The Global Waste Trafficking and Its Correlates

Abstract

Our understanding of illicit waste trafficking (IWT) is in its embryonic stages; most notably, the transnational nature of this phenomenon has hitherto been neglected in extant empirical research. This study provides the first analysis of the possible coorrelates of transnational IWT at a global level. Through recourse to information extracted from the official Basel Convention National Reports, we constructed a network of the most relevant IWT connections between 148 countries. Next, we quantitatively investigated the role of specific potential factors that influence the structure of this transnational network. Our results indicate that illicit waste is trafficked toward poorer and more insecure countries, primarily via former colonial connections. As such, IWT poses a direct threat to the sustainable development of these countries. Mere adherence to international treaties and promulgation of environmental laws does not in and of themselves explain whether a country is part of the global IWT network, although the establishment of dedicated courts and tribunals does reduce the risk of being a recipient of trafficked waste. Solid anticorruption measures and a strong rule of law increased the likelihood of being a source country in the IWT network, which, in turn, calls for a more global approach to the management of environmental issues.

Keywords

transnational crime trafficking network environmental crime logistic regressions IWT waste crimes

Introduction

Illegal waste activities can be broadly defined as “any movement of waste which is not in accordance with environmental regulations” (Tompson & Chainey, 2011, p. 180). These activities include illegal trafficking, which can occur both domestically and internationally (Calderoni et al., 2014; Liddick, 2010; Tompson & Chainey, 2011). At a domestic level, waste can be moved from one region to another (Massari & Monzini, 2004). At an international level, waste is usually trafficked from developed to developing countries where regulation is lax and waste can be sold as fraudulent secondary-raw material (Bisschop, 2012; Rucevska et al., 2015). Although difficult to map, the most likely trade flows are from the global North (the European Union, Japan, the United States, and Australia) to the global South (Africa, Asia, and South America) (Elliott & Schaedla, 2016; Klenovšek & Meško, 2011).

Within this spectrum of crimes, cross-border illicit waste trafficking (IWT) can be classified as a form of transnational environmental crime, which involves the illegal trade, shipping, and processing of waste by a broad range of actors (Elliott, 2012; R. White, 2011).¹ For example, it can involve entrepreneurs in the legitimate economy who decide to turn to the black market, or, alternatively, organized crime groups operating in the waste sector who are seeking to reduce costs (e.g., avoiding the treatment costs) and generate further profit (e.g., illegally reselling the waste) by avoiding the duties associated with the correct and environmentally sound treatment of waste. Offenders smuggle, either directly or via the help of intermediaries, their waste through false declarations and documentation, concealment, and other illicit methods that deleteriously affect the environment, as well as jeopardizing human health, and weakening countries’ legitimate economies (Andreatta & Favarin, 2020; Morganti et al., 2020; Sahramäki et al., 2017). Both academics and international organizations have recognized the threat posed by IWT, along with highlighting the importance of studying its transnational dimension (Elliott & Schaedla, 2016; United Nation Office on Drugs and Crime, 2009). However, empirical research on waste crimes has thus far predominantly considered the national dimension of IWT. Indeed, despite the growing scientific interest in IWT as a transnational form of crime, the topic remains largely unexplored, especially from an empirical perspective.

One of the principal obstacles to conducting empirical research on IWT is the dearth—and poor quality—of publicly available data on waste crimes, illegal waste shipments, the prices that are charged on the illegal market as well as on the licit waste market (e.g., production, treatment, import, and export). Many authors and institutions have noted this problem and stressed the need to improve both data collection and data availability within the field of environmental crime to aid future research (Bisschop, 2012; Gibbs & Simpson, 2009; Greife & Maume, 2020; Stassen & Ceccato, 2020). However, more recently, Nobles (2019) has argued that this aforesaid point about the lack of suitable data sets has been overstated, and that in fact there are a variety of secondary data readily accessible through which to investigate many aspects of different justice systems’ responses to environmental crimes. While this is indeed true with respect to the United States, it is not necessarily the case in other countries across the world (Andreatta & Favarin, 2020; Stassen & Ceccato, 2020). Both Lynch et al. (2017) and Lynch and Pires (2019) have recently lamented the scarce attention that has been directed toward quantitative analyses of relevant environmental crimes. The nonquantitative tendencies within green criminology, in particular, have limited the generalizability of many of these studies, and distanced green criminology from “orthodox criminology,” which, in turn, has weakened the debates on environmental crimes within the criminological community.²

Since 2016, the Basel Convention National Reports (BCNRs) include detailed information on single cases of IWT by each of the countries that are parties to the Convention. These data are a unique source of information through which to study IWT and its characteristics at a transnational level. The present study systematized the data pertaining to the 2016 and 2017 reported cases of IWT to reconstruct a network of connections among the countries that illegally exchange waste. Subsequent to this, the study proceeded to analytically investigate the correlates of this global IWT network.

Background Studies

There is an emergent scientific interest in waste crimes; this interest is strictly connected to a nascent awareness of environmental crimes and their manifold deleterious consequences at a global level (Klenovšek & Meško, 2011). Despite this, the extant empirical research on waste crimes has predominantly focused on the domestic dimension of these aforesaid crime types. Both quantitative and qualitative studies that analyze the illegal management, disposal, and dumping of waste have been mainly conducted at a national or subnational level (Almer & Goeschl, 2015; Biotto et al., 2009; Crofts et al., 2010; D’Amato et al., 2018; Dorn et al., 2007; Ichinose & Yamamoto, 2011; Kim et al., 2008; Liu et al., 2017; Matos et al., 2012; Sahramäki & Kankaanranta, 2017; Sigman, 1998). Fewer studies have investigated the dynamics and determinants of within-border IWT (Germani et al., 2015, 2017; Massari & Monzini, 2004).

However, there is a relative dearth of studies that have empirically investigated the transnational dimension of IWT. Bisschop (2012) examined the transnational trafficking of electronic waste from the port of Antwerp, in Belgium, to third countries, primarily in Africa and Asia. The research disclosed information pertaining to the various actors involved and their roles as well as the push, pull, and facilitating factors that motivated people to illegally transport e-waste from Belgium to distant countries. The study was based on a document analysis of various primary and secondary sources, as well as interviews with key informants and field research (e.g., in the port of Antwerp [Belgium], port of Tema [Ghana], Agbogbloshie dumpsite [Ghana]). The results of the study showed how the actors (e.g., waste collectors, waste transporters) involved in the e-waste trafficking equally walked on a thin line between legal and illegal and how push, pull, and facilitating factors on individual, organizational, and societal levels together provided the motivations and opportunities for illegal transports of e-waste.

Sahramäki and colleagues (2017) analyzed actors, their modi operandi, and facilitating factors in 13 IWT cross-border cases from the Netherlands, Italy, and Finland to third countries around the world. This comparative study highlighted the similarities and differences between these countries through conducting a crime script analysis of judicial files. The findings reveled that (a) offenders involved in cross-border IWT usually had great knowledge of the waste market and of its regulations and perceived the risks and costs of being caught as fairly low, (b) the complexity of the legislation combined with a variety of authorities addressing the illicit activities at different levels created a complicated regulatory and judicial system frameworks that weakened prevention capabilities, and (c) the lack and misuse of appropriate licenses and falsified documentation was a prevailing method to conduct cross-border IWT.

Spapens and colleagues (2018) described the modi operandi of traffickers who illegally ship waste from the Netherlands to China. They qualitatively analyzed the regulatory framework for waste shipments from the European Union to third countries, particularly China, the different actors involved in the trafficking, and the difficulties experienced by law enforcement agencies in tackling transfrontier IWT. Finally, building upon the work of Sahramäki and colleagues (2017), Andreatta and Favarin (2020) deepened extant knowledge about the different characteristics of the various stages of cross-border IWT from Italy to third countries. Through conducting a crime script analysis of five judicial cases, the authors shed light on the characteristics of the different stages of transnational IWT (i.e., creation, collection, storage, transport, treatment, and disposal). Moreover, they applied a situational crime prevention approach to propose preventive measures for tackling this crime. In their findings, the authors highlighted the presence of opportunities for crime in each of the IWT stages and they confirmed the importance of forgery and falsification of documents and licenses as recurrent features across all the stages of IWT. In contradistinction to some other literature on waste crime, they did not find evidence of corruption acting as a facilitator of IWT and the involvement of mafia groups in the trade. The offenders were companies or entrepreneurs working in the legal sector. Brokers and intermediaries also intervened at times to give advice on illicit practices. A core group of offenders usually managed the entire illicit process, but they were aided and abetted by a network of more peripheral offenders that conduct specific illegal activities. The offenders were Italian, African, and Chinese. African offenders usually became involved in the trade when waste was illegally transferred to Africa, while Chinese offenders when waste was sent to China.

All in all, empirical research on transnational IWT has hitherto primarily focused on delineating the actors, modi operandi, and problems associated with, as well as proposals for, tackling the phenomenon. Given the aforementioned paucity of data on waste crimes and illicit waste shipments, the analyses conducted in prior studies either only considered the national dimension of IWT (Germani et al., 2015; Massari & Monzini, 2004) or were narrow in their scope (Bisschop, 2012; Spapens et al., 2018) and relied mainly on case studies (Andreatta & Favarin, 2020; Sahramäki et al., 2017). Despite the undoubted value of this research, the results of these analyses failed to grasp the transnational dimension of IWT at a global level. This study aims to overcome this limitation by conducting a quantitative analysis of the main correlates of the transnational IWT at a global level. By exploiting information on reported cases of illegal trafficking included in the BCNRs, this research recreates a network of connections among countries that illegally exchange waste, so as to be able to empirically identify the sociodemographic, economic, geographic, and regulatory factors that shape this network. The results corroborate the findings of the extant research (e.g., the role of developed countries as exporters and of developing countries as importers, the general ineffectiveness of the regulation in tackling IWT) and present new insights into many different aspects of the illegal waste trade (e.g., the centrality of colonial ties in the trade, the importance of geographical proximity, the role of rule of law and control of corruption, and the relevance of establishing dedicated environmental courts and tribunals). Only by examining IWT as a transnational crime will it be possible to disentangle the main drivers of this crime and shed light on the dynamics between countries that have hitherto largely been unexplored. Moreover, the study maps for the first time transnational IWT flows and the proposed empirical approach generates new quantitative knowledge on cross-border IWT in an attempt to incite debate within the scientific community and to aid green criminology to get the attention of “orthodox criminology.”

Problem Formulation

Transnational IWT, as with any other form of transnational trafficking, constitutes a network that connects countries of origin and countries of destination. The network exists for the sole purpose of exchanging illegal waste: countries of origin send illicit waste to countries of destination in exchange for payment—selling it—or to simply get rid of their waste. Other transnational crimes, such as drug trafficking, have well-known countries of origin (e.g., Colombia for cocaine or Afghanistan for heroin) and destination (e.g., the United States or Europe). In the case of IWT, waste can be trafficked from more developed to less developed countries, but also along legal trade channels that take advantage of geographical proximity (e.g., from France to Germany). The present study aims to empirically identify the sociodemographic, economic, geographical, and regulatory factors that influence the connections between those countries in the transnational IWT network. Building upon the work of recent studies that have utilized advances in statistical modeling of social networks to identify the factors that influence the structure of the illicit trafficking networks for goods, such as drugs (e.g., Berlusconi et al., 2017; Boivin, 2014; Giommoni et al., 2017), or radiological and nuclear material (e.g., Sin & Boyd, 2016), the main research question underpinning this study is: What are the main correlates of the transnational IWT network at a global level?

The proposed analysis assesses the possible influence of several factors on the presence of a trafficking connection between any two countries in the network according to several hypotheses. These hypotheses were developed in relation to extant literature on IWT as well as literature on other illicit transnational crimes.

Hypothesis 1: Countries that are politically and socioeconomically insecure are more likely to receive trafficked waste from other countries.

The mechanism through which underdeveloped countries in the South are being turned into a reservoir of garage, toxic waste, and hazardous products is known as environmental injustice (Adeola, 2000; Clapp, 1994). Increased cognizance of environmental issues has led most governments in the global North to introduce more stringent regulations for waste management. Resultantly, the increased costs of safe and legal waste disposal have contributed to the development of an illegal export trade to many of the world’s least developed countries. It is widely believed that illegal waste imports cross-national borders relatively easily, particularly in countries with either weak or nonexistent inspection systems and technology (Klenovšek & Meško, 2011; Liddick, 2010; Pereira, 2015). According to the literature, the main destination countries for IWT are Ivory Coast, Ghana, Guinea, Nigeria, Sierra Leone, Tanzania, Togo, Benin, and Senegal in Africa as well as China, Hong Kong, Indonesia, India, Malaysia, Pakistan, and Vietnam in Asia (Nellemann et al., 2016; Rucevska et al., 2015). Notwithstanding the lax regulations and inadequate enforcement systems, these countries are also the victims of a more general form of exploitation by developed countries. Due to high levels of poverty and their weak status in the global economy, these countries are forced to accept hazardous waste in exchange for their extractive and agricultural products (Adeola, 2000). Underdeveloped societies that are relatively powerless due to their subordinate position in the capitalist world economy are more prone to receive illicit waste.

Hypothesis 2: Countries that produce more waste, that have a larger population, and are less able to recycle are more likely to unlawfully export their waste to third countries.

Simply put, more people produce more waste. Currently, the disposal facilities in the global North are unable to dispose of the total amount of waste produced (Liu et al., 2017). Moreover, population density directly affects the extent of anthropic pressures on the waste system (D’Amato et al., 2018). For this reason, we expect that exporter countries that produce a higher amount of waste and have a larger population—and population density—are more likely to unlawfully traffic their waste. On the contrary, countries that recycle large quantities of waste should be negatively correlated with being source countries for IWT.

Hypothesis 3: Countries that share a colonial relationship are more likely to exchange illicit waste.

Studies in macroeconomics have identified the importance of social proximity between countries for reducing the barriers involved in the legal trade of goods. Migration, common language, and past colonial relationships have all been noted as social factors that enhance legal trade (Prashantham et al., 2015; Rauch, 1999; Sgrignoli et al., 2015). Giommoni and colleagues (2017) found that social proximity is also of critical importance in the illegal trade of heroin. The results of their study showed that both migration flows and common language affect the way that heroin moves between countries. Hence, preexisting social and economic ties, such as a prior colonial relationship, can also be a determinant factor in shaping the transnational IWT network. On one hand, language and cultural affinity may facilitate cross-border IWT, in conjunction with established and robust trade channels between countries. On the other hand, colonizers tend to maintain hegemonic control over their former colonies even after their independence, which, in turn, increases the probability that they also send them their illegal waste.

Hypothesis 4: Close countries are more likely to engage in IWT.

Geographical proximity between countries reduces the costs associated with legal trade. Indeed, studies have highlighted the negative effect of geographical distance on the legal trade of goods (Disdier & Head, 2008). With all else held equal, heroin and marijuana trafficking have also been shown to experience the same negative effect, insofar as traffickers operating in the illegal economy face a far higher exposure to risk of interception and arrest over long distances (Caulkins & Bond, 2012; Giommoni et al., 2017). In the case of transnational IWT, we thus expect a negative correlation between geographical distance and the trade of illicit waste. Indeed, our hypothesis is that although countries in the global North do export their waste to countries in the global South that are ordinarily far away from the countries of origin, physical distances nevertheless play a role in terms of defining the actual exporter–importer pairs. Therefore, we expect that close countries are more likely to engage in IWT with all else held equal.

Hypothesis 5: The more relevant a country’s role in global trade, the more relevant its role in IWT.

The illegal trade of waste is much more closely connected to the operations of the legal supply chain than other transnational crimes such as drug trafficking (Lieselot, 2016). In light of this, countries that exchange large amounts of commodities are also more likely to trade/traffic in waste. Moreover, businesses operating in the legal waste sector who have excellent knowledge of the waste market and its complex licensing system, also ordinarily manage the supply of illicit services (Andreatta & Favarin, 2020; Sahramäki et al., 2017). For this reason, preexisting legal and economic connections between countries play an important role in shaping illegal trade. We assume that countries that are more active in the global trade of legal products are thus more likely to also exchange illicit waste.

Hypothesis 6: International treaties, specific national environmental regulations, and the establishment of ad hoc courts and tribunals do not shape the IWT network.

In recent decades, considerable effort has been dedicated to developing international and national legislation that seeks to both strengthen the waste regulatory framework and bolster enforcement capabilities. Nevertheless, extant research has reported on the general inability of the enforcement and regulatory frameworks to effectively fight and tackle IWT (Baird et al., 2014; Klenovšek & Meško, 2011; R. White, 2011). This results in a general sense of impunity (Pereira, 2015), insofar as both the perception of the risk of being caught and the probability of a criminal prosecution are extremely low for environmental crimes (Lynch et al.,2016, 2019; Morganti et al., 2020; Sahramäki et al., 2017). In addition, the financial losses stemming from potential punishment are often negligible for waste traffickers (Bisschop, 2012). For all these reasons, we expect that international treaties, specific national environmental regulations as well as the establishment of ad hoc courts and tribunals are insufficient measures for effectively tackling IWT at a global level. Within our empirical analytical framework, we expect that none of these measures will influence the likelihood of a connection between two countries in the trafficking network, when controlling for the other aforementioned factors.

Data and Method

We downloaded and systematized the data from the BCNRs in 2016 and 2017 on reported cases of IWT, to create a network of connections among countries that illegally exchange waste.³ Every time that a country—importer or exporter—was indicated as being involved in an illegal trafficking case with another country a link was established between the two. The binary variable—active dyad versus inactive dyad—that identifies trading partnerships and describes the structure of the trafficking network is our dependent variable.⁴ Next, independent variables were collected from a variety of sources to test the hypotheses delineated in the previous section. Tables 1–3 present the dependent variable and the independent variables with their related sources. Our analysis aims to understand what factors (independent variables) affect upon the presence of connection (dependent variable) in our IWT network.

The empirical test relies on a statistical model inspired by the gravity models commonly used in international trade studies (Greaney & Kiyota, 2020; Kabir et al., 2017; Yotov et al., 2016). Besides international trade, the gravity model has proved to be a useful framework in which to explain various illegal cross-border relations, as human trafficking (Akee et al., 2014), trade-based money laundering (Ferwerda et al., 2013; Walker & Unger, 2009), suspicious bank transfers (Cassetta et al., 2014), and firearm trafficking (Kahane, 2013). We also rely on this class of models as a workable empirical strategy to control for characteristics of both importing and exporting countries as well as for relational factors.

Dependent Variable and the Operationalization of Transnational IWT

For each case of IWT registered in a country, the BCNRs present information on the country of export, the country of import, the waste code, the type of waste, the amount in metric tons, the identification of the reason for the illegality, attribution of responsibility for the illegality, and the measures taken, such as any punishment that is imposed.⁵ Yet, plausible differences between countries’ interception capacities and quality of statistical reporting call for prudence in the use of these data for econometric analyses. In consideration of this, following the examples set by other scholars who faced similar challenges (Boivin, 2013; Giommoni et al., 2017), we exclusively focus on the presence or absence of a connection. In other words, information is used solely for the purposes of identifying the pairs of countries exporting and importing illicit waste with each other and, in turn, to establish the position of each country in the global IWT network. Rather than being a detailed and comprehensive schematization of the actual global trafficking network, then, the emerging graph is solely intended as a representation of the—likely—most trafficked routes. The underlying assumption is that the connections emerging from the BCNRs are, on average, more relevant than those that do not emerge, albeit others are very likely to exist.

In the 2-year period between 2016 and 2017, 114 different countries around the world produced their annual BCNRs (104 reports concerning 2016 and 91 concerning 2017, as by June 2019). Out of 114 countries, only 25 reported transnational cases of IWT; these were mostly European and—to a lesser extent—South-Eastern Asian countries. Yet, the information they reported allows for the identification of trafficking connections—imports and or exports of waste—with other 88 countries, 34 of which were nonreporting countries (Figure 1). Therefore, the available BCNRs allow for the observation of the potential connections between 148 countries. Countries that did not provide their BCNRs and that were not identified as either the origin or destination for illicit waste shipments by reporting countries were not considered in the analysis.

Figure 1.

Information reported in the BCNRs, years 2016 and 2017.

As countries were considered to potentially be both importers and exporters, the total number of possible pairs in our models is 21,756 (Table 1). Of these possible dyads, the data indicate direct connections between 131 pairs of countries; within the network comprising 148 countries, 109 belong to a single larger clique, 2 belong to a second smaller clique, and 37 do not have any connections (Figure 2).

Table 1.

Descriptive Statistics for the Dependent Variable, Years 2016 and 2017.

Dependent variable	Number of observations^a	M	SD	Minimum	Maximum
Dyads in the IWT network	21,756	0.02	0.13	0.00	1.00

Note. IWT = illicit waste trafficking.

The number of observations refer to importer–exporter dyads; the same 148 countries are both importers and exporters of IWT. The maximum number of pairs is, therefore, 148 × 147 = 21,756.

Figure 2.

Main connections in the transnational IWT network, years 2016 and 2017.

France, the United Kingdom, Germany, Sweden, and Belgium are the main illegal waste exporters in our network and China, Poland, Nigeria, and Ghana are among the main illegal waste importers. These results are in line with the previous literature that highlighted Western Europe as region of origin and China, Nigeria, Ghana, and Eastern Europe as main destinations (Nellemann et al., 2016, pp. 62–63). Within the European Union, France, Germany, Belgium, and the Netherlands appear to be also importers at times receiving illegal waste from other European Union countries such as Italy, Austria, Luxemburg, and the United Kingdom. Other destination countries present in our network and confirmed also by previous literature are Benin, Cameroon, Guinea, Senegal, Congo, Côte d’Ivoire, and Gambia in West Africa and India, Pakistan, Malaysia, and Thailand in South and Southeast Asia (Nellemann et al., 2016). The United States and Japan are usually listed as origin countries in the literature, which is also the case in our network, but these countries are not central as others in terms of number of connections. Table BI in Appendix B provides information on the network statistics.

Independent Variables

Based on the previous studies that have investigated transnational and international crimes at the macrolevel, we selected variables capable of representing the insecure political and socioeconomic status of importing countries (operationalization of factors related to Hypothesis 1) (e.g., Aziani, 2020; Goel & Saunoris, 2019), the presence of a prior colonial relationship between countries (Hypothesis 3) (e.g., Berlusconi et al., 2017; Giommoni et al., 2017), geographical distance between countries (Hypothesis 4) (e.g., Aziani et al., 2019; Boivin, 2014), and a synthetic index—that is, KOF Trade Globalization (de facto) Index—of the relevance and heterogeneity of the international trade in goods and services (Hypothesis 5) (e.g., Landman & Silverman, 2020).

Hypotheses 2 and 6 concern the production and recycling of waste and the regulation and judiciary system that provides oversight over the waste industry, respectively; their testing relies on the use of information emerging from the web portal of the Basel Convention (UNEP & Basel Convention, 2019), from the What a Waste project by the World Bank (2019b), on a report carried out by the Environmental Law Institute (2019) on behalf of the UN Environment Programme, and on more classical structural variables—that is, population density, population size, control of corruption, and rule of law—whose data come from the World Bank (2019a). In particular, we collected data on solid waste generated and the percentage of recycling after waste treatment at a national level, to take in consideration both the amount of waste generated in each country and their respective ability to correctly recycle their waste. Furthermore, we created a dummy variable where 1 represents those countries in which the Basel Convention has come into force, and a dummy variable where 1 represents those countries with national environmental framework laws, as per the Environmental Law Institute’s (2019) report. From the same report, we collected information on countries with specialized national environmental courts (a), specialized national environmental tribunals (b), and specialized national environmental courts and tribunals (c). We also produced a categorical variable synthetizing the level of information reported for each pair of countries—that is, high, if both countries reported cases of IWT; medium, if one of the two reported cases, but the other did not; low, if neither of the two countries reported any case during the years 2016 and 2017. The use of this control contributes to mitigate potential biases emerging from countries’ disparities in information reporting. Finally, we exploited a categorical variable representing the macroregions of the world to mitigate the potential biases caused by the omission of key variables. Table 2 presents the operationalization of the hypothesized correlates of transnational IWT, as well as the relationship between them and the formulated hypotheses; Table 3 provides the descriptive statistics for the independent variables and their related sources.

Table 2.

Hypothesized Structural Correlates of Transnational IWT.

Type of variable	Hypotheses	Variable	Description of the variable	Direction of the correlation
Characteristics of the importer country	Hypothesis 1. Countries that are politically and socioeconomically insecure are more likely to receive trafficked waste from other countries	Infant mortality	Mortality rate, infant (per 1,000 live births)	Positive
		Conflicts index	Measure of general security and vulnerability to political violence	Positive
		GDP growth	Annual growth rate of GDP (%)	Negative
		Shadow economy	Market-based productions evading detection in the official estimates of the GDP (GDP share)	Positive
Characteristics of both importer and exporter countries	Hypothesis 2. Countries that produce more waste, that have a larger population, and are less able to recycle are more likely to unlawfully export their waste to third countries	Population density	People per square meter	Positive
		Population size	Millions of inhabitants	Positive
		Generated waste	Quantity of generated waste	Positive
		Recycling	Percentage of recycling	Negative
Relational characteristics—Features of the dyads	Hypothesis 3. Countries that share a colonial relationship are more likely to exchange illicit waste	Colonial relationship between two countries	Colonial relationship post-1945	Positive
Relational characteristics—Features of the dyads	Hypothesis 4. Close countries are more likely to engage in IWT	Distance in km	Bilateral distance (km) between the biggest cities in the two countries	Positive
Characteristics of both importer and exporter countries	Hypothesis 5. The more central a country’s role in global trade, the more likely they are to exchange illicit waste	Trade globalization	Index measuring the exchange of goods and services over long distances	Positive
	Hypothesis 6. International treaties, specific national environmental regulations, and the establishment of ad hoc courts and tribunals do not shape the IWT network	Basel Convention	Ratification of the Basel Convention	Nonsignificant
		Environmental law	Presence of an environmental law at a national level	Nonsignificant
		Courts and tribunals	Establishment of ad hoc courts and tribunals for environmental matters at a national level (both dichotomous variable—presence/absence—and categorical variable—0 to 3 depending on the absence of ad hoc courts/tribunals (0), presence of a court (1), a tribunal (2), or both (3)	Nonsignificant
	Controls	Control of corruption	Perceptions of the extent to which public power is exercised for private gain	Control
		Rule of law	Extent to which agents have confidence in and abide by the rules of society	Control
		Information reporting	Categorical variable indicating whether the source country, the destination country, or both provide information on IWT cases in the years 2016 and 2017	Control
		Macroregion	Categorical variable identifying the seven macroregions of the world	Control

Note. GDP = gross domestic product; IWT = illicit waste trafficking.

Table 3.

Summary and Operationalization of Structural Hypothesized Correlates, Last Available Years.

Variable	Number of observations^a	Availability (%)	M	SD	Minimum	Maximum	Source
Socioeconomic factors (Hypothesis 1)
Infant mortality, importer	21,315	98.0	21.00	20.83	1.67	89.47	UN Inter-Agency Group
Conflicts index, importer	21,315	98.0	1.58	1.51	1.00	8.00	Marshall (2019)
GDP growth, importer	21,021	96.6	25.78	2.61	10.51	35.04	World Bank (2019a)
Shadow economy (original)^b	19,110	87.8	26.88	11.81	6.94	53.07	Medina and Schneider (2017)
Shadow economy	21,756	100.0	26.92	11.42	6.94	53.07	Medina and Schneider (2017)
Waste production and management and demography (Hypothesis 2)
Population density	21,609	99.3	0.21	0.69	0.00	7.88	World Bank (2019a)
Population size^c	21,609	99.3	46.93	161.03	0.04	1,378.76
Population size (ln)^d	21,609	99.3	2.52	1.48	0.04	7.23
Generated waste, per capita^c	21,756	100.0	0.36	0.24	0.07	1.59	World Bank (2019b)
Generated waste, per capita (ln)^d	21,756	100.0	0.29	0.16	0.07	0.95
Recycling (original)^b	14,112	64.9	16.87	14.48	0.00	64.60
Recycling	21,756	100.0	14.67	12.90	0.00	64.60
Colonial ties (Hypothesis 3)
Colonial relation^e	21,279	97.8	0.01	0.09	0.00	1.00	Mayer and Zignago (2011)
Geography (Hypothesis 4)
Distance^c,e	21,279	97.8	7,192.20	4,249.64	114.64	19,650.13	Mayer and Zignago (2011)
Distance (ln)^e	21,279	97.8	8.64	0.78	4.74	9.89	Mayer and Zignago (2011)
Trade ties (Hypothesis 5)
Trade globalization^c	21,168	97.3	56.45	19.73	12.39	99.16	Gygli et al. (2019)
Trade globalization (ln)	21,168	97.3	3.96	0.41	2.51	4.60	Gygli et al. (2019)
Regulation and judiciary system (Hypothesis 6)
Basel Convention	21,756	100.0	0.98	0.14	0.00	1.00	UNEP & Basel Convention (2019)
National environmental law	21,756	100.0	0.91	0.28	0.00	1.00	Environmental Law Institute (2019)
Courts and tribunals dummy	21,756	100.0	0.25	0.43	0.00	1.00
Courts and tribunals categories	21,756	100.0	0.39	0.77	0.00	3.00
Enforcement controls
Corruption control	21,756	100.0	2.89	1.04	1.19	5.14	World Bank (2019a)
Rule of law	21,756	100.0	3.35	1.03	1.00	5.35	World Bank (2019a)
Additional controls
Information reporting	21,756	100.0	2.25	0.50	1.00	3.00	BCNRs 2016–2017
Macroregions	21,756	100.0	10.7	2.13	1.00	7.00	World Bank (2019a)

Note. Unless specified otherwise, the summary statistics refer to both importer and exporter countries as the import–export matrix is squared. Data referring to both importer and exporter countries are averages of original data referring to 2016 and to 2017. Data referring to regulation and judiciary system depict the situation as in 2017. Relational data (distance, colonial relation) do not to change in time. GDP = gross domestic product; BCNRs = Basel Convention National Reports.

The number of observations refers to importer–exporter dyads; the same 148 countries are both importers and exporters. The maximum number of pairs is, therefore, 148 × 147 = 21,756 as possible loops were not examined. ^bThe statistics are reported here to provide a comparison, but the variable was only included in the econometric models in the version with imputed data. Missing values were imputed using estimates on macroregional averages. ^cThe statistics are reported here to provide a comparison, but the variable was only included in the econometric models in the log-transformed form. ^dIndependent variables were translated by 1 unit prior to performing the log transformation in case they had original values lower than 1. ^eThe variable is relational, insofar as it refers to the importer–exporter dyad.

Statistical Analysis

In a context of international trade, the traditional gravity model usually has the following form:

\ln (Y_{i j}) = β_{0} + β_{1} {\vec{X}}_{i} + β_{2} {\vec{Z}}_{j} - β_{3} l n D_{i j} + β_{4} {\vec{W}}_{i j} + ε_{i j}

(1)

where Y_ij is a measure of trade between countries $i$ and $j$ , ${\vec{X}}_{i}$ and ${\vec{Z}}_{j}$ are two vectors of characteristics of the considered countries (in particular their economic masses), $D_{i j}$ is the distance between countries, ${\vec{W}}_{i j}$ are possible special preference relationships, $ε_{i j}$ is the residual error, and $β_{n}$ are the parameter estimates (Ward & Hoff, 2007). Specificities related to the structure of available data—for example, binary dependent variable—and to the topic under investigation—for example, illegality of the traffic—suggest to adjust the statistical features and the considered variables of this general gravity model.

Both logistic regression and discriminant function analysis allow researchers to determine the categorical probability of an event occurring. However, discriminant function analysis assumes that the data describing the independent variables are a normally distributed sample. The use of both continuous and categorical independent variables negates this assumption; it follows that logistic regression is the recommended functional statistical tool for when both continuous (e.g., population density) and categorical variables (e.g., colonial relationship) are used (Sharma, 1995), as it is in our case. Moreover, logistic regression does not require independent variables to be linearly related, nor does it require homoscedasticity, which, ultimately, makes it a less stringent procedure for statistical analysis in comparison to discriminant function analysis and ordinary least squares regressions (Press & Wilson, 1978; Tabachnick & Fidell, 2012).

In consideration of this, we used multiple logistic regression to determine the probability of a pair of countries being connected within the global IWT network. The following logistic model was used to calculate the odds ratio (OR):

l o g i t (Y_{i j}) = n a t u r a l \log (o d d s) = l n (\frac{π_{i j}}{1 - π_{i j}}) = β_{0} + β_{1} {\vec{X}}_{i} + β_{2} {\vec{Z}}_{j} + β_{3} {\vec{W}}_{i j}

(2)

From Equation 2, it is possible to derive an Equation 3 to estimate the probability of the occurrence of our outcome of interest:

π_{i j} = \frac{e^{β_{0} + β_{1} {\vec{X}}_{i} + β_{2} {\vec{Z}}_{j} + β_{3} {\vec{W}}_{i j}}}{1 + e^{β_{0} + β_{1} {\vec{X}}_{i} + β_{2} {\vec{Z}}_{j} + β_{3} {\vec{W}}_{i j}}}

(3)

where πij is the pair of countries’ predicted probability of being connected by illicit flows of waste; $Y_{i j}$ is the binary variable (i.e., $Y_{i j} = 1$ IWT from exporter to importer, $Y_{i t} = 0$ absence of trafficking) of the observed connection; α is the $Y_{i j}$ intercept; $β_{n}$ are the regression coefficients corresponding to the ${\vec{X}}_{i}$ , ${\vec{Z}}_{j}$ , and ${\vec{W}}_{i j}$ vectors of independent variables referring to importing countries, exporting countries, and dyads; $e$ is the base of the system of the natural logarithm. The $β_{n}$ regression coefficients express both the direction and the intensity of the relationship between $X_{i}$ , $Z_{j}$ , $W_{i j}$ and the logit of $Y_{i j}$ , as it is the estimated increase in the log odds of the outcome per unit increase in the value of the independent variable. Consequently, the exponential function of the regression coefficient ( $e^{\hat{β_{n}}}$ ) is the OR associated with a one-unit increase in $X_{i}$ , $Z_{j}$ , $W_{i j}$ . An OR of 1 indicates that the considered independent variable has a neutral impact on the outcome; OR greater than 1 indicates a positive effect; reversely, an OR below 1 shows a negative impact.

Although logistic regression does not assume that the dependent and independent variables are related linearly, this analysis required the independent variables to be linearly related to the log odds. We tested the validity of this assumption with respect to the considered variables by using the Box–Tidwell method (Box & Tidwell, 1962; Hosmer & Lemeshow, 1980). Through recourse to this approach, we logistically regressed our dependent variable on each of the continuous predictors, and on their interaction terms, which consist of the continuous predictors and their natural logs. The statistical significance of any of the added interactions indicates that the linearity assumption is violated and a transformation of the offending independent variable is needed to better meet the linearity of the logit assumptions (Tabachnick & Fidell, 2012). The natural logarithm proved to be effective for solving issues concerning the four variables identified as problematic through the Box–Tidwell method (i.e., trade globalization, population size, distance, and generated waste). It follows:

\begin{array}{l} l n (\frac{π_{i j}}{1 - π_{i j}}) = β_{0} + β_{1} {\vec{\tilde{X}}}_{i} + β_{2} {\vec{\tilde{Z}}}_{j} + β_{3} {\vec{\tilde{W}}}_{i j} \\ where \tilde{X_{i}} = \ln (X_{i}), \tilde{Z_{j}} = \ln (Z_{j}), \tilde{W_{i j}} = \ln (W_{i j}) \end{array}

(4)

Accordingly, the interpretation of the correlation between these variables changes. In particular, for instance, the OR associated to a relative change in ${\tilde{X}}_{i}$ would be $q^{{\hat{β}}_{1}}$ , where q is the relative change in ${\tilde{X}}_{i}$ and $\hat{β_{1}}$ is the estimated coefficient. However, for small changes in $X_{i}$ , it is possible to use an approximation based on the first-order Taylor series expansion. Thus, a 1% increase in the log-transformed independent variable ${\tilde{X}}_{i}$ is associated with an OR ≈ 1 + $\hat{β} / 100$ of $Y_{i j}$ (Barrera-Gómez & Basagaña, 2015).

Since logistic regression requires independent variables not to be too highly correlated with each other, we verified that this assumption was not violated by analyzing the correlation between our independent variables (Appendix A). As additional assumption, logistic regressions require the observations to have mutually exhaustive categories that are independent—that is, the observations should neither be matched nor come from repeated measurements—(Tabachnick & Fidell, 2012). The data exploited here do not pose any issues with respect to longitudinality as they do not come from repeated measures; on the contrary, they refer to a single moment in time—the period 2016–2017. Yet, different observations might not be fully independent of each other as the same country may report links with several others. Therefore, to relax the independence of errors assumption, we produced logistic regressions with robust standard errors clustered at the exporter-country level (Arellano, 1987; Shepherd, 2016; H. White, 1984).

Results and Discussion

The results of our multivariate logistic analyses are presented in Table 4, which focuses on the investigation of Hypotheses 1 to 5, and in Table 5, which is intended to complete the analysis by including Hypothesis 6. The models presented in Table 5 were constructed by adding to the Model 6 a subset of variables that measured the regulatory and judicial system.⁶ Table 6 summarizes the implications of the obtained empirical results with respect to the formulated hypotheses.

Table 4.

Structural Predictors of Transnational IWT.

	Cases = 0 if no trafficking, 1 if trafficking
	(m1)	(m2)	(m3)	(m4)	(m5)	(m6)	(m7)
Socioeconomic factors (Hypothesis 1)
Infant mortality, importer	1.02*** [1.01, 1.03]	1.03*** [1.02, 1.04]	1.03*** [1.03, 1.04]		1.03*** [1.02, 1.04]	1.03*** [1.02, 1.04]	1.02** [1.01, 1.03]
Conflicts index, importer		0.85*** [0.78, 0.92]	0.89* [0.81, 0.97]	0.93 [0.85, 1.02]	0.89** [0.82, 0.96]	0.89* [0.82, 0.97]	1.00 [0.91, 1.09]
GDP growth, importer		1.03 [1.00, 1.07]	1.03 [1.00, 1.06]	1.05 [1.00, 1.09]	1.03 [1.00, 1.07]	1.03 [1.00, 1.07]	1.05*** [1.02, 1.08]
Shadow economy, importer				1.04*** [1.02, 1.06]
Shadow economy, exporter				0.95 [0.92, 0.99]
Waste production and management and demography (Hypothesis 2)
Generated waste pc (ln), importer	0.05*** [0.01, 0.19]	0.11*** [0.03, 0.36]	0.14** [0.04, 0.47]	0.03*** [0.01, 0.15]	0.07** [0.01, 0.40]	0.07** [0.01, 0.43]	0.06 [0.01, 0.56]
Generated waste pc (ln), exporter	27.40*** [3.88, 193.68]	30.63*** [4.32, 217.11]	33.46*** [4.46, 250.95]	6.77 [0.42, 108.43]	0.98 [0.08, 12.22]	1.08 [0.08, 15.34]	0.24 [0.01, 4.51]
Recycling, importer				1.02 [1.00, 1.03]	1.01 [1.00, 1.03]	1.01 [1.00, 1.03]	1.01 [1.00, 1.03]
Recycling, exporter				1.01 [0.98, 1.04]	0.99 [0.97, 1.02]	0.99 [0.96, 1.02]	1.00 [0.97, 1.03]
Population density, importer			0.87 [0.55, 1.37]	0.84 [0.48, 1.46]	0.73 [0.38, 1.39]	0.74 [0.39, 1.41]	1.04 [0.76, 1.43]
Population density, exporter			0.84 [0.67, 1.07]	0.81 [0.66, 1.00]	0.91 [0.76, 1.09]	0.89 [0.75, 1.07]	1.28 [0.98, 1.66]
Population size (ln), importer	1.79*** [1.63, 1.97]	1.92*** [1.74, 2.12]	2.18*** [1.98, 2.40]	2.01*** [1.81, 2.24]	2.21*** [2.01, 2.44]	2.22*** [2.01, 2.46]	2.42*** [2.15, 2.73]
Population size (ln), exporter	1.64*** [1.31, 2.04]	1.65*** [1.33, 2.05]	1.98*** [1.55, 2.53]	1.69*** [1.28, 2.22]	1.88*** [1.51, 2.35]	1.94*** [1.54, 2.44]	2.18*** [1.77, 2.69]
Colonial ties (Hypothesis 3)
Colonial relation	10.11*** [5.92, 17.26]	10.51*** [6.12, 18.06]	11.20*** [6.71, 18.70]	13.18*** [8.44, 20.59]	9.96*** [6.74, 14.72]	9.74*** [6.55, 14.49]	10.15*** [6.71, 15.36]
Geography (Hypothesis 4)
Distance (ln)	0.29*** [0.24, 0.34]	0.27*** [0.23, 0.32]	0.30*** [0.25, 0.36]	0.32*** [0.26, 0.38]	0.30*** [0.24, 0.37]	0.29*** [0.23, 0.36]	0.27*** [0.21, 0.36]
Trade ties (Hypothesis 5)
Trade globalization (ln), imp			2.34*** [1.60, 3.41]	1.56 [1.07, 2.29]	2.11** [1.33, 3.33]	2.16*** [1.38, 3.40]	2.14*** [1.41, 3.23]
Trade globalization (ln), exporter			3.59 [1.12, 11.57]	2.65 [0.71, 9.87]	1.94 [0.62, 6.07]	2.38 [0.79, 7.24]	2.33 [0.93, 5.83]
Enforcement controls
Rule of law, importer					1.10 [0.89, 1.37]
Rule of law, exporter					3.08*** [2.18, 4.36]
Corruption control, importer						1.09 [0.89, 1.33]	1.16 [0.94, 1.43]
Corruption control, exporter						2.78*** [2.08, 3.73]	2.55*** [2.00, 3.25]
Additional controls
Control for information reporting	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Macroregion control, importer							Yes
Macroregion control, exporter							Yes
Observations	20,842	19,972	19,418	19,560	19,418	19,418	19,134
Wald χ²	657.1	687.4	1,118.7	1,375.5	1,297.5	1,258.8	1,376.5
Probability > Wald χ²	.000	.000	.000	.000	.000	.000	.000
HL χ²	7.26	3.30	4.65	3.42	5.09	6.99	7.35
Probability > HL χ²	.508	.914	.794	.905	.748	.538	.499
McFadden’s R²	.412	.415	.424	.430	.461	.463	.485
Adjusted McFadden’s R²	.406	.409	.415	.419	.449	.451	.467
AUC	0.948	0.948	0.951	0.954	0.959	0.960	0.964
BIC	2,186.3	2,155.8	2,153.8	2,167.1	2,066.7	2,059.1	2,086.4

Note. The table reports the estimated OR for the presence of IWT between a pair of countries based on a series of structural characteristics for the two countries and their relationship. Robust 95% asymmetrical confidence intervals are provided in square brackets. The proposed Wald test works by testing the null hypothesis that all coefficients of interest are simultaneously equal to zero. The Hosmer–Lemeshow (1980) goodness-of-fit test assesses whether the observed event rates match expected event rates in subgroups of the model population (i.e., the deciles of fitted risk values). Both the McFadden (1973) and the adjusted McFadden’s (1973) pseudo R² provide indications on the proportion of the variance in the dependent variable that is predictable from the independent variables. The area under curve (AUC), referred to as an index of accuracy, is a perfect performance metric for ROC curve, which summarizes the model’s performance by evaluating the trade-offs between true positive rates (sensitivity) and false-positive rates (1—specificity). The Bayesian information criterion (BIC) values provide two measures of the relative quality of the models. GDP = gross domestic product; OR = odds ratio; IWT = illicit waste trafficking.

***

, **, and * indicate significance at 0.1%, 0.5%, and 1%, respectively.

Table 5.

Relationship Between Regulation, Judiciary System, and Transnational IWT.

	Cases = 0 if no trafficking, 1 if trafficking
	(m8)	(m9)	(m10)	(m11)
Regulation and judiciary system (Hypothesis 6)
Basel, importer	1.00 [1.00, 1.00]
Basel, exporter	0.37 [0.13, 1.03]
National environmental laws, importer		0.84 [0.52, 1.36]
National environmental laws, exporter		0.64 [0.30, 1.34]
Courts and tribunals (dummy), importer			0.68 [0.49, 0.94]
Courts and tribunals (dummy), exporter			0.70 [0.38, 1.27]
Courts and tribunals (categories), importer				0.67*** [0.55, 0.81]
Courts and tribunals (categories), exporter				0.92 [0.67, 1.27]
Enforcement controls
Corruption control, importer	1.08 [0.89, 1.31]	1.09 [0.89, 1.33]	1.15 [0.92, 1.45]	1.23 [0.98, 1.54]
Corruption control, exporter	2.82*** [2.09, 3.80]	2.81*** [2.08, 3.79]	3.06*** [2.23, 4.20]	2.92*** [2.11, 4.06]
Socioeconomic factors (Hypothesis 1)
Infant mortality, importer	1.03*** [1.02, 1.04]	1.03*** [1.02, 1.04]	1.03*** [1.02, 1.04]	1.03*** [1.02, 1.04]
Conflicts index, importer	0.88** [0.81, 0.96]	0.89** [0.81, 0.96]	0.90* [0.83, 0.97]	0.91 [0.84, 0.99]
GDP growth, importer	1.03 [1.00, 1.07]	1.03 [1.00, 1.07]	1.04* [1.01, 1.08]	1.05* [1.01, 1.09]
Waste production and management and demography (Hypothesis 2)
Generated waste pc (ln), importer	0.09 [0.01, 0.57]	0.07** [0.01, 0.42]	0.06** [0.01, 0.38]	0.07** [0.01, 0.41]
Generated waste pc (ln), exporter	0.80 [0.05, 12.91]	1.16 [0.09, 14.65]	1.02 [0.08, 12.67]	1.10 [0.09, 14.13]
Recycling, importer	1.01 [1.00, 1.03]	1.01 [1.00, 1.03]	1.01 [1.00, 1.03]	1.01 [1.00, 1.03]
Recycling, exporter	0.99 [0.96, 1.02]	0.99 [0.96, 1.02]	0.99 [0.96, 1.01]	0.99 [0.96, 1.02]
Population density, importer	0.73 [0.38, 1.40]	0.72 [0.36, 1.43]	0.83 [0.50, 1.39]	0.85 [0.55, 1.32]
Population density, exporter	0.91 [0.75, 1.09]	0.87 [0.75, 1.00]	0.90 [0.76, 1.08]	0.91 [0.76, 1.09]
Population size (ln), importer	2.24*** [2.02, 2.48]	2.24*** [2.02, 2.49]	2.29*** [2.07, 2.55]	2.31*** [2.09, 2.54]
Population size (ln), exporter	1.94*** [1.54, 2.45]	1.98*** [1.56, 2.50]	1.99*** [1.59, 2.50]	1.98*** [1.57, 2.51]
Colonial ties (Hypothesis 3)
Colonial relation	9.79*** [6.60, 14.53]	9.74*** [6.49, 14.62]	9.59*** [6.51, 14.12]	11.11*** [7.47, 16.52]
Geography (Hypothesis 4)
Distance (ln)	0.29*** [0.23, 0.36]	0.29*** [0.23, 0.36]	0.30*** [0.24, 0.39]	0.30*** [0.24, 0.38]
Trade ties (Hypothesis 5)
Trade globalization (ln), importer	2.09** [1.33, 3.28]	2.18*** [1.39, 3.43]	2.00* [1.23, 3.25]	1.92* [1.18, 3.11]
Trade globalization (ln), exporter	2.68 [0.87, 8.25]	2.31 [0.80, 6.69]	2.22 [0.79, 6.24]	2.38 [0.81, 7.01]
Additional controls
Information reporting	Yes	Yes	Yes	Yes
Observations	19,276	19,418	19,418	19,418
Wald χ²	1,236.0	1,285.2	1,360.4	1,294.6
Probability > Wald χ²	.000	.000	.000	.000
HL χ²	7.50	6.28	4.78	3.62
Probability > HL χ²	.484	.616	.781	.890
McFadden’s R²	.464	.464	.466	.468
Adjusted McFadden’s R²	.451	.451	.453	.455
AUC	0.960	0.960	0.960	0.960
BIC	2,063.3	2,074.0	2,068.1	2,060.7

***

, **, and * indicate significance at 0.1%, 0.5%, and 1%, respectively.

Table 6.

Summary of the Empirical Analyses of the Investigated Hypotheses.

Hypotheses	Confirmation or rejection
Hypothesis 1. Countries characterized by political and socioeconomic insecurity are more likely to receive trafficked waste from other countries	Partially confirmed
Hypothesis 2. Countries that produce more waste, that have a larger population, and are less able to recycle are more likely to unlawfully export their waste to third countries	Partially rejected
Hypothesis 3. Countries sharing a colonial relationship are more likely to exchange illicit waste	Confirmed
Hypothesis 4. Close countries are more likely to exchange illicit waste	Confirmed
Hypothesis 5. Countries more relevant within global trade are more likely to exchange illicit waste	Confirmed
Hypothesis 6. International treaties, specific national environmental regulations, the establishment of ad hoc courts and tribunals do not shape the IWT network	Partially rejected

Note. IWT = illicit waste trafficking.

A higher infant mortality—interpreted here as a measure of extreme poverty—in an importing country is always significantly correlated with the increased likelihood of a country being a destination for illegal waste. On the contrary, the economic growth of the recipient country is not significantly correlated with the import of illegal waste. The higher the relevance of the shadow economy in a given country, the higher the probability of importing illicit waste. This potentially pertains to the fact that unregulated waste management is a primary source of daily sustenance within the marginalized urban economies of developing countries (Gidwani, 2015; Lassou & Hopper, 2016). Finally, in most econometric specifications, conflicts in recipient countries are also significantly negatively correlated with importing of illicit waste. Overall, the statistical results only partially confirm our first hypothesis since more socioeconomically insecure countries tend to suffer the most from transnational IWT, but more politically insecure countries do not. Although we can speculate that the countries currently riven by conflicts are underreported by the BCNRs, the fact the Conflicts index is negatively correlated to the presence of trafficking connections indicates that IWT shares some characteristics with legal trades. Indeed, more dangerous countries, as the conflicting ones, are less likely to be involved in it. These results provide empirical support and further corroborate previous findings that suggested how trade flows tend to go from more developed to the less developed countries (Elliott & Schaedla, 2016; Klenovšek & Meško, 2011).

We partially reject Hypothesis 2. Countries with a higher per capita production of solid waste appear to more frequently operate as exporters of illicit waste; yet, the correlation loses its significance when controlling for the recycling capacity—that is, from m4 to m7. Per capita waste production is, instead, a significant predictor of IWT with respect to importing countries. The lower the level of waste production, the higher the probability of incoming connections in the global IWT network. Neither the share of recycled waste nor population density correlates with the structure of the trafficking network, in regard to either exporting or importing countries. The size of exporting countries, in terms of inhabitants, as hypothesized, is positively correlated with being exporters of illicit waste. Nonetheless, the most common importing countries also tend to have a larger population; this result is not surprising considering that we are examining either the presence or absence of trafficking connections, as opposed to, say, the rates of illicit waste per person. Yet a further explanation for this result might be that larger countries also offer more opportunities for traffickers to illicitly dispose of their waste—especially when considering that population density does not affect trafficking—or to reinsert the waste within secondary markets as either secondary raw materials or used goods.

Former colonial ties are significant predictors of trafficking connections in all of the econometric specifications performed, and, hence, this result validates Hypothesis 3. Holding other variables at a fixed value, on average, the models proposed indicate a 900% increase in the odds of observing IWT among countries, which had colonial relationships. Owing to the specificities of our econometric models, the positive correlation between colonial relations and IWT can be interpreted in two ways. Cultural and social proximity—that is, speaking the same language, and having prior and current political and institutional relationships—increase the level of trust between criminal partners and facilitate access to valuable information between countries, while, simultaneously, reducing transaction costs (Aziani et al., 2019; Kleemans & van de Bunt, 1999). At the same time, when observing the results on a prior colonial relation in conjunction with those on countries’ economic and—formal—institutional development, a complementary interpretation emerges: Former colonizers are continuing the environmental exploitation of their former colonial territories that began during the colonial era (Heger, 2017).

The results provide strong support for Hypothesis 4; as expected on the basis of geographical reasoning, the closer countries are the more likely one is to observe illicit trades between them. The strong relevance of geographical factors also emerges from the observation of the trafficking network (Figure 2). China stands out for its high indegree centrality; the country has numerous inward connections—signaling the import of illegal waste—from close-by Asian countries, such as Japan, India, Malaysia, Philippines, and South Korea. Similar geographical clusters emerge in Central Europe—centered on Germany and Poland—and in Western Africa—centered on Nigeria. At the samwe time, it is worth noting that sometimes geographical proximity partially overlaps with migration flows, which partially confounds the analysis of the relevance of social and geographical proximity.

Centrality in both the global trade network and the global provision of services over long distances is associated with a higher number of inward connections in the global IWT network. When including the regressor variables that describe the regulatory environment, the correlation between the presence/absence of trafficking connections and the level of trade globalization of the importing country remains significant. These empirical results corroborate the Hypothesis 5, which hypothesizes a functional overlap between legal and illegal movements of waste, as observed and described by previous studies (Andreatta & Favarin, 2020; Lieselot, 2016; Sahramäki et al., 2017).

Econometric results pertaining to regulatory and judiciary systems suggest that our Hypothesis 6 is only partially confirmed. The sole adhesion to international treaties on environmental issues as well as the promulgation of specific national environmental laws are not effective in curbing transnational IWT, as shown by m8 and m9 (Table 5). This is in line with the previous literature that has highlighted the inability of current international and national regulations to effectively fight national and transnational IWT (Baird et al., 2014; Klenovšek & Meško, 2011; Morganti et al., 2020; R. White, 2011). Yet, countries with the joint presence of specialized national environmental courts and tribunals (m11) were shown to have an impact on inward waste trafficking, which underscores the importance of setting up ad hoc judiciary bodies at a national level. Finally, a stronger rule of law and more effective control of corruption in exporting countries are positively correlated with the presence of IWT connections, whereas these factors are not relevant for importing countries. This suggests that higher difficulties associated with the illegal management of waste in source countries push unscrupulous entrepreneurs to look for cheap solutions overseas, thus fostering transnational IWT.

Overall, the results of the Hosmer–Lemeshow (1980) goodness-of-fit test indicate no evidence of poor fit, and, hence, that the proposed models are correctly specified. At the same time, the measures of the area under curve of each model indicate the models’ satisfactory capacity to strike a balance between true positive rates and false-positive rates. Viewed together, this underlines the robustness of the econometric results obtained in this study. Nonetheless, two main possible limitations must be taken into account when considering the results of our analysis. The network construction suffers from the fact that countries are known to have different inclinations and capacities with which to intercept illicit waste shipments, and in terms of reporting them to the Basel Convention. However, this limitation is mitigated to some extent by the fact that the role of a country in the network is also inferred by information provided from other countries, as well as the fact that the data refer to multiple years—that is, 2 years—as per other studies on transnational trafficking (e.g., Aziani et al., 2019; Berlusconi et al., 2017), and the inclusion in the analysis of a specific control expressing information reporting. Nevertheless, the final network still likely underestimates the complexity of actual transnational IWT. In consideration of this, the analysis is likely to depict the most relevant trafficking connections, rather than all the possible connections. The second class of limitations pertain to econometric specifications, in particular, the risk of suffering from an omitted variable bias and the presence of a reverse causality issue. Yet, the fact that the high number of considered independent variables and controls were carefully selected on the basis of indications from previous studies, together with the use of macroregional fixed effects, alleviates the risk of biases stemming from the omission of relevant independent variables. Hence, adopting a cautious interpretation of the econometric results inspired by previous findings and sound theoretical reasoning is the optimal way to deal with the impossibility of identifying indisputable causal relationships between the considered factors and our measure of IWT.

Conclusion

This study maps for the first time the transnational IWT flows and presents the first ever empirical quantitative analysis of the correlates of cross-border IWT at a global level. It provides evidence of the fact that more fragile countries are at a greater risk of paying the deleterious costs of global waste production. Conversely, and somewhat paradoxically, countries with stronger control of corruption and a more solid rule of law tend to export more illicit waste likely as a consequence of national illicit disposal being less practicable. At the same time, the adoption of ad hoc international and national environmental laws is shown to be ineffective in curbing IWT, if they are not supported by agencies capable of enforcing them. All in all, this study has shown that the current global waste management system implicitly—and partially—relies on the mistreatment of vulnerable communities. This situation calls for understanding that IWT is a common and transnational issue that needs to be enforced at a global level.

Currently, countries too often limit the investigation of intercepted illicit waste shipments to their own territories, while, simultaneously, their own waste is flowing directly to weaker jurisdictions. For this reason, the enforcement measures must also be common and transnational, while cross-border cooperation must be promoted and improved. The existence of different justice response systems (based on the classification of the relevant actions/inactions as misdemeanors or crimes) encourages the shift toward less stringent regulatory frameworks. Moreover, within the European Union, for example, legislation requires member states to implement proportionate repressive measures, while, simultaneously, granting them flexibility over determining the quality (either administrative or criminal) and amount of sanctions. To coordinate investigative and prosecutorial activities at a global level, the sanctions against waste legislation violations should be aligned as much as possible. Without harmonization of sanctions and certitude that penalties will be carried out, we will continually come up against misalignments between jurisdictions. To achieve this, it would be central to strengthen the rule of law of vulnerable countries and to politically address this phenomenon at a global level.

Together with these policy implications, the current study—despite its limitations—allows for the identification of different avenues of inquiry that scholars might want to explore in the future. First, the literature would benefit from studies capable of differentiating between various types of waste, as different costs of disposal and potential secondary markets are all associated with different categories of waste. By differentiating the trafficking by-product, it will become possible to better disentangle the manifold mechanisms and motivations that govern these different types of trafficking. Second, conducting longitudinal studies will enable researchers to trace the evolution of IWT while, simultaneously, allowing for the use of statistical strategies that are more effective in dealing with the potential biases stemming from omitted variables. Finally, those criminals that are involved in the trafficking of illicit waste internationally are themselves worthy of greater academic attention. While it is well known that organized crime groups and white-collar criminals manage this business, gaining a greater understanding of the typologies of criminals involved in IWT, as well as their connections at a global level, is essential for designing new and more effective policies to combat this crime.

Footnotes

Appendix A

Appendix B

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Serena Favarin is an assistant professor at the Università Cattolica del Sacro Cuore di Milano and research fellow at Transcrime, Milan. She holds a PhD in Criminology and a MSc in Applied Social Sciences. Her main research interests are environmental crimes, spatial analysis of urban crimes, perception of insecurity, and organized crime. She coordinated and collaborated to several national and international research projects on these topics.

Alberto Aziani is an assistant professor at the Università Cattolica del Sacro Cuore di Milano and research fellow at Transcrime, Milan. He holds a PhD in Criminology and an MSc in Economics and Social Sciences. His main research interests are transnational and organized crimes. On these topics, he has conducted numerous studies and developed projects for international organizations and funding institutions.

References

Adeola

F. O.

(2000). Cross-national environmental injustice and human rights issues: A review of evidence in the developing world. American Behavioral Scientist, 43(4), 686–706. https://doi.org/10.1177/00027640021955496

Akee

Basu

A. K.

Bedi

Chau

N. H.

(2014). Transnational trafficking, law enforcement, and victim protection: A middleman trafficker’s perspective. The Journal of Law and Economics, 57(2), 349–386. https://doi.org/10.1086/675404

Almer

Goeschl

(2015). The Sopranos Redux: The empirical economics of waste crime. Regional Studies, 49(11), 1908–1921. https://doi.org/10.1080/00343404.2013.854323

Andreatta

Favarin

(2020). Features of transnational illicit waste trafficking and crime prevention strategies to tackle it. Global Crime, 21, 130–153. https://doi.org/10.1080/17440572.2020.1719837

Arellano

(1987). Computing robust standard errors for within-group estimators. Oxford Bulletin of Economics and Statistics, 49, 431–434.

Aziani

(2020). Violent disequilibrium: The influence of instability in the economic value of cocaine markets on homicides. Crime, Law and Social Change, 52, 275–284. https://doi.org/10.1007/s10611-020-09894-2

Aziani

Berlusconi

Giommoni

(2019). A quantitative application of enterprise and social embeddedness theories to the transnational trafficking of cocaine in Europe. Deviant Behavior. https://doi.org/10.1080/01639625.2019.1666606

Baird

Curry

Cruz

(2014). An overview of waste crime, its characteristics, and the vulnerability of the EU waste sector. Waste Management & Research, 32(2), 97–105. https://doi.org/10.1177/0734242X13517161

Barrera-Gómez

Basagaña

(2015). Models with transformed variables: Interpretation and software. Epidemiology, 26(2), Article e16.

10.

Berlusconi

Aziani

Giommoni

(2017). The determinants of heroin flows in Europe: A latent space approach. Social Networks, 51, 104–117. https://doi.org/10.1016/j.socnet.2017.03.012

11.

Biotto

Silvestri

Gobbo

Furlan

Valenti

Rosselli

(2009). GIS, multi-criteria and multi-factor spatial analysis for the probability assessment of the existence of illegal landfills. International Journal of Geographical Information Science, 23(10), 1233–1244. https://doi.org/10.1080/13658810802112128

12.

Bisschop

(2012). Is it all going to waste? Illegal transports of e-waste in a European trade hub. Crime, Law and Social Change, 58(3), 221–249. https://doi.org/10.1007/s10611-012-9383-0

13.

Boivin

(2013). Drug trafficking networks in the world-economy. In Morselli

(Ed.), Crime and networks (pp. 182–194). Routledge.

14.

Boivin

(2014). Risks, prices, and positions: A social network analysis of illegal drug trafficking in the world-economy. International Journal of Drug Policy, 25(2), 235–243. https://doi.org/10.1016/j.drugpo.2013.12.004

15.

Box

G. E. P.

Tidwell

P. W.

(1962). Transformation of the Independent Variables. Technometrics, 4(4), 531–550.

16.

Calderoni

Favarin

Garofalo

Sarno

(2014). Counterfeiting, illegal firearms, gambling and waste management: An exploratory estimation of four criminal markets. Global Crime, 15, 108–137. https://doi.org/10.1080/17440572.2014.883499

17.

Cassetta

Pauselli

Rizzica

Tonello

(2014). Financial flows to tax havens: Determinants and anomalies. Quaderni Dell’antiriciclaggio, 1, 1–29.

18.

Caulkins

J. P.

Bond

B. M.

(2012). Marijuana price gradients: Implications for exports and export-generated tax revenue for California after legalization. Journal of Drug Issues, 42(1), 28–45.

19.

Clapp

(1994). The toxic waste trade with less-industrialised countries: Economic linkages and political alliances. Third World Quarterly, 15(3), 505–518.

20.

Crofts

Morris

Wells

Powell

(2010). Illegal dumping and crime prevention: A case study of Ash Road, Liverpool Council. Public Space: The Journal of Law and Social Justice, 5, 23. https://doi.org/10.5130/psjlsj.v5i0.1904

21.

D’Amato

Mazzanti

Nicolli

Zoli

(2018). Illegal waste disposal: Enforcement actions and decentralized environmental policy. Socio-Economic Planning Sciences, 64, 56–65. https://doi.org/10.1016/j.seps.2017.12.006

22.

Disdier

A.-C.

Head

(2008). The puzzling persistence of the distance effect on bilateral trade. The Review of Economics and Statistics, 90(1), 37–48.

23.

Dorn

van Daele

Vander Beken

(2007). Reducing vulnerabilities to crime of the European waste management industry: The research base and the prospects for policy. European Journal of Crime, Criminal Law and Criminal Justice, 23, 23–36.

24.

Elliott

(2012). Fighting transnational environmental crime. Journal of International Affairs, 66(1), 87–104.

25.

Elliott

Schaedla

W. H.

(2016). Handbook of transnational environmental crime. Edward Elgar.

26.

Environmental Law Institute. (2019). Environmental rule of law. First global report. United Nation Environmental Programme. https://wedocs.unep.org/bitstream/handle/20.500.11822/27279/Environmental_rule_of_law.pdf?sequence=1&isAllowed=y

27.

Ferwerda

Kattenberg

Chang

H.-H.

Unger

Groot

Bikker

J. A.

(2013). Gravity models of trade-based money laundering. Applied Economics, 45(22), 3170–3182. https://doi.org/10.1080/00036846.2012.699190

28.

Germani

A. R.

Pergolizzi

Reganati

(2015). Illegal trafficking and unsustainable waste management in Italy: Evidence at regional level. Journal of Security and Sustainability Issues, 4(4), 369–389.

29.

Germani

A. R.

Pergolizzi

Reganati

(2017). Le determinanti del traffico organizzato di rifiuti in Italia: Un’analisi empirica a livello regionale. Rivista economica del Mezzogiorno, 1, 269–304.

30.

Gibbs

Simpson

S. S.

(2009). Measuring corporate environmental crime rates: Progress and problems. Crime, Law and Social Change, 51(1), 87–107. https://doi.org/10.1007/s10611-008-9145-1

31.

Gidwani

(2015). The work of waste: Inside India’s infra-economy. Transactions of the Institute of British Geographers, 40(4), 575–595. https://doi.org/10.1111/tran.12094

32.

Giommoni

Aziani

Berlusconi

(2017). How do illicit drugs move across countries? A network analysis of the heroin supply to Europe. Journal of Drug Issues, 47(2), 217–240. https://doi.org/10.1177/0022042616682426

33.

Goel

R. K.

Saunoris

J. W.

(2019). Does variability in crimes affect other crimes? The case of international corruption and shadow economy. Applied Economics, 51(3), 239–258. https://doi.org/10.1080/00036846.2018.1494378

34.

Greaney

T. M.

Kiyota

(2020). The gravity model and trade in intermediate inputs. The World Economy. Advance online publication. https://doi.org/10.1111/twec.12947

35.

Greife

Maume

M. O.

(2020). Stealing like artists: Using court records to conduct quantitative research on corporate environmental crimes. Journal of Contemporary Criminal Justice, 36(3), 451–469. https://doi.org/10.1177/1043986220931613

36.

Gygli

Haelg

Potrafke

Sturm

J.-E.

(2019). The KOF Globalisation Index—Revisited. The Review of International Organizations. KOF Swiss Economic Institute.

37.

Heger

(2017). Trafficking in hazardous wastes (Article 28L) and illicit exploitation of natural resources (Article 28Lbis). In Werle

Vormbaum

(Eds.), The African criminal court: A commentary on the Malabo protocol (pp. 125–136) T. M. C. Asser Press. https://doi.org/10.1007/978-94-6265-150-0_8

38.

Hosmer

D. W.

Lemeshow

(1980). Goodness of fit tests for the multiple logistic regression model. Communications in Statistics—Theory and Methods, 9(10), 1043–1069.

39.

Ichinose

Yamamoto

(2011). On the relationship between the provision of waste management service and illegal dumping. Resource and Energy Economics, 33(1), 79–93. https://doi.org/10.1016/j.reseneeco.2010.01.002

40.

Kabir

Salim

Al-Mawali

(2017). The gravity model and trade flows: Recent developments in econometric modeling and empirical evidence. Economic Analysis and Policy, 56, 60–71. https://doi.org/10.1016/j.eap.2017.08.005

41.

Kahane

L. H.

(2013). Understanding the interstate export of crime guns: A gravity model approach. Contemporary Economic Policy, 31(3), 618–634. https://doi.org/10.1111/j.1465-7287.2012.00324.x

42.

Kim

G.-S.

Chang

Y.-J.

Kelleher

(2008). Unit pricing of municipal solid waste and illegal dumping: An empirical analysis of Korean experience. Environmental Economics and Policy Studies, 9, 167–176.

43.

Kleemans

E. R.

van de Bunt

(1999). The social embeddedness of organized crime. Transnational Organized Crime, 5(1), 19–36.

44.

Klenovšek

Meško

(2011). International waste trafficking: Preliminary explorations. In Meško

Dimitrijević

Fields

C. B.

(Eds.), Understanding and managing threats to the environment in South Eastern Europe (pp. 79–99). Springer. https://doi.org/10.1007/978-94-007-0611-8_5

45.

Landman

Silverman

B. W.

(2020). Globalization and modern slavery. Politics and Governance, 7(4), 275–290. https://doi.org/10.17645/pag.v7i4.2233

46.

Lassou

P. J. C.

Hopper

(2016). Government accounting reform in an ex-French African colony: The political economy of neocolonialism. Critical Perspectives on Accounting, 36, 39–57. https://doi.org/10.1016/j.cpa.2015.10.006

47.

Liddick

(2010). The traffic in garbage and hazardous wastes: An overview. Trends in Organized Crime, 13(2–3), 134–146. https://doi.org/10.1007/s12117-009-9089-6

48.

Lieselot

(2016). Illegal trade in hazardous waste. In Elliott

Schaedla

W. H.

(Eds.), Handbook of transnational environmental crime (pp. 190–211). Edward Elgar.

49.

Liu

Kong

Santibanez Gonzalez

E. D. R.

(2017). Dumping, waste management and ecological security: Evidence from England. Journal of Cleaner Production, 167, 1425–1437. https://doi.org/10.1016/j.jclepro.2016.12.097

50.

Lynch

M. J.

Barrett

K. L.

Stretesky

P. B.

Long

M. A.

(2016). The weak probability of punishment for environmental offenses and deterrence of environmental offenders: A discussion based on USEPA criminal cases, 1983–2013. Deviant Behavior, 37(10), 1095–1109. https://doi.org/10.1080/01639625.2016.1161455

51.

Lynch

M. J.

Barrett

K. L.

Stretesky

P. B.

Long

M. A.

(2017). The neglect of quantitative research in green criminology and its consequences. Critical Criminology, 25(2), 183–198. https://doi.org/10.1007/s10612-017-9359-6

52.

Lynch

M. J.

Pires

S. F.

(2019). Quantitative studies in green and conservation criminology: The measurement of environmental harm and crime. Routledge.

53.

Lynch

M. J.

Stretesky

P. B.

Long

M. A.

(2019). Environmental crime prosecutions in Ireland, 2004–2014. International Journal of Comparative and Applied Criminal Justice, 43(4), 277–293. https://doi.org/10.1080/01924036.2019.1615520

54.

Marshall

M. G.

(2019). Major Episodes of Political Violence (MEPV) and conflict regions, 1946-2018. Center for Systemic Peace. https://www.systemicpeace.org/warlist/warlist.htm

55.

Massari

Monzini

(2004). Dirty businesses in Italy: A case-study of illegal trafficking in hazardous waste. Global Crime, 6(3–4), 285–304. https://doi.org/10.1080/17440570500273416

56.

Matos

Oštir

Kranjc

(2012). Attractiveness of roads for illegal dumping with regard to regional differences in Slovenia. Acta Geographica Slovenica, 52, 431–451.

57.

Mayer

Zignago

(2011, December). Notes on CEPII’s distances measures: The GeoDist database (CEPII, Working Paper No. 25), 2011–2018. SSRN. https://ssrn.com/abstract=1994531

58.

McFadden

D. L.

(1973). Conditional logit analysis of qualitative choice behavior. In Zarembka

(Ed.), Frontiers in econometrics (pp. 105–142). Academic Press.

59.

Medina

Schneider

(2017). Shadow economies around the world: New results for 158 countries over 1991-2015 (No. 6430; CESifo Working Paper Series). CESifo Group Munich. https://ideas.repec.org/p/ces/ceswps/_6430.html

60.

Montgomery

D. C.

Peck

E. A.

Vining

G. G.

(2012). Introduction to linear regression analysis (5 ed.). John Wiley.

61.

Morganti

Favarin

Andreatta

(2020). Illicit waste trafficking and loopholes in the European and Italian legislation. European Journal on Criminal Policy and Research, 26(1), 105–133. https://doi.org/10.1007/s10610-018-9405-2

62.

Nellemann

Henriksen

Kreilhuber

Stewart

Kotsovou

Raxter

Mrema

Barrat

(Eds.). (2016). The rise of environmental crime: A growing threat to natural resources, peace, development and security. United Nations Environment Programme.

63.

Nobles

M. R.

(2019). Environmental crime and contemporary criminology: Making a difference. American Journal of Criminal Justice, 44(4), 656–669. https://doi.org/10.1007/s12103-019-09483-7

64.

Pereira

(2015). Environmental criminal liability and enforcement in European and international law (Vol. 21). Brill Nijhoff. https://brill.com/view/title/19665

65.

Prashantham

Dhanaraj

Kumar

(2015). Ties that bind: Ethnic ties and new venture internationalization. Long Range Planning, 48(5), 317–333. https://doi.org/10.1016/j.lrp.2015.02.005

66.

Press

S. J.

Wilson

(1978). Choosing between logistic regression and discriminant analysis. Journal of the American Statistical Association, 73(364), 699–705.

67.

Rauch

J. E.

(1999). Networks versus markets in international trade. Journal of International Economics, 48(1), 7–35. https://doi.org/10.1016/S0022-1996(98)00009-9

68.

Rucevska

Nellemann

Yang

Liu

Sandnæs

Olley

McCann

Devia

Bisschop

Soesilo

Schoolmeester

Henriksen

Nilsen

(2015). Waste crime —Waste risks, gaps in meeting the global waste challenge: A rapid response assessment. https://gridarendal-website-live.s3.amazonaws.com/production/documents/:s_document/27/original/RRA_WasteCrime_screen.pdf?1483646300

69.

Sahramäki

Favarin

Mehlbaum

Savona

Spapens

Kankaanranta

(2017). Wasting opportunities-prevention of illicit cross-border waste trafficking. European Journal of Policing Studies, 5(2), 1–61. http://www.maklu-online.eu/en/tijdschrift/ejps/volume-5/issue-2-november-december-2017/wasting-opportunities-prevention-illicit-crossbord/

70.

Sahramäki

Kankaanranta

(2017). Waste no money—Reducing opportunities for illicit waste dumping. Crime, Law and Social Change, 68, 217–232. https://doi.org/10.1007/s10611-016-9674-y

71.

Sgrignoli

Metulini

Schiavo

Riccaboni

(2015). The relation between global migration and trade networks. Physica A: Statistical Mechanics and Its Applications, 417, 245–260. https://doi.org/10.1016/j.physa.2014.09.037

72.

Sharma

(1995). Applied multivariate techniques (1st ed.). John Wiley.

73.

Shepherd

(2016). The gravity model of international trade: A user guide. United Nations. https://www.unescap.org/sites/default/files/GravityUserGuide-REVISED-02.pdf

74.

Sigman

(1998). Midnight dumping: Public policies and illegal disposal of used oil. The RAND Journal of Economics, 29(1), 157–178.

75.

Sin

S. S.

Boyd

M. A.

(2016). Searching for the nuclear silk road: Geospatial analysis of potential illicit radiological and nuclear material trafficking pathways. In Volders

Sauer

(Eds.), Nuclear terrorism: Countering the threat. Routledge.

76.

Spapens

Mehlbaum

Neve

(2018). Preventing illicit waste-exports from the Netherlands to China. In Mitsilegas

Hufnagel

Moiseienko

Yanan

Mingxiang

(Eds.), Transnational crime: European and Chinese perspectives (1st ed., pp. 215–234). Routledge. https://doi.org/10.4324/9781351026826

77.

Stassen

Ceccato

(2020). Environmental and Wildlife Crime (EWC) in Sweden 2000-2017. Journal of Contemporary Criminal Justice. https://doi.org/10.1177/1043986220927123

78.

Tabachnick

B. G.

Fidell

L. S.

(2012). Using multivariate statistics (6th ed.). Pearson College Div.

79.

Tompson

Chainey

(2011). Profiling illegal waste activity: Using crime scripts as a data collection and analytical strategy. European Journal on Criminal Policy and Research, 17(3), 179–201. https://doi.org/10.1007/s10610-011-9146-y

80.

UNEP & Basel Convention. (2018). Basel Convention. On the Control of Transboundary Movements of Hazardous Wastes and their Disposal. Protocol on Liability and Compensation for Damage Resulting from Transboundary Movements of Hazardous Wastes and their Disposal. Text and annexes. Secretariat of the Basel Convention. http://www.basel.int/TheConvention/Overview/TextoftheConvention/tabid/1275/Default.aspx

81.

UNEP & Basel Convention. (2019). Status of ratification. Secretariat of the Basel Convention. http://www.basel.int/Countries/StatusofRatifications/tabid/1341/Default.aspx

82.

United Nation Office on Drugs and Crime. (2009). Transnational trafficking and the rule of law in West Africa: A threat assessment. United Nation Office on Drugs and Crime. http://www.unodc.org/documents/data-and-analysis/Studies/West_Africa_Report_2009.pdf

83.

Walker

Unger

(2009). Measuring global money laundering: “The walker gravity model.” Review of Law & Economics, 5(2), 821–853.

84.

Ward

M. D.

Hoff

P. D.

(2007). Persistent patterns of international commerce. Journal of Peace Research, 44(2), 157–175. https://doi.org/10.1177/0022343307075119

85.

White

(1984). Asymptotic Theory for Econometricians. Academic Press.

86.

White

(2011). Transnational environmental crime: Toward an eco-global criminology. Routledge.

87.

World Bank. (2019a). DataBank—World development indicators. https://databank.worldbank.org/data/home.aspx

88.

World Bank. (2019b). What a waste. http://datatopics.worldbank.org/what-a-waste/

89.

Yotov

Y. V.

Piermartini

Monteiro

J.-A.

Larch

(2016). An advanced guide to trade policy analysis: The structural gravity model. United Nations and World Trade Organization Publications.