard et al., 2004; Hsieh, 2004), to define a list of domains and indicators for each of these domains. This task was completed in Phase II.  4.3.  Phase II: Quality of Life Model  Phase II consisted of the validation of domains and definition of the indicators. This was accomplished through qualitative analysis of on-site interviews, supported by participant observation notes taken by the researcher, and the literature review on quality of life studies and subjective wellbeing.  76  The process of defining domains and indicators was an iterative one. Based on the literature review on mining/single industry/resource communities, quality of life (QOL) and QOL measurement tools, 4 initial domains were chosen: economic, social, environmental and health. With the content analysis of the initial exploratory interviews on site (Canaã dos Carajás, Parauapebas and Núcleo Urbano de Carajás), the initial four domains were validated, and one additional domain was added: the work domain. The work domain was included in the QOL model for this research because of the unique nature of the communities studied (single-industry communities with populations largely composed of migrants from other towns or states). It is well documented that in life satisfaction studies, individuals living in single-industry or “occupational communities” such as mining towns usually consider work a very important part of their lives. For these individuals, work often influences their self-image, their choice of reference social groups and of friends (Near et al., 1980). Interviews in Phase II and the field observations support this theory. The literature on quality of life in the fields of organizational behaviour and mangement also provides discussions on what is referreed to as quality of work life (QWL) (e.g., O’brien, 1990; Tait et al., 1989). Quality of Work Life (QWL) has also been found to be significant predictor of life satisfaction (Sirgy et al., 2006).  Figure 4.2: A Preliminary Quality of Life Conceptual Model  77  The work domain was therefore included in the conceptual model for this research and added to the quality of life questionnaire. This hypothesis was tested in Phase III. Considering the five conceptual model domains, literature regarding indicators was reviewed once again. During the first fieldwork period (25 days), the researcher was able to visit all 3 case study communities and conduct unstructured interviews with employees and other residents. Interview content analysis was also critical in the validation of indicators from the literature as well as in determining new indicators appropriate to the context of the 3 case study communities.  4.3.1. Phase II: Methods Even though this research’s sampled population is mostly composed of employee-residents, it was considered valuable to include subjects who did not work for Vale in order to increase the researcher’s ability to identify as many local QOL issues as possible in a relatively short period of field work. These individuals were considered key informants and were mostly critical service providers (e.g., doctors, school teachers). The initial unstructured interviews were conducted with key informants working in the 3 case study communities (Portuguese terms are explained in the Glossary at the beginning of this document):  ƒ  Community Affairs staff at the Non-ferrous Division (Diretoria de Operações de Nãoferrosos de Carajás—DIOC), and Ferrous Metals Division (Diretoria de Ferrosos—DIFN) in the Núcleo Urbano de Carajás and Canaã dos Carajás  ƒ  Coordinator of Health and Education Services, Núcleo Urbano de Carajás  ƒ  Coordinator and Transportation and Food Services, Núcleo Urbano de Carajás  ƒ  Head Nurse, Special Programs, Hospital Yutaka Takeda, Núcleo Urbano de Carajás  ƒ  Executive Director, Hospital Yutaka Takeda, Núcleo Urbano de Carajás  ƒ  Doctor, Hospital Yutaka Takeda, Núcleo Urbano de Carajás  ƒ  Social Worker, DIFN  ƒ  Director, Agência de Desenvolvimento (Sustainable Development Agency), Canaã dos Carajás  ƒ  Primary School Principal; Vila Bom Jesus, Canaã dos Carajás 78  ƒ  Elementary School Principal, Canaã dos Carajás  ƒ  Senior geologist, DIFN, Núcleo Urbano de Carajás  ƒ  Junior geologist, DIOC, Canaã dos Carajás  A convenience sample 29 of 17 residents was also interviewed. This sample included interns, Vale employees and other residents of the 3 case study communities. A group of 3 workers participated in a focus group regarding the Sossego Camp, located in Canaã dos Carajás. Participants from interviews and the focus group were contacted by the researcher and were interviewed in either their places of work or their homes. The initial unstructured interviews were conducted to identify issues, likes and dislikes, concerns, and thoughts on quality of life in the 3 case study communities. Some interviews also aimed at understanding the company’s approach, policies and programs, or at understanding the particularities of the 3 mining operations: Carajás Iron-ore Mine, Azul Manganese Mine and Sossego Copper Mine. The interviews and the focus group were tape recorded with informed consent from all participants; following the requirements of the University of British Columbia’s Behavioural Research Ethics Board (confirmation of approval is available in Appendix B). . Table 4.1: Phase II Data Collection Focus of Research  Communities Nucleo Urbano de Carajás  Canaã dos Carajás  Parauapebas  1. Investigating factors (domains and facets) of quality of life  11 interviews  4 interviews 1 focus group with 3 participants  2 interviews  2. Understanding the company approach, policies and programs  6 interviews  2 interviews  2 interviews  Coding of the initial interviews and the focus group (Table 4.1) provided an initial list of broad themes or concepts (i.e., factors affecting quality of life), which were categorized under broader themes (i.e., quality of life domains) and considered as indicators. This list was instrumental in  29  A convenience sample is composed of easily accessible individuals and is not representative of the entire population (Shipman, 1997).  79  reaching the objectives of Phase II. Examples of the concepts are difficult access to goods and services, lack of privacy, and difficult access to health services. While several interviews were completed with Núcleo Urbano de Carajás and Canaã dos Carajás residents, only 2 interviews were completed with Parauapebas residents. This was due to the initial research design and scope, which included only Núcleo Urbano de Carajás and Canaã dos Carajás as case studies. During interviews and researcher’s observations on site, it became clear that the Núcleo Urbano de Carajás had physical growth constraints due to being located in a National Park and will, in the near future, accommodate exclusively higher-ranking Vale employees and their families. It will not be able to accommodate the growing influx of employees relocating to the region as mine operations expand. As a result, Vale employees are increasingly taking up residence in Parauapebas. In Phase II, the direct observation data assisted with the interpretation of results from the qualitative data collected through interviews and the focus group. These results indicated that all of the conceptual domains identified—i.e., social, economic, environmental, work and health— seem to be relevant to the quality of life in the case study communities. Phase III included further qualitative data collection as well as a significant quantitative data collection effort and aims at validating both the domains and the indicators chosen in Phase II.  4.3.2. Quality of Life Domains and Indicators Quality of life (QOL) indicators are standardized measures with which to monitor and compare the QOL of a group or population. The approach taken in this dissertation was to select a list of formative indicators from each of the selected domains and include them in a QOL questionnaire that will be described later in this chapter. What follows is the list of concepts that were brought up during Phase II interviews, and therefore considered as indicators, organized by domains. For each domain, a short discussion follows on how these indicators are aligned with the current literature on QOL and subjective wellbeing research.  80  a. Social Domain The social domain includes the social relationships and social aspects of life in a community. The following are the social issues identified during the Phase II interviews and focus group:  ƒ  ƒ  Personal relationships: o  relationships with friends  o  relationships with family and spouse  o  ability to form long-lasting friendships  Community connectedness/participation: o  identification with specific groups within a community (e.g., religious, cultural, ethnic)  o  community participation  ƒ  Freedom of expression  ƒ  Privacy levels  ƒ  Personal safety  In the QOL literature, the use of personal relationships and community connectedness or participation concepts as indicators are strongly supported (Pibernik-Okanovic et al., 1996; Michalos, 1997; Zumbo and Michalos, 2000; Bramston et al., 2002; Sirgy and Cornwell, 2002; Chen et al., 2006). Freedom of expression, privacy and personal safety have been considered of particular interest for studies of quality of life at work for mining and silviculture camps (White, 2004; Costa et al., 2005). These issues are also prominent in existing literature on resource communities and remote single-industry communities (e.g., Robinson, 1962; Lucas, 1971; Riffel, 1975; Barbedo, 2000). b. Economic Domain As is well articulated in the literature on resource communities and fly-in-fly-out mining camps (e.g., Robinson, 1962; Lucas, 1971; Riffel, 1975; Sibbel et al., 2006), economic issues such as income and wealth are significant in these settings. The prospect of a higher income has been identified as a major factor in motivating people to move to such communties (Riffel, 1975; Sibbel et al., 2006). The economic domain includes aspects of one’s financial life and material wealth. What follows are economic issues identified during the interviews and focus groups in Phase II.  81  ƒ  Salary and benefits  ƒ  Cost of living  ƒ  Material belongings  ƒ  Financial security  All of the above concepts have been used as indicators in QOL research (Cummins, 1996; Michalos, 1997; Sirgy and Cornwell, 2001; Bowling and Gabriel, 2004). c. Health Domain Quality of life researchers extensively agree on the definition of health as a QOL domain. QOL and health care have been linked for centuries (Sirgy et al., 2006). The health domain in this research includes the aspects of one’s physical and mental health status as well as the availability of health services. Health-related issues identified during the Phase II interviews and by the focus group were:  ƒ  Levels of stress  ƒ  Level of physical activity  ƒ  Food and nutrition  ƒ  Work-life balance  ƒ  Availability of recreation and entertainment options as a means to maintain mental health  ƒ  Access to good-quality health services and facilities  These indicators have also been identified as significant issues in remote resource communities, including fly-in-fly-out mining camps (e.g., Robinson, 1962; Lucas, 1971; Riffel, 1975; Sibbel et al., 2006). The indicators above have also been used as health indicators of quality of life, sustainability and community health (e.g. Michalos, 1997; Sirgy et al., 2001; Government of New Zealand, 2003; Bernhard et al., 2004). d. Environment Domain The environment domain can be defined as both natural and built environment and surroundings. Environment-related issues identified during the Phase II interviews and by the focus group were:  ƒ  Quality of one’s home or residence  82  ƒ  Quality and integrity of the natural environment  ƒ  Order and safety  ƒ  Infrastructure (e.g., transportation, sewage, water, electricity, internet services) quality and availability  ƒ  Shopping, education and recreation facilities  These issues have all been identified as indicators of QOL in previous studies (e.g., PibernikOkanovic et al., 1996; Michalos, 1997; Cummins, 2000; Noronha, 2001; Sirgy and Cornwell, 2001) and have been widely identified as evalution factors for remote resource or singleindustry communities (e.g., Robinson, 1962; Lucas, 1971; Riffel, 1975; Storey, 2001; Sibbel et al., 2006). For this research, ecreation facilities was included in the health domain because it addresses the physical need for exercise to maintain good health. e. Work Domain Phase II interviews and the focus group made it clear that work and their own perceptions of their work life are very significant for employees living in the 3 case study communities. Workrelated issues identified as affecting QOL were:  ƒ  Number of hours of work per week  ƒ  Amount of time off  ƒ  Salary and benefits  ƒ  Relationships at the work place  ƒ  Safety of the workplace  ƒ  Opportunities for growth in the organization  ƒ  Empowerment and ability to make decisions at work  These indicators are prominent isses in the Work Satisfaction literature (Rodgers, 1977; Kim and Cho, 2003; Requena, 2003). QWL research has identified participation in decision making and compensation packages as strong factors that impact QWL (Sirgy et al., 2006). The indicators number of hours of work per week, salary and amount of time off were tested and validated by the Australian Centre on Quality of Life PWI Survey 7; “The Wellbeing of Australians—The Effects of Work” (Cummins et al., 2003).  83  4.3.3. Quality of Life Survey Questionnaire Development The survey questionnaire developed for Phase III is divided into 3 parts. Part 1 includes an adaptation of an existing satisfaction measurement tool, the Personal Wellbeing Index (PWI). Part 2 includes questions organized under the life domains identified in Phase I: work, social, health, environment, and economic domains. Part 3 includes personal and demographic questions intended to characterize respondents. The questionnaire was applied in the local language, Portuguese. An English version of the questionnaire is available in Appendix B. For the first two parts of the questionnaire, an end-defined, Likert-like scale was used. The sociologist Rensis Likert, who was concerned with measuring psychological attitudes scientifically, originally developed this scale. Likert sought a method that would produce attitude measures that could reasonably be interpreted as measurements on a proper metric scale (equally spaced), in the same sense that we consider inches or degrees Celsius true measurement scales (Uebersax, 2006). Sets of Likert items are often used to form indices which are expected to pass the Cronbach’s alpha test 30 or some other test of inter-correlation to assure a common meaning of items comprising a latent variable. Likert-like scales are commonly used in quality of life survey instruments (Table 3.1). Usual response categories are associated with levels of agreement: "strongly disagree," "disagree," "don't know,” "agree," and "strongly agree” (e.g., Satisfaction with Life Scale) or levels of satisfaction: “extremely dissatisfied”, “dissatisfied”, “not satisfied nor satisfied”, “satisfied” and extremely satisfied” (e.g., Personal Wellbeing Index). These values are ordinal within any given Likert item, but sets of items are not necessarily ordinal with respect to each other. The 11-point scale provides five levels of satisfaction above the neutrality point of the scale. A detailed discussion that supports the use of a 0–10 point end-defined scale, with only 3 labels (at the ends and in the middle of the scale), for QOL tools is provided by Dr. Robert Cummins (2002) (Australian Centre for Quality of Life). In a discussion paper on the caveats of the Comprehensive Quality of Life Scale (ComQol), Cummins argues that while there is always uncertainty regarding the respondents’ discriminative capacity, five degrees of choice is not likely to exceed this capacity. Cummins further asserts that responses to 0–10 scales demonstrate equivalent reliability to scales with fewer choice points and that a 0–10 scale is  30  For the Cronbach Alpha test to be applicable, it is necessary to have at least four items (Uebersax, 2006).  84  intuitively meaningful. Moreover, Cummins argues that labeling each choice point on a Likertlike scale introduces error variance since individuals seem to often differ in their allocation of a numerical value to such labels (Cummins, 2002a). Based on Cummins’ observations, 11-point (0–10) end-defined scales were used in the quality of life questionnaire designed for this research. Similar to a Likert scale, the items were equally spaced. Only the two ends and the central items were defined with labels. The lowest end (0) was labeled as not at all satisfied, the middle point (5) was labeled as neutral, and the positive end of the scale (10) was labeled as extremely satisfied. f.  Questionnaire Part 1: The Subjective Wellbeing Index  The Subjective Wellbeing Index (SWI) tested in this research is a modified version of the Personal Wellbeing Index (PWI). This adaptation included a change in the list of items, but maintained the wording of the questions: “Thinking about your life and current situation, how satisfied are you with…?” To assure that questions were easily understood and therefore introduced few cognitive problems (Groves et al., 2004), the SWI questionnaire was first tested with a sample of 13 volunteers from diverse genders, job classes and communities of residence. Testing results suggested that 3 of the PWI questions presented cognitive problems 31 on site. Thus, the following questions were excluded from the final version of the questionnaire:  ƒ  “How satisfied are you with your standard of living?”  ƒ  “How satisfied are you with your financial security?”  ƒ  “How satisfied are you with how safe you feel?”  Questions about satisfaction with health, achievements in life and personal relationships were maintained. Based on the literature review and preliminary interviews conducted on site during Phase II of this research, other items of satisfaction were added: surrounding environment, economic situation, community and work (Figure 4.2). Through statistical analysis, a first attempt to validate the SWI construct included the seven SWI domains (Table 4.2). This was accomplished with linear regression. Validation of domains  31  Testing of these questions determined that there were differences on how some respondents understood the meaning of the questions.  85  occurs when satisfaction with each domain (independent variables) contributes with unique variance when regressed against the satisfaction with QOL (dependent variable). Table 4.2: Subjective Wellbeing Index and Personal Wellbeing Index Subjective Wellbeing Index (SWI) Items  Personal Wellbeing Index (PWI) Items (2006)  Economic situation  Standard of living  Health  Health  Achievements in life  Achievements in life  Personal relationships  Personal relationships  Physical environment  How safe you feel  Community  Community connectedness  Work  Future security  The original PWI items are a first-level deconstruction of the global question about satisfaction with Life as a Whole (LAW). Therefore, PWI questionnaires also include this single question: “Thinking about your own life and personal circumstances, how satisfied are you with your life as a whole?” The SWI items are intended to represent a first-level deconstruction of the global question about satisfaction with Quality of Life (QOL). Endeavouring to explore conceptual distinctions between LAW and QOL, the questionnaire used in this research included both reflective questions: ƒ  “Thinking about your own current life and personal circumstances, how satisfied are you with your quality of life in general?”  ƒ  “Thinking about your own life and personal circumstances, how satisfied are you with your life as a whole?”  These questions were added as dependent variables to test the construct validity of both the SWI items and the second part of the questionnaire, which includes questions on the five QOL domains. These single questions were administered first to ensure that domains and indicators could not influence a global response (The International Wellbeing Group, 2006). The distinction between the concepts of “life as a whole” (LAW) and “quality of life” (QOL) is acknowledged in the literature and yet is not clearly defined. The inclusion of the two single questions (about LAW and QOL) in the questionnaire allows for the investigation of the  86  suitability of the SWI items as a first-level deconstruction of both LAW and QOL. It also allows for the investigation of LAW and QOL predictors using the domains and indicators of the questionnaire. This investigation is intended to provide insights on the distinctions between these two constructs and contribute to the discussion about conceptual distinctions among wellbeing concepts such as quality of life, happiness, and life satisfaction. A study by Zumbo and Michalos (2000) in Jasper, Alberta presented an interesting discussion on the distinctions between LAW and QOL. The researchers found that in 1997, Jasper residents’ satisfaction with personal worth, interpersonal relations, finances, police protection and weather were predictors of life satisfaction, while personal worth, finances, Jasper as a place to live, sanitation/waste management and recycling were predictors of satisfaction with quality of life. Predictors of happiness were personal worth, finances, interpersonal relations, and recycling (Zumbo and Michalos, 2000), thus revealing differences in how the study population defined predictors of QOL, LAW and happiness. It is hoped that results from the analyses on the LAW data in this research will contribute to this discussion. g. Questionnaire Part 2: QOL Domains The second part of the questionnaire contains five sections, one for each domain, and includes questions about satisfaction with items in each domain (Table 4.3). h. Questionnaire Part 3: Demographic Data The third and final part of the questionnaire includes questions about the respondents and their families. These questions (e.g., community of residence, gender, age, and job class) were chosen to classify the respondents into groups of relevance to the research objectives. The variables and categories were initially defined by the researcher but were slightly modified, in collaboration with Vale personnel, once on site. A list of items addressed in Part 3 of the questionnaire is provided in Table 4.4.  87  Table 4.3: Questionnaire Items: QOL Domains and Indicators Domains  Indicators What you can buy and do with your money Material belongings  Economic  Economic situation of surrounding community Ability to save for the future Financial stability Number of hours of work per week Relationships with co-workers Autonomy and decision-making ability  Work  Relationship with immediate supervisor Opportunities for professional growth Occupational safety at the workplace Job security Salary and benefits Place of residence (house/apt/camp) Quality and integrity of the natural environment around you  Environment  General order, cleanliness and security in the community Basic infrastructure in the community Availability of goods and services Transportation options and quality in your community Stress level in your life Physical activity level  Health  The food you eat Entertainment and recreation options in your community Health services and facilities in your community Work-life balance Relationship with family Love life Relationship with friends  Social  Privacy Participation in community events Participation in decision making in the community Educational opportunities available  88  Table 4.4: Questionnaire Items: Demographic Information Domains  Indicators Community of residence  Your Home and Community  Length of residence in the community Housing status Region of origin Job class  Your Work  Area of work Length of employment with Vale Gender Age  You and Your Family  Education level Marital status Number of dependent children  4.4.  Phase III: Interviews and Questionnaire  Phase III took place during the third site visit and consisted of two major data collection efforts: a) in-depth structured interviews, and b) a Quality of Life survey questionnaire. Phase III began with the researcher participating in a planning and communication meeting with administrative co-ordinators working in several different Vale departments. These individuals were generally administrative assistants whose responsibilities included communicating with all employees about human resources, social events and other issues of broad interest. The support of these individuals for this research was effective because co-ordinators generally have wide access to all Vale employees in their respective departments, and they had the ability to access employees without the “boss-employee” relationship that could result in a feeling of coercion to participate in the research project and, consequently, biased responses to the questionnaire or during interviews. The meeting with the company support staff was facilitated and organized by a Corporate Communications Manager and a Human Resources Manager. Meeting participants included representatives of the Mining, Mineral Processing and Administration departments of the Carajás Iron-ore Mine, Azul Manganese Mine and Sossego Copper Mine.  89  4.4.1. Interviews Employee co-ordinators were responsible for communicating with managers and other employees about the research and for inviting them to answer the questionnaire and participate in the interview process. Reliability of information obtained though individual interviews can be assessed by asking for a confirmation of information later in the interview, or by re-interviewing the respondent about key points (Kvale, 1996). For this research, reliability was ensured by asking interviewees for agreement or confirmation of responses at the end of each interview. Interviewees therefore had the opportunity to confirm or clarify information. a. Structure and Content The interviews were designed to collect qualitative data regarding the 3 case study communities. Structured, in-depth interviews were conducted with a total of 32 workers residing in the 3 communities and 8 residents who at the time of the interview were not Vale employees (Table 4.5). Interviews included structured, open-ended questions about quality of life and the most and least satisfying aspects of quality of life in each community. Participants were also asked to make suggestions for quality of life improvements or identify priorities for improvement in their communities. Other comments on quality of life in the communities were encouraged. The list of interview questions is available in Appendix B. b. Sampling Frame The sampling frame included all full-time Vale employees or interns who were residents in Parauapebas, Canaã dos Carajás or Núcleo Urbano de Carajás and were 18 years of age or older (age of majority in Brazil). The call for participants was also advertised on the company’s printed and online newsletters. The invitation to participate indicated that the interviews were anonymous and confidential and were expected to last between 30 and 45 minutes. The sampling aimed at including subjects from the 3 different communities and with diversity in length of residence and employment. A balance of female and male subjects was also intended.  90  The interview population sample also included diverse job classes. The lowest job class was assistant mechanic, and the highest class was senior geologist. During the interview process, some participants suggested consultation with key informants in their communities, people who were not Vale employees. These key informants were invited to participate in an interview, and those who accepted the invitation were asked the same questions prepared for Vale employees. Respondents had to be 18 years of age or older and a resident in one of the case study communities in order to participate. The goals and objectives of the research project were explained to all interview participants, and everyone interviewed signed a consent form, following all requirements of the University of British Columbia Behavioral Research Ethics Board. c. Population Sample A total of 40 structured interviews were completed in Phase III. Interviews included participants from the 3 case study communities. Details of the total sampled population are illustrated in Table 4.5. More than one-third of the interviews were conducted with females (15), and 25 interviews were conducted with males. A total of 18 interviews were conducted with residents of the Núcleo Urbano de Carajás, 13 interviews were conducted with Canaã residents and 9 with Parauapebas residents. The population sample for interviews was fairly young—the mean age of participants was 30. The oldest participant was 56 years old, while the youngest was 19. Slightly more than half of the sampled population was married or living in common law (55%), while 44% of participants were single. Because the researcher was interested in discussions with both newcomers and long-term residents, the variance in length of employment with Vale was wide, from 3 months to 25 years of service.  91  Table 4.5: Interview Population Sample Variables  Mean  Median  Max.  Min.  Age  30.2 years  29 years  56 years  19 years  Length of Employment  5.85 years  2.5 years  25 years  About 3 months  Frequency  % of Total Sample  Núcleo Urbano de Carajás  18  45  Canaã dos Carajás  13  32.5  Parauapebas  9  22.5  Female  15  37.5  Male  25  62.5  Manager or supervisor  5  12.5  Engineer, geologist or analyst  8  20  Technicians  15  37.5  Interns  4  10  Not Vale employee  8  20  High school completed*  7  17.5  Technical or trades training completed  20  50  University level  13  32.5  Married or common law  22  55  Single  18  45  No children  23  57.5  1 or 2 children  14  35  3 or more children  3  7.5  North  15  37.5  Northeast  16  40  Southeast  9  22.5  Community of Residence  Gender  Job Class  Educational Level  Marital Status  Children  Region of Origin  *With no technical or trades training  The population sample also includes a variety of job classes—12.5% (5) were managers or supervisors, and 20% (8) were engineers, geologists or analysts. Technicians, who comprised 37.5% (15) of the sample, were the best-represented group. Only 4 interns (10% of the  92  population sample) participated in interviews, making them the least-represented group. A total of 20 interviews were conducted with local residents who did not work directly for Vale. As illustrated in Table 4.5, half of the population sample had completed a technical degree at high school or Technical or trades training. Only 17.5% had only a high-school education, and 32.5% had completed a university degree. The majority of the population sample (57.5%) did not have any dependent children, while 42.5% had one to two dependent children. Only 3 interview participants (7.5% of the population sample) had 3 or more children. The population sample did not include any participants from the South or Central-West Regions of Brazil. 40% of participants were originally from the Northeast Region, while 37.5% were from the North Region and 22.5 % were from the Southeast Region. d. Núcleo Urbano Population Sample A total of 18 structured interviews were completed with Núcleo Urbano de Carajás residents. Details of the Núcleo population sample are illustrated in Table 4.6. In the Núcleo Urbano, all employees live in company-provided accommodations, thus, all of the respondents who were Vale employees (83%) lived in a company house. The other 17% of the respondents lived in repúblicas 32—shared company houses—in the Núcleo. All employees were associated with the Carajás Iron-ore Mine. Of the eighteen structured interviews completed with residents of the Núcleo Urbano de Carajás, 8 were conducted with females and 10 were conducted with males. The average age of Núcleo respondents was 34.22 (median age = 32 years). The oldest participant was 56 years old and the youngest was 20 years old. While the Núcleo’s population  32  Repúblicas are shared living arrangements, usually owned by a third party. In the Núcleo Urbano de Carajás and Canaã dos Carajás, these are regular company houses that are shared among single employees. The density within the repúblicas varies between one person per bedroom to two persons sharing a bedroom, so the number of people living in a república will vary according to the number of bedrooms in the house and the need to accommodate singles.  93  sample was fairly young, the average ages of both the Parauapebas and Canaã dos Carajás population samples were lower. Almost 2/3 of the Núcleo Urbano de Carajás population sample were married or living in common law (72.2%), while 27.8% of participants were single. 44% of the sampled population did not have any dependent children, while another 44.4% had one to two dependent children. Only 2 participants (11.2%) had 3 or more children. Among the population samples for the 3 communities, the Núcleo’s sampled population had the largest variance in length of employment with Vale, ranging from 6 months of employment to 25 years. Of the total Núcleo population sample, 16.7% (3) were managers or supervisors, and 27% (5) were engineers, geologists or analysts. Technicians, who comprised 33.3% (6) of the sample population, were the best-represented group. The least-represented group were interns—only 2 interns (11.1% of the population sample) participated. An additional 2 interviews (11.1% of the population sample) were conducted with local residents who did not work directly for the company. Among the population samples for the 3 case study communities, the Núcleo’s had the highest levels of education. Half of the population sample had completed a technical degree at high school or Technical or trades training, and the other half had completed a university degree. Of the Núcleo’s population sample, 44.4% were originally from the Northern Region of Brazil, 27.8% were from the Northeast Region and 27.8% were from the Southeast Region.  94  Table 4.6: Interview Population Sample: Núcleo Urbano de Carajás Variables Age Length of Employment  Mean  Median  Max.  Min.  34.22 years  32 years  56 years  20 years  9.7 years  6.5 years  25 years  6 months  Frequency  % of Total Sample  Female  8  44.4  Male  10  55.6  Manager or supervisor  3  16.7  Engineer, geologist or analyst  5  27.8  Technician  6  33.3  Interns  2  11.1  Not Vale employee  2  11.1  Technical or training completed  9  50  University level  9  50  Married or common law  13  72.2  Single  5  27.8  No children  8  44.4  1 or 2 children  8  44.4  3 or more children  2  11.2  North  8  44.4  Northeast  5  27.8  Southeast  5  27.8  Gender  Job Class  Educational Level  Marital Status  Children  Region of Origin  *Refers only to Vale employees  e. Canaã dos Carajás Population Sample A total of 13 structured interviews were completed with Canaã dos Carajás residents. At the time of the interviews, 5 respondents (38.5% of the population sample) lived in a company house with their families, 3 respondents (about 23%) lived in shared company homes, 3 lived in the Sossego Camp 33, and 2 respondents (15.4 %) lived in their own homes with their families.  33  The Sossego Mine Camp is a temporary facility located in Canaã dos Carajás that accommodates company employees and business visitors.  95  Table 4.7: Interview Population Sample: Canaã dos Carajás Variables  Mean  Median  Max.  Min.  Age  28.5 years  29 years  40 years  19 years  Length of Employment  1.8 years  1.8 years  3 years  About 3 months  Frequency  % of Total Sample  Female  3  77  Male  10  23  Manager or supervisor  2  15.4  Engineer, geologist or analyst  3  23.0  Technicians  6  46.2  Interns  1  7.7  Not Vale employee  1  7.7  High school completed*  1  7.7  Technical or trades training  8  61.5  University level  4  30.8  Married of common law  6  46.2  Single  7  53.8  No children  8  61.5  1-2 children  4  30.8  3 or more children  1  7.7  North  3  23.0  Northeast  6  46.2  Southeast  4  30.8  Gender  Job Class  Educational Level  Marital Status  Children  Region of Origin  *With no technical or trades training  Table 4.7 presents more details on Canaã’s population sample. One participant from the Núcleo Urbano de Carajás, who had lived in Canaã for one year, was also able to answer questions about both of those communities. Canaã’s sample population was younger than the Núcleo’s. The average age of respondents was 28.5, with the oldest participant being 40 years old and the youngest being 19 years old.  96  Regarding marital status, Canaã’s population sample was quite balanced: 46.2% of the sampled population was married or living in common law, while 53.8% of participants were single. Canaã’s population sample had the smallest variance in length of employment with Vale, from about 3 months of employment to only 3 years of employment. All employees were associated with the Sossego Copper Mine, which has only been operating since 2004. Only 15.4% of the Canaã population sample (2) were managers or supervisors, while 23.1% (3) were engineers, geologists or analysts. The best-represented group was also technicians (6), who made up almost half (46.2%) of the sample population. Only 1 intern was interviewed (representing 7.7% of the population sample), and 1 interview was conducted with a local resident who did not work directly for Vale. The majority (61.5%) of participants from Canaã had completed a technical qualification program at high school or Technical or trades training, while 30.8% had a university degree; 1 participant (7.7%) had completed high school only. Most (61.5%) of the Canaã population sample did not have any children, while 30.8% had one to two dependent children, only 1 participant (7.7%) had 3 or more children. Different from the Núcleo’s sampled population, almost half (46.2%) of the Canaã population sample was originally from the Northeast Region of Brazil. Surprisingly, 30.8% participants were from the Southeast Region, and only 23% were from the Northern Region. f.  Parauapebas Population Sample  A total of 9 structured interviews about Parauapebas were conducted. Five of these interviews were conducted with individuals residing in the city. Four participants from Núcleo Urbano and two from Canaã were also able to answer questions about Parauapebas. While eight participants were renters at the time of the interview, only one was a homeowner. Table 4.8 presents the sampled population of respondents who resided in Parauapebas at the time of the interview.  97  Parauapebas’ population sample was the youngest of the 3 sample populations. The average age of respondents was only 25, with the oldest participant being 35 years old and the youngest being 20 years old. Just under half of the interviews with Parauapebas residents were conducted with females (4), and 5 interviews were conducted with males. Two-thirds of the Parauapebas population sample were married or living in common law, while one-third of participants were single. The vast majority of the participants from Parauapebas did not have any dependent children (77.8%), while 22.2% had one to two dependent children. Regarding the average length of employment, Parauapebas’ sampled population had the least variance. The average was 2.4 years, and participants’ total length of employment ranged from 1.5 years to 4 years. All employees were associated with the Carajás Iron-ore Mine. As indicated in Table 4.8, 2/3 of the Parauapebas population sample had had a high school degree only, and one-third had completed a technical degree at high school or Technical or trades training. None of the participants had completed a university degree at the time of the interviews. Most participants were not directly employed by Vale when they were interviewed. Of the Vale employees interviewed, 3 were technicians (33.3% of the population sample), and one (11.1%) was an intern. More than half of the Parauapebas population sample was originally from the Northeast Region (55.6%) of Brazil, while the remaining 44.4% were from the Northern Region.  98  Table 4.8: Sampled Population for Interviews: Parauapebas Residents Variables Age Length of Employment  Mean  Median  Max.  Min.  25 years  23 years  35 years  20 years  2.4  2  4 years  1.5 years  Frequency  % of Total Sample  Female  4  44.4  Male  5  55.6  Technician  3  33.3  Interns  1  11.1  Not Vale employee  5  55.6  High school completed*  6  66.7  Technical or trades training  3  33.3  Married of common law  3  33.3  Single  6  66.7  No children  7  77.8  1-2 children  2  22.2  North  4  44.4  Northeast  5  55.6  Gender  Job Class  Educational Level  Marital Status  Children  Region of Origin  *Does not include respondents who do not work for Vale  4.4.2. Administration of the Quality of Life Survey Questionnaire a. Sampling Frame and Technique The sampling frame for the quality of life (QOL) questionnaire included all full-time Vale employees or interns who resided in the Núcleo Urbano de Carajás, Parauapebas or Canaã dos Carajás and were 18 years of age or older at the time they completed the questionnaire. Vale employees from other parts of the country working on short-term projects in Carajás were excluded from the sample.  99  As with the interviews, communications/HR liaison persons were responsible for communicating with employees about the QOL questionnaire. A schedule for the survey implementation was set, and rooms were reserved for employees to come and answer the survey questionnaire. The questionnaire was administered in all 3 mines in the following areas: ƒ  Carajás Iron-ore Mine—mine, processing plant, maintenance/mechanical shop, and administrative offices  ƒ  Azul Manganese Mine—mining, maintenance/mechanical shop and administrative offices  ƒ  Sossego Copper Mine—mine and engineering/administration office  The areas/shifts were chosen at the convenience of department co-ordinators or managers, or at specific times of high traffic, such as the beginning or end of a shift. The approach used to administer the questionnaire was efficient and adequate since the researcher was able to personally introduce the idea of the research project, speak about the objectives and methods used, give instructions on how to fill out the questionnaire, answer any questions and address any concerns. It was also made clear to employees that the surveys were collected directly by the researcher; this was done to mitigate fears associated with issues of confidentiality and the anonymity of responses. A total of 455 paper questionnaires were collected in 4 days. The questionnaire was also made available to employees with access to the Internet through Survey Monkey, an online survey tool (www.surveymonkey.com). Vale’s Human Resources department advertised the online questionnaire via email notices advising employees that they had two weeks to complete the questionnaire online. One week after receiving the first email message about the online questionnaire, employees were reminded by another email of the deadline for completing the online survey. Unfortunately, even after the above efforts to encourage online participation, only six questionnaires were completed online; this is believed to have been due to a lack of familiarity with online questionnaire tools, as well as the extremely limited access Vale workers in the Carajás region have to the Internet.  100  b. Population Sample A total of 461 completed questionnaires were collected. Only 2 questionnaires were eliminated because of a high rate of non-response items (questions not answered) 34. Of the remaining questionnaire respondents, 168 were from the Núcleo Urbano de Carajás (36.7%), 75 were from Canaã dos Carajás (16.4%), and 215 were from Parauapebas (46.9%) 35. Not surprisingly, the majority (83.2%) of the population sample were males. Only 16.8% of the respondents were female. The majority (59.4%) of the questionnaire respondents were had moved to their communities of residence quite recently (5 years or less of residency). Only 8.3 % of respondents had lived in their communities between 5 and 10 years, but 32.5% of the respondents were long term residents who had lived in their communities for more than 10 years. Almost half (48%) of the sampled population resided in company-provided accommodations (house, shared house, camp, hotel). Homeowners made up 23.6 % of the sampled population, while 28.4 % were renters. Details of the questionnaire population sample are presented in Table 4.9. The population sample illustrates the cultural diversity of the Vale workforce in the Carajás region: there were questionnaire respondents came from each of Brazil’s five major regions. However, the population sample also highlights that a significant contingent (47.2%, or almost half, of the population sample) of Vale’s Carajás workforce is from the Northeast Region. 32.5% of the sampled population were from the North Region, 12.7% were from the Southeast, and employees from the Central and South Regions represented only 4.4 % and 3.3%, respectively.  34  In 2 of the 461 questionnaires, entire pages of answers were missed, resulting in a high rate of nonresponse items (questions not answered). This is probably due to the fact that the questionnaire booklet was printed double sided (the pages missed were the back pages). In order to address this risk, instructions were given to all respondents highlighting the need to answer questions on both sides of each page. Only 2 respondents (0.4%) did not follow this particular instruction to a point that made their questionnaires unusable. 35 One respondent did not identify a community of residence.  101  Table 4.9: Frequencies of Demographic Variables Variables  Frequency  Valid %*  Núcleo Urbano  168  36.7  Canaã dos Carajás  75  16.4  Parauapebas  215  46.9  Total  458  100.0  1  -  Less than 1 year  81  17.7  More than 1 and less than 6 years  191  41.7  More than 6 and less than 10 years  38  8.3  More than 10 years  148  32.3  Total  458  100.0  1  -  Rental  130  28.4  Ownership  108  23.6  Company-provided  220  48.0  Total  458  100.0  1  -  North  149  32.5  Northeast  216  47.2  Southeast  58  12.7  Central  20  4.4  South  15  3.3  Total  458  100.0  1  -  Mine operations  327  72.3  Support  109  24.1  Other  16  3.5  Total  452  100.0  7  -  Community of Residence  Missing** Length of Residence  Missing Housing Status  Missing Region of Origin  Missing Area of Work  Missing  *Valid % excludes missing values (number of non-response items). **Missing = number of respondents who did no answer the item ***Does not include any trades or technical program. Technical and trades training might be a high school degree with a trades/technical program or a certificate from a post-high school technical/trades institution. Graduate level includes all university education. Note: For Housing Status, categories that had low frequencies were grouped with others in order to simplify analysis.  102  Table 4.9: Frequencies of Demographic Variables (CONTINUED) Variables  Frequency  Valid %*  Six months or less  105  23.2  More than 6 months and less than 1 year  106  23.2  More than 1 and less than 3 years  97  21.4  More than 3 years and less than 8 years  61  13.5  More than 8 and less than 15 years  21  4.6  More than 15 and less than 20 years  35  7.7  20 years or more  28  6.2  Total  453  100  6  -  Manager or Supervisor  26  5.7  Engineer/geologist/analyst  30  6.6  Technicians/trades people/other  399  87.7  Total  455  100.0  4  -  18-30 years of age  288  63.6  31-40 years of age  101  22.3  Over 40 years of age  64  14.1  Total  453  100.0  6  -  Female  76  16.8  Male  377  83.2  Total  453  100.0  6  -  Elementary to high school***  256  56.5  Technical or trades training  129  28.5  University level  68  15.0  Total  453  100  6  -  Length of Employment  Missing Job Class  Missing Age  Missing Gender  Missing Educational Level  Missing  *Valid % excludes missing values (number of non-response items). **Missing = number of respondents who did no answer the item *** Does not include any trades or technical program. Technical and trades training might be a high school degree with a trades/technical program or a certificate from a post-high school technical/trades institution. Graduate level includes all university education. Note: For Job Class, Age, and Educational Level categories that had low frequencies were grouped with others in order to simplify analysis.  103  Table 4.9: Frequencies of Demographic Variables (CONTINUED) Variables  Frequency  Valid %*  Single  168  37.1  Married /common law  274  60.5  Divorced/separated  11  2.4  Total  453  100  6  -  No children  199  43.9  1 to 2 children  204  45.0  3 or more children  50  11.0  Total  453  100  6  -  Marital Status  Missing Number of Dependent Children  Missing  *Valid % excludes missing values (number of non-response items). **Missing = number of respondents who did no answer the item *** Does not include any trades or technical program. Technical and trades training might be a high school degree with a trades/technical program or a certificate from a post-high school technical/trades institution. Graduate level includes all university education. Note: For Marital Status, categories that had low frequencies were grouped with others in order to simplify analysis.  The population sample also suggests a young workforce: more than half of respondents (63.6%) were 30 years old or younger, while 22.3% were middle-aged (31 to 40 years of age), and 14.1 % were over 40. No minors (i.e., persons younger than 18 years of age) answered the questionnaire 36. Despite their youth, most employees were married or living in common law (60.5 %—includes those who had re-married), while 37.1% were single (never married) and only 2.4% was divorced or separated. The majority of the sampled population (56%) had at least one dependent child at the time they answered the questionnaire; 11% had 3 or more dependent children, and 43.9 % had no dependent children. A large majority of respondents (87.7%) were technicians (e.g., OHS technicians, IT technicians, mechanics, and electricians) or other workers (e.g., machine operators, drillers,  36  Vale workforce includes only adults (18 years of age or older).  104  blasters). Most respondents (72.3%) were working in operations, while 27.6% were working in professional, administrative or other leadership and support roles in company offices. The population sample included a significant number of recently employed individuals—the great majority of respondents had been working for Vale for 3 years or less. While 18.1% of respondents had been with the company between 4 and 8 years, 13.9 % had been a Vale employee for 15 years or more. While only 15% of the respondents had a university degree, the majority (56.5%) had a high school diploma or less. A significant number of respondents (129, representing 28.5%) had a Technical or trades training/high school degree or had been exposed to some kind of trades training.  4.5  Analytical Strategy  The analytical strategy consists of the triangulation of quantitative and qualitative results. The quantitative analysis includes the statistical analysis of the questionnaire. The qualitative analysis includes the coding of qualitative interviews in Phase III and the consideration of participant observation notes taken in the field.  4.5.1. Quantitative Analysis The statistical analysis of the questionnaire data was mostly performed using the Statistics Package for Social Sciences (SPSS) for Windows Software (Version 12, SPSS Inc.). The objectives of the statistical analysis are summarized in Table 4.10. a. Investigating Questionnaire Reliability The first step in the statistical analysis of the questionnaire data was the internal reliability test. The test used was the calculation of the Cronbach alpha, also known as the coefficient alpha. Introduced by Cronbach (1951), the Cronbach alpha is a frequently used statistical parameter in  105  empirical research involving tests with numerous items. The Cronbach alpha tests whether items are sufficiently interrelated to justify their combination in an index or a scale 37. Cronbach alphas were calculated to estimate reliability of the correlation among quality of life (QOL) items and domains and thus investigate the adequacy of the questions chosen for each domain. The use of this statistical test allows for the confirmation that the indicators (questions) chosen under each domain indeed collectively tapped into the same concept. These factors were a result of a comprehensive literature review and interviews with Vale’s employees and community members. The Cronbach Alpha ( α ) is calculated as:  α=  N×R [1 + (N − 1)× R ]  Where N is the number of factors and R is the average inter-item correlation among the factors(Garson, 2006). For social sciences research, the widely accepted cut-off for Cronbach alpha values is 0.70 or higher for a set of items to be considered a coherent scale. This is because when the alpha is 0.70, the standard error of measurement will be slightly over half (0.55) of a standard deviation (Garson, 2006). All the Cronbach alphas were calculated using SPSS for Windows Reliability Analysis Tool. b. Calculating Mean Scores Once reliability of the domains was established, the mean scores for each of the QOL Domains —economic, social, work, health and environment—as well as the Subjective Wellbeing Index (SWI) were obtained. The mean is the simple measure of average level of satisfaction:  M =∑  Yi n  37  Under the assumption that the items are parallel, measurements have identical true scores and uncorrelated errors have equal variances (Cronbach, 1951; Lord and Novick, 1968; Cortina, 1993; Miller, 1995).  106  Where n is the sample size, Yi is the score for the i  th  item. The scores in the questionnaire vary  from 0 to 10. The QOL means considered were the average scores with quality of life calculated using all the questions under every domain—an average score of the 31 QOL Domain questions in the questionnaire. c. Comparing Mean Scores Another important outcome of this research is the comparison of the questionnaire results of each of the 3 case study communities with one another, as well as comparisons among subpopulations (characterized by demographic variables such as age, job class, region of origin, etc.) of respondents. These comparisons were mostly achieved by comparing means using analysis of variance (ANOVA) or the Welch’s F Test (used when Levene’s test indicates that there are unequal variances among groups). For the variable gender —the only variable with only two categories—independent t tests were used to determine statistical significance of mean differences. A detailed discussion of these results, indicating when Welch F tests were needed, is presented in Chapter 5: Results. ANOVA is a statistical technique for comparing means of more than two variables. ANOVA results indicate if the null hypothesis—that there are no statistically significant differences between the variables’ mean scores (p values < alpha)—should be rejected. The confidence level desired for each test defines alpha values. The choice of the confidence level is discussed further in this chapter. The 3 assumptions on which ANOVA is based are independence, normality and homogeneity of variance (HOV) (Sirkin, 2006). For this specific questionnaire, independence is guaranteed by the fact that each questionnaire was answered individually, and normality 38 was assumed. The homogeneity of variance assumption deals with the population variance, not with the population  38  Normality is the assumption that all variables and all combinations of the variables are normally distributed. When the assumption is met, the residuals (the difference between an observed value of the response variable and the value predicted by the model) are normally distributed and independent, the differences between predicted and obtained scores (the errors) are symmetrically distributed around a mean of 0 and there is no pattern to the errors. Screening for normality may be undertaken in either statistical or graphical methods (Moore and McCabe, 1993).  107  sample, but because there are unequal numbers of subjects in each group (e.g., different number of respondents from each community of residence), this assumption needs to be tested (Sirkin, 2006). In order to assess the applicability of ANOVA or of the Independent t tests, Levene’s tests of Equality of Error Variances were run along with ANOVA. Levene’s tests are used to determine if variances among groups are homogeneous. Unlike ANOVA, Levene’s test is a “retain support hypothesis”, which means that significance (p values < alpha) indicates a significant difference among variances, and therefore the homogeneity of variance assumption is not satisfied 39. If the variance of means among groups is significantly different, ANOVA is not applicable. Therefore, when Levene’s tests indicated p values larger than alpha values, the Welch's F Test was used. Welch’s Tests is a multiple group version of the Welch-Satterthwaite t test 40, and because it does not assume variance equality, it is more robust than ANOVA when variances are not equal among groups (Hayes, 2005). The t test (independent or dependent) is the most commonly used method to evaluate the differences in means between two groups (Sirkin, 2006). The groups can be independent (e.g., different individuals answering each question) or dependent (e.g., the same individual answering a question before and after an event). For the variable gender, the independent t test was used since a different individual answered each questionnaire. This study is of an exploratory nature, and therefore hypotheses on the way different categories in each variable would potentially affect responses were not established. However, decisions were made regarding which variables would be tested with each demographic variable; it was important to do this in order to reduce the number of tests, since the larger the number of tests, the less reliable results are. Relevance of tests was heuristically 41 ascertained through assumptions based on participant observation and common sense. For instance, region of origin was considered of relevance to the environment domain because one’s previous experience of residence can impact one’s perception of the current surrounding environment. 39  For this reason, confidence levels for Levene’s test were set at 95%, not at 97.5%. The Welch-Satterthwaite t test is an alternative to the two sample t-test with equal variance, and is used when one cannot assume that the two populations have equal variances (Chung, 2004) 41 The term “heuristic” refers to a set of common-sense rules. “Heuristically” then means it is based on common sense or previous experience. 40  108  For the comparison of means in this research, the confidence level for ANOVA was established at 97.5% (alpha = 0.025), to minimize the incidence of Type I errors, since even though tests were not run for every possible combination of variables, the number of tests run was still quite high. The level of significance denoted by alpha is the probability of a Type I error. Type I errors occur when the null hypothesis is rejected when it is in fact true (Jaisingh, 2000). Table 4.11 exhibits the variables that were deemed of interest for each domain. Most variables were considered of interest or relevance to 2 or more quality of life (QOL) domains. The variable area of work was considered of relevance only for the work domain. Four variables—community of residence, gender, educational level and dependent children—were considered of interest for all of the domains. The importance of investigating the variable community of residence is directly related to one of the main objectives of this research: to investigate if the mining community model impacted perceptions of quality of life, and if so, how. Gender and age were considered of importance because these are critical variables for human resources management, and any corporate policy needs to address differences and design strategies to reach and retain diversity. Educational level was considered of importance because it is directly related to the quality of the workforce. The number of dependent children variable was chosen because it can be interpreted as “having a family”, which is expected to change one’s perception of most domains of quality of life. Both ANOVA and Welch’s F Tests can determine if there is at least one inequality among sample mean scores, but they do not indicate where the difference or differences lie, since only one inequality is enough to negate the null hypothesis. This is because these tests examine all sample mean scores at once, and it is possible that some pairs of mean scores may not be significantly different from one another. In order to establish inequalities, it is necessary to complete a follow-up step—a post-hoc procedure. Only post-hoc tests of multiple comparisons can determine where inequalities lie (Sirkin, 2006).  109  Table 4.10: Quality of Life Questionnaire Statistical Analysis Analysis Goals  Statistical Analyses  Interpretation and Comments  Establishing reliability: Are the domains and facets chosen at the conceptual level adequate?  Calculate Cronbach’s Alpha (α) values for domains and facets.  Cronbach’s α is the common test of whether items are internally consistent, and sufficiently interrelated to justify their combination in an index. Alphas of .7 or higher would be satisfactory to confirm if domains defined in this proposal’s conceptual framework are indeed “contributors” to overall QOL.  Comparing domains’ mean scores: What appear to be the most and the least satisfying aspects of QOL in the case study communities? How do these vary in sub-groups by gender, age, etc.?  ANOVA, Welch F tests or independent t tests were run for mean scores for satisfaction level regarding the domains of quality of life: social, economic, health, environment, and work.  These tests will determine if there are statistically significant differences among the different communities and other variables.  Comparing QOL and LAW scores: In which of the 3 communities do employees report higher satisfaction with general QOL and LAW?  ANOVA, Welch F tests or independent t tests for satisfaction level with overall quality of life.  Tests will determine if there are statistically significant differences among the mean scores for QOL and LAW in the different communities.  Defining LAW and QOL predictors: What appear to be the most important factors—the predictors—of satisfaction with overall QOL/ LAW?  Calculate standardized regression coefficients (Betas) for mean scores for satisfaction various domains and facets.  The domains associated with significant Betas (p < 0.05 for confidence level at 95%) are determined to uniquely predict QOL.  Comparing predictors within populations: How differently do the 3 communities define the predictors of overall QOL?  Comparison of standardized regression coefficients (Betas).  The domains associated with significant Betas (p<0.05 for confidence level at 95%) are determined to uniquely predict QOL.  Investigating possibility of using SWI as a proxy for the QOL questionnaire  Determine the correlation between SWI and Average QOL as a way to test the adapted SWI scale’s ability to measure quality of life as compared to the 31-item QOL survey.  A high correlation would mean that there is an overlap in measures, and therefore a justification to use the shorter scale in case there is a need for a short instrument.  110  Table 4.11: Variables of Interest for QOL Domains Domains Variables  Economic  Environment  Work  Social  Health  Community of Residence  X  X  X  X  X  Length of Residence  X  Housing Status  X  Region of Origin  X  X  Job Class  X  X  Length of Employment  X  X  Area of Work  X  X  X  Age  X  X  X  X  X  Gender  X  X  X  X  X  Educational Level  X  X  X  X  X  Marital Status  X  X  X  Dependent Children  X  X  X  X  X  Note: Bold variables indicate indicators that were tested for all categories.  For variables with 3 categories, such as community of residence (Núcleo Urbano de Carajás, Parauapebas or Canaã dos Carajás), the Fisher Least Significant Difference (LSD) post-hoc test was chosen. The LSD test is simply based on the rationale that if an omnibus test is conducted and is significant, the null hypothesis is incorrect (If the omnibus test is nonsignificant, no post-hoc tests are conducted). For sub-groups with more than 3 categories, Tukey’s post-hoc tests were preferred. Also known as Tukey’s HSD for “honest significant difference”, Tukey’s test calculates a new critical value that can be used to evaluate whether differences between any two pairs of means are significant. The critical value is somewhat different because it involves the mean difference that has to be exceeded to achieve significance. So it is necessary to calculate one critical value and then the difference between all possible pairs of means. Each difference is then compared to the Tukey critical value. If the difference is larger than the Tukey value, the comparison is significant (Newsom, 2006).  111  The critical value ( d T ) is as follows:  d T = qT  MS s / A n  Where qT is the standardized range statistic, similar to the t-critical values. MS s / A is the mean square error from the overall F-test, and n is the sample size for each group. Studies have shown that Tukey’s test has greater power than the other tests under most circumstances when group sizes ( n ) are unequal (Newsom, 2006). d. Identifying QOL Predictors One of the most important outcomes of the quantitative analysis in this research is the identification of predictors of quality of life for each community: the Núcleo Urbano de Carajás, Parauapebas and Canaã dos Carajás. In the social and natural sciences, multiple regression procedures are widely used in research (Wonnacott and Wonnacott, 1985; Levin and Fox, 2000). The general purpose of multiple regression is to learn more about the relationship between several independent or “predictor” variables and a dependent or criterion variable. Generally, multiple regression allows for the establishment of the best predictor (dependent variable) of an independent variable. However, it is important to note that a major conceptual limitation of all regression techniques is that one can ascertain relationships, but never be sure about the underlying causal mechanism. In order to define predictors, multiple linear regression analysis is used to define the coefficient of determination while analyzing the separate and the combined influences of two or more independent variables on a dependent variable. Tests of the assumption of correctly specified form (linearity) were completed with the use of ARC software and are available in Appendix C. Generally, multiple linear regression procedures estimate a linear equation of the weighted sum:  Y = β  0  + β1X  1  + β  2  X  2  + K + βiX  i  + ε  112  Where Y is the predicted score on the dependent variable, each X is one of the independent variables, β is the value of Y when X = 0 (the intercept), and each β i is a standardized regression coefficient and is interpreted as the change in outcome for each standard deviation unit 42, and ε is error. The standardized regression coefficients β are the Beta weights, and these indicate the relative importance of the corresponding independent variable in determining the predicted value of the dependent variable (Colman and Pulford, 2006). Statistically significant values of Betas indicate that the independent variable uniquely predicts variation of the dependent variable. However, interpretation is not complete when standardized regression coefficients (Betas) are calculated. Interpretation needs to take into consideration the coefficient of multiple determination—the R squared. The R squared indicates how much the regression equation can predict the dependent variable as it determines the proportion of variation in Y that is explained by the independent variable X in the regression model (Krehbiel, 2004). The coefficient of determination is equal to the regression sum of squares (explained variation) divided by the total sum of squares (total variation):  R2 =  SSR SST  Where SSR is the regression of sum squares and SST is the total sum of squares. SST is the sum of the squared differences between each observed Y and the average value of Y :  (  SST = ∑i=1 Y1 − Y n  )  2  Where Y is the average value of Y (Krehbiel, 2004).  42  The standardized regression coefficients (Betas) are often used instead of the regression coefficients (b) in an attempt to solve the problem of units of measurement. All variables (response and predictors) are standardized by subtracting the mean and dividing by the standard deviation. The standardized regression coefficients then represent the change in response for a change of 1 standard deviation in a predictor.  113  The regression sum of squares (explained variation) is equal to the sum of the squared differences between the predicted value of Y and the average value of Y :  (  ) n SSR = ∑i =1 Y1 − Y  )  2  )  w  Where Y is the predicted value of Y and Y is the average value of Y (Krehbiel, 2004). In this research, to define predictors of the dependent variable satisfaction with quality of life, linear regression was used to define the coefficients of determination while analyzing the separate and the combined influences of the independent variables (satisfaction with quality of life domains) on the dependent variable (satisfaction with quality of life). A multiple regression equation was defined for the total sampled population (all QOL questionnaire respondents) to determine how satisfaction with each domain predicts satisfaction with quality of life in the sampled population 43. The regression equation therefore would be:  Y QOL = β social X social + β health X health + K + β env . X env . + β econ . X econ . + β work X work + ε Where YQOL is satisfaction with quality of life, β social is the standardized b coefficient (beta) for the social domain, X social is the mean score for social domain (the same is applicable to other domains), and ε is error. Even though the equation above suggests a specific order of independent variables, the predictors in the regression equation have no order, and one cannot be said to enter before the other. Regression equations were also defined for 3 different sampled populations: residents of the Núcleo Urbano de Carajás, Parauapebas residents and Canaã dos Carajás residents.  43  Once reliability is confirmed with the calculation of Cronbach alphas, domain mean scores will be calculated as the average of means for domain indicators. Standardized b Coefficients (Betas) will be defined for each QOL domain.  114  Calculated R 2 of each regression equation determine the proportion of variation in satisfaction with QOL that is explained by satisfaction with each of the 5 domains: social, economic, environment, health and work. The standardized regression coefficients and regressions equations’ coefficients of determination were calculated using the SPSS Linear Regression tool. e. Investigating Life as Whole (LAW) Similar to QOL, in order to compare Life as a Whole (LAW) scores from the 3 case study communities, analysis of variance (ANOVA) was performed only for the variable community of residence. Similar to the method to identify satisfaction of quality of life predictors, multiple regression equations were established for LAW as the dependent variable, with the independent variables being the scores for the 5 QOL domains.  4.5.2. Qualitative Analysis Analysis of qualitative data is not a structured, static or rigid process. Indeed, it is a free-flowing and creative one in which researchers need to move back and forth between types of coding, using analytic techniques and procedures freely and in response to the data being analyzed (Strauss and Corbin, 1998b). The qualitative analyses in this research include the coding of interview data as well as the use of participant observation field notes. Analysis of observational data required the researcher to code and label field notes, while these are examined for patterns and relationships (Jorgensen, 1989). Based on the conceptual model for this research, the researcher used field notes to support the interpretation or explanation of quality of life (QOL) factors in the case study communities through the participants’ as well as the researcher’s perspectives. The qualitative data from interviews were also analyzed with open coding, “the analytic process through which concepts are identified and their properties are discovered in data” (Strauss and Corbin, 1998). Coding of the first question about QOL classified data in the categories  115  (domains) established in the conceptual model developed for this study: environmental, social, health, work, and economic. Coding of the answers to the 2nd, 3rd and 4th questions about the best and worst things about the communities and the priorities for intervention resulted in data being classified in concepts within more fluid categories, since answers usually presented concepts that could fall under more than one of the 5 categories (i.e., environmental, social, health, work, and economic). When coding, the researcher looked for answers to questions such as “Why?”, “Where?”, “When?”, and “With what results?”, in order to uncover relationships among categories. Extra content volunteered by the interview participants (i.e., information not related to the 4 interview questions) was also coded and considered in the qualitative analysis.  116  CHAPTER 5: RESULTS This chapter presents the results of the quantitative and qualitative analyses of the data collected in Phase III of this research. In the first section of this chapter, the quantitative results obtained with the statistical analysis of the Quality of Life (QOL) questionnaire are presented. In the second section, the results from the qualitative interviews are presented and supported with participant observation data. A detailed discussion and the triangulation of results from both quantitative and qualitative data analyses are presented in Chapter 6: Discussion.  5.1.  Quantitative Analysis Results  The quantitative analysis refers to the data from the Quality of Life (QOL) questionnaire, which was examined through statistical analysis using the software Statistical Package for the Social Sciences (SPSS).  5.1.1. Establishing Reliability The first level of statistical analysis in this research was the investigation of the reliability of the items in the QOL questionnaire. This was a logical first step since it ascertained if the indicators —i.e., the questions chosen for each QOL domain—tapped into the same concept and could therefore be considered as part of the same domain. Reliability was also established for Part 1: Subjective Wellbeing Index (SWI) and for Part 2: QOL questionnaire. Reliability was tested with the calculation of Cronbach alphas. The Cronbach alphas calculated for the QOL domains in the survey were between 0.73 and 0.83. The Cronbach alpha for Part 2 (31 items) was 0.82, and the alpha for the SWI (7 items) was 0.71 44. These results confirm that a) the items identified and questions for the domains indeed are cohesive enough to justify their use in an index; and b) the questions chosen for the first part of the questionnaire—the SWI adequately represent a single concept. The results achieved for the Cronbach alphas are illustrated in Table 5.1.  44  Since an acceptable minimum value for Cronbach alphas in the social sciences in 0.70 (Garson, 2006).  117  Table 5.1: Domain Reliability Tests – Cronbach Alphas Section  Item  Part 1  Subjective Wellbeing Index (SWI)  Indicators  Cronbach Alpha  Economic situation Health Achievements in life Personal relationships  0.71  Physical environment Community Work Part 2  Work Domain  Number of hours of work per week Relationship with co-workers Autonomy and decision making ability Relationship with immediate supervisor  0.83  Opportunities for professional growth Occupational safety at the workplace Job security Economic Domain  What you can buy and do with your money Material belongings Economic situation of surrounding community  0.83  Ability to save for the future Financial stability Environment Domain  Place of residence (house/apartment/camp) Quality and integrity of the natural environment around you General order, cleanliness and security in the community Basic infrastructure in the community  0.82  Availability of goods and services Transportation options and quality in your community Health Domain  Stress level in your life Physical activity level The food you eat Entertainment and recreation options in your community  0.74  Health services and facilities in your community Work-life balance Social Domain  Relationship with family Love life Relationship with friends Privacy  0.73  Participation in community events Participation in community decision making Educational opportunities  118  5.1.2. Calculating Mean Scores Once reliability of the domains was established, mean scores for all items in each of the QOL domains—i.e., economic, social, work, health and environment—as well as the SWI items were calculated. Scores were from 0 to 10, with 0 being “completely dissatisfied”, 5 being “neutral”, and 10 being “completely satisfied”. Table 5.2 presents all means and standard deviations for each of the 31 domain items. Table 5.2: Indicators’ Mean Scores  Item  Núcleo Urbano de Carajás  Canaã dos Carajás  Parauapebas  Mean  Std. Dv.  Mean  Std. Dv.  Mean  Std. Dv.  What you can do and buy with your money  5.58  2.47  5.55  2.19  5.90  2.46  Material belongings  6.42  2.31  5.93  2.20  6.02  2.47  Your community's economic situation  5.94  2.29  4.61  2.14  5.08  2.65  Your ability to save for the future  5.50  2.44  5.85  2.65  5.69  2.97  Your financial stability  5.35  2.42  5.28  2.40  5.27  2.48  The number of hours you work per week  6.12  2.75  7.01  2.55  7.09  2.78  Relationship with co-workers  8.07  1.70  8.34  1.72  8.77  1.75  Relationship with supervisors  7.57  2.25  8.15  2.09  8.25  2.17  Opportunities to grow professionally  6.25  2.80  7.14  2.74  7.47  2.59  Safety at work  7.93  2.25  8.57  1.76  8.46  2.15  Ability to make decisions on your own  6.82  2.41  6.70  2.51  7.55  2.14  Job stability  6.42  2.37  6.26  2.31  6.86  2.76  Salary and benefits  5.60  2.47  5.70  2.49  6.00  2.92  Your home  7.78  2.18  5.80  3.03  5.87  3.20  Access to nature  7.95  2.08  5.52  2.94  6.26  2.82  Order, cleanliness and security  8.40  1.94  5.15  2.67  5.86  3.19  Basic infrastructure  9.02  1.56  3.69  2.81  4.55  3.05  Services and goods available  5.46  2.77  3.88  2.34  5.40  2.71  Transportation options  6.03  2.70  4.12  2.75  5.05  3.08  Economic Domain  Work Domain  Environment Domain  119  Table 5.2: Indicators’ Mean Scores (CONTINUED)  Item  Núcleo Urbano de Carajás  Canaã dos Carajás  Parauapebas  Mean  Std. Dv.  Mean  Std. Dv.  Mean  Std. Dv.  Stress levels  5.60  2.43  5.47  2.50  6.22  2.57  Physical activity level  4.83  2.89  5.22  2.84  5.43  3.08  The food you eat  6.70  2.28  6.57  2.44  6.97  2.58  Entertainment options in your community  5.27  2.85  3.64  2.77  3.99  2.85  Health services in your community  5.29  2.77  3.24  2.49  4.36  2.92  Ability to balance life/work  6.71  2.39  7.10  2.52  7.81  2.12  Relationship with your family  8.66  1.50  8.77  1.79  9.20  1.61  Love life  8.35  1.96  7.89  2.74  7.98  2.79  Relationship with friends  8.37  1.52  8.18  2.04  8.91  1.53  Privacy  6.93  2.52  7.50  2.56  7.96  2.13  Participation in community events  6.36  2.29  5.84  2.58  5.88  2.64  Participation in community decision-making  4.73  2.44  4.31  2.52  4.69  2.69  Educational opportunities  6.49  2.89  4.16  2.92  5.50  2.89  Health Domain  Social Domain  The means are a simple average measure—the sum of all scores divided by the number of scores. Comparisons of the domain mean scores by community of residence were completed for each domain individually. Comparisons of mean scores for individual indicators were only investigated in specific cases and will be discussed further in this chapter. However, it is noted that residents from the 3 communities consistently rated 2 items of the SWI scale similarly: personal relationships with the highest scores and economic situation with the lowest scores (Table 5.2). Because Cronbach alpha results confirm that all the indicators in each domain are cohesive enough to represent one concept, the average mean scores for each domain were calculated as the average of the means for the indicators listed under each domain. The means for average quality of life (AQOL) were calculated as the average scores of all of the 31 questions under every QOL domain. All of the mean scores for all 5 domains, AQOL, the 120  Subjective Wellbeing Index (SWI), the single question of QOL, and the single question on life as a whole (LAW) are displayed in Table 5.3. The results in Table 5.3 highlight the most and least satisfying domains in each community and are only the first step in comparing the perceptions of quality of life in the 3 sample populations: the Núcleo Urbano de Carajás, Parauapebas and Canaã dos Carajás. The examination of significant differences among mean scores required a variety of tests.  5.1.3. Comparing Means As discussed in detail in Chapter 4, when deciding which tests to run, 5 variables were considered of relevance for all of the QOL domains, for the AQOL and for the Subjective Wellbeing Index (SWI): community of residence, age, gender, educational level, and number of dependent children (Table 4.10). These variables were heuristically ascertained through assumptions based on participant observation and common sense. Therefore, when running t tests, analysis of variance (ANOVA) or Welch Tests for AQOL, SWI and each of the 5 domains, the 5 chosen variables were tested, as well as other variables that were considered of specific interest for each domain. These variables are listed and discussed under each domain. Only the variables that proved, through post-hoc tests, to have significant mean score differences within categories are discussed. Because a large number of tests were necessary (all 5 domains, life as a whole—LAW—and SWI in relation to 13 demographic variables), the confidence level was set at 97.5 % (alpha = 0.025) to minimize Type I errors. The variable gender was the only variable with only 2 categories: male and female. For this reason, a simple independent sample t test was run for gender for all domains, SWI, AQOL and LAW. Results indicate that there were no significant differences between mean scores for satisfaction of males and females for any of the domains, SWI, AQOL or LAW. Results of the independent sample t tests (also at confidence level 97.5 %) are available in Appendix C.  121  Table 5.3: Mean Scores Núcleo Urbano  Std Dv.  Canaã dos Carajás  Std Dv.  Parauapebas  Std. Dv.  All  Std Dv.  7.88  1.62  7.59  1.65  7.79  1.95  7.79  1.78  Single question Quality of Life (QOL)  7.28  1.80  6.93  1.90  7.03  2.08  7.11  1.95  Average Quality of Life (AQOL)  5.84  1.16  5.33  1.21  5.73  1.26  5.72  1.23  Economic Domain (5 items)  5.76  1.99  5.44  1.73  5.59  1.93  5.63  1.92  Work Domain (8 items)  6.85  1.63  7.23  1.58  7.56  1.59  7.24  1.63  Environment Domain (6 items)  5.58  1.12  3.52  1.43  4.12  1.61  4.56  1.63  Social Domain (7 items)  6.24  1.25  5.83  1.37  6.26  1.30  6.19  1.30  Health Domain (7 items)  4.86  1.49  4.60  1.50  5.13  1.41  4.95  1.47  Subjective Wellbeing Index (SWI)  7.30  1.26  6.84  1.46  7.21  1.28  7.18  1.31  Personal Relationships  8.12  1.52  8.00  1.71  8.40  1.79  8.23  1.69  Environment  7.70  1.94  6.29  2.38  6.68  2.31  6.99  2.26  Work  7.20  2.04  7.73  2.35  8.13  1.96  7.72  2.10  Economic Situation  6.04  2.37  5.59  2.19  6.08  2.31  5.98  2.32  Accomplishments in Life  7.41  1.81  7.64  2.07  7.85  1.71  7.65  1.82  Community  6.96  2.20  5.60  2.09  5.75  2.69  6.17  2.50  Health  7.62  1.98  6.96  2.74  7.57  2.63  7.49  2.43  Items Life as a Whole Single Question Life as a Whole (LAW) Quality of Life  Subjective Wellbeing  122  5.1.4. Comparing Means: Average Quality of Life (AQOL) In addition to the 5 variables considered of interest for all domains (i.e., community of residence, age, gender, educational level and number of dependent children), the following variables were also tested: length of residence, region of origin, job class, and marital status. Tests determined that inequalities exist only among the variables community of residence [ANOVA: F (2, 447) = 4.537, p = 0.011] and educational level [Welch’s Test: F (2,442) = 4.27, p =0.014]. Tukey post-hoc tests indicate where these inequalities lie: residents of the Núcleo Urbano de Carajás and Parauapebas rated their satisfaction with quality of life similarly, while Canaã residents rated their quality of life slightly lower. Table 5.4: Statistical Tests for AQOL Category  Mean  Std. Deviation  n  5.33  1.21  72  5.73  1.26  212  5.85  1.16  166  Community of Residence ANOVA: F (2, 447) = 4.537, p = 0.011  Canaã dos Carajása b  Parauapebas  Núcleo Urbano de Carajás  b  Educational Level Levene’s test: F (2, 442) = 5.63, p = 0.004. Welch’s test F (2,442) = 4.45, p = 0.012 Technical or trades traininga University levela  5.47 5.59  1.19 0.98  126 67  Elementary to high schoolb  5.85  1.29  252  a, b  Groups in each row that share a superscript are not significantly different from each other. Confidence level: 97.5%.  As illustrated in Table 5.4, respondents with a high school degree or less rated their satisfaction with quality of life slightly higher than respondents with higher educational levels.  5.1.5. Comparing Means: Economic Domain Besides community of residence, age, gender, educational level and number of dependent children, the variables job class and length of employment were also tested for the economic  123  domain because of the direct relationship between salary and benefits and the ability to accumulate wealth while working towards economic wellbeing. Table 5.5: Satisfaction with Economic Domain by Education Level Educational Level a  Technical or trades training b  Elementary to high school University level  b  Mean  Std. Deviation  n  5.23  1.85  129  5.71  1.96  256  5.99  1.80  68  ANOVA: F (2,450) = 4.32, p=0.014 a, b  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: Fisher Least Significant Difference (LSD). Confidence level: 97.5%  Analysis of variance (ANOVA) suggested that in fact these variables do not significantly impact responses about satisfaction with the economic domain. At a confidence level of 97.5%, ANOVA indicate that differences exist only within the educational levels categories (p < 0.025). Fisher Least Significant Difference (LSD) post-hoc tests indicate that respondents with an elementary school education or high school degree and respondents with a university education reported higher levels of satisfaction than respondents with technical or trades training (Table 5.5).  5.1.6. Comparing Means: Work Domain In addition to the 5 variables considered of interest for all domains (i.e., community of residence, age, gender, educational level and number of dependent children), other variables investigated for the work domain were job class, length of employment, and area of work. These additional variables were investigated because they are related to the respondents’ type of work, experience and professional accomplishments. At a confidence level of 97.5%, the ANOVA showed that differences exist within categories of the following variables: community of residence [F (2,452) = 9.172, p < 0.001], educational level [F (2,447) = 6.708, p < 0.001] and length of employment [F (6, 444) = 7.831, p < 0.001]. Results of post-hoc tests for the work domain are discussed below in relation to each of these variables.  124  a. Satisfaction with Work Domain by Community of Residence Fisher Least Significant Difference (LSD) post-hoc tests indicate that significant differences in satisfaction with the work domain exist between the respondents that live in the Núcleo Urbano de Carajás and Parauapebas. Statistically, the mean score for Canaã dos Carajás respondents is not significantly different from the Núcleo Urbano de Carajás or Parauapebas sample populations. Table 5.6: Satisfaction with Work Domain by Community of Residence Community Núcleo Urbano de Carajás  Mean  Std. Deviation  n  6.85  1.64  167  7.23  1.58  74  7.56  1.59  214  a  Canaã dos Carajása,b b  Parauapebas  ANOVA: F (2,452) = 9.17, p < 0.001 a, b Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: Fisher Least Significant Difference (LSD). Confidence level: 97.5%  As depicted in Table 5.6, results show that Parauapebas residents reported higher satisfaction with the work domain than residents of the Núcleo. b. Satisfaction with Work Domain by Educational Level LSD post-hoc tests determined that statistically significant differences in reported satisfaction with the work domain exist between the respondents that have technical or trades training or a university-level education, and those who have an elementary-level education or a high school diploma. Respondents without any post-secondary education reported the highest levels of satisfaction with the work domain (Table 5.7). Table 5.7: Satisfaction with Work Domain by Job Class Job Class a  Technical or trades training University level  a b  Elementary to high school  Mean  Std. Deviation  n  6.92  1.66  128  6.94  1.44  68  7.49  1.62  254  ANOVA: F (2,447) = 6.708, p < 0.001 a, b  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%  125  c. Satisfaction with Work Domain by Length of Employment Table 5.8 shows the means of satisfaction with the work domain by length of employment. Table 5.8: Satisfaction with Work Domain by Length of Employment Length of Employment Between 8 and 15 years  a  More than 3 years and less than 8 years Between 1 and 3 years  a  a,b  More than 15 years and less than 20 years  a,b  a,b  20 years or more  More than 6 months and less than 1 year 6 months or less  b  b  Mean  Std. Deviation  n  6.57  1.81  21  6.70  2.29  61  7.40  2.13  97  7.46  2.24  35  7.61  1.77  28  8.21  1.94  105  8.48  1.79  104  ANOVA: F (6, 444) = 7.831, p < 0.001 a, b  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: Tukey HSD. Confidence level: 97.5%.  According to Tukey post-hoc tests, the significant differences in satisfaction with work exist among the respondents who had been with Vale for less than 1 year and employees who had been with the company for more than 3 but less than 15 years at the time they completed the questionnaire. Newer employees (i.e., those who had been working for the company for less than 1 year) reported higher satisfaction with the work domain than employees who had been working for Vale for longer periods of time.  5.1.7. Comparing Means: Environment Domain In addition to the 5 variables considered of interest for all domains (i.e., community of residence, age, gender, educational level and number of dependent children), other variables tested for the environment domain were housing status, length of residence and region of origin. The Analysis of Variance (ANOVA) showed that differences exist within several variables: housing status [F (2, 455) = 36.24, p < 0.001], job class [F (2, 452) = 3.52, p = 0.03], age [F (2, 450) = 13.554, p = 0.005], and number of dependent children [F (2,450) = 5.427, p = 0.005]. The Welch test indicates that differences exist within the community of residence variable [F (2.445) = 59.22, p < 0.001]. Results of post-hoc tests for the environment domain are discussed in relation to each of these variables.  126  a. Satisfaction with Environment Domain by Community of Residence Welch’s F test was used to examine the effect of the community of residence variable on satisfaction with the environment domain 45. The test indicates that the mean scores for satisfaction with the environment for residents of the 3 case study communities are statistically significantly different from one another (Table 5.9). The highest reported level of satisfaction is associated with residents of the Núcleo Urbano de Carajás. Parauapebas residents’ mean scores follow, while the lowest level of satisfaction is associated with Canaã dos Carajás residents. Table 5.9: Satisfaction with Environment Domain by Community of Residence Community Canaã dos Carajás  a  b  Parauapebas  Núcleo Urbano de Carajás  c  Mean  Std. Deviation  n  3.52  1.43  75  4.12  1.61  215  5.58  1.12  168  Levene’s test: F (2, 445) = 12.94, p< 0.001. Welch’s test F (2, 445) = 73.78, p < 0.001 a,b,c  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%.  b. Satisfaction with Environment Domain by Housing Status According to the Fisher Least Significant Difference (LSD) post-hoc tests, respondents within all 3 housing status categories (i.e., rental, ownership and company-provided accommodations) reported different levels of satisfaction with the surrounding environment (Table 5.10). Overall, respondents were most satisfied with accommodations provided by Vale and least satisfied with rental housing.  45  The Welch F Test was used here because Levene’s test indicated that ANOVA was not a robust test in this case. Levene’s tests are used to determine if variances among groups are homogeneous. In this case, F (2, 445) = 12.94, p< 0.001, indicating that the differences of homogeneity of variance were significant among groups.  127  Table 5.10: Satisfaction with Environment Domain by Housing Status Housing Status a  Rental  Ownership  b  Company-provided accommodations  c  Mean  Std. Deviation  n  3.76  1.62  130  4.32  1.56  108  5.15  1.43  220  ANOVA: F (2, 455) = 36.24, p < 0.001 a,b,c  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%.  c. Satisfaction with Environment Domain by Job Class The LSD post-hoc tests indicate that managers and supervisors reported statistically significantly higher levels of satisfaction than respondents who were technicians or working in trades. The levels of satisfaction reported by engineers, geologists and analysts were not statistically different from the other respondents, as showed in Table 5.11. Table 5.11: Satisfaction with Environment Domain by Job Class Job Class  Mean  Std. Deviation  n  a  4.50  1.31  399  4.56  1.37  30  5.38  1.66  26  Technicians or trades  Engineer/geologist /analysta,b Manager or supervisor  b  ANOVA: F (2, 452) = 3.52, p = 0.030 a, b  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%  d. Satisfaction with Environment Domain by Age The LSD post-hoc tests determined that all the 3 different age groups reported statistically different levels of satisfaction with the environment domain. As illustrated in Table 5.12, levels of satisfaction with the environment domain steadily increased with respondents’ age groups.  128  Table 5.12: Satisfaction with Environment Domain by Age Groups (Estimated Marginal Means) Age Group 18-30 years old  a  31-40 years old  b  More than 40 years old  c  Mean  Std. Deviation  n  4.30  1.59  288  4.69  1.64  101  5.42  1.50  64  ANOVA: F (2, 450) = 13.554, p = 0.005 a,b,c  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%.  e. Satisfaction with Environment Domain by Number of Dependent Children According to LSD tests, all 3 categories (i.e., no children, 1–2 children and 3 or more children) demonstrated statistically different means for satisfaction with the environment domain. As illustrated in Table 5.13, reported levels of satisfaction increased slightly with the number of dependent children respondents had. Table 5.13: Satisfaction with Environment Domain by Number of Dependent Children Number of Dependent Children No childrena 1 to 2 children  b  3 or more childrenc  Mean  Std. Deviation  n  4.27  1.60  199  4.76  1.59  204  4.83  1.76  50  ANOVA: F (2,450) = 5.427, p = .005 a,b,c  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%.  5.1.8. Comparing Means: Social Domain In addition to the 5 variables considered of interest for all domains (i.e., community of residence, age, gender, educational level and number of dependent children), the only other variable investigated for the social domain was length of residence. The analyses of variance showed that differences exist within categories of several variables: educational level [F (2,449) = 7.726, p = 0. 001], number of dependent children [F (2,449) = 10.002, p < 0.001] and community of residence [F (2,449) = 4.064 p < 0.001]. Results of post-hoc tests for the social domain are discussed below in relation to each of these variables  129  a. Satisfaction with Social Domain by Educational Level Fisher Least Significant Difference (LSD) post-hoc tests results suggest that there are statistically significant differences between the mean of satisfaction with the social domain reported by respondents with a high-school diploma or less and by those who have a technical or trades school diploma or a university degree, as illustrated in Table 5.14. Table 5.14: Satisfaction with Social Domain by Educational Level Educational Level a  Elementary to high school  b  Technical or trades training  University level or graduate school  b  Mean  Std. Deviation  n  6.38  1.28  255  5.90  1.35  129  5.91  1.18  68  ANOVA: F (2,449) = 7.726, p = 0.001 a,b,c  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%.  Respondents with a high-school degree or less seem to be more satisfied with the social domain than respondents with higher educational levels. b. Satisfaction with Social Domain by Number of Dependent Children According to LSD post-hoc tests, significant differences in social domain exist between respondents who have at least one child and those who do not have any children. As shown in Table 5.15, respondents with at least one dependent child reported slightly higher satisfaction with the social domain when compared to those who did not have children. Table 5.15: Satisfaction with Social Domain by Number of Dependent Children Number of Dependent Children No childrena 1 to 2 children  b  3 or more childrenb  Mean  Std. Deviation  n  5.88  1.27  198  6.35  1.24  204  6.63  1.49  50  ANOVA: F (2,449) = 10.002, p < 0.001 a,b,c  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test:  LSD. Confidence level: 97.5%.  130  5.1.9. Comparing Means: Health Domain For the health domain, analyses of variance were run only for the community of residence, age, gender, educational level, and dependent children variables. Results show that differences exist within the categories of the variables community of residence [F (2,449) = 4.064 p < 0.001] and educational level [F (2,444) = 14.328 p < 0.001]. Results of post-hoc tests for the health domain are discussed below in relation to each of these variables a. Satisfaction with Health and Community of Residence According to LSD post-hoc tests, there are no statistically significant differences in the health domain between mean scores of respondents from the Núcleo and mean scores of respondents from Parauapebas. The mean for satisfaction with the health domain for the Parauapebas sample population is slightly higher than the mean from the sample populations of Canaã dos Carajás. Statistically, the mean for the Núcleo sample populations are not significantly different from the other 2 communities (Table 5.16).  Table 5.16: Satisfaction with Health Domain by Community of Residence Community Núcleo Urbano de Carajása,b Canaã dos Carajás  a  Parauapebasb  Mean  Std. Deviation  n  4.86  1.49  167  4.60  1.50  72  5.13  1.42  213  ANOVA: F (2,449) = 4.064 p < 0.001 a, b  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%  b. Satisfaction with Health Domain by Educational Level The LSD post-hoc tests suggest that the statistically significant difference in levels of satisfaction within different educational level groups lies between the group of respondents with an elementary education or high-school diploma and the group of respondents who have technical training or a university-level education.  131  Table 5.17: Satisfaction with Health Domain by Educational Level Educational Level a  Technical or trades training University level  a b  Elementary to high school  Mean  Std. Deviation  n  4.46  1.45  126  4.94  1.35  67  5.25  1.43  254  ANOVA: F (2,444) = 14.328 p < 0.001 a, b  Groups in each row that share a superscript are not significantly different from each other. Post-hoc test: LSD. Confidence level: 97.5%  Even though the differences are not large, results suggest that individuals with basic education—elementary to high-school level —are the most satisfied with the health domain. 5.1.10. Defining QOL Satisfaction Predictors One of the important outcomes of this research is the identification of quality of life (QOL) predictors—those domains that seem to impact the perceptions of QOL in each community and for the respondents as a whole. Regression analysis was used to define predictors. The independent variables were defined as the QOL domains: social, economic, environment, work and health. These were entered concurrently in the equation using SPSS Regression Analysis Tool. Table 5.18 presents the regression model coefficients for the QOL domains. Table 5.19 illustrates how each community defines QOL predictors. Table 5.18: Model Coefficients: Regression against QOL, all communities  QOL Domain  Unstandardized Coefficients  Standardized Coefficients  Significance (p)  B  Std. Error  Beta ( β )  Economic  0.29  0.05  0.29  0.00  Work  0.03  0.06  0.02  0.66  Environment  0.12  0.05  0.10  0.03  Social  0.16  0.07  0.10  0.03  0.34  0.07  0.26  0.00  Health 2  Note: R = 0.38. Confidence level: 95%.  132  The simplified regression equation for the entire sample population would therefore be:  Y QOL  − all  = 0 . 29 X economic + 0 . 26 X  health  + 0 . 10 X  social .  + 0 . 10 X environmen  t.  +ε  Because the p value for the standardized coefficient ( β ) for the variable work was larger than 0.05 46 ( β = .024, p = 0.66), it can be inferred that the variance in satisfaction levels with the work domain does not uniquely predict variance in satisfaction with QOL in the sample population. The economic ( β = 0.29, p < 0. 001), health ( β = 0.26, p < 0. 001), social ( β = 0.10, p = 0. 03), and environment ( β = 0.10, p = 0. 03) domains were identified as predictors of QOL in the sample population. This equation predicts 38% of the variance in satisfaction with QOL scores (R2 = 0.38). The simplified regression equation for the Núcleo Urbano de Carajás sample population would be:  Y QOL  − Nucleo  = 0 . 17 X social + 0 . 23 X economic + ε  This equation indicates that QOL predictors in this community are the social ( β = 0.17, p = 0.028) and economic ( β = 0.23, p = 0.004) domains. Beta values for the other domains were not statistically significant (p values were above 0.05), and therefore these domains were not considered in the equation above. This equation predicts 42% of the variance in QOL scores (R2 = 0.42).  46  If the p value is larger than the alpha value (0.05 since the confidence level is 95%), the standardized coefficient (beta) is not considered statistically significant.  133  Table 5.19: QOL Model Coefficients per Community  Independent Variables  Standardized Coefficients  Unstandardized Coefficients B  Std. Error  Beta ( β )  Significance (p)  R2  Núcleo Urbano de Carajás Economic  0.21  0.07  0.23  0.004  Work  0.19  0.10  0.17  0.053  Environment  0.20  0.12  0.13  0.101  Social  0.25  0.11  0.17  0.028  Health  0.17  0.10  0.14  0.090  Economic  0.34  0.09  0.31  0.000  Work  -0.05  0.11  -0.04  0.643  Environment  0.96  0.10  0.08  0.323  Social  0.06  0.12  0.04  0.582  Health  0.45  0.12  0.31  0.000  Economic  0.21  0.13  0.23  0.096  Work  -0.11  0.15  0.17  0.466  Environment  0.32  0.15  0.13  0.832  Social  0.37  0.17  0.17  0.030  Health  0.50  0.16  0.14  0.002  0.42  Parauapebas  0.35  Canaã dos Carajás  0.46  Confidence level: 95% (alpha = 0.05)  The regression equation for Canaã dos Carajás is:  Y QOL  − Canaa  = 0 . 14 X  health  + 0 . 17 X  social  + ε  This equation indicates that QOL predictors in this community are the health (ß = 0.14, p = 0.03) and social (ß = 0.17, p = 0.002) domains. Beta values for the other domains were not statistically significant (p values were higher than 0.05). and therefore these domains were not considered in the equation above. This equation predicts 46% of the variance in satisfaction with QOL scores (R 2 = 0.46).  134  The regression equation for Parauapebas is:  Y QOL  − Parauapeba  s  = 0 . 31 X  health  + 0 . 31 X economic + ε  This equation indicates that QOL predictors in this community are the health ( β = 0.31, p < 0.001) and economic ( β = 0.309, p < 0.001) domains. Beta values for the other domains were not statistically significant (p values were higher than 0.05), and therefore these domains were not considered in the equation above. This equation predicts 35% of the variance in QOL scores (R 2 = 0.35). 5.1.11. Defining LAW Satisfaction Predictors Life as a whole (LAW) was analyzed as a separate variable. The purpose of investigating LAW was to contribute to the subjective wellbeing literature by analyzing the predictors of LAW in the sample population. To identify predictors of LAW in the sample population, an approach similar that taken in determining predictors of quality of life (QOL) for this research was taken. Table 5.20 presents the LAW regression model coefficients for the QOL domains. The LAW regression equation for the entire sample population is:  Y  LAW  − all  = 0 . 27 X  work  + 0 . 36 X  economic  + ε  The work domain ( β = 0.27, p = 0.004) and the economic domain ( β = 0.363, p <. 0.001) are the predictors of satisfaction with life as a whole (LAW) in the sample population. The other domains did not prove to uniquely predict variance in satisfaction scores with LAW ( β values for the other domains were not statistically significant if p values were above 0.05), and therefore were not considered in the equation above. This equation predicts 22% of the variance in satisfaction with LAW scores (R2 = 0.22).  135  Table 5.20: LAW Model Coefficients per Community Unstandardized Coefficients  Independent Variables  B  Standardized Coefficients  Std. Error  Beta ( β )  Significance (p)  R2  Núcleo Urbano de Carajás Economic  0.158  0.075  0.194  0.037  Work  0.122  0.101  0.123  0.231  Environment  0.433  0.128  0.300  0.001  Social  -0.036  0.117  -0.028  0.756  Health  -0.052  0.101  -0.048  0.609  Economic  0.178  0.121  0.187  0.146  Work  0.145  0.147  0.140  0.326  Environment  -0.239  0.145  -0.207  0.104  Social  0.123  0.161  0.097  0.449  Health  0.480  0.152  0.433  0.002  Economic  0.373  0.087  0.363  0.000  Work  0.332  0.115  0.271  0.004  Environment  -0.171  0.098  -0.141  0.082  Social  -0.118  0.116  -0.080  0.312  Health  0.124  0.117  0.090  0.288  Economic  0.373  0.087  0.363  0.000  Work  0.332  0.115  0.271  0.004  Environment  -0.171  0.098  -0.141  0.082  Social  -.0118  0.116  -0.080  0.312  Health  0.124  0.117  0.090  0.288  0.22  Canaã dos Carajás  0.38  Parauapebas  0.26  All 3 Communities  0.22  Confidence level: 95% (alpha = 0.05)  The life as a whole (LAW) regression equation for the Núcleo Urbano de Carajás is:  Y LAW  − Nucleo  = 0 . 30 X  environmen  t  + 0 . 19 X  economic  + ε  136  This equation indicates that QOL predictors in this community are the environment ( β = 0.30, p = 0.001) and economic ( β = 0.194, p = 0.037) domains. Beta values for the other domains were not statistically significant (p values were above 0.05), and therefore these domains were not considered in the equation above. This equation predicts 22% of the variance in LAW scores (R2 = 0.22). The LAW regression equation for Canaã dos Carajás would be:  Y LAW  − Canaa  = 0 . 43 X  health  + ε  This equation indicates that the health domain is the only domain that uniquely predicts variation in satisfaction with LAW in this population ( β = 0.433, p = 0.002). Beta values for the other domains were not statistically significant (p values were above 0.05), and therefore these domains were not considered in the equation above. This equation predicts 38% of the variance in LAW scores (R2 = 0.38). The LAW regression equation for Parauapebas would be:  Y LAW  − Parauapeba  s  = 0 . 27 X  work  + 0 . 36 X  economic  + ε  This equation indicates that LAW predictors in this community are the work ( β = 0.27, p = 0.004) and economic ( β = 0.363, p < 0.001) domains. Beta values for the other domains were not statistically significant (p values were above 0.05), and therefore these domains were not considered in the equation above. This equation predicts 26% of the variance in satisfaction with LAW scores (R2 = 0.26)  5.1.12. Subjective Wellbeing Index The purpose of pursuing a pilot test of a Subjective Wellbeing Index (SWI) with 7 questions in Part 1 of the quality of life (QOL) questionnaire was to investigate the possibility of using a smaller questionnaire instrument as an alternative to the 31-item QOL scale.  137  The first step of this investigation was involved confirming that the items defined as part of the SWI (Table 5.21) were cohesive enough to be part of one index. These items were presented in the first page of the questionnaire as single items. Confirmation was obtained with an internal reliability test by calculating the Cronbach alpha. The Cronbach alpha obtained for the SWI is 0.71 (Table 5.1), which is considered acceptable and confirms the cohesion of the items. Table 5.21: QOL Model Coefficients for SWI  SWI Items  Unstandardized Coefficients  Standardized Coefficients  β  (Beta)  Sig. (p)  B  Std. Error  Personal relationships  0.09  0.02  0.13  0.000  Surrounding Environment  0.05  0.02  0.10  0.003  Work  0.10  0.02  0.17  0.000  Economic situation  0.14  0.02  0.28  0.000  Accomplishments in life  0.05  0.02  0.08  0.031  Community  0.14  0.02  0.29  0.000  Health  0.14  0.02  0.27  0.000  R2  0. 64  Notes: R2 indicates that the percentage of the variance in QOL scores that the regression equation predicts is 64%. Confidence level 95%.  The second step was the confirmation that the items in the SWI could be collectively interpreted as a measurement of satisfaction with quality of life. Multiple regression analysis was employed and all SWI items were expected to contribute unique variance to confirm that these items uniquely predict variance in scores of the single question on satisfaction with quality of life. Table 5.20 shows that all the SWI items (independent variables) are predictors of QOL  47  since p  values are smaller than 0.05 when the confidence level is 95%, indicating statistically significant standardized coefficients. Moreover, the complete regression equation (including all 7 items) predicts 64% (R2 = 0.64) of the variance in scores for the single question, “How satisfied are you with your quality of life in general?” The correlation between the scores for the SWI and for average quality of life (AQOL) was calculated as a way to test the ability of the SWI to measure quality of life as compared to the 31  47  The dependent variable is the single question of satisfaction with quality of life.  138  items in the QOL domains. A high positive correlation indicates an overlap in measures and provides a rationale for the further investigation of the SWI instrument as a shorter alternative to the 31-item QOL instrument. Table 5.22: Pearson Correlations for SWI and Average QOL Pearson Correlations SWI (7 items) AQOL (31 items in 5 domains)  SWI  AQOL  1  0.79*  0.79*  1  *Correlation is significant at the 0.01 level.  Table 5.22 displays the correlations values for the adapted SWI and QOL, indicating a strong positive correlation. A detailed discussion on the implications of the results of this pilot test is presented in Chapter 6: Discussion.  5.2.  Qualitative Analysis Results  Phase III of this research included structured qualitative interviews with mine employees residing in the Núcleo Urbano de Carajás, Parauapebas and Canaã dos Carajás and. This Phase also included participant observation during 2 site visits. Concepts were defined through coding of the interviews and supported by participant observation notes. Through this analytic process, concepts were identified and their properties discovered in the qualitative data. Categories were identified from the conceptual framework defined for this project: social, environmental, economic, health and work domains of quality of life. With the coding process, it was also observed that some issues were brought up outside of the frame of positive and negative nature. These issues are presented in this section, under sub-sections called “Other Major Themes”. As with the quantitative data, the qualitative findings presented in this chapter are further interpreted and discussed in Chapter 6.  5.2.1. Interview Results: Defining Quality of Life The first question, “What is quality of life for you?”, was analyzed at the broader level, with all interviews considered together, independent of community of residence. The following questions, about the best and worst aspects of each of the 3 case study communities, as well 139  as priorities and suggestions for improvement, were coded and analyzed with each individual community considered separately. Quality of life (QOL) concepts were identified in open coding responses to the question “What is quality of life for you?” The following are quotes that illustrate the variety of views of QOL within the sample population: “Quality of life is about doing what you enjoy, be it at work or within the community; for me, it also includes helping others.” (Male employee living in Canaã dos Carajás) “Quality of life is peace and knowing you have access to good health services, to live well with family and to be able to improve yourself.” (Male employee living in Canaã dos Carajás) The main themes extracted from open coding of the interviews were then classified into categories following the framework of the conceptual model for this study. The categories and corresponding concepts obtained in interviews are illustrated in Table 5.23. Table 5.23: Major Concepts – How Do You Define Quality of Life? Category (Domain) Environmental Aspects  Concepts  ƒ ƒ ƒ  Adequate access to good housing Adequate access to services and facilities for: o Health care o Entertainment and recreation o Community interaction Basic infrastructure (water, sanitation, power) Security Quietness  Economic Aspects  ƒ  Having financial ability to meet basic needs  Health Aspects  ƒ ƒ ƒ ƒ ƒ ƒ ƒ ƒ ƒ  Life-work balance Spiritual wellbeing Mental wellbeing Physical health Access to good health care services Hygiene: water, sewage infrastructure, garbage collection Access to exercise facilities or space Access to entertainment and recreational options Freedom from stress and fatigue  ƒ ƒ  140  Table 5.23: Major Concepts – How Do You Define Quality of Life? (CONTINUED) Category (Domain)  Concepts  Work Aspects  ƒ ƒ ƒ ƒ  Professional achievements Employment opportunities Good relationships at work Good compensation  Social Aspects  ƒ  Family: o Family relationships, o Proximity to family members Good friendships Community involvement Access to education Access to social spaces  ƒ ƒ ƒ ƒ  5.2.2. Núcleo Urbano de Carajás a. Most Satisfying Things about the Núcleo Urbano de Carajás “Sometimes I feel like I am part of a big family—in many places in the Núcleo I feel at home. It is a weird relationship we end up having with the place.” (Female employee living in the Núcleo Urbano de Carajás) “I think that because you live inside the company, you don’t stress too much— this is quality of life. It’s really convenient.” (Female employee living in the Núcleo Urbano de Carajás) The positive aspects of the Núcleo Urbano de Carajás were identified through a structured and open-ended question. Participants were asked to list the 3 “best things” about living in the Núcleo. Table 5.24 depicts the concepts identified during interviews and classified in the QOL domains, as well as the frequency with which each concept was mentioned during the interview process (all Phase III interviews). Environmental qualities were the most frequently identified as the best things about living in the Núcleo Urbano de Carajás (51). Social aspects were also often mentioned as best things about the Núcleo (9), as well as work-related issues (6) like employee benefits (e.g., housing, health care, school for children, etc).  141  Table 5.24: Best Things about Living in the Núcleo Urbano de Carajás QOL Domain Environment  Social  Work  Concept  Frequency*  Access to nature  13  Good security  12  Quietness  10  Convenience of facilities and work close by  9  Quality of the urban space  5  Quality of basic infrastructure  2  Total  51  Ease to socialize and make friends  4  Cultural and socio-economic diversity  3  Educational opportunities close by  1  Sense of community  1  Total  9  Work related benefits  4  Opportunities for professional growth  2  Total  6  *Refers to number of times a concept was mentioned during interview process  b. Least Satisfying Things about living in the Núcleo Urbano de Carajás “This feels like Big Brother 48—people are watching you all the time, everybody is watching you, judging. Lots of silly gossip comes up from nothing, really.” (Female employee living in the Núcleo Urbano de Carajás) The negative aspects of the Núcleo Urbano de Carajás were also identified through a structured and open-ended question in the interview: Participants were asked to list the 3 “worst things” about living in the Núcleo. Table 5.25 depicts the concepts identified during the interviews, classified into the QOL domains (i.e., social, health, economic and environment), as well as the frequency with which each concept was mentioned during the interview process. As indicated in Table 5.25, the majority of negative issues identified by participants are fall under the social (38) and health domains (29), and it is interesting to note that these issues are  48  Big Brother is a popular television show in which a group of people lives in a house with cameras in every room that operate 24 hours a day.  142  closely related. Some of the negative aspects offal under the economic (04) and environment (03) domains, but these were mentioned significantly fewer times than the social and health challenges encountered in the Núcleo. Table 5.25: Worst Things about Living in the Núcleo Urbano de Carajás QOL Domain Social  Health  Economic Environment  Concept  Frequency*  Lack of privacy  12  Blurred boundaries between personal and professional lives  8  Lack of entertainment options  6  Perceived over-regulation and control  5  Isolation and disconnection from the world, services and resources  2  High turnover of people  2  Limited educational opportunities after K–12  2  Social stigmatization: elite  1  Total  38  No work-life balance  12  Monotony and boredom  12  Work related-stress  3  High incidence of depression and unhappiness  2  Total  29  High cost of living  4  Total  4  Problems with housing  2  Transportation challenges  1  Total  3  *Refers to number of times a concept was mentioned during interview process  c. Priorities for the Núcleo Urbano de Carajás “Sometimes I am frustrated with the isolation… Even the Internet does not work half the time, so I feel isolated and as I have been denied [certain] things… And I feel powerless to do anything about it.” (Male employee living in Núcleo) At the end of each interview, participants were asked to identify priorities for improvement and to make suggestions for improving the quality of life in the Núcleo Urbano de Carajás. Table 5.26 depicts the priority issues and suggestions identified during interviews, as well as the frequency with which suggestions were mentioned during the interview process.  143  Table 5.26: Areas of Priority: Núcleo Urbano de Carajás Priority Issues to be Addressed Boredom and depression  Feelings of isolation  Privacy and freedom  High cost of living  Social inequality  Suggestions  Frequency*  Increase entertainment and recreational opportunities and variety  5  Create more professional development and training opportunities for family members who are not Vale employees  3  Improve access to culture  1  Total  9  Improve access and ability to come and go  2  Improve communication systems for residents  2  Increase transportation and access to/from Parauapebas to increase access to education, services and goods, and a larger part of the community  1  Promote social integration with Parauapebas  1  Total  6  Increase range of modifications one can make to company houses to address privacy issues  2  Reduce control (resident rules and regulations, bylaws)  2  Total  4  Improve salary and benefits for employees to mitigate costs  1  Establish strategy to increase competition between local businesses in order to increase quality of goods and services and reduce prices  1  Total  2  Eliminate housing inequalities within the community  1  Total  1  *Refers to number of times a concept was mentioned during interview process  It is clear from the issues and frequencies listed in Table 5.25 that the majority of participants feel that intervention efforts should focus on the types of social and psychological issues, such as feelings of isolation and boredom, which often result from living in a remote, difficult-toaccess location. Lack of privacy and freedom, the perceived social segregation within the community, and the high cost of living are other issues that seem to warrant intervention. Through the interview coding process, other relevant concepts outside of the discussion of the best and worst aspects of the community were identified. The recurring concepts are regarded as “other major themes”.  144  d. Other Major Themes  •  Social Segregation between the Núcleo Urbano de Carajás and Parauapebas  There is a perceived social segregation between the Núcleo Urbano de Carajás and Parauapebas, which causes discomfort for some Núcleo residents, and perceived social discrimination against Parauapebas residents. As stated by a male employee who had been a resident of the Núcleo for 6 years: “I worry about getting more benefits from Vale—it sounds nice, but I am afraid that more benefits would only increase the disparity between the 2 communities (the Núcleo and Parauapebas). Our privileges—urban services, basic infrastructure—should be available to everybody in this region.” For residents of the Núcleo, the disparities that exist between their community and Parauapebas create a certain level of fear; many residents of the company-controlled town feel that if they do not dedicate themselves to their work, they will lose their jobs and residence benefits. This fear seems to have resulted in further stress, pressure to work overtime and a lack of work-life balance for residents of the Núcleo. As stated by a male resident living in Parauapebas, who previously lived in the Núcleo for 14 years: “These great inequalities only increase the fear of job loss and reinforce the belief that you need to dedicate your life to your job.” Because some residents of the Núcleo regard Parauapebas as an undesirable place to live, the prospect of losing a job with Vale and having to work for contractors and live in Parauapebas may sound terrifying. The fear of job loss is in fact a very serious concern in the Núcleo. Because employees have benefits that include housing, schooling for their children and health care, the fear of loosing a job is amplified because of the consequences for the individual and his/her family. Job loss means more than the loss of a paycheque from Vale—it also potentially means the loss of decent housing, quality schooling for children, social networks and friends, and possibly also the loss of the spouse’s job.  145  •  Raising Children in a Company Town  Generally, participants regard the Núcleo Urbano de Carajás as a great place to raise young children. Residents appreciate the tranquility and security of the Núcleo and the fact that children can play on the streets without direct supervision. However, it appears that life in this community poses challenges for pre-teens and teenagers, who often need more variety and discovery not available in small communities like the Núcleo Urbano. Some participants expressed fear that this might increase the risks of teen pregnancy and drug use among the youths of the company-controlled town. As discussed by a female employee who had lived in the Núcleo for 22 years: “This place is beautiful and quiet, great for raising small children. But there is an age limit—about 15 or 16 years old. At this age they are looking for something they can’t find here anymore, and what happens is that their sexual life starts too early. There’s little to do, so they start dating too early, at 10, 11, 12 [years old]” Because children in the Núcleo generally grow up with such high levels of security and protection, they are often not “street smart” and have a hard time when they need to face the reality of a larger community. When young people leave the Núcleo to attend university or college elsewhere (which occurs because of the lack of options in the community itself), it is not unusual for them to have trouble adjusting to life outside of the safe and controlled atmosphere in which they grew up. In addition, some interview participants regard the security and isolation of the community as a barrier for child and youth development in regards to learning important self-protection and survival skills. “The teenagers who grew up here are a little out of it and spoiled—I have noticed that. There are many teenagers who have a hard time when they need to leave to go to university. It’s different and difficult because there’s no violence or muggings here.” (Female employee living in the Núcleo Urbano de Carajás)  •  Negative Changes since Privatization “Vale has changed a lot since the privatization—it used to be a lot better. I wouldn’t say we would have been in heaven living here, but it would have been pretty close to it.” (Male employee living in the Núcleo Urbano de Carajás)  There seems to be a general feeling that the quality of the urban environment and community services provided in the Núcleo Urbano de Carajás, and salary and benefits as well, have declined since the privatization of Vale. Longtime residents tend to speak with melancholy about  146  better housing, infrastructure maintenance and general morale in the community before the company was privatized. For example, in the past, the company was financially supportive of individuals renovating their houses in the Núcleo. Current company housing policies do not encourage such remodeling as the company will not reimburse employees for any personal investment in improving company homes. Hedges are also remembered as healthy and trimmed, marking off property lines and creating a sense of intimacy. The current lack of maintenance of hedges has contributed to the lack of privacy of homes, since these hedges were the only dividing elements between lots.  •  Adaptation and Coping Strategies  Several interviewees living in the Núcleo Urbano de Carajás discussed strategies for adaptation and coping with the challenges of living in a small, paternalistic, company-controlled town. The Núcleo residents brought forward several strategies: o  Addressing Feelings of Isolation  One of the suggestions for mitigating feelings of isolation involved leaving the Núcleo as much as possible: “I am always going away with my family—we don’t live locked up here… I see this place more like my workplace, Carajás is the place where we have the work and study routine, but every long weekend and vacation time we go somewhere [out of town].” (Female employee living in the Núcleo Urbano de Carajás) It was also noted that perceiving the Núcleo as a neighborhood of Parauapebas, and therefore a part of a larger community, might encourage individuals living in the Núcleo to breach the perceived barriers between the 2 communities: “I see Parauapebas as a cool place, and when I have time off I always go there. I go [to Parauapebas] on Sundays, have lunch, and even sleep over there. I love Parauapebas—my sister lives there… I see several people from Carajás there, at the town square, at the ice cream parlor. There are nice things in Parauapebas.” (Female employee living in the Núcleo Urbano de Carajás)  147  o  Addressing Work-Life Balance Challenges  To mitigate the challenge of achieving a healthy work-life balance, some employees suggested that aiming to make friends outside of the workplace and refraining from “talking shop” outside of working hours might be a good strategy: “Even though I work a lot, when I leave work I meet with people outside of my work, and I have conversations that are not work related—I discuss other sorts of issues, I take part in different activities.” (Male employee living in the Núcleo Urbano de Carajás) Keeping priorities straight was also identified as a strategy to mitigate the difficulties of balancing work and private life: “When I came as an intern, my supervisor told me what he thought the rules of this place are: your first and foremost priority is the company, then family, and then the other “stuff”. At that moment I knew I had to take the opposite approach to be able to survive here—my family comes first, I came here for them, so that we could have a better life and future.” (Male employee living in the Núcleo Urbano de Carajás) Moreover, choices regarding friends and companions are thought to impact levels of depression and boredom. Three employees suggested that spending leisure time with co-workers might make it difficult to separate private life and work: “You have to choose your friendships well so that you are not working 24 hours a day.” (Male employee living in the Núcleo Urbano de Carajás) o  Addressing Boredom and Depression  Interview participants living in the Núcleo Urbano de Carajás offered several thoughts on how to address boredom and feelings of isolation. One of the suggestions (referring to wives or husbands of mine employees) involved being open to trying different things and training for the employment opportunities available in the community: “The company helps… There’s a lot of training available. There is a job bank… And you see lots of employees’ wives working… For the company, it is better if they find work— this reduces [the company’s] costs because these wives already have a home to live in and are already familiarized with the community. [Vale] give[s] you priority, and I think there’s a lot of value in that. You come here to accompany your husband, but they also value you. However, you need to continue to deserve that position that was given to you. You need to be proactive and can’t just  148  wait for things to happen. Nowadays, everybody in Brazil and in the world has to work hard for these things— there are a lot of people wanting to take our place here.” (Female resident living in the Núcleo Urbano de Carajás) Another suggestion was to use time wisely: “I think you need to occupy yourself with constructive things, not with the activities that are exhaustive. You can only spend all your time going to parties for so long before you get tired of it, and then you have to find meaningful things to do with your time… You need to find those people who are similar to you in a way, who have similar interests.” (Male employee living in the Núcleo Urbano de Carajás) Varying the weekly routine by taking advantage of facilities available and events in the Núcleo as well as in Parauapebas was also identified as a way to avoid feeling bored and isolated: “I have a varied schedule: once a week I go to the movies, once a week I rent a DVD, twice a week I go to the Spiritualist Centre 49… I try to diversify my activities. Because there’s no real shopping here, once a week I go to the community market in Parauapebas.” (Male employee living in the Núcleo Urbano de Carajás) Finally, one resident of the Núcleo suggested that personalizing one’s home or space— i.e., treating the company house as if it were your own and modifying it to bring you increased comfort and wellbeing—is a good strategy for addressing depression: “Many people say, “I am not going to modify the house because one day I will leave”. Imagine if I had had this attitude until this day—for 23 years I would have been kept awake at night, annoyed, upset, complaining… So I think I need to be able to transform this place in the best way [I can]… So that I can be happy, and my family can grow happily.” (Female resident living in the Núcleo Urbano de Carajás) o  Addressing Peer Pressure and Privacy Issues  Several interview participants living in the Núcleo Urbano de Carajás mentioned privacy issues as a significant problem in their community; some were able to offer their own strategies to reduce the impact of the lack of privacy on their quality of life.  49  Spiritualism is a religion widely popular in Brazil. The main difference between Spiritualism and other religions is that it is founded on the belief that there is life after death and that there is a path of eternal progression for all mankind with the acceptance of responsibility for all one's own actions.  149  One of the suggestions was to make an effort to ignore what others think or say about one’s personal life or personal choices: “This Big Brother thing does not bother me anymore because I no longer care about what others say about me”. (Female employee living in the Núcleo Urbano de Carajás) Another suggestion on how to mitigate the privacy challenge in the Núcleo was to be protective of one’s privacy: “We try to keep our voices down at home—we can hear the neighbors. We don’t want the entire community to know what happens in our home.” (Male employee living in the Núcleo Urbano de Carajás) 5.2.3. Canaã dos Carajás a. Most Satisfying Things about Canaã dos Carajás “For me, the greatest advantage of living in Canaã is the peace that this town still offers. We can still go out without fear of being mugged; we can sleep without the fear that our home will be broken into; we can meet up with friends at the town’s square for a leisurely chat”. (Male employee living in Canaã) During the qualitative interviews, participants living in Canaã were asked to list the 3 best things about their community. The answers to the question “What are the best things about living in Canaã?” were analyzed and coded. Table 5.27 depicts the concepts identified during interviews, classified into the QOL domains, as well as the frequency with which each concept was mentioned during the interview process. For residents of Canaã, aspects associated with the environment (12), social (12) and health (10) domains were the most frequently identified as the most positive aspects of the community. Interestingly, these issues are closely related and associated with the fact that Canaã is a small and quaint town, experiencing dramatic change and growth.  150  Table 5.27: Best Things about Living in Canaã dos Carajás QOL Domain Environment  Social  Health  Work Economic  Concepts  Frequency*  Quaintness/rural atmosphere  07  Continuous development  03  Environmental qualities  02  Total  12  Sense of connection with the community  06  Quality of social interaction/networks  05  Cultural diversity of the population  01  Total  12  Tranquility  06  Sense of security  04  Total  10  Employment and professional growth opportunities  03  Total  03  Business opportunities in the booming community  01  Total  01  *Refers to number of times a concept was mentioned during interview process  b. Least Satisfying Things about Canaã dos Carajás “I don’t have a family, but I can imagine that it is much more complicated for those who do—if you have a stay-at-home wife and kids, it’s harder to have good quality of life… You work all week and what happens to your wife? And what about your kids? What options do they have? There’s nothing around here.” (Male employee living in Canaã) The negative aspects of Canaã dos Carajás were also identified through a structured and openended question about the “three worst things” about Canaã. Table 5.28 depicts the concepts identified and classified into the QOL domains, as well as the frequency with which each concept was mentioned during the interview process. Interestingly enough, the most frequently identified negative issues during interviews were also associated with the health (14), social (12) and environment (10) domains. While positive aspects seem to be mostly about the character of the community, negative aspects seem to be  151  associated with a lack of basic infrastructure—due to Canaã’s “growing pains”—and the social implications of Canaã’s remoteness, boom phase and high cost of living (3). Table 5.28: Worst Things about Living in Canaã dos Carajás QOL Domain Health  Social  Environment  Economic  Concepts  Frequency *  Lack of entertainment and recreational facilities  5  Inadequate health care services and facilities  5  Monotony and boredom  4  Total  14  Isolation and disconnection from the world  5  Inadequate social services  3  No educational opportunities after K–12  2  High turnover of residents  1  Lack of job opportunities for spouses of Vale employees  1  Total  12  Deficient basic infrastructure  5  Pollution from field burns for cattle raising  2  Lack of transportation options  2  Overcrowding in shared company homes (repúblicas)  1  Total  10  High cost of living  3  Total  3  *Refers to number of times each concept was mentioned during interview process  c. Priorities for Canaã dos Carajás “I think we are missing a well-structured recreation center, and the constant power blackouts are annoying.” (Male employee living in Canaã) At the end of each interview, participants were asked to identify priorities for improvement and/or to make suggestions for improving the quality of life in Canaã. The priority issues and suggestions brought forward during the interviews are displayed in Table 5.29. Similar to the Núcleo Urbano de Carajás, feelings of isolation resulting from the remoteness of the community and boredom (the feeling that there is “nothing to do” in the community) were also strongly identified as priorities for improvement in Canaã dos Carajás. As illustrated in  152  Table 5.29, suggestions to reduce the remoteness of the community were made. These suggestions include improving access to Canaã and the community’s communication systems. Suggestions to mitigate the boredom problem focused on improving entertainment and recreational opportunities (and increasing the variety of options) as well as opportunities to participate in community events. Table 5.29: Areas of Priority: Canaã dos Carajás Priority Issues to be Addressed Feelings of isolation and boredom  Suggestions  Frequency*  Increase entertainment/recreational opportunities and variety  4  Improve communication systems  3  Encourage employee engagement with the community  2  Improve access to and from the community  1  Total  10  Lack of basic infrastructure  Improve basic services and infrastructure  5  Total  5  Lack of basic community services  Improve basic community services such as banking, post office, transportation, etc.  5  Total  5  Education  Improve educational opportunities for the community  1  Total  1  *Refers to number of times a concept was mentioned during interview process  Several participants also view the infrastructure challenges and lacking community services as main priorities for intervention. One respondent highlighted the need to improve the education opportunities for the community as a whole. Through the Canaã interviews’ coding process, other relevant concepts were identified and categorized. The recurring concepts are regarded as “other major themes”. d. Other Major Themes  •  Vale’s Responsibilities towards the Development of the City  A recurring topic brought up by Vale employees related to the lack of clarity of the differences between the corporate (Vale) and municipal government responsibilities towards the  153  development of Canaã dos Carajás. Vale has participated in building a hospital, an elementary school and a community cultural center; laid asphalt in parts of the city, paid for the lighting of the major access street, etc. Some perceive that there is a general expectation that the company will solve all of the problems created by the rapid growth of the population and increasing need for basic infrastructure and services. There is also the perception that the municipal government has counted on the company to be the major player in the development of the community. Interview participants often pointed at the municipal government for not maintaining the infrastructure built by Vale: “The relationship between Vale and Canaã is positive, not negative. On the other hand, the municipal government thinks that Vale has to do everything. Yesterday I was arguing about that with the guy from the Municipal Chamber of Commerce —they keep waiting for stuff. I told him, ‘You need to start walking with your own feet.’ ” (Male employee living in Canaã)  •  Socio-economic Impacts of the Sossego Copper Mine  Long-term residents—people who witnessed the boom of the city and the impacts of this boom on the existing population—brought up interesting concepts. Perceived socio-economic impacts of mining activity are both positive and negative. Some of the positive impacts were the increase of goods and services available, a perceived reduction in serious crimes and more job opportunities: “The city has improved a lot—we now have asphalt, a bank… We have water distribution systems—before we needed to have our own wells in the backyard. [And} there were tonnes of kids without schools.” (Male resident living in Canaã) Negative impacts relate to pressure on community services, such as schools and health facilities, and the significant increase in the cost of living: “Before the mine, we did not have violence or child prostitution. Moreover, we could sleep with our windows and doors open without fear of being robbed. We did not have homeless problems—everyone had a place to live, food to eat… We would never see anybody on the streets asking for food, and nowadays we see all of that in Canaã. Violence levels have increased, the problem[s] of hunger, housing shortages, informal and illegal settlements… With this mine], the inflation has gone off the roof… Land prices have increased… A housing lot you could buy for 400 Reais is now 5,000 to 6,000.” (Male resident living in Canaã) Canaã is a booming town. As with any fast-growing city, both positive and negative changes have occurred during the economic boom. The economic benefits of the mine development to  154  the city are undeniable. As the city has grown, social indicators, including health and employment levels, have improved (Chapter 3). However, with the growth and economic development of Canaã, the cost of living has increased, and social and physical infrastructure are severely lacking.  •  Sossego Camp  Three of the Phase III interview participants lived at the Sossego Camp. The camp is located on the outskirts of Canaã (not at the mine site) and includes dormitories, a cafeteria and some administrative offices. The camp was built as a temporary facility to house Sossego Mine employees while company homes were built and basic infrastructure of the city was completed. An area near the camp has been allocated for a recreation centre which will include gym facilities, sports fields and a swimming pool. In of August 2006, construction of this facility was underway. Once the recreation centre in completed, current camp residents will have immediate access to the sports and recreational facilities it will offer. Permanent Sossego Camp residents are Sossego Copper Mine interns and single employees at the technician level and above, including engineers and analysts. The only differentiation is that employees with a university degree have individual rooms while all others have shared rooms. Visitors to the mine site and consultants on short assignments also use the Sossego Camp facility on a temporary basis. With the analysis of the qualitative data from the focus group in Phase II, it became apparent that for young interns, the Sossego Camp seems to be appreciated for the brotherhood bonds that are created by living there. However, while some appreciate the comfort and convenience of living in a hotel-like facility, others are dissatisfied with the lack of privacy and distance from the city centre and its services. As stated by a male employee living in the Sossego Camp at the time he was interviewed: “I lived in a military camp before and I will tell you, compared to that this is a great camp! It’s like living in a hotel. I don’t have to cook, clean or do laundry— it’s all done for me… I don’t have a problem going to town, so I hang out with people from the community, not only Sossego employees. I see that the interns though, they stay in the camp all the time [and] only make friends with other interns for the year they spend here.” However, another male employee offered a different perspective: 155  “I hate living in this camp. I am a single adult, and an engineer and I have a Masters degree—I was promised a company house but they changed the rules on me when I got here. Sure it’s nice to live in a hotel-like place for a while, but I like to cook my own food and have my friends over for homemade dinners. This is not a way to live—I don’t have [the] privacy or freedom that I would have in my own home.” 5.2.4. Parauapebas a. Most Satisfying Things about Parauapebas “The best thing about Parauapebas is the interaction with the neighbors—I have met so many people from different places.” (Male employee living in Parauapebas) Interview participants from Parauapebas were also asked to list the 3 best things about their community. Their answers were analyzed and coded. Table 5.29 depicts the concepts identified during interviews, classified into the QOL domains, as well as the frequency with which each concept was mentioned during the interview process. In comparison with the other 2 communities, Parauapebas respondents identified only a few positive aspects of their community. As shown in Table 5.30, the most recurrently mentioned positive aspects of Parauapebas were the social aspects of the community (8), followed by the abundance of work opportunities (6) Respondents who had lived in Parauapebas for a number of years also mentioned the fact that the city has evolved in the last decade or so, the availability of jobs and the vibrant economy.  156  Table 5.30: Best Things about Living in Parauapebas QOL Domain Social  Work Environment Economic  Concept  Frequency*  Profusion of concerts and dancing parties  4  Cultural diversity  2  Social networks and personal relationships  2  Total  8  Abundance of work opportunities  6  Total  6  Continuous development of the city  5  Total  5  Vibrant economy  1  Total  1  *Refers to number of times a concept was mentioned during interview process  b. Least Satisfying Things about Parauapebas “One thing that really bothers me in Parauapebas is the issue of sanitation— there is none. The infrastructure is very deficient. Our treated water often has too much chlorine, and the sewage system does not work—sometimes you can see the sewage draining on the street, and this is really bad because there’s this bad smell right in front of your home. There’s a broken sewage pipe in front of my house right now—it’s been a month and nothing has been done about it.” (Female employee living in Parauapebas) The negative aspects of Parauapebas were also identified through a structured and open-ended question about the “three worst things” about the city. Table 5.31 depicts the concepts identified during the interviews, classified into the QOL domains, as well as the frequency with which each concept was mentioned during the interview process. As depicted in Table 5.31, issues related to the physical environment (29) seem to be the most problematic in Parauapebas, followed by health issues (15), particularly the inadequacy of health care services (4) and monotony and boredom (4), which can be said to affect the mental health of Parauapebas residents. These are followed by social issues (9) such as violence and crime (8), limited educational opportunities (1), and economic challenges (4) such as the high cost of living (3) and poverty and inequality in the community (1).  157  Table 5.31: Worst Things about Living in Parauapebas QOL Domain Environment  Health  Social  Economic  Concept  Frequency*  Inadequate public/community services  9  Inadequate urban and basic infrastructure  9  Garbage, dirt and unpleasant odours  4  High temperature  2  Environmental degradation  2  Inadequate housing  2  Noise and chaos of the urban environment  1  Total  29  Inadequate Sanitation  9  Inadequate health services  4  Monotony and boredom  4  Total  15  Violence and crime  8  Limited educational options/opportunities  1  Total  9  High cost of living  3  Poverty and inequality in the community  1  Total  4  *Refers to number of times a concept was mentioned during interview process  c. Priorities for Parauapebas “It’s always the same story: you go to the hospital the night before, get a number, sleep in the line up to be seen by a doctor the next day.” (Male resident living in Parauapebas) Similar to the process used with participants from the Núcleo Urbano de Carajás and Canaã dos Carajás, Parauapebas interview participants were also asked to identify priorities for improvement and/or to make suggestions for improving the quality of life in their community. The priority issues and suggestions brought forward during the interviews are displayed in Table 5.32. As illustrated in Table 5.32, suggestions were quite varied, and several priorities were identified. These were related to issues of health care (6); violence and crime (4); basic infrastructure and sanitation problems (3); lack of educational opportunities (3); poverty and inequality in the  158  community (2); inadequate housing quality and availability (2); lacking community services (2); and the lack of support for culture (1). Table 5.32: Areas of Priority: Parauapebas Priority Issues to be Addressed Health care  Suggestions  Frequency*  Improve health care services and facilities  6  Total  6  Improve security and contain violence  4  Total  4  Improve basic community services such as banking, post office, transportation, etc.  3  Total Improve education and training opportunities  3 3  Total  3  Improve living conditions for the poor  1  Reduce inequality  1  Total  2  Inadequate housing  Address housing availability, quality and affordability  2  Total  2  Lacking community services  Improve community and public services  2  Total  2  Lack of support for culture  Increase access to cultural events and facilities  1  Total  1  Violence and crime Basic infrastructure and sanitation Lack of educational opportunities Poverty and inequality  *Refers to number of times a concept was mentioned during interview process  d. Other Major Themes A great variety of other major themes were also identified with the qualitative data collection. These issues include corruption, social segregation and discrimination of the Parauapebas residents, the fact that many individuals come to Parauapebas as a sacrifice to succeed financially, and the city’s overcrowding. What follows is a discussion, supported by participant observation data, of these themes.  159  •  Corruption  Several interview participants (from all 3 case study communities) mentioned the issue of corruption of politicians in Parauapebas, and one participant mentioned the corruption of police officers. Parauapebas receives large amounts of money in the form of royalties paid by the company, and many participants pointed out that these royalties are inadequately used or not used at all to develop the city. Interview participants generally blamed politicians for misusing royalty money and accused the former mayor and other municipal staff of corruption and incompetence: “Public funds in Parauapebas are very badly managed; this has always been the case—it’s a cultural issue related to the political process.” (Male employee and former resident of Parauapebas now living in Canaã dos Carajás) It is quite disturbing to observe the poverty levels in Parauapebas and the quality (or lack thereof) of its urban environment considering that the city is the second wealthiest city in the State of Pará. Even though a new mayor has been in power for the last 2 years, the fact that the same family had controlled the political power in the city for over 15 years also adds to the perception of corruption and the municipal government‘s lack of transparency.  •  Social Segregation and Discrimination  The issue of social segregation of and discrimination against Parauapebas residents was mentioned in interviews with residents of both Parauapebas and the Núcleo Urbano de Carajás. These discussions revolved around the perception that the “elite” lives in the Núcleo, while the poor and disadvantaged population lives in Parauapebas. It is important to note here that in Brazil, the economic elite is more often than not also the social elite—education and wealth are closely associated in Brazilian society. Expressed concerns relate to the perception that Parauapebas residents are stigmatized, discriminated against and regarded as inferior by Núcleo’s residents: “The people from up here (the Núcleo), they discriminate against people from Parauapebas, and vice versa.” (Male employee living in the Núcleo Urbano de Carajás)  160  Municipal statistics clearly show the inequality in Parauapebas (Atlas do Desenvolvimento Human Brasil, 2006). The segregation is indeed very apparent. Higher job classes (with consequently higher salaries) are eligible to live in a company-provided house in the Núcleo Urbano de Carajás. All others need to fend for themselves in expensive and low-quality housing in Parauapebas, which increases differences in available income and access to basic services. The physical barrier and distance between the 2 communities intensifies this separation.  •  Life of Sacrifice  Most Parauapebas residents are migrants from the Northeast or other northern states. They seem to often come alone to find work at the mines. Some participants said that they have lived a life of sacrifice: “I went through a lot of challenges to get here. Since I got here I have faced many others… Not financial problems, but being far from friends and from home… I miss home a lot. For 10 months I had to suffer, missing home… I am divided—I came here to make money, to accomplish something for myself, [things that] I would not have been able to do back at home. The work here gives me the opportunity to improve myself, to take courses… and home is where I left my loved ones, my origins. I miss that tremendously, but I am surviving.” (Male contractor’s employee living in Parauapebas) The attitude and feeling exemplified by this quote seem to be widely held by residents of Parauapebas. Migrants who travel to the Carajás region to find work are often responsible for supporting family elsewhere. Many of these workers send away part of their salaries to families in other parts of the state or country, and as a result are not able to spend their earnings to improve their living conditions in Parauapebas.  •  Overcrowding  It became clear through the interviews that service providers and contractors’ workers were dealing with the housing affordability issue by living in overcrowded conditions: “We share a rented house. I work at Intertec, and we pay the rent ourselves. We are 7 men in a 4-bedroom house.” (Male contractor’s employee living in Parauapebas)  161  Indeed, with the shortage of housing and the rapidly increasing need for accommodations for employees of both Vale and outside contractors, all sorts of units are coming onto the Parauapebas housing market. Rental housing units are usually very small and offer minimal amenities, or are larger renovated homes. These larger units are often shared by large groups of workers, who usually work shifts and therefore do not refrain from sharing homes and rent with many others. The housing challenges in Parauapebas, and also in Canaã dos Carajás, are very significant. Vale is not investing in housing construction in Parauapebas (except for camp accommodations for short-term interns), and the local government has failed to provide decent housing options to the community as a whole. It has become very clear that this problem is in fact getting worse, and that significant interventions are needed to avoid overcrowding and to provide housing options that respect human rights and provide healthy living environments for mineworkers and other residents.  162  CHAPTER 6: DISCUSSION This chapter aims to discuss the findings presented in Chapter 5 in light of the objectives set out for this research. This deliberation includes a discussion of the results of the triangulation of the qualitative and quantitative results, and an attempt to provide answers to the research questions. It also includes the identification of opportunities for interventions that could result in improved quality of life (QOL) in the communities investigated in this research and perhaps in other remote mining communities around the world. The first objective of this research was to develop a subjective quality of life framework that can be integrated into lifecycle planning of new and existing mining operations in remote locations. The framework suggested here is one that takes into account the factors defined as the predictors of satisfaction with QOL—those factors that seem to most impact mineworkers’ perception of quality of life in remote mining communities. The definition of these factors was accomplished with the findings of the first research question: “What appear to be the main factors impact the quality of life of mineworkers in Canaã dos Carajás, Parauapebas and the Núcleo Urbano de Carajás?” All domains identified as predictors for the population sample as a whole were considered: i.e., the economic, environment, social and health domains. In designing new mines in remote areas or addressing challenges in existing remote mining communities, these domains should therefore be used as parameters for design, evaluation and intervention. As it was done in these 3 case studies, context-relevant indicators could be developed in order to evaluate and monitor performance more closely and identify specific areas of concern or success. The second objective of this research was to identify opportunities to improve the quality of life of mineworkers in the 3 case study communities: Canaã dos Carajás, Parauapebas and Núcleo Urbano de Carajás. This was accomplished by using the subjective quality of life framework to identify mineworkers’ perceptions of quality of life in the 3 case study communities, and by investigating the way in which the communities’ and the workforce’s characteristics impact these perceptions. Specifically, once the 4 QOL domains (predictors of satisfaction with QOL) were defined, answers to the next 3 research questions were sought: 2. How differently do mineworkers living in these three communities perceive their quality of life?  163  3. What appear to be the most and least satisfying aspects of the quality of life in these communities? 4. How do the mineworkers’ characteristics appear to impact their perceptions of their own quality of life? Based on the research findings, opportunities for interventions that would result in improved QOL in remote mining communities were addressed by the formulation of suggestions and recommendations directed at mining companies in general, at Vale and at local decision makers. These opportunities for intervention were identified as a potential means to improve the quality of life in remote mining communities and the level of satisfaction of mine employees in remote mining communities. The chapter is therefore divided into 3 distinct sections to a) address research questions, b) identify opportunities for intervention and c) discuss methodological findings (Figure 6.1).  Figure 6.1: Structure of Chapter 6  164  Within the first section of the chapter, the intent is to propose answers to the 4 research questions initially established for this research. The 2nd section of this chapter includes the discussion of opportunities for intervention directed at mining companies planning and operating mines in remote areas, local governments and policy makers. The 3rd and final section of this chapter includes a discussion of the findings in regards to the methodology used for this research. This section also makes methodological recommendations regarding measuring and monitoring quality of life in remote mining communities.  6.1.  Addressing Research Question 1: Main Factors That Appear to Impact Perceived Quality of Life  The process to define the main factors affecting perceived quality of life in the case study communities was initiated in Phase I and completed in Phase III of the research. It included the triangulation of results from mixed-methods. Refer to Figure 6.2 for a diagram summarizing the overall approach used to answer Research Question 1.  Figure 6.2: Approach to Answering Research Question 1  165  Phase I of the research strategy included a literature review and the first conceptual quality of life (QOL) model development. The definition of the 5 quality of life domains (deconstruction of the QOL concept) and indicators (further deconstruction of QOL domains) was finalized with qualitative on-site interviews during Phase II of this research. Phase II generated a list of 5 domains (Figure 4.2), as well as indicators that make up each domain, all of which are assumed to affect the perception of quality of life in remote mining communities (Table 4.2). In Phase III, 461 mineworkers at the Carajás Iron Ore Complex, the Azul Manganese Mine and the Sossego Copper Mine completed a quality of life questionnaire. Of specific relevance to the first research question are the results of the statistical analysis of this questionnaire, which determined, through linear regression, the predictors of quality of life—the aspects that seem to uniquely predict variance in satisfaction with quality of life in the sample population. The quantitative evidence was supported by qualitative results of this research. Phase III further investigated the mineworkers’ quality of life and their communities through 40 in-depth interviews with employees and residents living in the 3 communities. The qualitative and quantitative results were triangulated with participant observation to identify the factors that most significantly affect perceptions of quality of life in the sampled population. Specifically, these factors are the qualit