The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries ECONOMICS ACKNOWLEDGEMENTS This report was prepared by the World Bank under the The report was prepared by a World Bank team composed West Africa Coastal Areas Management Program (WACA). of Susmita Dasgupta (Lead Environmental Economist), It was funded by the Nordic Development Fund (NDF), a joint Subhendu Roy (Consultant), and David Wheeler (Consultant) © 2023 The World Bank Group Nordic-international finance institution that focuses on the nexus under the guidance of Maria Sarraf (Practice Manager), between climate change and development in lower income Peter Kristensen (Lead Environmental Specialist), and 1818 H Street NW, Washington DC 20433 countries and PROBLUE, an umbrella multi-donor trust fund Sarah Jung (Environmental Specialist). Editing was done by administered by the World Bank that supports the sustainable Jennifer Statsny and design by The Ethical Agency. Telephone: 202-473-1000 Internet: www.worldbank.org and integrated development of marine and coastal resources in healthy oceans. This work is a product of the staff of The World Bank Group, with external contributions. The World Bank Group refers to the legally separate organizations of the International Bank for Reconstruction and Development (IBRD), the International Development Association (IDA), the International Finance Corporation (IFC), and the Multilateral Investment Guarantee Agency (MIGA). The World Bank Group does not guarantee the accuracy, reliability or completeness of the content included in this work, or the conclusions or judgments described herein, and accepts no responsibility or liability for any omissions or errors (including, without limitation, typographical errors and technical errors) in the content whatsoever, or for reliance thereon. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank Group concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the organizations of the World Bank Group, their respective Boards of Executive Directors, and the governments they represent. The contents of this work are intended for general informational purposes only and are not intended to constitute legal, securities, or investment advice, an opinion regarding the appropriateness of any investment, or a solicitation of any type. Some of the organizations of the World Bank Group or their affiliates may have an investment in, provide other advice or services to, or otherwise have a financial interest in, certain of the companies and parties named herein. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of any of the organizations of The World Bank Group, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank Group encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given and all further permissions that may be required for such use (as noted herein) are acquired. The World Bank Group does not warrant that the content contained in this work will not infringe on the rights of third parties and accepts no responsibility or liability in this regard. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. Cover photo: ©Mel D. Cole for World Bank © Mel D. Cole for World Bank ABOUT THIS REPORT KEY MESSAGES This report aims to help decision-makers better understand the The Africa region is currently the second-largest source of ocean plastic pollution from rivers, with a share of 7.8 percent. By 2060, Africa could become the world’s largest economics of marine plastic-waste generation and its cleanup, contributor of mismanaged plastic waste. with a focus on West African coastal countries. It was carried out under the remit of the West Africa Coastal Areas Management Program (WACA), which In 14 out of 17 West African coastal countries, more than 80 percent of plastic is addresses coastal degradation—including from plastic pollution—in 17 West and Central coastal African mismanaged, increasing the risk of plastic waste entering the oceans. There is an urgent countries and island states spanning from Mauritania to Gabon. need to improve plastic waste management systems in the region. This report is part of a series that includes: The overall economic cost of marine plastics to society is estimated at between US$10,000 and US$33,000 per ton of plastic. These costs are concentrated in four sectors: fisheries and aquaculture; marine-linked tourism; value of waterfront property; and biodiversity and ecosystems sectors. More research is needed to determine the costs of plastic pollution for other sectors. Econometric analysis indicates that import taxes on polyethylene sheets could play a role in reducing marine pollution by driving down single-use plastics waste. Plastic Pollution in Coastal West Africa: West Africa Circular Economy: Realizing However, given that single-use plastics are widely used for safe drinking water, the Synthesis paper the Potential of Plastics potential public health impacts of such measures will need to be carefully considered. Similar consideration needs to be given to the distributional effects of such import taxes. Cleanup efforts before seasonal rains in pollution hotspots should be better targeted to optimize the volumes of plastic pollution that is prevented from reaching oceans. There is no one-size-fits-all solution. West African coastal countries do not have sufficient data for estimating country- and sector-specific costs. In this context, location-specific analyses are needed to determine the most cost-effective policy mix for plastic waste remediation, with the most practical policy solutions likely entailing some combination of quantity- and price-based approaches balanced by highly targeted cleanup strategies. Producer Responsibility Organisation WACA Plastic E-book to manage Polyethylene Terephtalate bottles in Senegal Public awareness, stakeholder participation in policy and strategy design, and access to environmentally friendly alternatives will be key to effective waste management. CONTENTS Abbreviations ..................................................................................................................................... 2 6. How to Implement Cost-Effective Cleanups: Examples from Accra and Lagos ......................................................................................................................................... 25 Executive Summary ......................................................................................................................... 3 Incorporating plastic disposal hotspots ................................................................................................. 26 Incorporating rivers as conduits of plastic waste .................................................................................. 29 1. Introduction .................................................................................................................................. 9 Combining hotspot data with the likelihood of river transport ............................................................... 31 Incorporating seasonal rainfall cycles ................................................................................................... 32 2. The Economic Cost of Marine Plastic Waste to Society ........................................ 14 Damage to overall marine ecosystem services .................................................................................. 15 7. What we have Learned .............................................................................................................. 35 Aggregation of sector-specific costs ..................................................................................................... 16 References .......................................................................................................................................... 39 3. How the Economic Costs Compare with those of the Main Approaches to Pollution Mitigation ............................................................................................................. 17 Appendices ......................................................................................................................................... 41 Incentives ............................................................................................................................................. 18 Appendix A. The macroeconometrics of single-use plastic imports ...................................................... 41 Command and control .......................................................................................................................... 18 Appendix B. The health impact of plastic container use ....................................................................... 43 Removal of plastic waste through cleaning, recycling, and safe disposal ............................................ 19 Appendix C. Household income and plastic container use ................................................................... 44 Appendix D. Spatial clustering of income and plastic-waste generation ............................................... 45 4. How General Economic Measures could significantly Reduce Plastic Appendix E. Rivers as conduits for plastic waste .................................................................................. 47 Pollution ............................................................................................................................................ 20 5. The Public Health Risks: The Case of Plastic Water Containers in Ghana and Nigeria ...................................................................................................................................... 22 © Mel D. Cole for World Bank BOXES ABBREVIATIONS & DEFINITIONS Box 1. Classification of plastic waste generation for WACA countries .......................................................... 10 DHS Demographic health survey LIST OF TABLES ECOWAS Economic Community of West African States Table 1. Plastic waste in countries of West Africa Coastal Areas Management Program, 2010 ..................... 9 Table 2. Annual plastic waste intensity by dimension ..................................................................................... 21 GDP Gross domestic product Table 3. Real GDP growth in West Africa, 2005–19 ........................................................................................ 26 MIS Malaria indicator survey Table 4. Child health impacts with and without plastic water container use in Ghana and Nigeria ................. 24 NDF Nordic Development Fund Table 5. West African import demand model results ....................................................................................... 43 Table 6. Plastic drinking water container use and child health in Ghana and Nigeria ..................................... 45 SUP Single-use plastic Table 7. Household income and plastic container use in Accra and Lagos, 2003–19 .................................... 46 TPS Thin polyethylene sheet Table 8. Household use probabilities for plastic drinking water containers ..................................................... 46 UNEP United Nations Environment Programme Table 9. Population indicators for 1 km grid cells ............................................................................................ 47  WACA West Africa Coastal Areas Management Program LIST OF FIGURES WACA Figure 1: Plastics in a tributary of the Odaw River (Ghana), May 2015 ........................................................... 11 countries/region The 17 coastal and island states covered by the Figure 2. Decision-making process for setting policies to reduce marine plastic pollution ............................... 12 WACA program: Benin, Cabo Verde, Cameroon, Figure 3. Damage cost estimates from plastics for marine ecosystem services .............................................. 15 Côte d’Ivoire, Equatoria Guinea, Gabon, Figure 4. Damage cost estimate from presence of plastic in four sectors ........................................................ 15 The Gambia, Ghana, Guinea, Guinea-Bissau, Figure 5. TPS imports by 10 West African countries, 1995–2019 .................................................................... 21 Liberia, Mauritania, Nigeria, São Tomé and Figure 6. Imported TPS price for West Africa, 1995–2019 ............................................................................... 21 Príncipe, Senegal, Sierra Leone, and Togo Figure 7. Child health impacts of plastic water container use in two countries ................................................ 23 Figure 8. Incidence of plastic container use, by household income per capita, 2003–19 ................................ 26 Figure 9. Maps showing residential income clustering ..................................................................................... 27 Figure 10. Maps showing plastic water container use and disposal over time ................................................... 28 Figure 11. Maps of river systems in Accra and Lagos ........................................................................................ 29 Figure 12. River transport of plastic waste ......................................................................................................... 30 Figure 13. Maps showing the likelihood of river transport of plastic waste ......................................................... 31 Figure 14. Plastic hotspots, with and without adjustment for river transport likelihood ....................................... 32 Figure 15. Median monthly rainfall, 2010–20 ...................................................................................................... 33 Figure 16. Plastic pollution measurement area for Accra ................................................................................... 33 Figure 17. Monthly plotting results for the Accra offshore area .......................................................................... 34 Figure 18. Real income per capita in Ghana and Nigeria, 2000–19 ................................................................... 44 Figure 19. Per capita income distribution ........................................................................................................... 44 Figure 20. Maps of net elevations ....................................................................................................................... 47 © Freepik 1 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 2 Rapidly growing, unregulated plastic litter has created a multitude of environmental and economic problems worldwide. EXECUTIVE The rapid rise in global plastic production, which had true economic cost of plastics is scarce. This cost is difficult reached 368 million metric tons by 2019, is expected to to estimate, as persistent post-use environmental damage is double over the next two decades (Geyer, Jambeck, and hard to monetize. Also, the pros and cons of various market- Law 2017). Durability, one of the main attributes accounting based policy instruments for remediation are lacking. In for plastic’s popularity, also poses serious hazards for the addition, the spatial distribution and timing of plastic-waste SUMMARY unregulated disposal of plastic waste. generation is poorly understood. This issue is especially key because it can affect the relative importance of With an estimated lifetime of centuries, plastic waste policy instruments. has become a major stressor in marine ecosystems (Díaz-Mendoza et al. 2020; Gallo et al. 2018; Jeftic et al. This study aims to help decision-makers better 2009; UNEP 2005). Plastic ocean debris, first observed in understand the economics of marine plastic-waste the 1960s, now affects all of the world’s oceans. Recent generation and its cleanup, with a focus on West studies estimate the ocean entry of plastic waste at between African coastal countries. To aid the policy process to 4.8 million and 20 million metric tons annually (Jambeck reduce marine plastic pollution, it addresses the following et al. 2015; UNEP 2014). Each year, thousands of fish, key questions: seabirds, sea turtles, and other marine mammals die as a result of ingesting or becoming entangled in plastic debris. • What is the economic cost to society of marine plastic waste? In West Africa, the use of plastic products has proliferated with urbanization, and their unregulated • How does this cost compare with the pollution mitigation cost, using various incentive-based, command-and- disposal has created a host of terrestrial and marine- control approaches for pollution prevention and the cost related environmental problems. The Africa region is of plastic waste removal through cleaning, recycling, and currently the second-largest source of ocean plastic pollution safe disposal? from rivers, with a share of 7.8 percent (Lebreton et al. 2017; Ritchie and Roser 2018). Three African rivers figure among • Would general economic measures (for example, tariffs the world’s top 20 plastic pollution sources: the Cross River on imported polyethylene) significantly reduce pollution (Nigeria and Cameroon); the Imo River (Nigeria); and the from single-use plastics? Kwa Ibo River (Nigeria) (Lebreton et al. 2017). Projections for 2025 indicate that mismanaged plastic waste from the • Are there trade-offs between plastic pollution prevention Africa region will likely comprise 10.6 percent of the global and any other social objectives related to policymaking? total (Jambeck et al. 2015). With urbanization continuing How should cost-effective cleanups be implemented? in an unabated fashion, Africa could become the largest The study takes a location- and season-specific approach, contributor towards global mismanaged plastic waste by using detailed information for Accra (Ghana) and Lagos 2060 (Lebreton and Andrady 2019). (Nigeria) from household surveys, geographic and weather A survey of current literature reveals that, in 14 out data, and measures of marine plastic pollution. of 17 West African coastal countries, the share of Using a holistic approach, the economic cost of marine mismanaged plastic waste in proportion to the total plastics to society is estimated at between US$10,000 exceeds 80 percent.1 All coastal countries need to have and US$33,000 per ton of plastic. Sector-specific well-functioning plastic-waste management infrastructure, damages of between US$2,000 to nearly US$7,000 have policies, and practices in place to lower the risk of plastic been determined for four sectors. The study’s literature waste generated in coastal areas entering the oceans via review shows that two main approaches are currently wind, tidal transport, and/or transport to coastlines by inland being used to estimate the external cost of plastics in the waterways (Jambeck et al. 2015; Ritchie and Roser 2018). marine environment: damage to overall marine ecosystem Clearly, West African coastal countries are in urgent need services and aggregation of sector-specific costs. Using the of improved plastic waste management systems. first (holistic) approach, the annual damage cost estimate While the reduction of mismanaged plastic waste appears to be between US$10,000 and US$33,000 per ton has been recognized as an important development of plastic (Barrett et al.; Conservancy 2015; Costanza et al. objective, several key factors have hindered cost- 2014; Jang et al. 2015). effective remediation. For example, information on the 1 The figures are for 2010, which is the latest year for which the available data permit cross-comparison. © Mel D. Cole for World Bank 3 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 4 © Mel D. Cole for World Bank Using the second (partial) approach, the aggregate cost estimate that, for each 1 percent increase in national income, TPS imports of SUP container waste in hotspots during low-rainfall periods, require a widely distributed cadre of enforcement agents. Since for the four sectors where damage from the presence of plastic is increased by about 1 percent. For each 1 percent increase in followed by rapid river transport through flooding and runoff with a tariff may have a disproportionate impact on the poor, policy clearly visible—fisheries and aquaculture; marine-linked tourism; TPS price, TPS imports decreased by about 1 percent. These the return of heavier rainfall. makers should consider potential distributional implications value of waterfront property; and biodiversity and ecosystems— results have two major policy implications. First, without putting before implementing a tariff on polyethylene. More local case studies on sector-specific losses from ranges from more than US$2,000 to nearly US$7,000 per ton countermeasures in place, TPS imports and the waste generated plastic wastes are needed in West African countries. At Economic measures must avoid adverse health impacts. of plastic waste. These estimates are important elements of by SUP containers will likely keep pace with national income present, West African coastal countries do not have sufficient While the case for public intervention to reduce plastic waste the social cost of plastics and will be useful for future location- growth. Second, TPS price increases on the world market data for estimating country- and sector-specific costs. Better data seems clear, attention must also be paid to potential conflicts specific cost-benefit analyses for public and private interventions have produced rapid, proportionate reductions in West Africa on waste plastic externalities can play a key role in assessing the with public-health outcomes. Thus, measures to reduce the for waste management. import demand. Since producers are indifferent to the sources benefits and costs of policy options for plastic waste remediation. use of plastic sachets and bottles should be accompanied by of price change, the same will be true for price increases from The costs of reducing plastic pollution, using three programs designed to improve health outcomes for children, import duties. Thus, a TPS tariff could be a potent weapon in Location-specific analyses are needed to determine the main approaches, fall within the range of the estimated particularly in poor households. As an alternative, subsidies the struggle to reduce SUP pollution. However, policy makers most cost-effective policy mix for plastic waste remediation external costs. A study of global experience reveals three could be provided for use of biodegradable drinking-water should consider the distributional implications of this option since in each country. West African coastal countries require urgent main approaches for reducing plastic pollution: (i) incentives; containers, which are more costly to produce. the poor could be disproportionately impacted. intervention because mismanaged plastic waste in the marine (ii) command and control; and (iii) removal of plastic waste environment will continue to increase at high rates (Lebreton and Cleanup measures should be better targeted. Priority should through cleaning, recycling, and safe disposal. Incentive-based An econometric analysis was conducted for Ghana and Andrady 2019). However, there is no one-size-fits-all solution. be given to areas with a high incidence of plastic waste disposal approaches include levying production excise taxes or import Nigeria to assess the public health risks from policies As options for plastic waste management improve, the most near rivers, particularly more elevated areas with steeper duties on raw materials or taxing plastic products at the point of to reduce waste from the use of SUP drinking-water practical policy solutions will likely entail some combination of slopes. Cleanup resources should be concentrated in marine sale. Command-and-control approaches minimize the external containers. Using Demographic and Health Survey data for quantity- and price-based approaches balanced by cleanup plastic hotspot areas before the onset of the first-semester costs generated by plastic products by banning their use through these two countries, the analysis tested whether child morbidity strategies. Determining the most cost-effective policy mix for rainy season. regulation and enforcement. In principle, both incentive-based and mortality are lower in households that use SUP drinking- each country should involve location-specific analyses. and command-and-control approaches can reduce the use of water containers, after controlling for income, education, and many plastic products; however, complete elimination may not other socioeconomic factors widely cited in the literature. The Effective waste management requires awareness raising, be feasible for some, in which case the removal of plastic waste respective results showed notable declines in the median stakeholder participation in policy and strategy design, and through cleaning, recycling, or safe disposal will be beneficial. predicted rate of child mortality (42 percent and 20 percent) promoting the development of environmentally friendly These three approaches are not mutually exclusive. Rather, and incidence of diarrhea (21 percent and 10 percent) for all alternatives. Successful global experience indicates that they can be tailored to a particular country’s local economic children (0–5 years of age) attributable to SUP container use effective outcomes require broad-based awareness raising and political conditions to achieve the most cost-effective mix. across and within years. This means that general measures to about plastic pollution, including regular public consultations; reduce plastic use might also increase childhood illness and stakeholder engagement in designing mitigation policies Taxes and bans can reduce marine pollution from single-use death. These findings suggest the need to offset reduction and strategies; and the development of reasonably priced, plastic (SUP) waste, but targeted enforcement may prove and prevention policies with compensatory measures that environmentally friendly alternatives planned well in advance of difficult if producers, sellers, and consumers are widely promote other sources of clean drinking water, particularly implementing plastic reduction policies. West African countries dispersed. In West African coastal countries, SUP sachets, for poor households. can improve their waste management performance by learning bags, and other containers are fabricated from thin polythene from successful global experience. sheet (TPS), virtually all of which is imported. One appealing The study developed an illustrative cleanup strategy for policy would directly target bulk imports of polyethylene; marine plastic pollution in Accra and Lagos. Its focus was Import taxes on polyethylene sheets can play a key role in however, evaluating this policy also requires understanding the SUP drinking-water containers, given the potentially adverse reducing SUP waste, but understanding the distributional expected response of TPS imports to the imposition of tariffs. public-health effects of banning or severely restricting their use. A implications for the poor is critical. Taxation of the imported hotspot targeting strategy was developed for the two cities, using polyethylene that comprises most of the production feedstock The results of this study’s econometric analysis of TPS a methodology that combined georeferenced household survey for SUP in West Africa is a potentially effective, price-based import demand for seven West African coastal countries data on plastic use, measures of seasonal variation in marine policy with relatively low administrative costs. Plastic demand indicate a high degree of income and price responsiveness. plastic pollution from satellite imagery, and a model of plastic exhibits a very elastic response to changes in the price of This study addressed the responsiveness question by analyzing waste transport to the ocean using information on topography, imported polyethylene. Import taxes have a potentially major TPS imports and their prices over time for Benin, Cameroon, seasonal rainfall, drainage to rivers, and river transport to the cost advantage over directly targeted measures since the former Côte d’Ivoire, Ghana, Guinea, Nigeria, and Senegal. It found ocean. The results provide clear evidence of the accumulation can be administered at relatively few entry points while the latter © Mel D. Cole for World Bank 5 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 1. Plastic use has expanded rapidly since World War II.2 In recent years, global plastic production has been increasing at an average rate of more than 8 percent INTRODUCTION per year.3 By 2019, annual production had reached 368 million metric tons, and this figure is expected to double over the next two decades (Geyer, Jambeck, and Law 2017).4 The main attributes that account for plastic’s popularity are its low cost, convenience, and durability; however, the durability feature also poses serious hazards (for example, environmental, public health, economic, and aesthetic) for the unregulated disposal of plastic waste that is not incinerated, stored in landfills, or recycled.5 Plastics, which have an estimated lifetime of hundreds of years, have become major stressors in marine ecosystems (Díaz-Mendoza et al. 2020; Gallo et al. 2018; Jeftic et al. 2009; UNEP 2005). Plastic ocean debris, first observed in the 1960s,6 affects all of the world’s oceans; recent studies estimate ocean entry at between 4.8 million and 20 million metric tons annually (Jambeck et al. 2015; UNEP 2014).7 In the marine environment, plastic slowly degrades into microplastics over time, accumulates on shorelines, sinks to the seabed, or floats on the sea surface.8 Each year, thousands of fish, sea birds, sea turtles, and other marine mammals die as a result of ingesting or becoming entangled in plastic debris. Plastic waste generated in coastal areas is at high risk of entering the oceans via wind, tidal transport, and/or transport to coastlines by inland waterways (Jambeck et al. 2015; Ritchie and Roser 2018). However, management determines this risk, highlighting the need for all coastal countries to implement well-functioning plastic waste-management infrastructure, policies, and practices. Much of the world’s mismanaged plastic waste enters rivers and water systems before ending up in the ocean.9 The Africa region is currently the second-largest source of ocean plastic pollution from rivers, with a share of 7.8 percent (Lebreton et al. 2017; Ritchie and Roser 2018).10 2 The term plastic originally meant “pliable” or “easily shaped”. 3 https://www.statista.com/statistics/282732/global-production-of-plastics-since-1950/. Accessed February 2021. 4 In 2020, global production decreased by 0.3 percent, owing to the impact of Covid-19 on the industry. 5 Prior to 1980, virtually all plastic was discarded, with negligible incineration and recycling. After 1980 (for incineration) and 1990 (for recycling), the combined rate for incineration and recycling rates increased by about 0.7 percent per year (Geyer, Jambeck, and Law 2017). In 2015, an estimated 55 percent of global plastic waste was discarded, 25 percent was incinerated, and 20 percent recycled (Ritchie and Roser 2018). 6 https://www.sciencehistory.org/the-history-and-future-of-plastics Accessed August 2021. 7 Plastic accounts for between 61 and 87 percent of marine litter (Barboza et al. 2019; Galgani et al. 2019). 8 Over two-thirds of marine plastic litter ends up on the seabed. Half of the remaining third washes up on beaches, while the other half (i.e. sixth of total) floats near the surface (Gallo et al. 2018). 9 Plastic that ends up in the ocean also results from the disposal of solid waste, dumping of wastewater, direct littering, vehicular transport, and/or transport by wind and stormwater. 10 The largest share has its origin in Asia, which accounts for 86 percent of the global total. © Mel D. Cole for World Bank 7 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 8 BOX 1 Classification of plastic waste generation for WACA countries The 17 West African coastal countries can be separated per capita. Each of these indicators has been assigned a into low, medium, and high categories for plastic waste percentile category [low 0–33; medium 34–66; or high 67–100]. intensity across three dimensions: plastic waste per thousand A composite category is determined from a country’s scores in people; plastic waste per square kilometer; and plastic waste at least two dimensions. per US$100,000 purchasing power parity-adjusted GDP © Mel D. Cole for World Bank © Mel D. Cole for World Bank Table 2 Annual plastic waste intensity by dimension Country Plastic waste (t) /1,000 Plastic waste (t) /km2 Plastic waste Category Nigeria is Africa’s largest generator of plastic waste, and A survey of the current literature revealed that the status people (t)/US$100,000 among the top producers of the substance worldwide. Three of plastic waste in 17 West African coastal countries in PPP-adjusted GDP African rivers figure among the world’s top 20 plastic pollution 2010, the latest year for which the available data permitted Côte d'Ivoire 29.08 2.38 5.3 High sources: (i) the Cross River (Nigeria and Cameroon); (ii) the Imo cross-comparison (Table 1). Although the majority of Liberia 23.93 1.21 6.77 High River (Nigeria); and (iii) the Kwa Ibo River (Nigeria) (Lebreton these countries have instituted SUP reduction policies Nigeria 28.92 6.45 4.74 High et al. 2017). Projections for 2025 indicate that mismanaged by banning certain products, their share of mismanaged plastic waste from the Africa region will likely comprise 10.6 waste in proportion to the total still exceeds 80 percent São Tome and Principe 29.98 6.56 2.97 High percent of the global total (Jambeck et al. 2015). With the rate in 14 of them.11 Clearly, these countries need an urgent Senegal 29.00 2.47 5.83 High at which urbanization is taking place, Africa could become the improvement in their plastic waste-management systems Togo 16.34 2.39 7.65 High largest contributor to global mismanaged plastic waste by 2060 (Box 1). Benin 11.91 1.26 4.3 Medium (Lebreton and Andrady 2019). Cape Verde 21.44 2.96 1.41 Medium Table 1: Plastic waste in countries of West Africa Coastal Areas Management Program (WACA), 2010 Equatorial Guinea 35.63 1.78 1.02 Medium Gambia 12.27 2.55 5.02 Medium Country Total plastic waste Per capita plastic waste Share of mismanaged Relative share of plastic generation (metric tons) (kg/person/day) waste (% of global total) Guinea-Bissau 15.58 0.85 5.26 Medium plastic (%) Mauritania 12.75 0.06 3.98 Medium Nigeria 5,961,750 0.1 81 2.67 Sierra Leone 12.12 1.35 7.77 Medium Côte d’Ivoire 766,988 0.1 82 0.61 Cameroon 12.63 0.7 2.51 Low Senegal 485,586 0.1 82 0.8 Gabon 14.53 0.12 0.54 Low Ghana 357,877 0.04 81 0.29 Ghana 11.52 1.5 1.35 Low Cameroon 335,305 0.05 81 0.09 Guinea 9.00 0.48 2.92 Low Benin 144,382 0.04 83 0.14 Togo 135,294 0.06 84 0.11 Source: Calculations based on data from Jambeck et al. 2015. Liberia 121,050 0.08 84 0.18 Guinea 118,196 0.03 84 0.06 Sierra Leone 96,655 0.04 84 0.11 Mauritania 59,287 0.05 82 0.04 Equatorial Guinea 49,990 0.14 30 0.02 Gabon 32,329 0.05 34 0.02 Guinea-Bissau 30,666 0.05 83 0.06 PROJECTIONS FOR 2025 INDICATE THAT MISMANAGED Gambia 29,646 0.05 84 0.06 PLASTIC WASTE FROM THE AFRICA REGION WILL LIKELY Cape Verde 11,919 0.07 74 0.03 COMPRISE 10.6 PERCENT São Tome and Principe 6,571 0.1 81 0.02 OF THE GLOBAL TOTAL. 11 Most plastic waste in these countries results from domestic use, and the disposal of plastic bags, grocery bags, water sachets, straws, and beverage/water bottles. © Freepik 9 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The rapid expansion of unregulated plastic disposal has Figure 1: Plastics in a tributary of the Odaw river (Ghana), May 2015 Figure 2. Decision-making process for setting policies to reduce marine plastic pollution created a multitude of local problems. Waste plastics clog drainage systems; contribute to widespread flooding and waterborne diseases during the rainy season; degrade sites with potential value for tourism; and contaminate both terrestrial What is the economic cost of marine • Damage to overall ecosystem Section 2 plastic waste to society? services and coastal marine ecosystems. Marginalized communities and those living near plastic waste sites are disproportionately • Aggregation of damage to affected, constituting an environmental injustice (UNEP 2021). individual sectors Plentiful photographic evidence documents the problem (Figure 1). Well-functioning plastic waste-management systems that can address the plastics pollution problem from both a global and a local perspective offer a strong win-win potential. While the reduction of mismanaged plastic waste has been How do these costs compare with mitigation • Incentives © Peter Kristensen Section 3 costs using different approaches? recognized as an important development objective, cost- • Command and control effective remediation has been hindered by the scarcity of information on the true economic cost of plastics. This This analysis, with its focus on the coastal countries of • Removal of plastic waste cost is difficult to estimate, as persistent post-use environmental West Africa (Figure 2), aims to help decision-makers through cleaning, recycling, damage is hard to monetize. In addition to incorporating post-use better understand the factors that should be considered and safe disposal environmental impacts, policy making for remediation requires when developing cost-effective policies for reducing weighing the pros and cons of various market-based policy plastic waste in the marine environment. It begins by instruments – as well as understanding the spatial distribution estimating the economic cost of the externalities generated and timing of plastic-waste generation. The latter issue especially by plastic waste in the marine environment, including damage Weighing the Trade-offs between Environmental and Social Objectives in West Africa is key, because it can affect the relative importance of the policy estimates for ecosystems and related sectors (Section 2). Such instruments. For example, locally targeted prevention and uncompensated external costs are compared with those of the collection may be cost-effective in cases where the generation three main approaches currently used to reduce plastic pollution of plastic waste is highly concentrated – such as in particular Can general economic measures (Section 3). Because of the current lack of data in the countries What are the public health risks (e.g., tariffs on imported areas and seasons. In other cases, a more uniform pattern may of interest, global examples have been used to estimate cost. Section 4 (e.g., from reduced use of plastic Section 5 shift the advantage toward general measures, such as taxation polyethylene) significantly reduce drinking-water containers)? and quantity restrictions. SUP pollution? Achieving the Right Balance between Economic and Waste-Collection Measures Need to identify How should cost-effective cleanup be Section 6 implemented? Examples from Accra and Lagos • Hotspot areas • Seasonal timing for cleanup • Location-specific analyses are needed. What have we learned? • Promotion of awareness, consultation with stakeholders, and development of alternatives are critical. • Import taxes can play a key role. • Economic measures must avoid adverse social impacts. Section 7 • Cleanup measures should be better targeted. © Mel D. Cole for World Bank 11 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 12 2. The economic cost of marine plastic waste to society Plastic products are popular because they are inexpensive, but their market price does not reflect their true environmental cost. As in other sectors, plastics production generates air and water pollution. But, unlike many other products, plastics persist in the environment and generate external costs for long periods before they disintegrate and are assimilated. Thus, the true cost of plastic equals its production cost, plus the costs of production externalities and post-use externalities. The post-use environmental externalities of plastic are the focus of this analysis. Although its impact may differ according to product category,14 plastics are treated as a generic product and average effects are considered. External costs are created in different environments, as plastics undergo life-cycle states after use. These include plastic litter accumulation near points of use, clogged drains, accumulation in landfills, contamination of inland water bodies (rivers and streams) (Van Emmerik and Schwarz 2019), and water transport to the marine environment. In this analysis, the focus of the post-use externality cost estimation is © Smart Edge the marine environment, which provides a very conservative and lower-bound estimate of the true economic cost of plastics. A survey of the literature finds that two broad approaches are currently used to estimate the external cost of plastics in the marine environment. The first approach estimates average cost across all marine The analysis then turns to the material source of the Even if waste generation is significantly reduced dimensions, while the second estimates separate costs for the most critical dimensions and aggregates them plastics pollution problem in West Africa: imported thin through import duties on TPS, or quantity restrictions to an estimated total cost. polyethylene sheets (TPSes) used mainly to manufacture on polyethylene use, appropriately targeted cleanup SUP sachets, bags, and other containers (Section 4).12 programs will remain important. Because plastic-waste The potential efficacy of import duties has been estimated, to cleanup requires significant resources, guidance is required on reduce SUP by raising the TPS price. This estimation covered cost-effective implementation. This study takes a location- and 10 countries for which relevant data could be accessed: Benin, season-specific approach, using detailed information from Cameroon, Côte d’Ivoire, Gambia, Ghana, Guinea, Nigeria, household surveys, geographic and weather data, and measures Senegal, Sierra Leone, and Togo. The study found that such of marine plastic pollution for Accra (Ghana) and Lagos (Nigeria) general economic measures would clearly have environmental (Section 6). Based on its findings, the study offers decision makers benefits; however, conflicts with other social objectives may lessons on achieving the most cost-effective policy solutions also arise. For example, because SUP containers are a major (Section 7). It is expected that this study’s results will contribute source of clean drinking water in West Africa, their reduced use to the development of evidence-based strategies for improved would likely have adverse public-health outcomes for children plastic-waste management and pollution prevention in the from poor families—including increased sickness and death countries of interest. from waterborne diseases. The potential severity of this problem has been estimated from household surveys for Accra (Ghana) and Lagos (Nigeria), since these river systems have been identified as major sources of marine plastic pollution in the region (Section 5).13 12 In virtually all cases, SUP containers (plastic bottles, bags, and packaging) are fabricated from imported polyethylene with the exception of Nigeria, which has 14 See Plastic in the Ocean Statistics 2020–2021 (https://www.condorferries.co.uk/plastic-in-the-ocean-statistics). some domestic production; however, that country is also West Africa’s largest importer of polyethylene. 13 Lebreton et al. (2017) estimate annual plastic emissions from the Odaw River in Accra, and the river systems that discharge waste into Lagos Harbor, at 2.3 million kg and 6.1 million kg, respectively. 13 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 14 Damage to overall marine ecosystem cost factors include medical care for marine debris–related in fish yield and damage to fishing gear caused by marine plastic Biodiversity and ecosystems accidents or illnesses; water in waste plastic containers where litter costs the fishing industry 1–5 percent of output value. services mosquitos breed; air pollution from the incineration of plastic Based on these estimates, the annual cost of plastic litter for When plastic litter contaminates marine ecosystems, Several recent studies have estimated the cost of plastic waste; mortality risks for households situated near garbage world fisheries is in a range of US$1.12–$5.6 billion. Using organisms can suffocate and die from its ingestion. litter in marine environments, as a reduction of social dumps, that are rendered unstable and prone to collapse due to the estimated marine plastic stock of 75–150 million tons, this In addition, plastic litter can reduce the growth rates of seagrass, benefits from marine ecosystems that are treated as natural plastic waste; and life-threatening floods from drainage channel translates to an annual cost of US$56–$279 per ton of plastic coral, and mangroves. Information scarcity in this context capital15 (Buonocore et al. 2020).16 On a global scale, a 2011 blockage caused by plastic waste. Most of these indirect cost litter. It should be noted that these estimates are conservative, hinders comprehensive valuation, but existing data is sufficient study estimated the annual social value of marine ecosystem factors are location-specific, and therefore difficult to value in the because they do not include market-price reductions from to permit external cost estimates for mangrove ecosystems. services at approximately US$49.7 trillion (Costanza et al. 2014). global context. Although estimation of the health externalities of perceived contamination. Plastic litter damages mangrove stands, mainly by A reduction of 1–5 percent in marine ecosystem service delivery, plastics is critical, a comprehensive estimation of these health preventing germination and growth of their seedlings. One related to plastic litter, has been estimated by Beaumont et al. effects on mortality and morbidity was not feasible because the recent study found a negative correlation between the density “risk ratios” are not yet available in peer-reviewed literature. (2019). However, using a recent estimate of microplastics on Marine-linked tourism of plastic debris in mangrove areas, and that of seedlings and the sea floor provided by Barrett et al. (2020), that is nearly trees, as well as the mean diameter and height of trees. Global However, losses are clearly visible in certain other sectors. twice an earlier estimate, a more realistic reduction in marine Plastic pollution of beaches and offshore waters The aggregate cost estimate for the four sectors highlighted studies indicate that each hectare of mangroves provides an ecosystem service delivery would be 3–5 percent. Given the can significantly reduce beach tourist visits and below, ranges from more than US$2,000 to nearly US$7,000 average of 17 tons of woody material per year. Wood loss from estimate of Costanza et al. (2014), this translates to an annual revenue. In response, the Global Tourism Plastics per ton of plastic waste (Figure 4).17 plastic litter can vary between 10 and 50 percent, depending on loss of US$1,500–$2,500 billion in social benefits. Initiative has been formed to reduce pollution, by litter density (Manullang 2020). With a conservatively estimated promoting the elimination of unnecessary plastic items Various studies have estimated the 2011 stock of Figure 4: Damage cost estimates in four sectors due to the presence loss rate of 10–20 percent, the annual loss in woody material and the development of reusable, recyclable, and/or plastics in the marine environment at 75–150 million tons of plastic per hectare could range from 1.7–3.4 tons. compostable plastics. (Conservancy 2015; Jang et al. 2015). Combining these $8,000 Losses can also be imputed from numerous studies that numbers with the annual loss estimate, puts the annual cost in Total leisure-tourism revenue in 2018 was US$840 billion, have ascribed value to mangroves as sources of timber and the range of US$10,000–$33,000 per ton of plastic. $7,000 of which about US$280 billion was linked to marine firewood, flood protection, prevention of shoreline erosion, tourism.19 Various studies indicate that beach litter can lower $6,000 carbon sequestration, water purification, fish spawning, and tourism revenue by as much as 40 percent, depending on the Cost per ton of plastic Figure 3: Damage cost estimates from plastics for marine other biodiversity-related benefits. The estimates vary widely, ecosystem services extent of the littering involved (Jang et al. 2014). A conservative $5,000 with a median value of about US$1,200 per ha (Salem and assumption of 5–10 percent loss in revenue from littering $35,000 Mercer 2012). Assuming 147,186 km2 in global coastal mangrove $4,000 produces economic losses in the range of US$14–$28 billion coverage22 and a range of 37.5–75 million tons of plastic litter $30,000 per year. It is estimated that shorelines are littered by nearly Cost per ton of plastic trapped in neighboring shoreline, the yield is 2.5–5.1 tons $3,000 20 million tons of marine plastic annually, yielding an external $25,000 of plastic litter per hectare of mangroves. The associated tourism cost in the range US$695–$1,390 per ton of plastic. $2,000 external cost could therefore reach US$473–US$946 per ton $20,000 of plastic litter. $1,000 It should be noted that these two approaches to computing $15,000 Value of waterfront property $0 the total economic costs of marine plastics to society are $10,000 Studies on the impact of plastic pollution on property values not fully comparable. The first approach provides a holistic Minimum Maximum are rare, but numerous studies find that real estate prices can estimate of overall costs in the marine environment. In contrast, $5,000 Fishery Tourism Mangrove Beach property be reduced by as much as 25 percent by nearby air or water the second approach only includes a few sector-specific costs $0 pollution (Liu et al. 2018).20 In light of these findings, it seems for which data is available, and the aggregation of these sector- reasonable to assume a 10 percent reduction in the value of specific costs only provide a partial estimate of the overall Minimum Maximum Fisheries and aquaculture a beachfront property for each ton of plastic beach litter found costs. Nevertheless, it is expected that these partial estimates around that location. Beachfront property values vary greatly by will still prove useful for critical sector-level analysis. Estimates Plastic litter can impact fisheries by lowering fish yields, Aggregation of sector-specific costs damaging fishing gear (for example, nets and boat country and locale. In the United States, for example, median presented in this section are important elements of the social beach house values range between US$250,255 and $885,086 cost of plastics and will be useful for future location-specific propellers), and lowering the market prices of products that It is difficult to monetize the externality costs of plastic across a range of locations (2020 figures).21 Assuming a cost-benefit analyses for public and private interventions for are considered contaminated by plastics and their associated by sector. The largest direct effects of plastic waste may be 10 percent valuation loss for plastic beach litter, a 30-year use waste management. chemicals. Among these, the price-reducing impact is the most the public-health impacts from consumption of microplastics period, and a 3 percent interest rate, the annual loss range could difficult to evaluate. in marine food products, but results to date are inconclusive. be US$1.314–$4,647. Converting this loss at purchasing power Nearly total absence of reliable secondary data on the Global fish output was US$159 billion in 2019, of which parity for 37 large countries yields an average global annual loss quantitative impacts of microplastics prevented the inclusion of marine output was estimated at US$112 billion.18 Various in the range of US$1,207–$4,269. the externalities of microplastics in the analysis. Indirect health- studies in different locations have estimated that the reduction 15 Defined as the world’s stocks of natural assets. 19 See https://www.statista.com/topics/962/global-tourism/#:~:text=Globally%2C%20travel%20and%20tourism’s%20direct,at%20580.7%20billion%20U.S.%20dollars. 16 Most of these studies are regional, and very few attempt a global-level evaluation. 20 Also see https://courses.lsa.umich.edu/healthy-oceans/group-1/group-1-sub-1/plastic-pollution-and-its-economic-damage/. 17 These widely varying sectoral and global estimates can provide benchmark-cost estimates for the 17 coastal countries of West Africa, and can be replaced by 21 See https://www.vacasa.com/top-markets/2021-best-place-to-buy-a-beach-house. country-specific estimates where appropriate supporting data becomes available. 22 https://www.mangrovealliance.org/30-years-of-global-forest-data/#:~:text=%E2%80%9CMore%20than%2040%20percent%20of,and%202020%20(Table%2031). 18 Details are available at http://www.fao.org/fishery/statistics/global-production/en. 15 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 16 3. How the economic costs Incentives Demand-side policy instruments can reduce plastic waste Command and control A second approach for minimizing the external costs compare with those of by adding external costs to the market prices of plastic generated by specific plastic products is to ban their use, products. One common practice is levying production excise through direct regulations prohibiting their use and by taxes or import duties on raw materials, such as polyethylene. imposing high enough penalties to enforce the ban. While Another incentive-based approach taxes plastic products at successful bans will eliminate the cost of external waste for such the main approaches to the point of sale. Section 2 showed that the external cost of plastic waste in the global marine environment was in the range of US$10,000–$33,000 per ton. The tax or duty imposed on products, they may also force consumers to forgo the associated conveniences or pay higher prices for biodegradable substitutes. pollution mitigation The production cost of biodegradable paper bags the production of plastic, or the extra price charged on its sale, is US$0.04–0.05 per unit, compared to the US$0.01 should be commensurate with this level of external cost. manufacturing cost.24 Using these values, replacing SUP bags The case of SUP bags is analyzed for illustration. Since with paper bags would involve an extra cost of US$0.05 per each kilogram of plastic yields, on average, 180 SUP bags, bag. Because each kilogram of plastic yields 180 plastic bags The above section shows that even conservative methods yield significant external cost estimates internalizing the external damage of US$10,000–$33,000 per on average, this replacement would translate into an extra cost for plastic waste in the marine environment, and action is clearly warranted in the majority of cases. ton of plastic would add US$0.06–0.18 to the price of each of US$9,000 per ton of plastic. Since this cost is less than the A study of global experience reveals three main approaches to reducing plastic pollution, all of whose costs bag. Adding this to the average retail cost of US$0.03 per external damage cost of US$10,000–$33,000 per ton of plastic, fall within the range of the estimated external costs for plastic waste. These approaches, which are described plastic bag would result in a final bag price of US$0.09–0.21.23 using paper bags as an alternative to support the ban would below, are not mutually exclusive. Rather, they can be tailored to a particular country’s local economic and Numerous studies have examined the impact of a higher price enhance its social benefit. political conditions, so as to achieve the most cost-effective mix. on demand for SUP bags. One study in Ireland revealed a use The use of reusable bags instead of plastic bags also reduction of nearly 100 percent at a per-bag price of US$0.15 involves extra cost, as it is often inconvenient to carry and (Convery, McDonnell, and Ferreira 2007). A number of other reuse bags. Also, reusable bags may pose health hazards.25 studies found major reductions with charges of US$0.05 or more Although the costs of inconvenience and health hazards are (Dikgang et al. 2012; Homonoff 2018). These findings suggest difficult to estimate directly, it is possible to obtain an indirect that internalizing all the costs associated with a plastic bag (its valuation of these costs from studies that have estimated the manufacturing costs and negative environmental externalities) costs of plastic bans using willingness-to-pay (WTP) analyses would probably reduce its use to zero. of consumer surveys. These studies report WTP values However, several caveats should be noted. Firstly, such in the range of US$0.03–0.08 for a plastic bag alternative, economic incentive schemes can entail significant operating depending on the location and nature of the population surveyed costs. If such schemes are carefully designed, experience (Convery, McDonnell, and Ferreira 2007). Based on these WTP shows that operating costs can be reduced to as low as values and converting to tons using average bags per ton in 3 percent of revenues (Convery, McDonnell, and Ferreira production yields, reusable bag costs appear to have a range of 2007). Secondly, the convenience of many plastic products may US$5,760–$14,220 per ton of plastic—which is lower than perpetuate their use, even when charges are significant. This the external damage estimate (US$10,000–$33,000 per ton could be particularly true when consumers are unaware of the of plastic). Thus, in view of the high external damage caused environmental-change component. Thirdly, price effects may by plastics, eliminating their use through legislative bans and erode over time as incomes change and preferences shift. switching to paper or reusable bags should enhance overall social welfare. One should also note that plastic bans may face significant implementation difficulties. These include direct implementation costs, opposition from the plastics industry, job losses in that sector, and diversion of demand to untaxed parallel markets that can only be suppressed with high enforcement costs. 23 Manufacturing a plastic bag costs about US$0.01, and it retails for about US$0.03 at the consumer level. 24 https://www.bmt.com/what-is-the-real-cost-of-paper-vs-plastic/ 25 For example, reusable grocery bags can transmit bacteria and viruses to other shoppers and store employees. 17 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 18 4. How general economic measures could significantly reduce plastic pollution Marine pollution from plastics can be reduced significantly by reducing single-use plastic waste. As noted above, direct economic measures (taxation at points of production, or sale, or product bans) can be used to reduce marine pollution from SUP waste. However, targeted policies imply their targeted enforcement, which can be difficult to achieve when producers, sellers, and consumers are widely dispersed.26 In West African coastal countries, all SUP sachets, bags, and other containers are fabricated from thin polythene sheets (TPSes), virtually all of which are imported.27 One appealing policy would directly target bulk imports of polyethylene, the essential feedstock for SUP containers, at relatively few ports of entry. Tariffs are already familiar in all West African coastal countries, and implementation could be limited to relatively few port areas and a small, more-easily-monitored cadre of agents. However, evaluating this policy option also requires understanding the expected response of imports of TPS to the imposition of tariffs. The responsiveness question could be addressed by analyzing polyethylene imports and their prices over time. Figure 5 displays the TPS imports trend over the 1995-2019 period for 10 countries in the region.28 Steady growth is evident, with substantial interim fluctuations. As shown, import prices for TPS have varied more than fourfold since 1996, and the growth of TPS imports has been accompanied by wide price fluctuations (Figure 6). © Smart Edge Removal of plastic waste through plastic bag and bottle waste, since this accounts for the bulk of plastic litter. They show that removal and recycling costs can cleaning, recycling, and safe disposal range between US$0.01 and US$0.08 per unit of plastic product In principle, both incentive-based and command-and- (Burnett 2013; Taylor and Villas-Boas 2016). Incorporating control approaches can significantly reduce the use of average product weight yields a cost range of US$1,920– many plastic products; however, for some types of plastic $14,220 per ton of plastic for plastic bags, and one-third to one- products, complete elimination may not be a feasible half of those estimates for plastic bottles, which are heavier. option. Thus, the removal of specific plastic products before With increased efficiency of removal and recycling operations, they enter the waste stream will be beneficial for protecting the these costs should decrease over time. Since they are much marine environment. Plastic-waste removal includes collection less than the external damage estimate (US$10,000–$33,000 for recycling from users, targeted removal of plastic litter from per ton of plastic), removal of plastic litter should also lead to identified natural traps, and reuse or safe disposal. The costs enhanced social welfare. at each step vary widely with the volume of plastic waste and its spatial dispersion. This approach is cost-effective, if its cost is less than the external cost of plastic waste. 26 Small-scale producers that are widely dispersed can command large output shares for products. If both sales outlets and consuming households are widely dispersed, targeted quantity or price policies will require a large cadre of low-wage agents – who are both disciplined and incorruptible. In practice, meeting such Numerous cost studies for rivers and lakes have analyzed conditions has generally proven difficult. 27 The UN Comtrade database, under code 57111, classifies TPS as “polyethylene sheets, with a specific gravity of less than 0.94.” 28 Since 1995, West Africa has sourced TPS mainly from four supplier regions ¬– the European Union (principally France and Belgium), East Asia (Republic of Korea and China), the United States, and the Middle East (Saudi Arabia and Qatar). Over time, the supplier shares of the European Union and East Asia have declined, while the shares of the United States and the Middle East have increased. 19 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 20 Figure 5. TPS imports by 10 West African countries, 1995–2019 Figure 6. Imported TPS price for West Africa, 1995–2019 5. 80 2 The public health risks: Real import price (US$/kg), SITC3 57111 Import quantity index, SITC3 57111 60 1.5 40 1 The case of plastic water 20 0.5 containers in Ghana and 0 1995 2000 2005 2010 2015 2020 0 1995 2000 2005 2010 2015 2020 Nigeria Year Year Note: The 10 countries are Benin, Cameroon, Côte d’Ivoire, Gambia, Source: Comtrade. Import duties and other measures to reduce plastic waste have clear environmental benefits, but Ghana, Guinea, Nigeria, Senegal, Sierra Leone, and Togo. The Comtrade decision-makers must also take other social objectives into consideration. For example, SUP containers database has no TPS import entries for Liberia. (water sachets and bottles) are a major source of clean drinking water in West Africa. Thus, reducing the use Source: Comtrade. of these containers may increase sickness and death from waterborne diseases. If so, public health may suffer significantly from the use of general price- or quantity-based instruments that limit plastic consumption This study performed an econometric analysis of TPS Second, the evidence shows that price increases on the and waste. import demand for seven countries: Benin, Cameroon, world market have produced rapid, proportionate reductions Since this is a potentially critical policy question, an econometric analysis was conducted in two Côte d’Ivoire, Ghana, Guinea, Nigeria, and Senegal in West African TPS import demand. Since producers are countries – Ghana and Nigeria – to test the child-health impact of plastic container use. A database (Appendix A).29 The results indicate a high degree of income indifferent to the sources of price change, the same thing will was constructed from demographic health surveys (DHSes) in Ghana (2003, 2008, 2014) and Nigeria (2003, and price responsiveness: Each 1 percent increase in national be true for price increases from import duties. Thus, a TPS 2008, 2013, 2018), which reported caretaker responses for 12,500 and 99,500 children, respectively. For each income increases TPS imports by about 1 percent, and each tariff could be a potent weapon in the struggle to reduce SUP child, caretakers reported mortality status; recent incidence of diarrhea; gender; age in months (age at death 1 percent increase in TPS price reduces imports by about pollution. However, a tariff may have a disproportionate impact for mortality); years of mother’s education; real household income; and the primary source of their drinking 1 percent. From a policy perspective, these results have two on the poor, and policy makers should consider the distributional water, including plastic drinking containers (sachets and/or bottles). major implications. Firstly, because the real TPS price exhibits implications if they choose this option. no trend over time, TPS imports and the waste generated by SUP containers should keep pace with national income growth without putting countermeasures in place (Table 2). For Ghana, the region’s fastest-growing economy, this would mean nearly doubling the amount of SUP waste over the coming decade. Table 2: Real GDP growth in West Africa, 2005–19 Country Growth rate (%) Ghana 6.5 Togo 5.2 Côte d'Ivoire 5.2 Senegal 4.5 Nigeria 4.4 Benin 4.3 Cameroon 4.2 Guinea 4.1 Gambia 2.8 Sierra Leone -0.9 Source: World Development Indicators. 29 Data problems prevented the inclusion of Gambia, Liberia, Sierra Leone, and Togo. 21 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 22 The econometric analysis tested whether child morbidity Figure 7 and Table 3 summarize the strong effects revealed by Table 3: Child-health impacts with and without plastic water container use in Ghana and Nigeria and mortality were lower in households that used plastic the analysis. containers for drinking water, all else being equal. After Country Mortality rate (per 1,000) Incidence of diarrhea (per 1,000) The box plots in Figure 7 display the distributions of With Without With Without With Without With Without controlling for income, education, and other socioeconomic predicted child mortality (Figure 7a) and diarrhea (Figure factors, the analysis found significantly lower mortality rates and 2003 2014 2003 2014 7b), scaled to rates per 1,000 children. The figures display incidence of diarrhea for children in households that used plastic Ghana 50 81 26 45 116 145 92 117 similar patterns—that is, notable declines attributable to plastic water containers (Appendix B). To explore the implications, water container use for comparable measures, both across and 2003 2018 2003 2018 the econometric results were used to predict mortality rates within years. Nigeria 113 137 72 90 113 124 89 99 and diarrhea incidences for all children (0–5 years of age) in the sample, with and without use of plastic water containers. Source: Demographic and Health Surveys. Figure 7: Child health impacts of plastic water container use in two countries Table 3 summarizes the median values in Figure 7. The most In summary, the message from these results seems clear recent results for Ghana (2014) show that with plastic water and highly-relevant for formulating plastic-waste reduction a. Mortality rates container use, the median predicted child mortality rate fell by policies in both Ghana and Nigeria: Analyses of large samples, 42 percent (from 45 to 26). The equivalent result for Nigeria in drawn over extended periods from high-quality household Without plastic container use With plastic container use 2018 was a 20 percent decline (from 90 to 72). For the median surveys, provide strong evidence that use of plastic sachets Ghana Nigeria predicted child incidence of diarrhea in the most recent years and bottles for drinking water significantly reduces mortality under analysis (2014 and 2018), the two countries’ respective and incidence of diarrhea among children—after controlling 2003 2014 2003 2018 rates declined by 21 percent and 10 percent when households for the other determinants of child morbidity and mortality 200 300 sourced drinking water from plastic containers. Clearly, these that are widely cited in the literature. Therefore, reducing are not small effects. the use of plastic drinking-water containers may significantly increase childhood illness and death.31 This suggests that 200 The econometric results align with the widespread belief policy makers who opt for reducing SUP containers should among West Africans that water in plastic containers 100 also consider countervailing health policies, such as targeted is cleaner and safer, than water from other sources.30 measures to compensate for the potential impacts on child However, the benefits of cleaner water may be underestimated 100 health—particularly in poorer households. In light of these results, if the sample data includes contaminated containers. Some assessing potential conflicts with public-health objectives, contamination has been revealed by sample-based analysis of using DHS data from other countries, is clearly a domain for water sachets in Accra (Kwakye-Nuako et al. 2007) and Lagos 0 0 additional research. (Omolade and Gbadamosi 2017). The larger estimated benefits (Excludes outside values) (Excludes outside values) for Ghana, shown in Table 3, suggest that the issue of water contamination may be greater in Nigeria. b. Incidence of diarrhea Without plastic container use With plastic container use Ghana Nigeria 2003 2014 400 2003 2018 NOTE 300 In each plot, boxes are bounded by first and third 300 quartile values, medians 200 are identified by interior horizontal lines, and limits 200 for non-outlier values 100 are identified by top and 100 bottom horizontal lines. For further discussion, see Tukey (1977). 0 0 (Excludes outside values) (Excludes outside values) 30 These results do not indicate that plastic water containers are cleaner and safer in all cases. Actual quality in specific cases may be problematic, in the absence of consistent public testing and certification. 31 Although these estimates may fully or partially reflect the influence of unobserved variables that are correlated with plastic container use, the results are cautionary, given the size of the estimated impacts on and stakes for public health. © Smart Edge 23 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 24 6. Incorporating plastic disposal hotspots Knowing with precision where plaster litter accumulates areas, during the 2003–19 period, used household survey data on plastic container use and household economic status from multiple rounds of demographic health surveys (DHSes) and is critical for cost-effective cleanups. In the absence malaria indicator surveys (MISes).36 The findings indicate that How to implement of geocoded data by type of litter, this study estimated household income has a large effect on plastic-container use location-specific waste accumulation by combining in both areas. In addition, a highly significant time trend in both population maps with three components of the relationship cities indicates that plastic container use has spread rapidly cost-effective cleanups: between income and plastic container use: (i) the effect of across income groups over time. overall income growth on demand; (ii) the diffusion of demand for plastic sachets, which entered the market in the late 1990s;34 Figure 8 illustrates the implications of results from the econometric analysis (Appendix C). In both Accra and Examples from Accra and (iii) the spatial distribution of plastic-container use, which reflects the spatial distribution of household income. Lagos, the intensity of plastic container use over the 2003–19 period exhibited major increases. In Accra, use by the poorest households increased from less than 10 percent to nearly and Lagos 80 percent, with use by the richest households increasing from Household income and population data about 30 percent to nearly 100 percent (Figure 8a). Diffusion to Real income per capita since 2000 has approximately the poorest households was less pronounced in Lagos than in doubled in Ghana and Nigeria. Since plastic containers are Accra; even so, incidence of use among the richest households normal goods,35 one would expect income growth to have increased from less than 30 percent to about 70 percent Discussion in the previous sections shows that effective strategies for reducing plastic-waste pollution increased the demand for plastic-water containers – all else (Figure 8b). The econometric results thus highlight the importance in the marine environment need to balance pollution prevention, with cost-effective cleanup measures. being equal. This analysis of the Accra and Lagos metropolitan of residential income data, to identify areas with a higher incidence Scarce cleanup resources should target “hotspots”—that is, those areas with a particularly high incidence of plastic-container use and disposal in Accra and Lagos. of plastic-waste disposal.32 In the context of marine pollution, the most critical inland points are those near If households are strongly clustered by income, then plastic- rivers that carry plastic waste to the ocean.33 This section develops an illustrative cleanup strategy for marine waste hotspots will occur in higher-income areas. plastic pollution in Accra and Lagos. The focus, in this case, is plastic drinking-water containers, since the previous section shows that bans or severe use restrictions may be inadvisable because of their potentially adverse public-health effects. The results suggest that targeted policies may have a temporal, as well as a Figure 8. Incidence of plastic container use, by household income per capita, 2003–19 spatial, component. a. Accra 100 2019 HH 80 2016 HH 60 2014 HH 40 20 2008 HH 0 2003 HH 1000 2000 3000 Real income per capita b. Lagos 80 2003 HH 60 2013 HH 40 2010 HH 2008 HH 20 2015 HH 0 1000 2000 3000 2018 HH Real income per capita Source: DHS and MIS surveys. 32 At the outset, it will be noted that modeling leakages and quantification of location-specific environmental impacts (soil, groundwater etc.) were not possible as 34 Sachets account for the bulk of plastic drinking-water containers used in Ghana and Nigeria. Sachets first appeared on the market in the late 1990s, when high-resolution geocoded data required for such analyses are not yet available in the countries of interest. entrepreneurs in West African cities began using new Chinese machinery that heat-sealed water in plastic sleeves (Stoler et al. 2012). 33 Estimates indicate that approximately 80 percent of the world’s ocean plastics enter the ocean via rivers and coastlines (Li, Tse, and Fok 2016). 35 Normal goods are those for which demand rises with an increase in consumer income. 36 This analysis is based on five surveys conducted in Accra (DHS 2003, 2008, 2014; MIS 2016, 2019) and six surveys conducted in Lagos (DHS 2003, 2008, 2013, 2018; MIS 2010, 2015). 25 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 26 Residential clustering in Accra and Lagos was tested by vary from 32 to 94 in Accra, and from 20 to 82 in Lagos; clusters dividing the metropolitan areas into square cells, that each of high-income (brown) areas and low-income (blue) areas are Figure 10: Maps showing plastic-water container use and disposal over time measure 1 km on their sides. Household survey data from the clearly visible on the maps (Figures 9a and 9b). For both cities, past two decades were then used to compute mean income residential clustering by income has remained roughly stable a. Accra percentiles for each cell. Figure 9 shows that cell percentiles over a long period.37 Figure 9: Maps showing residential income clustering a. Accra b. Lagos Mean income percentile Mean income percentile 32-38 46-47 54-57 20-27 45-48 56-59 39-42 48-50 58-66 28-37 49-51 60-64 43-45 51-53 67-78 38-44 52-55 65-72 79-94 73-82 0-29 Disposal trends low use in 2003, the maps display the rapid onset of widespread plastic-container use after 2010. The second pattern shows the 30-43 Identification of hotspots also require information on the spatial variation in population density, which is reflected in the number of households and the population figures involved, 44-59 spatial gradations of plastic use. The third pattern relates to as total waste load depends on total use. Figure 10 combines population clustering by income, which differs markedly between 60-77 income and population data into estimates of plastic container the two cities. In Accra, the overall pattern is roughly concentric, 78-94 use and disposal trends over the 2003–19 period.38 Areas with with aggregate plastic use declining from the area of highest the least and greatest disposal are shown in blue and brown, 95-113 population density (Figure 10a). A small exception is posed by respectively. Three clear patterns are evident from the maps. a separate northeast cluster. By contrast, the pattern in Lagos 114-128 The first one reflects the previously mentioned dominance of is polycentric – with three visible clusters that exhibit increasing 129-145 changes in plastic-use intensity over time. Starting with uniformly density on a roughly north-south axis (Figure 10b).39 146-163 164-185 37 Appendix D provides technical details. 38 Population indicators for the 1 km grid cells were developed from data provided by the Worldpop project at the University of Southampton (Lloyd et al. 2019). 39 Appendix D, Section D2, provides a detailed technical presentation of the analyses in this section. © Smart Edge 27 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 28 Figure 11: Maps of river systems in Accra and Lagos b. Lagos a. Accra b. Lagos Numerous researchers have attempted to measure the The steeper the gradient (or the greater the net elevation of transport of plastic waste by river (Schmidt et al. 2017; van point A), the more likely that the waste discarded at point A will Calcar and van Emmerik 2019; van Emmerik and Schwarz be transported to the river during the runoff period.40 But this 2019; Windsor et al. 2019). For this study, their analyses have effect will be attenuated by the distance from point A to the river been extended to a simple geophysical model for identifying (Figure 12). spatial clusters whose plastic-waste generation has particular To summarize, the likelihood of river deposition of waste importance for ocean pollution. The model incorporates three discarded at point A increases with the net elevation of sources of plastic transport by rivers: (i) direct dumping by 0-36 point A, and decreases with the distance between points riverside households; (ii) downhill transport by rainfall runoff; and 37-66 A and B. What applies at a distance also applies for points (iii) transport by floods. In the vicinity of a river, the probability near a river: The transport likelihood is greatest at locations 67-97 that a plastic container discarded at point A will enter the river is on the riverbanks, where net elevation and distance are both modeled using three variables: (i) the elevation of point A; (ii) the 98-129 zero. Points along the river’s floodplain also have high likelihood elevation of river point B closest to point; and (iii) the distance 130-165 during periods of heavy rainfall, since both their net elevation from point A to point B. The model’s operation can be illustrated and distance are low. The likelihood of transport declines as 166-205 by imagining the impact of torrential rain, and its runoff, on a distance increases, but may be relatively high if areas with a plastic container where it is discarded (point A). The likelihood 206-265 high net elevation are not too distant.41 that runoff will transport the container to the river depends, in 266-347 part, on the gradient from point A to the river—which can be At all grid cells in Accra and Lagos, the model incorporates 348-428 approximated by the difference in elevation of points A and B the effects of distances from the closest river points, as well 429-510 (“net elevation” of point A). as their net elevations to display the relative likelihood that Figure 12: River transport of plastic waste a plastic container dumped in the cell will be transported by a river. The results are presented in Figure 13 (page 31), which d Incorporating rivers as conduits of plastic rivers into which plastic waste has been directly dumped – either identifies rivers by white lines; least-likely areas for riverine waste transported downhill by rainfall runoff, or picked up by seasonal transport in dark blue; and most-likely areas in dark brown. The waste floods (Lebreton et al. 2017). A spatial patterns for the two cities differ markedly, owing to their ne B The analysis presented above finds strong evidence, in both unique topographies and river systems. In Accra, the likelihood Figure 11 displays the river systems in Accra and Lagos. Accra and Lagos, of the spatial clustering of plastic waste. of plastic-container transport declines continuously with distance As shown, Accra is dominated by the Odaw system, River Local environmental impacts are undoubtedly significant, and from the Odaw River and its tributaries, with modifications for whose basin roughly bisects the metropolitan area along a comprehensive treatment of plastic-waste deposition should areas where higher elevation has greater runoff during rain and a north-south axis (Figure 11a). In Lagos, the rivers are more consider targeted measures to address this problem. However, flooding. In Lagos, areas of high transport likelihood are defined scattered, with the Owo system in the west, the Ogun system A = Point inland where plastic waste is discarded reducing marine-plastic pollution requires greater geographic by three riverine areas, while a large swath of interior territory in the northeast (which empties into the Lagos Lagoon), and B = River point nearest to point A focus. Unless plastic waste is deposited within the tideline, it ne = Net elevation has a low likelihood of waste transport. numerous waterways in the south, including Badagry Creek, cannot pollute the ocean without conduits. The main ones are d = Distance between points A and B Ajegunle Canal, and Lagos Harbor (Figure 11b). 40 Appendix E provides maps of “net elevations” (elevation of each point minus elevation of nearest river points) for Accra and Lagos. 41 The role of terrain in plastic-waste discharge to rivers has been little explored by empirical research. Full empirical analyses would be challenging, given the need to geolocate the initial positions – and subsequent locations – of a large sample of waste plastic items over an extended period. 29 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 30 Figure 13: Maps showing the likelihood of plastic waste being transported by river Combining hotspot data with the likelihood of river transport a. Accra b. Lagos Figure 14 combines the information on plastic-waste hotspots (subsection 6.1), and the likelihood of river transport (subsection 6.2), to produce marine plastic-pollution hotspots for Accra and Lagos.42 In Figures 14a and b, the unadjusted maps on the left-hand side show predicted plastic-container waste generation (previously shown in Figures 10a and b, respectively). The maps on the right-hand side, adjust the left-hand side maps, for the likelihood that plastic-waste containers will be transported by river to the ocean. In each case, the incorporation of river transport creates a marine pollution hotspot map, which differs significantly from the general (unadjusted) hotspot map. Figure 14. Plastic hotspots, with and without adjustment for rive-transport likelihood a. Accra Unadjusted, 2019 (Figure 10a) Adjusted Likelihood Score Likelihood Score 0-5 31-35 61-65 91-95 1-5 31-35 61-65 91-95 6-10 36-40 66-70 96 6-10 36-40 66-70 96 11-15 41-45 71-75 97 11-15 41-45 71-75 97 16-20 46-50 76-80 98 16-20 46-50 76-80 98 21-25 51-55 81-85 99 21-25 51-55 81-85 99 26-30 56-60 86-90 100 26-30 56-60 86-90 100 0-29 60-77 114-128 0-6 22-30 55-68 30-43 78-94 129-145 7-12 31-40 69-86 44-59 95-113 146-163 13-21 41-54 87-116 164-185 117-157 b. Lagos Unadjusted, 2018 (Figure 10b) Adjusted 0-36 98-129 206-265 0-12 42-61 99-126 37-66 130-165 266-347 13-24 62-80 127-158 67-97 166-205 348-428 25-41 81-98 159-237 429-510 238-341 42 Spatially distributed predicted plastic-container depositions are multiplied by spatially distributed likelihood scores for river transport. © Xavier Bourgois 31 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 32 Incorporating seasonal rainfall cycles waste transport requires an understanding of rainfall patterns, together with actual evidence of ocean pollution. The river-transport model used in this study follows the recent literature in assigning an important role to rainfall, Figure 15 displays monthly rainfall patterns for the two as a source of both flooding and runoff (Lebreton et al. cities. In Accra, rainfall increases from January to an annual 2017). Plastic waste disposal occurs at a steady rate, which is peak in June; declines through August; increases to a lower determined by daily or weekly plastic container use. Runoff from peak in September/October; and declines during the course rainfall transports plastic container waste to rivers at some rate, of December (Figure 15a). Lagos’s pattern is more compact, even during months when the average rainfall is low. But plastic with an increase from January to July; a decline in August; an waste will tend to accumulate in the dry months if the dry-season increase to another peak in September; followed by a decline rate of transport is below the rate of plastic waste accumulation, during the course of December (Figure 15b).43 until the next rainy season increases transport to the nearby rivers once again. Understanding the temporal dynamics of Figure 15. Median monthly rainfall, 2010–20 a. Accra 150 b. Lagos Mean precipitation 2010–2020 (mm) Median precipitation 2010–2020 (mm) 120 100 100 80 60 50 40 20 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec © Mel D. Cole for World Bank Figure 16: Plastic pollution measurement area for Accra Assessing the relationship between seasonal rainfall Figure 17: Monthly plotting results for the Accra offshore area cycles, and the timing of riverine deposition and transport of plastic waste, requires a model of their interaction. a. Floating debris index b. Plastics index As no such systematic study of this relationship is available in the literature, actual evidence of ocean plastic pollution 700 44 in coastal waters was analyzed using satellite images. Relevant spectral images, from the European Space Agency’s 600 Sentinel-2 satellite platform, were downloaded for the period 42 Floating debris index Ocean plastic index December 2019 to November 2020. Sufficiently clear images for the ocean area immediately abutting the mouth of the Odaw 500 40 River were available for the Accra analysis (Figure 16), but images for Lagos were unfortunately too cloudy to perform a 400 comparable analysis.44 38 These Accra satellite images were used to compute a floating-debris index adjusted for plastic content from the methodology 300 of Biermann et al. (2020) (Figure 17a), and a plastics index based on the methodology of Themistocleous et al. (2020) 36 (Figure 17b). The plotting results paint a picture quite similar to the seasonal plastic pollution cycle in the coastal waters abutting 200 the Odaw River: a rapid increase from December to a peak early in the first-quarter rainy season, followed by a steady decline in mid-to-late summer and consistently higher levels in November.45 Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov 2019 2020 2019 2020 43 Considerable uncertainty surrounds measurement of rainfall in Accra and Lagos in any given year, because daily reports from the relevant ground stations are Source: European Space Agency. relatively sparse. Records for Accra could be accessed on average for 72 reporting days per year, and for Lagos for 44 reporting days, which was clearly insufficient for tracking daily or monthly precipitation in any given year. For this study, monthly median precipitation is therefore computed from These results provide clear evidence of the accumulation of plastic-container waste in hotspots during low-rainfall periods, followed World Weather Online daily data downloads.This data takes the form of recordings of daily forecasts, not actual observations, but aggregation over the ten-year period 2010-2020 provides at least an approximation to the annual pattern. by rapid river transport by means of flooding and runoff with the return of heavier rainfalls. The potentially important policy implication 44 Clear imagery for Accra includes two images for December 2019 and January, February, March, August and November 2020, together with three images for April is that cleanup resources should be concentrated in marine plastic hotspot areas, before the first-quarter rainy season begins. 2020 and one image for May. 45 Both the indices mentioned are based on new methodologies that have not been extensively tested, beyond their original applications in other regions. 33 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 34 7. More local case studies on sector-specific losses from plastic wastes are needed in Awareness raising initiatives, stakeholder participation in policy and strategy West African countries. design, and access to environmentally What we have learned Although it is difficult to monetize the adverse coastal friendly alternatives are key to effective impacts of marine plastic litter, the global studies cited waste management. in Section 2 indicate that even conservative estimates Global experience indicates that both bans, and price- of externality costs are high. At the same time, global cost based strategies, can be effective methods to reduce Researchers estimate that 8.3 billion tons of plastic has been produced since the 1950s, with roughly estimates vary widely by location. At present, West African plastic waste. However, effective waste management also 60 percent ending up in landfills or the natural environment.46 In 2015, global production of mismanaged coastal countries do not have sufficient data for the estimation requires broad-based awareness about plastic pollution and plastic waste totaled 60–99 million tons, and a business-as-usual scenario increases this range to of country-specific and/or sector-specific costs; and more local stakeholder engagement when designing mitigation policies and 155–265 million tons annually by 2060. Without waste-management improvement, plastic waste in the oceans case studies are needed for the computation of sector-specific strategies. Regular public consultations can also help to promote is predicted to increase by an order of magnitude. Mismanaged plastic waste unfortunately accumulates and losses resulting from plastic waste. Better data on waste plastic awareness and create the necessary political will. Experience persists in the oceans, with adverse consequences for marine ecosystems and potential damage to human externalities can play a key role in assessing the benefits and shows that people are more likely to accept a ban or price-based health. The stakes for coastal countries are also high, because marine plastic litter adversely affects fisheries costs of policy options for plastic waste remediation. strategy if they have access to suitable, environmentally friendly and aquaculture, biodiversity, coastal ecosystems, tourism, and waterfront property values. alternatives that are reasonably priced. The development and promotion of alternative or reusable products require planning Location-specific analyses are needed to well in advance of implementing plastic-reduction policies. determine the most cost-effective policy The jobs created in the alternative sectors can also mitigate the opposition that may arise from the potential loss of employment mix for plastic-waste remediation in in the plastics industry.47 Taking the extra time for advance each focus country. planning and publicity can help both the plastics industry, and the West African coastal countries require urgent intervention, general public, to adjust to a scenario with lower plastic usage. because mismanaged plastic waste in the marine West African countries can improve their waste-management environment will continue to increase at alarmingly high performance by learning from successful global experiences. rates (Lebreton and Andrady 2019). However, there is no one- size-fits-all solution. Currently, 12 of the 26 member countries in the Economic Community of West African States (ECOWAS) have some type of SUP policy. Of those, 11 have weakly enforced plastic bans, while one (Ghana) has a price-based strategy based on excise taxes (Adam et al. 2020). As options for plastic waste management improve, the most practical policy solutions will likely entail some combination of quantity- and price-based approaches balanced with cleanup strategies, as discussed in Section 3. Determining the most cost-effective policy mix for each country should involve a degree of location-specific analyses. In addition, practical problems, such as pressure from stakeholders and political viability, will likely be major considerations. At present, the countries of interest lack sufficient data for estimating country-specific costs – which suggests the need for more local case studies to compute sector-specific damages and losses from plastic waste. Better data on the externalities of plastic waste could also play a key role in assessing the benefits and costs of remediation policy options. © Smart Edge 46 It is estimated that approximately 80 percent of marine-plastic debris results from land-based sources (Lebreton et al. 2017). 47 In practice, enforcement of plastic-use reduction policies can be difficult because the plastics industry employs thousands of people – serving as a livelihood for many families and a significant source of government revenue (Behuria 2019; Death 2016; Jambeck et al. 2018). For example, the government of Côte d’Ivoire revoked its plastic ban in 2013 because of threats and demonstrations by the plastics manufacturers’ association and employees, who were provoked by the potential revenue loss of over 7,600 jobs. 35 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 36 © Mel D. Cole for World Bank Import taxes on polyethylene sheets Cleanup measures should be better can play a key role in reducing single-use targeted. plastic waste, but understanding the Priority should be given to areas with a high incidence distributional implications for the poor of plastic-waste disposal near rivers, particularly more is critical. elevated areas with steeper slopes. Focusing on waste cleanups before the onset of the first-semester rainy season Taxation of the imported polyethylene, which comprises will be most effective. most of the production feedstock for SUP in West Africa, is a potentially effective, price-based policy option with Although policies to reduce plastic waste are critical, realism also relatively low administrative costs. Ghana’s waste- dictates the need for improved waste-collection measures. For reduction strategy, which employs taxation of imported the marine environment, it is particularly important to develop plastics, provides a useful example (Adam et al. 2020). As collection strategies that target areas where high-waste volumes discussed in Section 4, plastic demand exhibits a very elastic also have a high likelihood of transport to the ocean via local response to changes in the price of imported polyethylene. rivers. Where feasible, these strategies should be informed by Import taxes have a potentially major cost advantage over continuous sampling to identify such areas. For more resource- directly targeted measures since the former can be administered constrained environments, the analysis in Section 6 shows how at relatively few entry points, while the latter require a widely readily available economic, demographic, and topographical distributed cadre of enforcement agents. Taxation of imported information can be combined to identify the “hotspot” areas plastic may be urgently needed because historical evidence where high-volume waste transport via rivers is most likely. This suggests that, without countervailing policy measures, the analysis also shows that the intensity of the waste-collection future income growth rate in West Africa will be matched by a activity should vary with the annual rainfall cycle, since the concomitant growth rate in plastic demand. However, a tariff may greatest river transport of plastic waste to the ocean occurs at have a disproportionate impact on the poor, and policy makers the beginning of the rainy season, when large volumes of waste should consider potential distributional implications before they accumulated during the dry season are flushed into rivers by the implement a tariff on polyethylene. onset of heavy rains. An additional lesson from the river-transport analysis is that policies to reduce marine-plastic waste should often be formulated for river basins, rather than operating at national Economic measures must avoid adverse or local levels. This suggestion may require the development health impacts. of institutions that foster coordinated implementation of waste- While the case for public interventions to reduce plastic management policies, across urban areas drained by the same waste seems clear, Section 5 suggests that attention river basin. must also be paid to potential conflicts with public-health outcomes. Empirical analysis strongly suggests that clean- water consumption from plastic sachets and bottles has significantly reduced sickness and death among West African children. Thus, measures to reduce the use of plastic sachets and bottles should be accompanied by programs designed to improve health outcomes for children, particularly in poorer households. As an alternative, subsidies could be provided for the use of biodegradable drinking-water containers that are more costly to produce than traditional plastic containers. 37 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 38 References Adam, I., Walker, T., Bezerra, J., and Clayton, A. 2020. Death, C. 2016. “Green states in Africa: beyond the usual Jeftic, L., Sheavly, S., Adler, E., and Meith, N. 2009. Marine Stoler, J., Weeks, J. R., and Fink, G. 2012. “Sachet drinking “Policies to reduce single-use plastic marine pollution in suspects.” Environ. Polit. 25 116–135. litter: a global challenge. water in Ghana’s Accra-Tema metropolitan area: past, West Africa.” Marine Policy 116. present, and future.” Journal of Water, Sanitation and Díaz-Mendoza, C., Mouthon-Bello, J., Pérez-Herrera, N. L., Kwakye-Nuako, G., Borketey, P. B., Mensah-Attipoe, I., Hygiene for Development, 2(4), 223-240. Barboza, L. G. A., Cózar, A., Gimenez, B. C., Barros, T. L., and Escobar-Díaz, S. M. 2020. “Plastics and microplastics, Asmah, R. H., and Ayeh-Kumi, P. F. 2007. “Sachet drinking Kershaw, P. J., and Guilhermino, L. 2019. “Macroplastics effects on marine coastal areas: A review.” Environmental water in Accra: the potential threats of transmission of Taylor, R. L., and Villas-Boas, S. B. 2016. “Bans vs. fees: pollution in the marine environment.” In World Seas: An Science and Pollution Research, 1-10. enteric pathogenic protozoan organisms.” Ghana Medical Disposable carryout bag policies and bag usage.” Applied Environmental Evaluation (pp. 305-328). Academic Press. Journal, 41(2). Economic Perspectives and Policy, 38(2), 351-372. Dikgang, J., Leiman, A., and Visser, M. 2012. “Elasticity Barrett, J., Z. Chase, J. Zhang, M. M. B. Holl, K. Willis, of Demand, Price and Time: Lessons from South Africa’s Lebreton, L., and Andrady, A. 2019. “Future scenarios of Themistocleous, K., Papoutsa, C., Michaelides, S., and A, Williams, ... & Wilcox, C. 2020. “Microplastic pollution Plastic-Bag Levy.” Appl. Econ. 44 (26): 3339-3342. global plastic waste generation and disposal.” Palgrave Hadjimitsis, D. 2020. “Investigating detection of floating in deep-sea sediments from the Great Australian Bight.” Communications, 5(1), 1-11. plastic litter from space using sentinel-2 imagery.” Remote Galgani, L., Beiras, R., Galgani, F., Panti, C., and Borja, A. Frontiers in Marine Science, 808. Sensing, 12(16), 2648. 2019. “Impacts of marine litter.” Frontiers in Marine Science, Lebreton, L. C., Van Der Zwet, J., Damsteeg, J. W., Slat, B., Beaumont, N. J., Aanesen, M., Austen, M. C., Börger, T., 6, 208. Andrady, A., and Reisser, J. 2017. “River plastic emissions Tukey, J. W. 1977. Exploratory Data Analysis (Vol. 2, pp. Clark, J. R., Cole, M., Hooper, T., Lindeque, P. K., Pascoe, to the world’s oceans.” Nature Communications, 8(1), 1-10. 131-160). Gallo, F., Fossi, C., Weber, R., Santillo, D., Sousa, J., C., and Wyles, K. J. 2019. “Global ecological, social and Ingram, I., Nadal, A., and Romano, D. 2018. “Marine Li, W. C., Tse, H. F., and Fok, L. 2016. “Plastic waste in the UNEP Regional Seas Programme, UNEP. Mediterranean economic impacts of marine plastic.” Marine Pollution litter plastics and microplastics and their toxic chemicals marine environment: A review of sources, occurrence and Action Plan, Secretariat of the Basel Convention on the Bulletin, 142, 189-195. components: the need for urgent preventive measures.” effects.” Science of the Total Environment, 566, 333-349. Control of Transboundary Movements of Hazardous Behuria, P. 2019. “The Comparative Political Economy of Environmental Sciences Europe, 30(1), 1-14. Wastes, Their Disposal, UNEP/GPA Coordination Office, Liu, R., Yu, C., Liu, C., Jiang, J., and Xu, J. 2018. “Impacts Plastic Bag Bans in East Africa: Why Implementation Has & Intergovernmental Oceanographic Commission. 2005. Geyer, R., Jambeck, J. R., and Law, K. L. 2017. “Production, of haze on housing prices: an empirical analysis based Varied in Rwanda, Kenya and Uganda.” GDI Working Paper Marine Litter: An Analytical Overview. use, and fate of all plastics ever made.” Science Advances, on data from Chengdu (China).” International Journal of 2019-037, The University of Manchester. 3(7), e1700782. Environmental Research and Public Health, 15(6), 1161. UNEP. 2014. “Valuing Plastics: The Business Case for Biermann, L., Clewley, D., Martinez-Vicente, V., and Measuring, Managing and Disclosing Plastic Use in the Homonoff, T. A. 2018. “Can small incentives have large Lloyd, C. T., Chamberlain, H., Kerr, D., Yetman, G., Topouzelis, K. 2020. “Finding plastic patches in coastal Consumer Goods Industry”. effects? The impact of taxes versus bonuses on disposable Pistolesi, L., Stevens, F. R., and Tatem, A. J. 2019. “Global waters using optical satellite data.” Scientific Reports, 10(1), bag use.” American Economic Journal: Economic Policy, spatio-temporally harmonised datasets for producing United Nations Environment Programme. 2021. “Neglected: 1-10. 10(4), 177-210. high-resolution gridded population distribution datasets”. Environmental Justice Impacts of Marine Litter and Plastic Buonocore, E., Donnarumma, L., Appolloni, L., Miccio, Big Earth Data, 3(2), 108-139. Pollution”. Nairobi Jambeck, J. R., Geyer, R., Wilcox, C., Siegler, T. R., A., Russo, G. F., and Franzese, P. P. 2020. “Marine Perryman, M., Andrady, A., ... and Law, K. L. 2015. “Plastic Manullang, C. Y. 2020. “Distribution of plastic debris Van Calcar, C. J., and Van Emmerik, T. H. M. 2019. natural capital and ecosystem services: An environmental waste inputs from land into the ocean.” Science, 347(6223), pollution and its implications on mangrove vegetation.” “Abundance of plastic debris across European and Asian accounting model.” Ecological Modelling, 424, 109029. 768-771. Marine Pollution Bulletin, 160, 111642. rivers.” Environmental Research Letters, 14(12), 124051. Burnett, S. 2013. “Do bans on plastic grocery bags Jambeck, J., Hardesty, B., Brooks, A., Friend, T., Teleki, Omolade, O. O., and Zanaib, G. O. 2017. “Parasitological Van Emmerik, T., and Schwarz, A. 2020. “Plastic debris in save cities money?” National Center for Policy Analysis. K., Fabres, J., and Baleta, T. “Challenges and emerging evaluation of sachet drinking water in areas of Lagos State, rivers.” Wiley Interdisciplinary Reviews: Water, 7(1), e1398. Policy Report. solutions to the land-based plastic waste issue in Africa.” Nigeria.” Electronic Journal of Biology, 13(2), 144-151. Watkins, E., and Ten Brink, P. 2017. Marine litter: Conservancy, O. 2015. “Stemming the tide: Land-based Mar. Pol. 96 (2018) 256–263. Ritchie, H., and Roser, M. 2018. “Plastic pollution.” Our socio-economic study. strategies for a plastic-free ocean.” Ocean Conservancy Jang, Y. C., Hong, S., Lee, J., Lee, M. J., and Shim, W. J. World in Data. and McKinsey Center for Business and Environment. Windsor, F. M., Durance, I., Horton, A. A., Thompson, R. C., 2014. “Estimation of lost tourism revenue in Geoje Island Salem, M. E., and Mercer, D. E. 2012. “The economic value Tyler, C. R., and S. J. Ormerod, S. J. 2019. “A catchment- Convery, F., McDonnell, S., and Ferreira, S. 2007. “The from the 2011 marine debris pollution event in South Korea.” of mangroves: a meta-analysis”. Sustainability 4, 359–383. scale perspective of plastic pollution.” Global Change most popular tax in Europe? Lessons from the Irish plastic Marine Pollution Bulletin, 81(1), 49-54. Sustainability Factor in Wonorejo Mangrove Ecotourism Biology, 25(4), 1207-1221. bags levy.” Environmental and Resource Economics, 38(1), Jang, Y. C., Lee, J., Hong, S., Choi, H. W., Shim, W. J., (Parmawati, et al.). 1-11. and Hong, S. Y. 2015. “Estimating the global inflow and Schmidt, C., Krauth, T., and Wagner, S. 2017. “Export of Costanza, R., De Groot, R., Sutton, P., Van der Ploeg, S., stock of plastic marine debris using material flow analysis: plastic debris by rivers into the sea.” Environmental Science Anderson, S. J., Kubiszewski, I., Farber, S., and Turner, a preliminary approach.” Journal of the Korean Society for & Technology, 51(21), 12246-12253. R. K. 2014. “Changes in the global value of ecosystem Marine Environment & Energy, 18(4), 263-273. services.” Global Environmental Change, 26, 152-158. 39 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 40 © Mel D. Cole for World Bank Table 4: West African import demand model results Appendixes All variables measured in logarithms Dependent variable: Import Volume [SITC3 57111] (1) (2) (3) (4) GDP 1.381 1.182 1.118 1.123 Appendix A. The macroeconometrics of The TPS data for each West African country has been drawn (11.29)** (10.66)** (10.95)** (10.64)** from the UN’s Comtrade database.51 TPS prices for each country single-use plastic imports are computed in two steps.52 Firstly, total current-dollar import Price [SITC3 57111] -1.006 -0.835 -0.707 -0.813 values are divided by import quantity. Secondly, corrections are Panel Estimation of Regional Import Demand made for interim dollar inflation, using the U.S. gross domestic Elasticities (9.10)** (4.39)** (7.45)** (4.55)** product (GDP) deflator. Our measure of national income, drawn Import duties are already familiar in all West African countries from the World Bank’s World Development Indicators, is GDP Interactions: and relatively easy to enforce, but how effective would they be measured in constant U.S. dollars (2010). in reducing demand for single-use plastics? Ghana x Price 0.125 0.116 Composite income and price elasticities for the West African The responsiveness of demand for thin polyethylene sheets countries are estimated by pooling the TPS and national (0.51) (0.51) (TPSes) to import duties is addressed with data from the income data for the 1990–2019 period. The TPS data is far natural experiment on TPS import prices for 10 West African from complete for Gambia, Sierra Leone, and Togo, so the Côte d’Ivoire x Price -0.083 -0.118 countries:48,49, Benin, Cameroon, Côte d’Ivoire, Gambia, econometric exercise simply employs data for Senegal, Ghana, Guinea, Nigeria, Senegal, Sierra Leone, and Togo. Guinea, Côte d’Ivoire, Ghana, Benin, Nigeria and Cameroon. (0.20) (0.31) As Figure 5 (page 21) shows, the import price has varied more Appropriate panel regression techniques are used to estimate the econometric model, and to test for significant differences in Cameroon x Price 0.608 0.597 than fourfold since 1995. All West African economies are price- takers, since they are too small to affect aggregate demand/ import-price elasticities across countries. (1.33) (1.40) supply relations in the global TPS market. Each country Table 4 presents the results, which indicate large and highly therefore faces an exogenously determined TPS price, and significant income and price elasticities. For the composite Senegal x Price 0.084 0.083 econometric estimation of the following import-demand model regression reported in column (1), the estimated income and is straightforward. In the regression specification, logarithms price elasticities are 1.4 and -1.0, respectively. By implication, (0.19) (0.21) of model variables are employed for two reasons. Firstly, each 1 percent increase in national income induces a 1.4 percent logarithmic models are less susceptible to distortions from a few Benin x Price 0.269 0.238 increase in imports of thin polyethylene sheets (TPSes), and “outlier” observations for the dependent variables. Secondly, and each 1 percent increase in price induces a 1 percent decrease (0.78) (0.74) equally importantly, the estimated parameters in a logarithmic in imports. model are easily interpreted because they measure the Guinea x Price -2.204 percentage change in the dependent variable that is attributable to a percentage change in an independent variable. (6.30)** lnmit= β0+β1 lnqit+ β2 lnpit+ εit50 1 ( ) Constant -16.867 -12.238 -10.636 -10.774 where, for country i in year t, mit equals TPS import quantity, qit equals national income, pit is the TPS import price, and εi stands (5.77)** (4.62)** (4.33)** (4.24)** for the effect of unobserved variables—which may be random Observations 158 158 141 141 or temporally correlated. R-squared 0.55 0.69 0.52 0.53 Note: Column (1) shows that the estimated composite income and price elasticities are 1.38 and -1.006, respectively, when all six countries are included in the estimation. Column (2) indicates no significant difference in price elasticities across the sample countries, with the exception of Guinea. Columns (3) and (4) repeat the exercise with Guinea excluded from the sample. In all cases, composite income and price elasticities are large in absolute value and highly significant statistically. Absolute value of t statistics are shown in parentheses. ** = a significance level of 1 percent. 48 The Comtrade database has no entries for TPS imports by Liberia. 49 The import quantity index in Figure 4.1 is calculated in two steps. Firstly, import quantities in each country are normalized to the range 0–100. Then, a mean From a policy perspective, the results have two major implications. Firstly, since the real TPS exhibits no trend over time, the waste normalized quantity for the 10 countries is calculated for each year. We prefer this index to the regional total, which would give most of the weight to the three countries that account for 80 percent of TPS imports (Nigeria, Côte d’Ivoire, and Ghana). Nigeria, alone, accounted for 46 percent of regional imports during the 2015–19 period. generated by SUP containers in West African countries is likely to grow at roughly the same rate as the country’s national income. 50 In this model, β1 is the income elasticity of import demand. The estimated parameter is expected to have a positive sign, because it measures the percentage Secondly, since price responsiveness is high, our results suggest that a TPS tariff could be a potent policy weapon in the struggle increase in TPS imports that will be induced by a one-percent increase in national income. Β2 is the price elasticity: It is expected to have a negative sign because it to reduce SUP pollution. measures the percentage decrease in TPS imports that will be associated with a one-percent increase in the price. 51 https://comtrade.un.org/data. 52 Separate import prices have been computed for each country, to allow for some difference in import composition not captured by Comtrade’s five-digit data. 41 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 42 Appendix B. The health impact of plastic child’s age in months (age at death for mortality). Unobserved To explore the implications of these results for both countries, the probit results in Table 5 55 are used to predict mortality rates and spatial and temporal factors were also controlled for, by including diarrhea incidences for all children [0–5 years of age] in the sample (12,500 for Ghana; 99,500 for Nigeria), with and without plastic container use dummy variables for DHS years and level-1 administrative water container use, which is reported in Figure 7 (page 23) and Table 5. Does drinking water from plastic sachets and bottles reduce the regions (12 in Ghana, 38 in Nigeria). Although these estimates may fully or partially reflect the influence of unobserved variables that are correlated with plastic container probability of mortality and morbidity (incidence of diarrhea), Table 5 presents probit-based econometric estimates that can use, the results are certainly cautionary given the size of the estimated impact and the stakes for public health. In both Ghana and among children living in Ghana and Nigeria? be interpreted as the change in dependent variable probability, Nigeria, reducing the use of plastic drinking-water containers may significantly increase childhood illness and death. Policy makers To test the impact of plastic drinking water containers on mortality for a one-unit change in the independent variable.54 For clarity, who opt for reducing SUP containers should also consider countervailing health measures, particularly for poorer households. and the incidence of diarrhea in children, an econometric the dummy variable results for time periods and administrative database was constructed from DHSes for Ghana (2003, 2008, regions were excluded. These are significant in all cases, 2014) and Nigeria (2003, 2008, 2013, 2018). The database suggesting that a host of temporal and local factors also have reports caretaker responses for more than 12,500 children in an important impact on child mortality and morbidity. Among Appendix C. Household income and Figure 18. Real income per capita in Ghana and Nigeria, 2000–19 Ghana and 99,500 children in Nigeria. Probit regressions were the reported regression variables, mother’s education and child’s age have consistently high significance. In the results, plastic container use 2500 used to analyze the data. The regression-dependent variables were dichotomous measures for mortality (child has died: 1 if income has a perversely positive, significant association with How significant is the effect of income on plastic container use yes, 0 if no) and diarrhea incidence (1 if had diarrhea recently, child mortality. Income has the expected sign for diarrhea and across households and over time? 2000 0 otherwise). The independent variables were child gender; is significant for Nigeria. Among the four results for plastic container use, all have the expected sign (container use reduces Real income per capita has approximately doubled in Ghana and years of mother’s education; real household income;53 plastic the dependent variable probability), and three of the results are Nigeria since 2000 (Figure 18). It is expected that this income drinking-container use (1 if plastic bottles or sachets are the statistically significant. growth has increased demand for plastic water containers, all 1500 household’s primary drinking water source, 0 otherwise); and else being equal. Plastic sachets account for the bulk of plastic Table 5: Plastic drinking-water container use and child health in Ghana and Nigeria water-container use in Ghana and Nigeria. They first appeared on the market in the late 1990s, when entrepreneurs in West 1000 Dependent variable: Probability of child death or recent diarrhea African cities began using new Chinese machinery that heat- Ghanaa Nigeriab sealed water in plastic sleeves (Stoler et al. 2012). Given the 2000 2005 2010 2015 2020 (1) (2) (3) (4) sachets’ novel status, one would expect a period of product Year Variable Death Diarrhea Death Diarrhea diffusion to produce some adjustment lag in the income/ Ghana Nigeria consumption relationship. Source: World Development Indicators, Income Per Capita (Constant $US 2010). Female 0.032 -0.001 0.006 0.008 To capture both income and diffusion effects for plastic containers at the household level, an econometric model for households (0.89) (0.02) (0.52) (0.72) surveyed by the DHS or MIS in the Accra and Lagos metro areas is estimated for the 2003–09 period. The dependent variable is a dichotomous measure (1 if plastic containers are the household’s primary drinking water source, 0 otherwise). The independent Mother’s education (Years) -0.022 -0.010 -0.033 -0.003 variables are household income and a time trend.56,57 A probit probability model is employed, which bounds model predictions within (4.49)** (2.60)** (21.11)** (2.17)* the range [0–1]. Figure 19. Per capita income distribution Income per capita 0.712 -0.129 0.397 -0.054 Accra (2019) Lagos (2018) ($US 2010) (8.24)** (1.49) (19.56)** (2.21)* .0015 .001 Plastic container use -0.250 -0.140 -0.117 -0.056 8.0e-04 (2.82)** (2.03)* (2.93)** (1.34) .001 6.0e-04 Density Density Age (months) 0.007 -0.008 0.009 -0.012 (6.63)** (10.00)** (28.50)** (34.66)** 4.0e-04 5.0e-04 Constant -1.651 -0.710 -1.384 -1.276 2.0e-04 (23.02)** (12.39)** (25.62)** (18.79)** 0 0 Observations 12,708 12,541 99,471 98,743 0 1000 2000 3000 4000 0 1000 2000 3000 4000 5000 Real Income Per Capita (US 2010) Real Income Per Capita (US 2010) Note: Absolute value of t statistics are shown in parentheses. * and ** equal significance levels of 5 percent and 1 percent, respectively. a. Data from Ghana DHS (2003, 2008, 2014). 55 The results for panel dummy variables in years and level-1 administrative units, that are excluded from Table 5, have now been incorporated. b. Data from Nigeria DHS (2003, 2008, 2013, 2018). 56 The DHS and MIS include a measure of relative economic status: a factor score derived from a principal components analysis of many dummy variables that record the presence, or absence, of household possessions. To estimate real household income per capita for each household, its factor score is transformed into a percentile [0–100]; divided by the total household members; the result is divided by its sample mean value; and that result multiplied by the World Bank estimate of 53 The measure of household income is derived from two variables. The first is the DHS measure of relative economic status, which is a score derived from an real income per capita in the relevant survey year. inventory of household possessions. This is standardized for each DHS by dividing by the mean score. The second is the World Bank measure of real income per 57 The focus is on income/container use dynamics here because the relationship is critical for the follow-on identification of plastic-waste “hot spots” in the two cities. capita in the DHS years. The World Bank measure of real income per capita for the DHS years is multiplied by the standardized DHS score, to produce an estimate The research also tested the role of a demographic variable that has a potentially significant role for health-related reasons: the household percentage of children of relative household income per capita. aged five or less. If drinking water in plastic containers is generally deemed safer, households with proportionately more vulnerable children might be expected to use 54 Probit estimation appropriately constrains model-based probability estimates to the range [0–1]. plastic containers at higher rates. However, the econometric tests of this proposition have not revealed a vulnerable-child effect that is consistently significant for in either Accra or Lagos. 43 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 44 Figure 19 shows that the distribution of household income is The dynamic implications of these results are explored by Tables 7 and 8 display yearly statistics for use intensity and time, most of the variation is temporal, not cross-sectional. highly skewed in both cities. In the regression specification, estimating the probability of plastic-container use, for the full population, with three highlights. Firstly, the statistics for To cite one example, the median use probability for plastic income is transformed to its logarithm to minimize outlier effects, range of incomes in each DHS/MIS year. Figure 6.3 displays household-use probabilities provide another illustration of the containers increases in Accra from 16 in 2003 to 95 in 2019, and and because it provides a better fit than untransformed income the results, which show the separate effects of income growth intertemporal pattern revealed by the econometric results in in Lagos from 20 to 63. At the same time, the ratio of maximum (which also fits very well in any case). Table 6 shows that the and product diffusion over the past two decades. figure 6.1. In Accra and Lagos, the table registers continuous to minimum value—a measure of cross-sectional variation— regression results are extremely robust. Household income upward shifts in distribution from 2013 to 2019. At the same declines steadily toward 1.0 in both cities. has a large and highly significant, statistically identical effect on plastic container use in Accra and Lagos. The time trend is Table 7: Household use probabilities for plastic drinking water containers: Distribution statistics for 1-km grid cells highly significant in both cities, but steeper in the case of Accra (0.158 versus 0.077). Survey Year Min P10 P25 Median P75 P90 Max Max/Min Table 6: Household income and plastic container use in Accra and Lagos, 2003–19 Accra DHS 2003 12 15 15 16 17 18 19 1.66 Dependent variable: Household uses plastic sachets or bottles as primary drinking water source (1 if yes, 0 if no) DHS 2008 35 41 42 44 45 46 48 1.36 Accra Lagos DHS 2014 73 78 78 80 81 82 83 1.13 DHS/MIS Year 0.158 0.077 MIS 2016 83 86 87 88 88 89 90 1.09 (28.09)** (17.68)** Log (Household Real Income Per Capita) 0.356 0.355 MIS 2019 92 94 95 95 95 96 96 1.04 (9.05)** (12.50)** Lagos Constant -320.780 -157.111 DHS 2003 12 18 19 20 21 22 27 2.23 (28.37)** (18.01)** DHS 2008 22 29 31 32 34 35 41 1.89 Observations 3,094 5,137 MIS 2010 26 35 37 38 40 41 47 1.78 DHS 2013 35 44 46 47 49 50 56 1.63 Note: Absolute value of t statistics are shown in parentheses. * and ** equal significance levels of 5 percent and 1 percent, respectively. MIS 2015 41 50 52 54 55 57 62 1.54 DHS 2018 50 60 61 63 64 65 71 1.42 Appendix D. Spatial clustering of income If spatial clustering by income were a transient phenomenon, Population provides a strongly contrasting case in Table 8, where most of the variation is cross-sectional – not temporal. one would expect to see a quasi-uniform spatial distribution The distributions for Accra and Lagos both shift upward over time as the urban population grows, but with less than doubling of median and plastic-waste generation of mean income percentiles that have been calculated from indicator values (56.0 to 94.8 in Accra; 68.4 to 108.6 in Lagos). At the same time, the maximum/minimum ratios vary from 8.1 to 8.9 How do income and population affect the spatial distribution of random surveys over nearly two decades. However, the results in Accra, and from 81.5 to 146.4 in Lagos. plastic-waste generation? presented in figure 6.2 are highly non-uniform. Mean percentiles across grid cells range from 32 to 94 in Accra, and from 20 to 82 Table 8: Population indicators for 1-km grid cells Income Clustering in Lagos – and spatially clustered high- and low-income areas are clearly visible. The conclusion is that both Accra and Lagos Survey Year Min P10 P25 Median P75 P90 Max Max/Min In order to assess the intertemporal stability of residential spatial have stable patterns of residential income-class separation. Accra clustering by income, 1 km grids are overlaid on the two cities DHS 2003 12.7 21.3 31.1 56.0 80.0 96.1 110.4 8.7 and mean household income percentiles computed for grid Spatial Clustering of Plastic-Waste Generation squares using 3,094 households in five surveys for Accra (DHS DHS 2008 15.5 22.9 35.4 63.3 96.8 111.4 125.4 8.1 2003, 2008, 2014; MIS 2016, 2019), and 5,137 households Appendix C documents the strong association between DHS 2014 17.7 27.8 41.4 75.0 120.6 137.6 158.2 8.9 in six surveys for Lagos (DHS 2003, 2008, 2013, 2018; MIS household income, and the probability of using plastic containers MIS 2016 19.8 30.5 45.9 80.5 129.4 149.6 170.0 8.6 2010, 2015). The methodology incorporates the random 2 km for drinking water. The above subsection shows that both MIS 2019 21.9 31.9 52.3 94.8 137.7 167.1 195.6 8.9 locational variation imposed on each urban survey cluster by Accra and Lagos have stable residential clustering by income DHS and MIS, to ensure respondent anonymity. Each survey stratum. It follows that the intensity of plastic-container use and Lagos cluster is treated as a high-resolution set of points, bounded plastic-waste generation would also be clustered spatially by DHS 2003 6.1 17.8 34.8 68.4 120.7 168.0 522.6 85.7 by a circle with a 2 km radius that is centered on the cluster income stratum. However, the econometric results illustrated in DHS 2008 7.4 21.2 40.6 79.3 141.8 201.1 603.4 81.5 coordinates recorded by the survey. For one cluster, the mean figure 6.1 also suggest that new-product diffusion of plastic water MIS 2010 7.8 22.4 42.5 85.0 151.9 217.8 650.0 83.3 household income percentile is assigned to all points within the sachets has reduced this spatial variation over time. To explore DHS 2013 6.7 23.7 45.9 93.8 167.9 230.7 715.0 106.7 circular bound. Then, all cluster circles are overlaid, and each the implications, the mean probability of plastic-container use in each grid cell and time period has been computed.58 This MIS 2015 7.1 24.9 50.4 100.7 177.2 246.5 770.0 108.5 point is assigned the mean value for all clusters represented at that point. A GIS raster has been created from these points, and captures the spatial distribution of use intensity, but aggregate DHS 2018 5.8 27.5 54.9 108.6 191.8 276.4 849.2 146.4 resampled to the scale consistent with our 1 km grid. use is the relevant measure for policy analysis. To proxy aggregate use, use intensity is scaled by population in each grid cell.59 Since the aggregate index for plastic-container use is the product of use probability and the population indicator, the preceding results have two clear implications. Firstly, for any given year, the spatial variation in population greatly outweighs spatial variation in 58 This methodology incorporates the random 2 km locational variation imposed on each urban survey cluster by DHS and MIS, to ensure respondent anonymity. Each plastic-use probability when determining aggregate indicator values. Secondly, also across years, temporal variation in plastic-use survey cluster is treated as a high-resolution set of points bounded by a circle, with 2 km radius centered on the cluster coordinates recorded by the survey. For one probability greatly outweighs spatial-population variation when determining indicator values. cluster, the mean probability of plastic container use is assigned to all points within the circular bound. Then, all cluster circles are overlaid and each point is assigned the mean value for all clusters represented at that point. A GIS raster has been created from these points, and then resampled to the scale consistent with our 1 km grid. 59 To produce population indicators for the 1 km grid cells, 100 m rasters were resampled from the Worldpop project at the University of Southampton (Lloyd et al. 2019). 45 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 46 © Mel D. Cole for World Bank Appendix E. Rivers as conduits for plastic waste Does proximity to urban rivers make a difference to marine-plastic pollution? To understand the risk posed by river transport of plastic waste to oceans, the simple geophysical model developed for this study assigns each 1 km grid cell a score for its likelihood of accumulated waste transport via river. The score incorporates net elevation (elevation of a point minus the elevation of the nearest river point), and the distance from the closest river point. Figure 20 presents maps of net elevations of all 1 km grid cells for Accra and Lagos. Figure 20: Maps of net elevations Accra Lagos Net Elevation (m) Net Elevation (m) 0-1 6-10 41-50 6-10 31-35 2 11-15 51-60 2 11-15 36-40 3 16-20 61-70 3 16-20 41-45 4 21-30 71-80 4 21-25 46-50 5 31-40 81-90 5 26-30 51-55 47 The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries The Economics of Plastic Use and Cleanup Priorities for West African Coastal Countries 48 www.wacaprogram.org | waca@worldbank.org