Corruption, sustainable development goals performance and modern slavery practices: an international evidence

In recent years, the topic modern slavery (MS) practice and firms’ accountability to curb such practices has received increased attention, particularly after the introduction of regulations in some countries such as the United States (US), the United Kingdom (UK) and Australia. According to a published report by the Australian Human Rights Commissions (2017), while slavery has been synonymous with historical human exploitation, contemporary slavery can be seen in many forms such as human trafficking, servitude, child labour, sex trafficking, forced labour and debt bondage. A recent report published by the International Labour Organisation (ILO, 2022) outlined that approximately 50 million people around the globe experienced MS in 2021, of which more than 50% (27.6 million) were victims of forced labour; the rest were in forced marriages. The report further stated that out of the 27.6 million, over 60% (over 17 million) were exploited in the private sector, over 22% (above 6.3 million) in forced commercial sexual exploitation, and nearly 4 million in forced labour imposed by the state. Women and girls accounted for 4.9 million of those in forced commercial sexual exploitation (ILO, 2022). In their recent report, the United Nations Development Program (UNDP, 2021) states that one in four victims of contemporary forms of slavery were children, and over 70% of victims were girls and women (UNDP, 2021). The US Department of Justice (2020) has added that essentially, contemporary forms of slavery have emerged in almost all countries and have become the second most lucrative illicit trade after the drug trade (US Department of Justice, 2020).

Against this backdrop, it is argued that addressing and mitigating MS is essential to achieve the United Nation’s sustainable development goals (SDGs) where multiple stakeholders, namely government, regulators, private sectors, NGOs and others, are required to work jointly to identify root causes and drivers of slavery, human exploitation and human rights violations (Bose et al., 2024; ILO, 2022; UNDP, 2021). In the current era, to end human exploitation and slavery practices, both public and private sectors should have a strict accountability and regulatory regime, as the private sector is often linked to slavery issues because of its global supply chain integration and business operations in developing countries (ILO, 2022). Jurisdictions, such as the US, UK and Australia, have devised regulations for firms to report whether they (the firms) comply with human rights in their business operations ^[1]. Other countries such as France, Germany, the Netherlands, and Qatar have adopted regulations to combat MS practices with different degrees of punishment and penalties in different jurisdictions. To illustrate, firms subject to France’s Duty of Vigilance Law-2017 must establish mechanisms to prevent human rights violations and environmental impacts throughout their supply chains, including for their subsidiaries’ companies. Applicable to firms with at least 5,000 employees, these requirements resemble the UN guiding principle on business and human rights and expect that companies demonstrate their capacity to mitigate slavery risks (Wolmen, 2024). In Germany, the 2021 Supply Chain Due Diligence Act has been in effect since 1 January 2023. Under the law, large German companies, as well as foreign companies with a branch office in Germany, are required to take certain measures to ensure that both firms' and their suppliers, from Germany and abroad, comply with a broad range of social and environmental standards along with reporting annually on their compliance with the law (Wolmen, 2024). Enacting the law with provision of a penalty of €800k for responsible managers and up to €8 million for enterprises for non-compliance, the law applies to firms having at least 1,000 employees. The Netherlands government has also joined the effort in 2019 and passed the Child Labor Due Diligence Law. Under the law, companies operating in the Netherlands will have to produce due diligence reports outlining and declaring that their supply chains are not involved in child labour and submit the report to the Dutch government (The Kingdom of the Netherlands, 2019). The report will have to meet internationally recognised due diligence standards; otherwise, firms will receive administrative fines and even criminal charges ^[2].

Academically, the motivations of the current study are manifold. To illustrate, the current study is aimed at contributing to academic literature by exploring the impact of country-level corruption on MS practices. That is, whether corruption in a country has an impact on MS practices. It is argued that a corrupt country perpetuates MS practices in its' diaspora since a country with higher levels of corruption cultivates opportunities to establish inequalities and violations of human rights, particularly for disadvantaged and marginalised people (Boersma, 2012; Figueiredo, 2017). Within accounting literature, following mandatory regulation, different academic studies have been conducted in different contexts such as the UK (Islam and Van Staden, 2022), Australia (Christ et al., 2020; Christ and Burrit, 2021) or in a conceptual study (Christ et al., 2023). These studies have provided increased understanding of issues, such as businesses’ responses to mandatory regulation in the Australian context (Christ and Burritt, 2023b), stakeholders’ narratives on the MS Act in the UK (Islam and Van Staden, 2022), the due diligence mechanism and assurance process in the MS report and the role of NGOs (Islam and van Staden, 2018), and the effectiveness of social audit in the MS reduction process (Islam et al., 2018). Nevertheless, previous accounting literature is virtually silent on exploring the relationship between country-level corruption and MS practices. We believe that our effort is academically pertinent and timely. We reason that countries with greater corruption practices will fuel more MS practices in different contexts because corruption creates vulnerable communities skewed towards exploitation and inequality. According to the UNDP (2021), as a root cause of contemporary forms of slavery, corruption accelerates the creation of a pool of vulnerable individuals in society where perpetrators reap the benefits of weak governance and accountability practices in different contexts by ignoring vulnerable people’s basic human rights. The report further adds that perpetrators rely on corrupt officials to perpetuate illegal activities, namely taking bribes, migrations and transportation, and engage in the trafficking chain (victim sourcing and exploitation) (UNDP, 2021).

Similarly, in the academic literature, while adoption of the UN’s SDGs has received increasing attention in recent years (Bose and Khan, 2022; Bose et al., 2024; Subramaniam et al., 2023), academic scholars have not yet explored any possible link between country-level SDGs’ performance and MS practices. This issue needs further attention and academic investigation because with the emergence of the UN’s SDGs and their subsequent adoption within different contexts globally (Bose and Khan, 2022; KPMG, 2020), it is essential to understand whether countries with better SDG performance can reduce MS practices than countries with low SDG performance scores. The motivation of the current study is also to address this gap.

Practically, understanding the above-mentioned issues have implications as well. To illustrate, in today’s stakeholder-driven business environment, if a country makes accountability and transparency practices a secondary priority, slavery practices will not only surge, but also will likely be negatively reacted to by stakeholders, namely local citizens, media and other stakeholders globally (UNDP, 2021). With the spread of globalisation and cross-border integration, and dependency of business activities, business firms have come under the 'global radar' where activities, particularly in developing countries, are closely monitored by global customers/stakeholders across their entire value chains. Foreign investors’ interest in, and positive attitudes towards local products are driven by country-level initiatives to curb slavery practices, reduce corruption practices and establish accountability by improving governance practices. Similarly, a country’s image towards international communities is largely established and enhanced when that country can reduce different non-economic performance indicators, such as perceived corruption, bribery practices, human rights violations, child and forced labour, and can improve workplace safety, accountability, and above all, law and governance. Empirical evidence on this pertinent issue is likely to inform regulators, governments and other actors in different contexts whether they need to take more pragmatic actions/strategies to reduce slavery practices. The findings are likely to provide additional input for a country’s regulators to take the necessary actions to reduce corruption practices, work with other international and national bodies and civil societies to monitor and take holistic action to reduce corruption, as well as slavery practices in different sectors in their local contexts. Stakeholders, such as media, civil societies and international audiences, are likely to monitor and suggest areas of slavery practices needing reduction, which a country’s powerful actors should prioritise. Additionally, the findings will have key take-home messages for different actors working with both MS practices and SDGs. Overall, the study aims to answer the following two key research questions:

Using a sample of 431 country-year observations from 146 countries, we examine the association between corruption and MS practices. We employ different proxies for measuring MS. We find a positive association between countries with higher corruption and MS practices. This finding suggests that when less accountability and transparency prevail in a country, it results in an increase in corruption and MS practices. We also examine the impact of SDG performance on MS practices. Our findings suggest that the SDG performance score has a positive impact on MS practices. These findings suggest that countries that actively embrace SDGs and improve their SDGs performance exhibit a more pronounced resilience towards the reduction of MS from that country. Our additional analysis finds that higher country-level debts serve as a moderator between corruption and MS practices. The findings suggest that corruption within highly indebted poor countries further accelerates and contributes to MS within those countries than in less indebted poor nations. Our results are robust using several sensitivity analyses, including alternative measures and instrumental variable analysis. The consistency of our findings across these robustness tests reinforces the validity and reliability of our conclusions.

The findings of the study contribute to the academic literature in several ways. First, to the best of our knowledge, the current study is the first that assesses the impact of country-level corruption on MS practices in different jurisdictions, thus offering new empirical evidence drawn from global data. While several earlier studies have been conducted and provided increased understanding of MS disclosure practices (e.g. Christ et al., 2020; Christ and Burrit, 2021; Islam and Van Staden, 2022), the contribution of our research is different from these studies in several ways. To illustrate, while previous studies contributed by focusing on MS issues using firm-level data collected from different nations, we examine country-level corruption practice scores and have explored their impact on country-level MS practices, thus filling a gap and contributing to the MS accounting and sustainability literature. Second, prior studies have contributed by examining MS practices within a single country context such as Australia, the UK or the USA (e.g. Birkey et al., 2018; Christ et al., 2020; Christ and Burrit, 2021; Islam and Van Staden, 2022; Ma et al., 2016). Yet there was a lack of focus in addressing MS practices at the global level. By considering global level data, our study also contributes to the accounting literature by embracing the interplay between MS and corruption and by empirically evidencing their associations across different jurisdictions. Third, our findings highlight the importance of the adoption of the UN’s SDGs by member nations in combating MS practices in various jurisdictions. The findings contribute to the larger debate on the reasons why a country’s government/regulators should consider adopting and implementing the SDGs as a priority basis and take the necessary initiatives to improve SDG performance. Ultimately, improvement of SDG performance in different countries will tend to reduce MS practices globally. This line of understanding is original in the SDGs literature.

Lastly, to the best of our knowledge, our study is the primary that not only develops an integrated theoretical model focusing on national-level corruption practices, national-level SDG performance and their efforts to alleviate MS practices but also provides empirical evidence on these issues relying on different jurisdictions’ data and multi-year data. While MS practices and SDGs have received increased academic attention in recent years (Bose et al., 2024 for details), in the sustainability and accounting field, to date, developing a single theoretical framework outlining a possible nexus between these emerging issues (e.g. country-level corruption, country-level SDGs score and MS practices), and exploring the nexus between corruption and MS practices, SDGs and MS practices were surprisingly absent in the literature. By developing a theoretical model and providing empirical evidence, the study’s findings contribute to the accounting and sustainability literature and provide a fresh understanding relating to the topics under investigation. Our findings highlight the necessity of reducing corruption practices from different countries as well as the implementation of the UN’s SDGs in combating MS practices around the globe. This type of initiative is primary in the MS and sustainability literature.

Our findings have practical value as well. To illustrate, it is evident from our findings that regulators, other national and international actors, and development partners are required to work collectively to curb corruption and to improve SDG performance in different countries; these efforts are crucial as, without such efforts, slavery practices in different jurisdictions will not be reduced. Our empirical findings have provided a rich business case for different governments and other actors to take this action seriously. Otherwise, those activities (corruption and MS) will perpetuate, and the SDGs will merely be a political promise, resulting in no change in the status quo of MS practices (for detailed practical implications, see the conclusion section).

The remainder of this paper is structured as follows. Section 2 presents the conceptual underpinning of key variables, discusses the literature review and hypotheses development. Section 3 explains the methodology of the study. Section 4 presents the empirical results and the additional analysis and robustness checks. The final section (section 5) concludes the paper.

Corruption is the abuse of discretionary power by bureaucratic officials in any country who advance their own interests by engaging in unauthorised rent-seeking (speed-money) activities (Houqe and Monem, 2016). Following previous studies (e.g. Blackburn et al., 2006, 2010; Everett et al., 2007; Houqe and Monem, 2016; Shleifer and Vishny, 1993), we broadly define corruption as the use of public office for unauthorised private gain; however, one may also refer to corruption as a general perception in a society. In this paper, we focus on perceived corruption, instead of the actual level of corruption in a society, an approach in line with others (Houqe and Monem, 2016).

We adopt the definition of SDGs given by the UN (UN, 2022, 2015) and later used in many academic studies (see Bose and Khan, 2022; Bose et al., 2024 for details). Announced in 2015 by the United Nations, the SDGs is treated as the global goals (through 17 specific goals) and a universal call to take action to end poverty and hunger, reduce inequalities, protect the planet and ensure that all people enjoy peace and prosperity by achieving these goals by 2030 (UN, 2022, 2015). Given the SDGs are integrated and well connected, it is argued that action in one area will affect outcomes in others, and that development must balance economic, social and environmental sustainability (Bose et al., 2024; UN, 2015). The SDGs are designed to end poverty and hunger, establish human rights, education and climate action, and reduce discrimination against women and girls (UN, 2015, 2017, 2022). The United Nations Brundtland Commission (1987) defines sustainability and sustainable development as the development of our current way of living, meeting our present needs and ensuring our better tomorrow, without compromising the chances of future generations to meet their future needs (UN, 1987, 2023). The UN (2023) in their recent report argued that unsustainable development happens when people pursue immediate rewards without thinking about harms to other people or the planet. Additionally, when people and organisations often prioritise short-term gains over longer-term costs, unsustainable development gets more deteriorated. For example, when an entire forest is chopped down to turn a quick profit, this action brings diverse bandwagon effects, namely causing an ecosystem collapse, endangering species to die off, and leaving local communities at permanent risk of devastating floods and other natural disasters (UN, 2023).

Following the Global Slavery Index (GSI) report published by the Walk Free, we define MS as “situations of exploitation that a person cannot refuse because of threats, violence, coercion, deception, abuse of power” and “is a hidden crime that affects every country in the world” (Walkfree, 2018, p. 7). The index, regulargly publish by the human rights organisation (Walk Free), and uses survey data, statistical modelling and other metrics to measure the prevalence of MS in different countries of the world (Walk Free, 2018).

Although limited, previous accounting studies have studied corruption and different accounting issues (EL-Helaly et al., 2020; Houqe and Monem, 2016; Kimbro, 2002; Wu, 2005). Within limited studies, the issues addressed include examinations of corruption and International Financial Reporting Standards (IFRS) adoption (EL-Helaly et al., 2020; Houqe and Monem, 2016); the development of a theoretical framework on the relationship between accounting and the fight against corruption (Everett et al., 2007); the role of accounting quality in corruption perception (Kimbro, 2002; Malagueño et al., 2010); and exploring the relationship between firm-level accounting practices on bribery incidence using samples from twelve Asian countries (Wu, 2005).

In terms of investigating historical forms of slavery, accounting literature is progressive with past studies providing related evidence from plantations and slave traders (Pinto and West, 2017; Rodrigues and Craig, 2018; Tyson et al., 2004). Recent developments in this line of literature flow from two key streams of research. The first stream of studies addresses the accounting practices of legitimate businesses, outlining general human rights issues without specific attention to addressing MS practices (Hofmann et al., 2018; O’Brien and Dhanarajan, 2016). O’Brien and Dhanarajan (2016), for example, explore recent tendencies toward governmental directions on MS practices, and adoption of human rights due diligence practices by business decision-makers in different countries (see also Hofmann et al., 2018). The second stream of studies emerge in the literature specifically to examine firm-level MS reporting, particularly in response to readiness for transparency in supply chain legislation (Birkey et al., 2018; Christ et al., 2019), or regulation in other parts of the world. Notable studies within this latter stream have been conducted in various contexts, including the UK (Islam and Van Staden, 2022), Australia (Christ et al., 2020; Christ and Burrit, 2021), the US (Birkey et al., 2018; Ma et al., 2016) or in a conceptual study (Christ et al., 2023). These studies have increased our understanding of multiple issues. These include businesses’ responses to mandatory MS regulation in the Australian context (Christ and Burritt, 2021); stakeholders’ narratives on the Modern Slavery Act in the UK (Islam and Van Staden, 2022); the due diligence mechanism and assurance process in the MS report and the role of NGOs (Islam and van Staden, 2018); the effectiveness of a social audit in the MS reduction process (Islam et al., 2018), the extent of compliance by US firms on the California Transparency Act (Birkey et al., 2018; Ma et al., 2016) and readability of MS disclosure information and its impact on firm value in the context of Australia (Saha et al., 2024). Nevertheless, one of the key limitations of the previous studies is, so far, previous studies have given limited attention to explore the impact of country-level corruption on MS practices. The exploration of this area represents a virgin territory within the broader context of sustainability accounting research. As mentioned previously, exploring a possible nexus between country-level corruption and country-level MS practices is important not only for advancing theoretical contribution, but also for validating national and international stakeholders’ claims that slavery and other forms of modern exploitations are driven by country-level corruption (UNDP, 2021).

With respect to SDGs research in the accounting and sustainability literature, there have also been some recent developments by accounting researchers exploring multiple issues (Bose et al., 2024; Bose and Khan, 2022; Lenort et al., 2023; Pizzi et al., 2021; Rosati and Faria, 2019a, b). These include country-level institutional factors (e.g. politics and law, economics and finance, society and culture, technology and innovation, education and labour, and sustainability) influencing SDG adoption (Rosati and Faria, 2019a); organisational factors and the role of SDG disclosure (Rosati and Faria, 2019b); SDG adoption by the worldwide automotive industry (Lenort et al., 2023); country-level institutional factors influencing SDG disclosure (Bose and Khan, 2022); determinants of SDG disclosure using data from European countries (Hummel and Szekely, 2022); and determinants and consequence of SDG disclosure by global firms (Bose et al., 2024). Some studies look at SDG adoption using single country data such as Japan (Ike et al., 2019), Italy (Pizzi et al., 2021) or Australia (Subramaniam et al., 2023).

One of the key drawbacks in previous SDGs studies is the absence of examination of the role of SDGs performance in reducing MS practices in different countries. In other words, exploring the impact of country-level SDG performance on MS practices in different countries is to be established empirically. Specifically, to date, there has been a limited understanding of whether the SDGs performance of a country reduces MS practices in that country. The motivation of the current study is to fill this gap.

It has been argued that corruption in different jurisdictions and MS is interconnected (OECD, 2016; UNDP, 2021; Verité, 2013). According to a report by the UNDP (2021), in the modern era, slavery and other forms of human rights exploitation in different countries are facilitated by corruption and a lack of accountability practices in different contexts. It is of no secret that higher levels of perceived corruption practices go hand in hand with poorer responses to human trafficking and other forms of exploitation for poor and powerless people in society (ILO, 2016; OECD, 2016; Verité, 2013). There are several underlying arguments indicating that country-level corruption practices increase MS practices in different ways. First, women and children, marginalised and poor citizens, are considered the most powerless people in society and continue to be treated as either unpaid workers (domestic tasks and childcare) or as part of the underpaid workforce (Peters, 2018; UNODC, 2019). Rampant corruption practices in each country give rise to inequalities, human rights violations for society’s powerless people (Freedom Collaborative, 2020; UNDP, 2021). This group of powerless people (e.g. marginalised and poor citizens) commonly suffer higher unemployment rates and are often deprived of access to other basic citizens’ rights (Peters, 2018; UNODC, 2019). Second, marginalised and poor people are targeted for bribes and/or sexual acts in exchange for employment, further impairing their ability to access basic public services, which gives rise to social inequality (Freedom Collaborative, 2020; UNDP, 2021). Given that poor and marginalised citizens are often unable to pay bribes, they are often denied receiving a public service (e.g. health care, education or other services) (Boersma, 2012; Figueiredo, 2017). These discriminatory and inequitable practices create more vulnerability for wide range of marginalised group (Boersma, 2012; Figueiredo, 2017; Peters, 2018; Figueiredo, 2017).

According to the UNDP (2021), corruption in a country is a key factor for fuelling a vicious cycle that creates further opportunities for violation of human rights and the establishment of inequality in society. The underlying argument is that different forms of MS in a country cannot take place on a large, sustained, profitable scale without corruption rooted in the context (Figueiredo, 2017; Peters, 2018). To illustrate, in countries with highly corrupt practices, institutional mechanisms to assist victims, such as reporting physical and/or sexual assault to the police or seeking health-related assistance, are often non-existent, inaccessible and dysfunctional (UNDP, 2021). Likewise, lack of accountability and pervasive corrupt practices in different jurisdictions will foster countries’ governments to become indifferent to mitigating slavery practices (Peters, 2018; UNDP, 2021). Several practitioners’ reports have argued that corruption in a country enables slavery and other forms of exploitation practices, such as facilitating payments to gain government approval, use of forced labour, low wages or maintaining poor working conditions (Freedom Collaborative, 2020; Liberty Asia and Freedom Fund, 2015).

Overall, from the above discussion, we argue that countries that involve themselves in unethical, discriminatory and unfair practices and have established records of unaccountability and lack of transparency practices through different corrupt practices, will be unresponsive to addressing MS practices seriously. These countries will fail to take robust initiatives and efforts within their own jurisdictions to eliminate corruption from their day-to-day activities. Subsequently, at the country-level, they will maintain not only dysfunctional and/or poor institutional structures for reducing slavery practices but also will create an ineffective culture of mitigating different modern forms of slavery practices (Peters, 2018; UNDP, 2021). From the above discussion, we have a priori reasoning that country-level corruption practices will increase MS practices in the country. Formally, we hypothesise:

In addition to the above hypothesis (H₁), in the current study, it is hypothesised that when country-level SDGs performance increases, MS practices of that country are likely to be lessened. In other words, country-level SDG performance will negatively influence MS practices in different countries. We offer several reasons for the arguments. To illustrate, it is obvious that without mitigating forced labour, child labour, and other forms of slavery practices, the UN’s aspirations of the SDGs will be impossible to achieve (UNDP, 2021; UN, 2022, 2023). It is important to mention that four of the UN’s SDGs specifically outline slavery and inequality issues (SDGs 5, 8, 10 and 16). To illustrate, SDG 5 “Gender equality” advocates for a reduction of gender inequality and posits that inequality directly influences MS risks. This SDG also argues that government efforts towards achieving this goal will naturally tend towards tangible efforts to reduce slavery practices (UN, 2017, 2022). In a similar vein, SDG 8 “Decent work and economic growth,” explicitly states that countries need to take “immediate and effective measures to eradicate forced labour, end MS and human trafficking and secure the prohibition and elimination of the worst forms of child labour, … by 2030” (UNDP, 2021, p. 15). Additionally, SDG 10 “Reduce inequality within and among countries” unpacks that all types of inequalities based on income, sex, age, disability, sexual orientation, race, class, ethnicity, religions and others are to be eradicated because inequality in a society not only threatens long-term social and economic development and obstructs poverty reduction, but also abolishes people’s sense of accomplishment and self-worth (UNDP, 2021). Lastly, SDG 16 “Peace, justice and strong institutions” fosters providing peace and justice for all citizens in a society, and strong institutions by establishing accountability and inclusive societies at all levels. All these elements of SDG 16 link with MS because in a context characterised by peace, justice and institutional transparency, there will be an observable reduction in MS practices (UN, 2023, 2022). Overall, the UN (2022) stresses that achieving the SDGs is impossible without establishing peace, stability, human rights and effective governance and, in effect, an accountability practice in different jurisdictions across the globe. Given that the UN’s goals promote the rule of law at the national and international levels and ensure equal access to justice for all citizens (UN, 2023), it is obvious that the achievement of these goals will also reduce MS practices. To illustrate, SDG 8, which promotes sustained, inclusive workplaces and sustainable economic growth, encompassing full and productive employment and decent work for all, is fundamentally linked to the issue of MS (https://sdgresources.relx.com/tags/modern-slavery, 2023). Likewise, SDG 16 includes the creation of legal frameworks that criminalise all forms of exploitation and strengthening law enforcement responses to these crimes (UN, 2017).

Furthermore, MS practices of a country are likely to be reduced when a country’s SDG performance is improved because when the national-level SDG performance score is calculated by the UN’s SDGs committee (specifically, the Sustainable Development Solutions Network-SDSN), they examine the performance of other key variables. These include the national-level SDG monitoring mechanism, taking into account of all stakeholders interests and government ongoing efforts towards achieving related social indicators namely reducing gender inequality, social inequality and discrimination, eradicating forced labour, reducing human trafficking, as well as access to justice by poor and marginalised people (Sachs et al., 2019; UN, 2017). Overall, a coordinated effort from a country’s government is required to fight against MS practices, as well as for the improvement of SDG performance (Sachs et al., 2019; UN, 2017). It is argued that country-level SDG performance will improve when it accommodates the diverse interests of stakeholders, including those of minority, marginalised and powerless stakeholders (Bose and Khan, 2022; Bose et al., 2024; Freeman, 2010). Otherwise, a country would not be able to secure, improve and progress towards better SDGs performance (Bose et al., 2024). Given that combating MS is critical, not only for establishing human rights but also for achieving sustainable development (UN, 2022), the lack of initiative by national regulators and other powerful local actors to reduce MS practices is directly detrimental and, in effect, contradicts the core principles of the SDGs. As a result, when a country achieves a higher SDG performance score, it is obvious that the country’s government has taken all-inclusive measures to reduce MS practices.

Overall, from the above discussion, we reason that country-level MS practices will be reduced depending on the home country’s SDG performance. When a country obtains a higher SDG performance score, this indicates that both public and private level initiatives and endeavours in a country to meet the SDGs are promising and satisfactory. It also reveals that a country’s government has improved different performance indicators directly associated with the slavery practices specified in the different SDGs (SDGs 5, 8, 10 and 16). Unquestionably, countries with higher SDGs performance will have greater endeavours to reduce inequality, establish human rights, improve workplace safety, work to improve gender inequality, above all, work to eliminate child labour in that country compared with countries with low SDG performance scores. MS practices in those countries are therefore expected to be reduced because of improving country-level SDG performance. Formally stated:

To test the study’s hypotheses, we collect data from several sources. To perform this analysis, our primary annual observation was 214 countries for the years 2016, 2018 and 2023. First, we excluded 67 countries whose MS score is not available on www.globalslaveryindex.org. Then, we excluded one country whose corruption score is not available in Kaufmann and Kraay (2023). After these exclusions, our final sample comprises 146 countries around the globe. Data for MS scores are collected from the global slavery database (www.globalslaveryindex.org). We select these years because our research variable of interest (e.g. country-level MS score) is available only for 2016, 2018 and 2023. Data on SDGs are collected from https://sdgs.un.org/goals. Data relating to the control variable shareholder versus stakeholder is collected from Dhaliwal et al. (2012). Finally, other control variables, such as gross domestic product, freedom of information, foreign direct investment, merchandise trade, mandatory MS regulation and highly indebted poor countries, are collected from the World Bank (www.worldbank.org).

The following model is proposed to test the study hypotheses:

Where $MSit$ is the MS score (dependent variable) for country i in year t. $CORRUPit$ is the control of corruption for a country i in year t. SDG_it is the SDGs for country i in year t. Controls_it refers to country-level control variables. We use year fixed effects to control for macroeconomic trends that may affect MS scores.

Our dependent variable is the MS score developed by Walk Free (www.globalslaveryindex.org) as part of the worldwide tracking of slavery practices. The global MS index is affected by a complex interaction of multiple factors related to the presence or absence of protection and respect for rights, physical safety and security, access to the necessities of life such as food, water and health care, and patterns of migration, displacement and conflict. Statistical testing has grouped 24 measures of vulnerability into four dimensions covering: (1) civil and political protection, (2) social health and economic rights, (3) personal security, and (4) refugee populations and conflict. It ranges from 0 to 100. Higher values indicate higher slavery and vice versa.

Our first independent variable is the control of corruption index developed by Kaufmann and Kraay (2023) as part of the Worldwide Governance Indicators (WGI) project. Under this project, data from 30 different sources are combined from six aggregate governance indicators namely, voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, rule of law and control of corruption. Aggregate governance indicators are developed by using the unobserved components model (Kaufmann and Kraay, 2023). The aggregate governance indicators are reported in two ways: (1) standard normal units ranging from −2.5 to 2.5 and (2) percentile rank ranging from 0 (lowest) to 100 (highest) among all countries globally (Kaufmann and Kraay, 2023). In this study, we use the control of corruption index based on standard normal units. Control of corruption measures how public authority is utilised for personal benefit, encompassing various levels of corruption, ranging from minor to significant instances. This evaluation also considers the influence of elites and private interests in manipulating or co-opting the state system (Kaufmann and Kraay, 2023). For simplicity, we multiply (−1) by the control of corruption index. Higher values indicate higher corruption and vice versa.

Our second independent variable is SDGs performance scores developed and published by the UN. The current study’s SDGs performance index comprised 17 items set by the UN, assigning equal weight to each goal, a practice in line with others (e.g. Sachs et al., 2023). It ranges from 0 (worst) to 100 (target) using multiple global indicators. Higher values indicate higher SDGs achieved and vice versa. To calculate the SDGs score, the UN collects data both from official statistics (two-thirds from their custodian agencies) and non-traditional sources (one-third from research centres, universities and non-governmental organisations). Peer-reviewed since 2015, the global edition of this performance data also underwent statistical auditing by the European Commission in 2019 ^[3].

We also consider the number of control variables in the study. The control variables employed in the tests have been found in the literature to correlate with the dependent variable. Our measure of the stakeholder versus stockholder (STAKE) is sourced from Dhaliwal et al. (2012). The justification for using this measure within our sample period is that governance frameworks and corporate social responsibility are primarily influenced by country-level macro governance, i.e. legal cultural, and institutional environments of individual jurisdiction. Using this proxy ensures methodological uniformity and comparability with previous literature. Moreover, country-level governance structure will not change overnight, making this measure a valid proxy for our test models (Houqe et al., 2012). It takes the value of 1 if the country is a stakeholder country and 0 otherwise. Economic development (GDP) is measured by the natural logarithm of gross domestic product per capita divided by midyear population (Corrigan, 2014; Jović et al., 2016; Papyrakis et al., 2017; Pitlik et al., 2010). The freedom of information (FOI) measures the extent to which a country’s citizens can participate in the selection of their government, as well as enjoy freedom of expression, freedom of association, and free media. It ranges from −2.50 to 2.50. For simplicity, we multiply (−1) with the FOI index, with higher scores indicating greater participation and freedom and vice versa (www.worldbank.org). The foreign direct investment (FDI) is measured by the natural logarithm of foreign direct investment in a country in a year. The merchandise trade (EXPORT) is the balance of merchandise exports and imports in a country in a year. The mandatory MS regulation is measured by the dummy variable that equals 1 if a country has devised regulations for firms to report whether they (the firms) comply with human rights in their business operations and else 0. Our choice of these control variables is motivated by strong evidence that stakeholder versus shareholder orientation, economic development, FOI, FDI, merchandise trade and mandatory MS regulations of a country are related to the level of slavery (Islam and van Staden, 2018). Consistent with previous literature, we expect a country’s slavery practices scoring to be negatively associated with STAKE, GDP, FOI, FDI and MAN. In contrast, we expect a country’s slavery practices score to be positively associated with EXPORT (see Table 1).

Table 3 presents the descriptive statistics of the sampled countries’ mean values of the variables. The MS scores (MS) range in our study from 6.31 (lowest MS scores in Denmark) to 80.11 (highest slavery scores in Congo). As Table 2 shows, 70 countries (48.28%) have a higher-than-average score of MS (30.6761), suggesting that slavery practices are present around the globe. The corruption scores (CORRUP) range in our study from −2.20 (least corrupt) to 1.44 (most corrupt). Table 2 shows that 85 (58.62%) countries have a higher-than-average score of corruption (0.1100), indicating that most of the countries around the globe are fighting to reduce corruption. The SDG mean score is 66.2949, which suggests that 60 countries (41.37%) have a below-average SDG score. The mean value of STAKE is 0.6756 i.e. 98 countries in our study sample are stakeholder orientation countries. The mean value of the natural logarithm of gross domestic product per capita is $4.0948 i.e. 89 (61.37%) countries have a lower-than-average gross domestic product per capita. The freedom of information (FOI) score ranges in our study from −1.71 (highest freedom of information in Norway) to 1.2 (lowest freedom of information in Lao PDR). Table 2 shows that 72 (49.65%) countries have below-average freedom of information scores. In Table 3, the FDI and EXPORT mean and median values are −0.3020 (−0.0545) and 9.8022 (9.8779), respectively, as consistent with our expectations. Finally, mandatory MS regulations of a country mean value is 0.0483. Seven countries in our sample are mandatory MS regulations countries. These values are consistent with evidence in the extant literature (Moses et al., 2021).

Table 4 presents Pearson’s correlation matrix for the study sample. Among all variables, MS has the strongest positive correlation with corruption (CPORRUP) (r = 0.4437). On the other hand, MS has the strongest negative correlation with sustainable development goals (SDG) (r = −0.4718), followed by gross domestic product per capita (GDP) (r = −0.4413), freedom of information (FOI) (r = −0.4168), merchandise trade (EXPORT) (r = −0.3035); and mandatory MS regulation (MAN) (r = −0.1740); all significant at the 1% level. In general, the correlation matrix does not show any potential serious multicollinearity problems.

The regression results are presented in Table 5. Panel A contains the results for the contemporaneous effect of corruption and MS. The coefficient of CORRUP is positive and highly significant (Coefficient = 4.2412, p < 0.01), indicating that countries with higher corruption are associated with higher MS. Put differently, the results imply that corruption increases the MS of the country and thereby supports our H₁. In terms of economic magnitude, as in Model (1), a one standard deviation increase in corruption (1.0318) is associated with a 4.37 (=1.0308 × 4.2412) percentage point increase in MS. On the other hand, the coefficient of SDG is negative and highly significant (Coefficient = − 0.3469, p < 0.01), indicating that countries with lower SDG scores are associated with higher MS and thereby support H₂. In terms of economic significance, as in Model (1), a one standard deviation increase in SDGs (10.1786) is associated with a 3.53 (=10.1786 × −0.3469) percentage point decrease in MS. Among the control variables, STAKE (Coefficient = −2.1855, p < 0.05) and FOI (Coefficient = −3.6063, p < 0.01) are negatively associated with MS. These results are expected and consistent with the extant literature.

Table 5 Panel B contains the results for the lagged effect of corruption and MS. The coefficient of CORRUP is positive and highly significant (Coefficient = 2.5924, p < 0.01), indicating that countries with higher corruption are associated with higher MS. Similarly, SDG is negative and significant (Coefficient = −0.3690, p < 0.01), highlighting that lower SDG scores are associated with higher MS. In terms of economic significance, as in Model (1), a one standard deviation increase in corruption (1.0308) is associated with a 2.67 (=1.0308 × 2.5924) percentage point increase in MS. Similarly, a one standard deviation decrease in SDGs (10.1786) is associated with a 3.76 (=10.1786 × −0.3690) percentage point increase in MS. Among the control variables, GDP and FOI are negatively associated with MS. These results are consistent with the extant literature and our main results.

Table 6 reports the additional analysis of estimating the relationship between corruption and MS. Table 6, column (1) considers the 2023 sample only. The coefficient of CORRUP is positive and statistically significant (Coefficient = 2.0276, p < 0.01), indicating that countries with higher corruption are associated with higher MS. Similarly, columns (2) and (3) consider the 2018 and 2016 samples, respectively. The coefficient of CORRUP is positive and statistically significant (Coefficient = 3.4205, p < 0.01; Coefficient = 4.1610, p < 0.01), indicating that countries with higher corruption are associated with higher MS. In economic terms, for the 2023, 2018 and 2016 samples, a one standard deviation increase in corruption (1.0308) is associated with a 2.09, 3.53 and 4.29% point increase in MS ^[4]. Table 6 column (3) presents the findings using another alternative estimation of MS (i.e. the number of people in MS in a country) and the association between corruption and MS. The estimated coefficient for CORRUP is statistically significant and positive (Coefficient = 0.4043, p < 0.01) at the conventional levels. Similarly, a one standard deviation increase in corruption (1.0308) is associated with a 0.42% point increase in MS ^[5]. We observe similar results for SDG as negative and significant for the 2023, 2018 and 2016 sample respectively (Coefficient = −0.4321, p < 0.01; Coefficient = −0.7634, p < 0.01; Coefficient = −0.5096, p < 0.01), and alternative measure of MS (Coefficient = −0.0156, p < 0.01). In terms of economic significance, for 2023, 2018, 2016 and the alternative estimation of MS, a one standard deviation decrease in SDGs (10.1786) is associated with a 4.40, 7.77, 5.19 and 0.16% point increase in MS ^[6].

Table 6 column (5) considers the OECD country sample only. The coefficient of CORRUP (Coefficient = 3.5865, p > 0.10) is not significant. The possible explanation for this finding is that in OECD countries more robust institutions and better enforcement of laws might mitigate the direct impact of corruption on MS. On the other hand, column (6) considers the non-OECD country sample only. The coefficient of CORRUP (Coefficient = 2.7333, p < 0.01) is negative and significant, i.e. highly corrupt countries, official statistics on MS are underreported, or corrupt officials might be involved in or tolerate human trafficking activities, influencing the recorded levels of MS. In economic terms, for the non-OECD sample, a one standard deviation increase in corruption (1.0308) is associated with a 2.81% point increase in MS ^[7]. Finally, column (7) shows an analysis of the relationship between corruption and MS by incorporating the Religious Diversity Index (RDI) (www.pewresearch.org). The index ranges from 0 to 9. Higher scores indicate higher religious diversity and vice versa. The coefficient of CORRUP and SDG is negative and significant. The coefficient of RDI (Coefficient = −0.6234, p < 0.01) is negative and significant, i.e. countries with low religious diversity might have low social cohesion, potentially leading to weaker collective action against practices like MS. We observe similar results for SDG as negative and significant for the non-OECD sample (Coefficient = −0.3576, p < 0.01) and incorporation in RDI index (Coefficient = −0.3986, p < 0.01). In terms of economic significance, for the non-OECD sample and incorporation of the RDI index, a one standard deviation decrease in SDGs (10.1786) is associated with a 3.64 and 4.06% point increase in MS ^[8]. Altogether, we document the evidence that country-level corruption increases MS and MS decreases due to SDG achievement of a country, which is consistent with our prediction of Hypotheses 1 and 2.

The consensus regarding heavily indebted poor countries (HIPC) is that the highest levels of poverty are associated with a higher level of MS. We therefore expect that heavily indebted poor countries will have higher levels of corruption. Table 7 shows the association between corruption and MS on highly indebted poor countries. To maintain consistency, we classify countries based on the International Monetary Fund (IMF) classification of the highest levels of poverty. A dummy variable that equals 1 is applied if a country is classified as a heavily indebted poor country and otherwise, 0. The key variable of interest is CORRUP × HIPC. We find that the coefficient of CORRUP × HIPC is positive and significant, implying that corruption within highly indebted poor countries contributes to an increase in prevalence of MS within those countries.

Following Cheng et al. (2014), we use inequality (INEQUALITY) as an instrument. The instrument satisfies the relevance and exclusion conditions, as it is most likely that inequality is positively associated with corruption (Moses et al., 2021). More importantly, the instrument will influence MS through the corruption channel only. Table 8 shows the results using this instrumental variable approach. Model (1) reports the first stage, where we find that the instrument is positively and significantly associated with CORRUP (Coefficient = 0.0329, p ≤ 0.01). In Model (2), we report the second-stage regression results of MS. The key variable of interest is the instrumented CORRUP (predicted corruption, PRED_CORRUP), and we find that the estimated coefficient of PRED_CORRUP and SDG is statistically significant as expected (Coefficient = 5.1149, p ≤ 0.01; Coefficient = −0.6019, p ≤ 0.01). These results further support our claim that corruption increases MS and decreases SDG achievement of the country.

We check the validity of our instrument in Table 8. First, the under-identification test, or Kleibergen-Paap rk LM statistic, reveals that the model is identified (p ≤ 0.01). Second, the result of the weak identification test or Kleibergen-Paap rk Wald F statistic (Kleibergen and Paap, 2006) shows that the F-statistic is very high in our sample (32.1296), which suggests our instrument is relevant and strong. Third, the Sargan statistic or the Hansen’s J statistic (Hansen, 1982) is used to check the overidentification concern. The p-value of Hansen’s J statistic is high for the model (p ≤ 0.01), suggesting that we cannot reject the null hypothesis that the instruments are exogenous. This means that other sources of endogeneity, such as omitted variables or simultaneity issues, are unlikely in the model. Overall, our post-estimation tests confirm both the relevance and the exclusion restrictions of our instruments.

The study examines the relationship between corruption and MS practices, taking country-level data as a sample. This study also examines the relationship between SDGs performance in different countries and MS practices. The study finds a positive association between country-level corruption and MS practices by the sample country. We also find that countries with higher SDGs performance practices are negatively associated with MS practices. Specifically, MS practices are lessened when country-level SDGs performance is higher. Our additional analysis reveals that higher country-level debt serves as a moderator/channel between corruption and MS practices. The findings of our research have both academic (discussed in the introduction section) as well as practical implications.

Practically, the findings have value in many ways. First, the findings contribute to the key debate on the reasons why countries’ policy makers should consider improving the transparency and accountability issues that are essential to combat MS practices. Corruption is a precursor for lack of transparency and accountability, and it is revealed that owing to corrupt practices at the country-level, MS is duly injected in the context. The findings thus establish and confirm that in modern days, only enacting mandatory regulation of MS is not sufficient for global leaders and national-level policymakers to curb such human rights violations and slavery practices. To receive benefits from mandatory laws for slavery practices, it is necessary for country-level policymakers to ensure transparency and greater accountability by reducing country-level corruption. Our findings reiterate that without reducing corruption, it is not possible to reduce MS practices in different countries. Second, as revealed in the study, the adoption and implementation of SDG practices reduce slavery practices. Regulators and governments in different jurisdictions should therefore monitor not only the corruption but also oversee to what extent the country is prepared and positioned to implement SDGs and its future commitment to achieving different SDGs. In association with multi-stakeholders’ engagement at the national level, they should also exert efforts to improve SDG performance continuously. Otherwise, the anti-slavery practices adopted in the given country will not be reduced and will not add any value to the local stakeholders, citizens, community people or civil societies, above all, international communities.

Additionally, it is appropriate to add that other stakeholders such as private sector managers, media and civil societies are required to perform their role, respectively. For example, managers in different firms can not only develop “no bribe and hush money” policy in their normal business transactions, but also can institutionalise “zero inequalities and discriminations” policies and practices within their organisations and should not tolerate any form of discriminatory and biased practices among marginalised employees. Similarly, the press and civil society's role in disclosing reports of corrupt practices within a country, or practices related to severe human rights violations, are inevitable. Their investigative reports and protests will assist governments in different countries to take actions against corruption and MS duly. Similarly, their informative report is likely to attract not only in the national media but also garner international attention. Overall, in association with multi-stakeholders’ contributions to eradicate corruption and contemporary forms of slavery, and their assistance to national governments towards achieving better SDGs performance, MS risk is likely to be reduced in the future from different nations.

Lastly, the findings of the study are applicable to international organisations such as the UN, World Bank, or others. To illustrate, these transnational organisations should develop a global monitoring and reporting system, particularly by focussing on corruption and its’ bandwagon effect in contemporary forms of slavery. This may take place in different forms, including publishing statistics about the number of related complaints, investigations and convictions relating to corruption practices, trends of MS practices over time, yearly budget allocations for reducing MS practices and SDGs implementation and developing multi-stakeholders’ engagement mechanisms. Without initiating a multi-stakeholders dialogue (government, civil societies, media, private sectors and NGOs) and/or consultations among other international organisations to explore and develop collective efforts to abate corruption and contemporary forms of slavery, such MS risk is unlikely to be reduced from the globe.

Despite the above implications, we acknowledge some of the limitations of the current study. First, the study investigates the impact of corruption on MS only in the context of selected countries (not all countries’ data were available), so findings of the study should not be applied beyond the sample contexts. Second, the findings of the study apply to MS practices using only three years of data. Future studies should capture a larger sample with more longitudinal time at the global level. Furthermore, the study has relied on secondary data and has established how country-level corruption and SDGs performance influence MS practices using econometric analysis. At the country level, we encourage future research using other methods of data collection. These may include interviews, large-scale surveys or a case study approach (taking cases from multiple countries), to explore the roles of different policymakers, regulators, and other stakeholder groups, including anti-slavery activists, journalists, suppliers, and workers. These efforts will open up an in-depth understanding of different actors’ roles towards reducing country-level corruption practices, SDG performance and MS practices in both developed and developing countries’ settings. Lastly, it appears that with this current divided world, the UN has not been able to mediate and stop wars and conflicts in different countries, particularly during recent years. It appears that power and business motives tend to replace humanity, justice and peace in different parts of the world. As a result, this conflicting situation, and social political instability create more inequalities, human rights violations, lack of justice, lack of education, food crisis, lack of accomodations and shelter, all these give rise to famine conditions in several countries. Future studies should examine the extent to which powerful nations accelerate and perpetuate slavery practices, violate human rights, compromise the standards of humanity, and sideline SDGs goals for their economic benefit, and what they do to create a solid trade-off between SDGs and MS practices.