Transnational Repression: International Cooperation in Silencing Dissent

Abstract

Why do some states assist other countries to reach across national borders and repress their diaspora, while others do not? Transnational repression involves host countries (including democracies) working closely with origin states (typically autocracies) to transfer their citizens living abroad into their custody and silence dissent. We expect international cooperation on transnational repression to rely on a host country’s domestic rule of law (opportunity to repress) and economic ties with the origin country (leverage to cooperate). To measure international cooperation on transnational repression, we present new data containing 608 direct physical cases of transnational repression from 2014 to 2020 involving 160 unique country dyads (79 host countries and 31 origin countries). We test our hypotheses using a dataset of 33,615 directed dyad-years that accounts for refugee flows between pairs of countries and find empirical support for our theoretical argument. Autocracies are better able to elicit cooperation on human rights violations from states that have shared economic interests and a weak rule of law. Our findings provide one of the first quantitative accounts of foreign complicity in extraterritorial repression and have policy implications for civil society activists that seek to prevent governments from committing future human rights abuses against foreign nationals living abroad.

¿Por qué algunos Estados ayudan a otros países a cruzar las fronteras nacionales y reprimir a su diáspora, mientras que otros no? La represión transnacional implica que los países de acogida (incluyendo las democracias) trabajen en estrecha colaboración con los Estados de origen (normalmente autocracias) con el fin de transferir a sus ciudadanos que viven en el extranjero de vuelta a su custodia y silenciar la disidencia. Se espera que la cooperación internacional en materia de represión transnacional esté basada en el Estado de derecho interno del país anfitrión (oportunidad de reprimir) y en los vínculos económicos con el país de origen (influencia para cooperar). Presentamos, con el fin de medir la cooperación internacional en materia de represión transnacional, nuevos datos que contienen 608 casos físicos directos de represión transnacional entre 2014 y 2020, los cuales incluyen 160 díadas de países únicos (79 países de acogida y 31 países de origen). Probamos nuestras hipótesis utilizando un conjunto de datos formado por 33.615 díadas-años dirigidas que registran los flujos de refugiados entre pares de países y, a través del cual, encontramos apoyo empírico para nuestro argumento teórico. Las autocracias están en mejores condiciones de obtener cooperación en materia de violaciones de derechos humanos por parte de Estados que comparten intereses económicos y un Estado de derecho débil. Nuestras conclusiones proporcionan uno de los primeros relatos cuantitativos en materia de la complicidad extranjera en la represión extraterritorial y tienen implicaciones políticas para aquellos activistas de la sociedad civil que buscan evitar que los Gobiernos cometan futuros abusos contra los derechos humanos de los ciudadanos extranjeros que viven fuera de sus fronteras.

Pourquoi certains États aident-ils d'autres pays à traverser les frontières nationales pour réprimer leur diaspora, tandis que d'autres non ? La répression transnationale implique une collaboration étroite entre les pays hôtes (notamment des démocraties) et les pays d'origine (généralement, des autocraties) pour leur remettre leurs citoyens vivant à l’étranger et faire taire les contestations. Nous postulons que la coopération internationale en matière de répression transnationale se fonde sur la loi nationale du pays hôte (possibilité de répression) et les liens économiques avec le pays d'origine (leviers de coopération). Pour mesurer la coopération internationale sur la répression transnationale, nous présentons de nouvelles données contenant 608 cas physiques directs de répression transnationale entre 2014 et 2020, impliquant 160 dyades de pays différentes (79 pays hôtes et 31 pays d'origine). Nous testons nos hypothèses à l'aide d'un ensemble de données de 33 615 années de dyades à sens unique, qui recense les flux de réfugiés entre paires de pays. Nous trouvons ainsi des éléments pour venir étayer notre argument théorique. Les autocraties sont davantage en mesure d'obtenir la coopération d’États avec lesquels elles partagent des intérêts économiques et à l’État de droit faible sur des questions de violations des droits de l'homme. Nos conclusions proposent l'une des premières représentations quantitatives de complicité étrangère en matière de répression extraterritoriale. Elles s'accompagnent d'implications politiques pour les militants de la société civile qui souhaitent empêcher les gouvernements de commettre de nouvelles violations des droits de l'homme à l’égard des ressortissants étrangers vivant en dehors de leurs frontières.

Introduction

In January 2016, Li Xin, a pro-democracy Chinese journalist, disappeared in Thailand after boarding a train from Bangkok to Laos to renew his visa. Prior to his disappearance, Li had posted articles online accusing China of censoring its journalists and forcing him to spy on his peers. Weeks later, Li reappeared in Chinese police custody, where he claimed to have returned voluntarily and was under investigation, though he did not know where he was, and his family suspects that the Chinese government had influenced his statements (Iyengar 2016). In October 2016, Joshua Wong, a prominent Hong Kong student activist and leader of the 2014 pro-democracy protests and Umbrella movement, was detained upon arrival at Bangkok Airport while on his way to address student activists at a Thai university event. Joshua was held in detention at the airport for 12 hours by Thai authorities before being deported to Hong Kong, allegedly without explanation and at the Chinese government’s request (Phillips and Malkin 2016). In November 2015, Jiang Yefei, a Chinese political cartoonist with official UNHCR refugee status, was arrested in Thailand and rendered to China on a plane with two other Chinese dissidents, Hong Kong Swedish publisher Gui Minhai and human rights activist Dong Guangping. Before seeking refuge in Thailand, Jiang had been imprisoned and tortured by the Chinese government. Upon his return to China, Jiang was held incommunicado and sentenced to six and a half years of prison in a secret trial (Reporters Without Borders 2022a).

These three cases are examples of transnational repression. However, they are not isolated incidents. For example, Turkey has been referred to the ICC for committing acts of torture, enforced disappearances, wrongful imprisonment, and persecuting numerous political opponents around the world, including individuals linked to the Gülen religious, social, and educational movement designated by Turkey as a terrorist organization and accused of organizing the 2016 coup attempt (Borger 2023). Rwanda has also been accused of being among the worst perpetrators of transnational repression, engaging in extrajudicial killings, disappearances, kidnappings, physical attacks, and harassment of Rwandans living abroad and manipulating foreign judicial and law enforcement processes to detain and return the Rwandan diaspora, including the dissident and former Kigali hotel manager whose story is told in the film Hotel Rwanda (Human Rights Watch 2023). Other well-known examples of transnational repression include the murder of dissident journalist Jamal Khashoggi by Saudi agents in the country’s consulate in Turkey, Russia’s poisoning of former intelligence officers Alexander Litvinenko and Sergei Skripal in the United Kingdom, Iran’s assassinations and renditions of political opponents that have fled to Europe, the Middle East, and North America, and China’s persecution of ethnic and religious groups in many overseas territories, including Uyghurs, Tibetans, and Falun Gong members (Freedom House 2022).

Since 2014, Freedom House has documented 608 physical cases of transnational repression involving 79 host countries and 31 origin countries, representing 160 unique country dyads (Freedom House 2021). This increasingly “normal phenomenon” often involves host countries (including many democracies) working closely with origin countries (typically autocracies) to transfer their citizens living abroad into their custody and silence dissent (Freedom House 2021, 2). Origin countries also carry out transnational repression through collaborations with non-state actors in the host country, including political parties, educational and religious groups, its own intelligence agents, diaspora, businesses, criminal gangs, etc. (Dukalskis et al. 2023). Origin countries physically target their diaspora using assassinations, assaults, detentions, renditions, unlawful deportations, and unexplained disappearances. The most frequent targets of transnational repression include national, ethnic, religious, and linguistic minorities, political activists, journalists, former insiders/government elites, and human rights advocates (Freedom House 2021).

Many instances of transnational repression would not be possible without international cooperation. To reach their diaspora abroad, origin countries often depend upon host countries to arrest, detain, and transfer their citizens back to their country of origin. However, because of the secret nature of cooperation, very little is known as to why states become involved in this deeply sensitive area of international politics. This raises the question, “Why do some states assist other countries to reach across national borders and repress their diaspora, while others do not?” This global pattern is especially puzzling given the political costs of foreign complicity in abusing human rights and the political, economic, and regional diversity of host countries.

While a considerable amount of research has examined why autocratic governments use transnational repression to silence dissent, relatively less attention has been paid to exploring the theoretical mechanisms that explain why foreign governments decide to cooperate. Our study advances existing explanations from qualitative case studies by conducting a cross-national test on the causes and dynamics of host country participation (Shain 2010; Cooley Heathershaw 2017; Glasius 2018; Dukalskis 2021; Furstenberg, Lemon, and Heathershaw 2021; Tsourapas 2021; Lessa 2022; Moss, Michaelsen, and Kennedy 2022).

We argue that international cooperation on transnational repression relies on a host country’s economic ties to its origin countries and domestic rule of law. The most desirable transnational repression partners should be those countries that are more willing to trade off civil liberties for economic cooperation. Countries that are less concerned about the domestic and international consequences of breaking commitments to the rule of law should be easier to co-opt and should require less persuasion to cooperate, especially when they have greater financial dependence on the origin country. Together, this provides origin countries with the opportunity to repress and the leverage to elicit cooperation.

We test our hypotheses using a dataset of 33,615 directed dyad-years that accounts for refugee flows between pairs of countries and find empirical support for our theoretical argument. Origin countries are better able to elicit cooperation on human rights violations from host countries that have shared economic interests (leverage) and a weak rule of law (opportunity). This article provides one of the first quantitative accounts of foreign complicity in transnational repression and contributes to emerging research on extraterritorial repression. Our results suggest that while domestic institutions upholding the rule of law can be an effective tool for deterring international cooperation on silencing dissent, economic interconnectedness can have the unintended negative consequence of incentivizing states to trade off human rights for material gains.

We make several contributions to debates in international relations. First, our findings complement normative and instrumental arguments on how domestic commitments, robust legal frameworks, and liberal democratic institutions can constrain government behavior on violations of international law (Davenport 2007; Kelley 2007; Lupu and Wallace 2019). Second, our results reveal a darker side of the relationship between trade and human rights, whereby autocratic states use their financial linkages and economic agreements to pressure foreign governments and buy support for suppressing dissent (Hafner-Burton 2005; Dreher, Nunnenkamp, and Thiele 2008; Cao, Greenhill, and Prakash 2013; Peterson, Murdie, and Asal 2016). Third, our study speaks to research on the cross-border security apparatus of authoritarian regimes, the dynamics of refugee rights violations, and limitations of extraterritorial human rights obligations (Cerone 2008; Heupel 2018; Abdelaaty 2021; Dukalskis 2021; Lessa 2022). Finally, we contribute to debates on whether globalization, the internet, migration, and institutions help or hinder the promotion of human rights (Brysk 2002; Simmons 2019; Gohdes 2020; Weidmann and Rød 2019).

International Cooperation on Transnational Repression

Transnational Repression

Prior research on the global scale and scope of transnational repression has predominantly focused on why origin countries aim to silence dissent among their citizens living outside of their national borders. Like intra-state repression, transnational repression is more common among autocracies than democracies. Authoritarian regimes use repression to exert political control over their citizens, weaken the opposition, and ensure that the ruling elite survive (Davenport 2007; Hassan, Mattingly, and Nugent 2022). The benefits of authoritarian regimes using repression to deter and punish dissent (at home and abroad) outweigh the costs of public backlash, as the institutions that enable citizens living in democracies to hold their governments to account for violating human rights are typically absent (Olar 2019). Individuals that engage in political activism abroad are particularly threatening to authoritarian regimes, as the origin country does not have direct control over the foreign institutions that allow the diaspora to publicly criticize the regime (i.e., the internet) (Moss 2016; Moss, Michaelsen, and Kennedy 2022). This alternative flow of information is harmful to origin countries as it provides a different account of events that can be used to mobilize dissent internally and apply international pressure on the regime to change its domestic policies (Dukalskis 2021). Dukalskis (2021) frames transnational repression as an “obstructive form” of external authoritarian image management designed to eliminate negative narratives from the public sphere that undermine the legitimacy of the regime back home and prevent domestic and international dissidents from communicating with one another (Dukalskis 2021, 40).

Tsourapas (2021) argues that transnational repression emerges because of migration, and the need for authoritarian regimes to maintain open borders for economic purposes while at the same time controlling citizens’ political activities once abroad. Additionally, advances in internet communication technologies provide opportunities for individuals to mobilize and share anti-government information with others on a global scale while simultaneously improving the ability of origin countries to surveil the political behavior of their diaspora and identify the location of their targets (Dukalskis 2021; Tsourapas 2021). While transnational repression has wide appeal across authoritarian regimes for its ability to silence dissenting voices abroad and avoid internal challenges back home, patterns of transnational repression tend to reflect the domestic politics of each country involved. In a cross-national analysis of transnational repression, Dukalskis et al. (2023) find a strong association between a country’s level of domestic repression and transnational repression since domestic crackdowns push dissidents abroad and can make international connections between domestic and international dissidents appear more threatening. For example, China’s targeting of the Uyghur diaspora mirrors its post-2014 domestic security crackdown in the Xingjiang province, and Turkey’s targeting of the diaspora following the 2016 coup reflects its domestic campaign against individuals that it suspected of being involved (Dukalskis 2021). Because of globalization, domestic politics and repression continue to extend beyond a country’s national borders as a diaspora’s “physical exit no longer necessarily implies exit from the national public sphere” (Glasius 2018, 81).

Cooley and Heathershaw (2017) outline three stages of transnational repression whereby origin states warn and intimidate targets in the first stage, arrest and seize property in the second stage, and pursue renditions, disappearances, attacks, and assassinations in the third stage. The multifaceted nature of transnational repression therefore requires origin countries to have a strong state security service capacity that can conduct extraterritorial violence toward a targeted diaspora and repress/coerce their associates back home (Shain 2010; Furstenberg, Lemon, and Heathershaw 2021). States with greater diplomatic relations should be better equipped to conduct repressive campaigns abroad since origin countries often depend on their diplomatic institutions overseas (e.g., embassies and consulates) and relations with foreign governments, intelligence services, and non-state actors to induce cooperation on transnational repression (Dukalskis et al. 2023). Michaelsen and Ruijgrok (2023) also find that networks play an important role in the occurrence of transnational repression, especially among countries in the same geographical region and regional organizations, where states are more likely to have existing ties between security agencies and similar ideologies and preferences for preserving authoritarianism in the region. More broadly, countries that share a similar history with “common and enduring cultural, economic and political” ties (like Russia and post-Soviet states in Central Asia) are more likely to regularly cooperate with one another on transnational repression because of formal intelligence, criminal justice, and immigration arrangements, even though they are not technically in the same geographical region (Cooley and Heathershaw 2017, 192). Finally, Shain (2010) suggests that we are more likely to observe a country pursuing transnational repression when they have less tolerance of the domestic opposition, a “greater ideological component,” fear external threats to their lives (e.g., in personalized regimes), and are less dependent on international support (Shain 2010, 162).

International Cooperation

While existing research has advanced our understanding of this topic, we still know comparatively less about what drives host countries to assist origin countries in repressing their diaspora. Most instances of transnational repression would not be possible without international cooperation (Cooley and Heathershaw 2017). Transnational repression often involves origin countries “co-opting” host governments to silence dissent among their diaspora (Freedom House 2021, 2).¹ Cooperation on transnational repression is predominantly bilateral between the origin country that seeks to target its citizens living abroad and the host country where the target is apprehended. In some scenarios, cooperation is inferred using reliable qualitative evidence and in-depth case study analysis, like the rendition of Swedish-Iranian political activist and exile Habib Asyud from Turkey to Iran, who disappeared from a Turkish airport in October 2020 and turned up in Iranian custody facing charges of terrorism and a death sentence, where a deep investigation of reputable media reports by Freedom House (2021) made it evident that the rendition “would have required cooperation from Turkish authorities.” In other scenarios, host countries are implicated by a paper trail or openly acknowledge cooperation, like the rendition of Nawaf al-Rasheed, a Saudi-Qatari poet and member of the Al Rashid family with a historic rivalry to the ruling al-Saud family, from Kuwait to Saudi Arabia in May 2018, where Kuwait’s Interior Ministry admitted in a tweet that they acted on behalf of the Saudi government “under bilateral mutual security arrangements” (Freedom House 2021, 33).

The most common form of physical repression identified by Freedom House (2021) in their Transnational Repression Database is (i) detentions, where an individual is held by the host country for over 12 hours at the request of the origin country, followed by (ii) renditions, where the host country holds an individual incommunicado, without access to a lawyer, and forcibly returns them to the origin country without protection, and (iii) unlawful deportations, where the host country forcibly deports an individual to the origin country where they are at risk of persecution with some violations of due process (Freedom House 2021).² The least common forms of physical repression include assassinations or assassination attempts, assaults, unexplained disappearances, and intimidation. These are cases where the origin country engages in direct attacks against individuals that have fled their territories with little or no involvement from the host country (Freedom House 2021). Michaelsen and Ruijgrok (2023) find that detentions, renditions, and unlawful deportations are more common in autocracies, whereas “direct attacks” (e.g., assassinations or assassination attempts, assaults, and intimidation) are more common in democracies since securing assistance in transnational repression is easier to obtain in the former than the latter.

The most frequent targets of transnational repression in the database include national, ethnic, religious, and linguistic minorities, followed by political activists, journalists, former insiders/government elites, and human rights advocates (Freedom House 2021). In many instances, origin countries apply the label of terrorism or extremism to diaspora that they pursue or accuse targets of anti-state actions or corruption to justify their repressive actions and persuade host countries to cooperate. In some instances, the origin country files a Red Notice, Diffusion Request, or Extradition Request through Interpol that results in the host country detaining the individual or rendering/unlawfully deporting them to the origin country, where they face the risk of persecution (Freedom House 2021). While Interpol abuse by origin countries accounts for less than 7 percent of all cases of transnational repression in the database, Interpol Red Notices (requests to law enforcement in other countries to arrest a person pending extradition, surrender, or similar legal action) and other notifications provide a “cheap and easy means of reaching exiles” by legitimizing attacks on diaspora and increasing pressure on host countries to transfer individuals into their custody (Freedom House 2021, 6). The political manipulation of Interpol is likely to be a particularly effective tool for securing cooperation on transnational repression from democracies where host countries may be more reluctant to transfer individuals into the custody of their origin countries without evidence that the individual is guilty of the crime that they are accused of committing (even if the allegations are later discovered as to be false). For example, Interpol notices and requests on behalf of origin countries such as Azerbaijan, Egypt, Kazakhstan, Russia, Tajikistan, India, Iran, Rwanda, and Turkey have all successfully led to the detention of individuals by democratic host countries including Cyprus, the Czech Republic, Finland, Germany, Italy, Poland, Portugal, Spain, and the United States.

Figure 1 displays thirty-one origin countries in the database that have targeted their citizens living abroad through transnational repression between 2014 and 2020. Unsurprisingly, most origin countries are autocracies where human rights violations are widespread. The top ten origin countries that have engaged in the most instances of transnational repression in the database are China (213 cases), Turkey (111 cases), Egypt (42 cases), Tajikistan (38 cases), Russia (32 cases), Uzbekistan (29 cases), Thailand (15 cases), Iran (12 cases), Azerbaijan (10 cases), and Rwanda (10 cases).

Figure 2 displays seventy-nine host countries where the origin country’s target has been apprehended on behalf of the origin country or by the origin country themselves. The political, economic, and regional diversity of host countries makes international cooperation in transnational repression particularly intriguing. While the top ten host countries that have engaged in the highest number of recorded cases of transnational repression include many autocracies and semi-democracies (e.g., Thailand, Russia, Saudi Arabia, Malaysia, Turkey, Kuwait, Egypt, the United Arab Emirates, Kazakhstan, and India), there are many democracies where multiple instances of transnational repression have taken place on their territory despite foreign complicity in human rights violations carrying greater political costs (e.g., Kenya, Ukraine, Georgia, Moldova, Germany, Poland, Bulgaria, Spain, France, and Austria).

Existing Explanations: International Cooperation and Human Rights

Why do some states assist other countries to reach across national borders and repress their diaspora, while others do not? Very little is known as to why states become involved in this “growing pattern” of transnational repression at the cross-national level (Abramowitz and Schenkkan 2021). International cooperation in this deeply sensitive area of international politics is especially puzzling when we find that core theories on international cooperation and human rights fail to fully explain why host countries would agree to participate.

For example, prior research suggests that we should be more likely to observe international cooperation between allies since defense pact members are obliged to come to the aid of their alliance partners in times of need (Leeds 2003). Additionally, countries that collaborate with allies on political and economic issues reduce security externalities and increase the strength of the whole alliance (Long and Leeds 2006). However, this argument fails to fully account for participation in transnational repression as many host countries do not have formal alliances with the origin country, with none of the host countries that assisted Iran with carrying out transnational repression having a formal military alliance with them (e.g., Turkey, Iraq, and the United Arab Emirates).

Alternatively, previous studies suggest that democracies are less likely to violate human rights because of the domestic costs that would result if detected (Davenport 2007). Host countries with liberal democratic institutions should be less likely to engage in repression as their behavior is constrained by domestic and international legal commitments (Conrad and Moore 2010). However, this also fails to account for cooperation in transnational repression, as half of the host countries that assisted Rwanda with targeting their diaspora overseas were democracies, such as Germany and Kenya.

Similarly, scholars would predict that host countries with strong human rights protections should be less likely to participate in transnational repression, as doing so puts the human rights of targeted individuals at risk (Cordell 2019). A country’s decision to assist a country in violating human rights is driven by the anticipated consequences of being caught, with public backlash being more likely in countries with core human rights norms (Efrat and Newman 2020). However, this is not the case as 62 percent of host countries that helped Russia engage in transnational repression have higher levels of human rights protections (e.g., the Czech Republic and the United States).

Findings from prior studies also indicate that countries with greater state capacity should be less likely to participate in transnational repression, as strong states should be better able to monitor and ensure compliance within the institutions that origin countries seek to co-opt (Young 2009).³ However, when we look at the level of state capacity among host countries that cooperated with Tajikistan on transnational repression, we observe twice as many strong states than weak states involved, such as Austria, the Netherlands, and Finland.

Opportunity and Leverage

Given these largely null findings, what drives host countries to assist origin countries in repressing their diaspora? We propose that a host country’s decision to engage in transnational repression with an origin country is based on a cost-benefit calculation. On the one hand, foreign complicity in transnational repression carries significant domestic political costs as it can lead to public backlash and political resistance in the host country (Efrat and Newman 2020; Cordell 2021). All forms of transnational repression are prohibited under international law (e.g., assassinations/assassination attempts, assaults, detentions, renditions, unlawful deportations, and unexplained disappearances) with countries that aid or assist an origin country in the commission of an internationally wrongful act considered internationally responsible if they have knowledge of the circumstances and do not take reasonable steps to prevent it (United Nations 2010). On the other hand, participation in transnational repression can be beneficial for the host country in maintaining or improving its bilateral relations with the origin country. Issue linkages in international relations enable countries to exchange favors on different issue areas and incentivize states to take great risks to uphold their international reputations and advance their national interests (Long and Leeds 2006; Mattes and Weeks 2024).

In order to maintain secrecy and successfully carry out transnational repression, it is important that origin countries approach host countries that are more likely to agree to cooperate in the first place. Approaching a country that refuses to cooperate increases the risk that the host country will expose the origin country’s plans to target their diaspora and draw attention to their repressive behavior, which is counterproductive to the goal of transnational repression (i.e., silencing dissent rather than amplifying it).⁴ This selection mechanism can be conceptualized as a screening process that identifies those host countries that are most likely to perceive the payoffs of cooperation in transnational repression as positive and eliminates those countries most likely to perceive the payoffs of cooperation as negative. We propose that a host country’s willingness to trade off international cooperation for civil liberties depends on two factors: the opportunity to repress and the leverage to induce cooperation.

Opportunity: The Rule of Law

First, we argue that international cooperation on transnational repression relies on a host country’s domestic rule of law in providing origin countries with the opportunity to repress. Countries with a strong rule of law are more likely to honor their commitments to laws protecting human rights, as disregarding the rule of law can have negative domestic consequences (Kelley 2007). Experimental research shows that citizens value the rule of law and are willing to punish states for violating international legal commitments (Lupu and Wallace 2019). The risk of public backlash incentivizes states to obey their legal commitments, as violating the rule of law can undermine their domestic legitimacy and threaten their survival in office, especially in countries where the normative value of the rule of law is high (Finnemore and Sikkink 1998). Prior research predicts that violating human rights should be more politically costly for governments that typically respect them, as it reveals a greater deviation between their public and private types that can lead voters to question their honesty and credibility as a government (Cordell 2021), whereas countries with a weak rule of law have less to lose in terms of their domestic and international reputation for upholding the rule of law, as the public and other actors are less likely to expect them to honor their international commitments and are more likely to be accustomed to the government disrespecting the domestic rule of law on a regular basis (Simmons 2009). Due to the interconnectedness of the rule of law and the quality of democracy, it is also unlikely that citizens in countries with weak rule of law will have the ability to effectively constrain the government’s behavior and punish them for violating commitments to the rule of law; therefore, reducing the likelihood that the public’s reaction to abuses will threaten the government’s survival in office (O’Donnell 2004).

In addition to being less likely to perceive transnational repression as politically costly, countries with weak rule of law should be easier to co-opt by origin countries, as they are more likely to have similar preferences on handling dissent and engaging in repression (Kelley 2007). Prior research indicates that countries with similar preferences make more reliable cooperation partners, as they are more likely to perceive cooperation as mutually beneficial and are less likely to disclose sensitive information that is detrimental to the group (Cordell 2019). From a logistical standpoint, origin countries should be better able to exploit the institutions of host countries with a weak rule of law and target their diaspora, as detentions, unlawful deportations, and renditions depend on the host country’s ability to forgo due process and act without constraints to abuses of civil liberties and state power (O’Donnell 2004; Freedom House 2021).

For example, Turkey has successfully secured cooperation in transnational repression from multiple countries with a weak rule of law, persuading targeted states to “hand over individuals without due process, or with a slight fig leaf of legality” (Freedom House 2021, 38 and 40). In Lebanon (where authorities are often not bound by the rule of law or held accountable when they violate it), Turkey was accused of manipulating authorities to detain journalist and socialist activist Ayten Öztürk for five days in Beirut airport and rendition her to Turkey in 2018, where she was held in secret detention for five months and reportedly tortured (Freedom House 2021). And in Ukriane (another country with a weak rule of law), Turkey was believed to have co-opted the country’s legal system and security services in the rendition of Yusuf Inan, a journalist and blogger accused of having ties to the Gülen movement, who was arrested and extradited all in the same day in 2018 following a sudden cancellation of his Ukranian residence permit, “leaving almost no time between detention and handover, and no clear legal process” (Freedom House 2021, 41). In settings with a weakened rule of law, origin countries have an easier time exploiting a host country’s institutions (e.g., courts, immigration systems, and security agents) and obtaining assistance from host countries to target their diaspora, as such practices pose less of a threat to rule of law traditions as well as the tenure of officials and political elites. In sum, we hypothesize that:

 

Leverage: Economic Interconnectedness

Second, we argue that international cooperation on transnational repression relies on a host country’s economic ties to the origin country, providing origin countries with the leverage to encourage cooperation. Most research on the link between economic interconnectedness and human rights has focused on the positive effect of trade on human rights, whereby human rights respecting states can use their financial linkages to pressure abusive states into improving their human rights practices (Hafner-Burton 2005; Cao, Greenhill, and Prakash 2013; Peterson, Murdie, and Asal 2016). However, few studies have focused on the way in which abusive states can leverage their position to induce cooperation from their trade partners on human rights violations. Countries with financial linkages are more likely to cooperate with one another on different issue areas, as refusing to cooperate could place valuable economic transactions in jeopardy (Nooruddin and Payton 2010). Recent survey experiments on public opinion, foreign aid, and human rights show that public attitudes toward international cooperation with repressive governments are conditional on the benefits of cooperation for the donor country in other areas of international politics (Heinrich, Kobayashi, and Long 2018). Countries that are dependent on one another are more likely to continue to cooperate in the future to avoid potential sanctions, ensure that their cooperation continues in other domains, and prevent domestic backlash for any losses that would result from refusal to cooperate. Together, these factors increase the likelihood that host countries with existing ties to origin countries will perceive cooperation in transnational repression as beneficial and find rewards in helping them for their own sake (Long and Leeds 2006; Mattes and Weeks 2024).

Additionally, host countries that are financially dependent on the origin country should be easier to co-opt by origin countries, as they can use their relative power to elicit cooperation in transnational repression through the threat or use of sanctions (e.g., a reduction in trade) (Nooruddin and Payton 2010). Previous studies suggest that powerful states can better use sanctions that are threatened or imposed to secure desired policy outcomes and persuade other countries to do something they would otherwise not do when the balance of power is unequal between the sender and the target (Dreher, Nunnenkamp, and Thiele 2008). Host countries with economic ties to the origin country should also make more trustworthy transnational repression partners, as states are more likely to exchange favors with friends as they know that they will be dealing with each other repeatedly in the future, reducing the incentives to exploit a situation and weakening the bilateral linkage (Leeds 2003; Mattes and Weeks 2024).

For example, China has frequently been accused of using its economic leverage to compel foreign governments to collaborate in transnational repression, with a leaked internal policy document outlining specific instructions for using “economic statecraft” abroad to secure cooperation in suppressing Uyghur dissent (Uyghur Human Rights Project [UHRP] and the Oxus Society for Central Asian Affairs [Oxus Society] 2021, 24). In Cambodia, China signed fourteen trade deals with the country worth $1.2 billion just days after it deported twenty Uyghur asylum seekers in 2009 (UHRP and the Oxus Society 2021), and in Turkey, arrests and deportations of Uyghurs dramatically increased after it joined the Belt and Road Initiative in 2015 and China’s economic investments in the country went up (Freedom House 2022). In both countries, China is an important trade partner (their largest importer), and trading off human rights presents new and continued strategic economic opportunities. Such relations of dependence allow origin countries to manipulate the incentives of their economic partners and coerce them into providing assistance to target their diaspora. In these scenarios, issuing direct economic threats is not a necessary condition for inducing cooperation, as the latent prospect of economic losses is enough to drive host countries to collaborate. In sum, we hypothesize that:

 

Research Design

We test our hypotheses using data on international cooperation in transnational repression, trade flows, and the rule of law for all countries in the world from 2014 to 2020. Our unit of analysis is directed dyad year. While the occurrence of transnational repression has become increasingly common in recent years, it is still a relatively rare event, resulting in many zeros on our dependent variable (a count of cooperation in transnational repression). We employ two methods for dealing with this research design issue.

First, we reduce our global sample of directed dyad-years (210,714 observations) to a sample of host countries that have admitted refugees or asylum seekers from the origin country that year according to annual refugee flow data from the UNHCR (2022). This case selection produces a sample of 33,615 directed dyad-years and ensures that only eligible cooperation partners are included in our data since the presence of a foreign citizen (from the origin country) in the host country is a necessary condition of cooperation in this issue area, with many transnational repression targets seeking asylum in the host country to avoid being returned to their country of origin.⁵ This preprocessing step removes pairs of countries from our data where there is no possibility for cooperation such as Belize and Russia (no Russian refugees or asylum seekers in Belize) but guarantees that Poland (a country hosting many Russian refugees and asylum seekers) remains in our dataset. The study’s main findings also hold when we relax this selection criteria, using the original global sample of 210,714 directed dyad-years (online appendix, table A1), and when we increase the eligibility threshold to include host countries with greater annual refugee flows from origin countries falling above the median value (see online appendix, table A2).

Second, we use a zero-inflated negative binomial model to account for excess zeros in the data where there is never cooperation in transnational repression (i.e., structural zeros) and true zeros in the data where there could be cooperation in transnational repression (i.e., random zeros). The first stage of the regression uses a binary model to determine whether an observation in the data is in the “excess zero” or “true zero” category. The second stage of the regression then uses the probabilities from the first stage to model the count of cooperation in transnational repression. A Vuong test confirms that estimating a zero-inflated component is more appropriate than using a single-equation count model. We use a negative binomial distribution rather than a Poisson distribution as we find significant overdispersion in the data.⁶  Online appendix, table A3 shows the findings from a standard negative binomial model instead of a zero-inflated model, and the paper’s main findings remain the same.

Dependent Variable: Cooperation in Transnational Repression

The dependent variable is a count of the number of instances of transnational repression between a pair of countries each year. To measure participation in transnational repression, we use the Transnational Repression Database from Freedom House (2021), which records 608 direct, physical cases of transnational repression from January 2014 through November 2020. Freedom House uses information from UN communications, court rulings, news articles, and other official documents and reported sources to identify episodes of transnational repression and code cases according to their certainty on whether the incident was state-driven and direct.⁷ When creating our dependent variable, we exclude assassinations/assassination attempts, assaults, unexplained disappearances, and intimidations from our count of transnational repression incidences (sixty-two incidents), as they represent cases where host countries had little to no involvement (and are presumed not to be complicit in the action). This increases the likelihood that our measure includes instances of transnational repression where host countries directly cooperated with origin countries on detentions, renditions, and unlawful deportations. This count variable ranges from 0 (e.g., the United States and China in 2014) to 121 (e.g., Thailand and China in 2015), with a mean value of 0.013.

Independent Variable: Rule of Law (Inverse)

To test the article’s first hypothesis that countries with a weaker rule of law should be more likely to cooperate in transnational repression, we use the “Rule of Law” index from Kaufmann, Kray, and Mastruzzi (2010) Worldwide Governance Indicators data. This index captures “perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence” (Kaufmann, Kray, and Mastruzzi 2010, 4). To make the interpretation of the results more straightforward, we reverse the scale of the Rule of Law index to create an inverse variable. A higher value indicates that the state has a weaker presence of the rule of law (e.g., Cambodia in 2015), and a lower value indicates that the state has a stronger presence of the rule of law (e.g., the United Kingdom in 2015).

Independent Variable: Total Trade (Log)

To test the article’s second hypothesis that countries with greater economic ties to origin countries should be more likely to cooperate in transnational repression, we use dyadic data from the Direction of Trade Statistics (International Monetary Fund 2023). This data presents the value of merchandise imports and exports that have been disaggregated by a country’s primary trading partners. We extract the variables “Goods, Value of Exports, Free on board (FOB), US Dollars” and “Goods, Value of Imports, Cost, Insurance, Freight (CIF), US Dollars,” which indicate the exports and imports value between countries. We then add the export and import values together for each country dyad-year and take the natural logarithm. A higher value on this continuous variable (e.g., Turkey and Iran in 2019) indicates more trade and stronger economic ties between two countries, and a lower value (e.g., Israel and Iran in 2019) indicates that the two countries have less trade and weaker economic ties.

Control Variables

To reduce the possibility that other factors affecting the likelihood of international cooperation in transnational repression may be correlated with the rule of law and trade variables, we control for several confounders. First, we create an Alliance dummy control variable using Leeds et al. (2002) data. Allied countries should be more likely to engage in transnational repression to preserve their international relationships and maintain future cooperation with states in other security issue areas (Long and Leeds 2006). Second, the Refugees (log) variable controls for a country’s refugee population size by country of origin using UNHCR (2022) data. Countries with high refugee populations might be more likely to participate in transnational repression, as origin countries are more likely to seek cooperation in locations where their targets are potentially located (Efrat and Newman 2020). Third, Minimum Distance (log) controls for the distance between two countries in km using Weidmann, Kuse, and Gleditsch (2010) data. Dissidents escaping persecution often seek refuge in nearby countries, and proximate states with existing ties and similar preferences may increase the potential for cross-border repression (Cooley and Heathershaw 2017; Efrat and Newman 2020; Michaelsen and Ruijgrok 2023). Fourth, we create a Democracy dummy control variable using Geddes, Wright, and Frantz (2014) and Chin, Wright, and Carter (2021) data. Democracies should be less likely to engage in transnational repression as democratic institutions increase the risk that governments will be held accountable and punished by voters (Conrad and Moore 2010). Fifth, we control for a country’s State Capacity using Kaufmann, Kray, and Mastruzzi (2010) “Government Effectiveness” index. Countries with greater state capacity should be better able to monitor and ensure compliance within the institutions that origin countries seek to co-opt, making cooperation less likely (Young 2009). Sixth, we control for a country’s GDPPC (log) and Population (log) using the World Bank (2023) data. Smaller countries and countries with a weaker economy may be more likely to experience an imbalance of power with origin countries and perceive the potential economic payoffs of cooperating as consequential (Davenport 2007; Dreher, Nunnenkamp, and Thiele 2008). Seventh, we create a Right Orientation dummy control variable using Cruz, Keefer, and Scartascini (2021) data. Right-wing governments could be more likely to trade off human rights for cooperation in other areas of international politics, especially security (Cordell 2021). Eighth, we control for a country’s level of respect for Civil Liberties using Coppedge et al. (2021)v2x_civlib index. Countries with stronger normative and legal commitments to this right should be less likely to trade them off for cooperation with origin countries (Cordell 2019). Ninth, we construct an Origin Country dummy control variable that measures whether a country has been an origin country themselves using the Freedom House (2021) database. Countries with similar preferences and experiences of transnational repression should be less likely to object to the principle of cooperation and cooperate to secure reciprocal cooperation in the future. Finally, we control for a country’s level of Press Freedom using the Reporters Without Borders (2022b) data. We may be more likely to observe transnational repression in countries with greater press freedoms, especially since the Freedom House (2021) database uses news articles as a source for identifying cooperation. The online appendix, table A4, displays descriptive statistics of the article’s independent and control variables.

Results

The theoretical framework in our paper predicts that countries with a weaker rule of law (greater opportunities to repress) and a higher volume of total trade with their origin countries (greater leverage for cooperation) are more likely to participate in transnational repression. We estimate the effect of the rule of law and bilateral trade flows (exports and imports) on the count of transnational repression using zero-inflated negative binomial regression models and a sample of 33,615 directed dyad-years. The results of our analyses are reported in table 1. Model 1 is the baseline model that includes the independent variables only, and model 2 is the full model that includes the independent variables and control variables.

The results provide empirical support for H1 and H2. First, countries with weaker rules of law are more likely to engage in transnational repression with origin countries, as anticipated. This positive relationship holds in all models and is statistically significant above the 99 percent confidence level. This result indicates that origin countries have fewer opportunities to elicit cooperation on transnational repression from countries with strong commitments to the rule of law, as they are less willing to trade off legal protections for their foreign citizens for international cooperation, whereas origin countries have much less difficulty in co-opting the institutions of host countries with a weak rule of law or convincing governments to participate. When we look at the most prominent host countries in our data, we find many countries with a low presence of the rule of law. Such as Russia (the second most frequent cooperator after Thailand), where the absence of rule of law traditions allows authorities and political elites to act outside of the rule of law without facing accountability, thereby reducing the constraints and political costs of assisting other countries to target their diaspora.

Second, countries with a higher volume of trade with origin countries are more likely to participate in transnational repression than countries with a lower volume of total trade, as expected. This positive relationship holds in all models and is statistically significant above the 99 percent confidence level. This finding suggests that origin countries can use their economic leverage over host countries to manipulate the incentives of their economic partners and coerce them into providing assistance to target their diaspora. Countries with greater economic ties make better cooperation partners as they are more willing to exchange favors to maintain or improve their economic relations with the origin country, even in deeply sensitive areas of international politics. Indeed, the most frequent cooperators of transnational repression in our data are the dyads with the highest volume of trade: Thailand and China. For Thailand, it is in their economic interest to comply with China’s requests, as refusal to cooperate could place valuable economic transactions in jeopardy.

The findings for the control variables also yield some interesting patterns.⁸ The Alliance and Refugees (log) variables have a positive and significant effect on international cooperation on transnational repression, indicating that states may be engaging in security as well as economic-civil liberties trade-offs and that cooperation depends on hosting potential origin country targets. Contrary to expectations, the State Capacity variable has a positive and significant effect, which suggests that stronger states may have greater capacity to control and override their agents and institutions at the request of origin countries. The Civil Liberties and Population (log) variables have a negative and significant effect, implying that host countries with stronger normative and legal commitments to human rights are less likely to cooperate, but smaller countries with an imbalance of power are.

To examine the substantive effects of the rule of law and bilateral trade flows on the frequency of cooperation on transnational repression, we present the predicted count of observing transnational repression from the full model (model 2) with 95 percent confidence intervals, while holding all control variables at their mean. The x-axis in figures 3 and 4 displays the range of values for the Rule of Law (inverse) variable and the Total Trade (log) variable, ranging from a strong rule of law and lower levels of trade to a weak rule of law and higher levels of trade. The y-axis in both figures displays the predicted count of cooperation in transnational repression between a country and an origin country each year, ranging from no instances of transnational repression to a greater number of instances of transnational repression.

In line with our expectations, the predicted count of transnational repression increases as a country’s rule of law weakens and dyadic trade flows grow.

In figure 3, moving from the minimum to the maximum value of the Rule of Law (inverse) variable increases the predicted count of transnational repression from 0 (countries with a strong rule of law) to 1.52 (countries with a weak rule of law). In figure 4, shifting the volume of dyadic trade from the minimum value to the maximum value in our data increases the predicted count of transnational repression from 0 (low levels of trade) to 1.22 (high levels of trade). While the expected number of transnational repression events predicted by the Rule of Law (inverse) and Total Trade (log) variables is quite low, the difference in the predicted count of transnational repression instances for countries with a strong rule of law and low levels of bilateral trade compared to countries with a weak rule of law and high levels of bilateral trade is substantively large. Because international cooperation in transnational repression is a rare event (occurring in less than 1 percent of our sample), even one occurrence of cooperation in transnational repression is a significant finding.

Together, these findings confirm H1 and H2. Host countries with weaker institutions and lower levels of governance related to the rule of law and those that are financially dependent on repressive origin countries are more likely to become targets of international cooperation in transnational repression as they are easier to co-opt and influence. Assisting origin countries to reach across their national borders and silence dissent among their diaspora carries fewer political costs for countries with weaker preferences on the rule of law and helps countries with greater origin country trade flows to avoid any disruption to their existing economic agreements.

Interaction Effects

The results presented in this article so far provide empirical support for the independent effects of the rule of law and bilateral trade flows on international cooperation in transnational repression. Next, we examine the interaction between these two variables. The logic of our theoretical argument suggests that the most likely targets of cooperation in transnational repression are those states where origin countries have greater opportunities to repress (host countries with a weaker rule of law) and greater leverage for inducing cooperation (host countries with a higher volume of origin country trade).

We test this claim by plotting the predicted counts of transnational repression conditional on both the rule of law and bilateral trade flows from the full model (model 4) with interaction effects as displayed in online appendix, table A5. The results in online appendix, table A5 show that the interaction between the rule of law and dyadic trade is positive and statistically significant at the 99 percent confidence interval. Figure 5 shows the effect of the rule of law on the predicted count of transnational repression at different levels of bilateral trade flows, one standard deviation below and above the mean (13.42 and 22.6, respectively), with 95 percent confidence intervals, while holding all control variables at their mean.

As figure 5 demonstrates, the predicted count of transnational repression changes under different combinations of the rule of law and dyadic trade. When host countries have low economic ties to origin countries, the effect of the rule of law on the predicted count of transnational repression does not increase. However, when host countries have high economic ties to origin countries and a weak rule of law, the predicted count of transnational repression dramatically increases from zero (no instances) to thirty-four occurrences. Origin countries are better able to exploit the institutional vulnerabilities of a host country when they can use their economic power as a tool to elicit cooperation in transnational repression. This is arguably why China has been so successful at securing support from its trade partners with a weakened rule of law (e.g., Egypt and Nepal), as such countries face fewer legal and political constraints and can see the benefits and opportunities of engaging in economic-civil liberties trade-offs.

Robustness

We systematically test the robustness of our results using a variety of different model specifications displayed in the Appendix. First, the online appendix, table A6, excludes instances of transnational repression where it was less clear that the incident was orchestrated by the origin country and had a political motive (as determined by Freedom House). While Freedom House (2021) codes most observations in the data as being driven by origin states and political in nature (e.g., 85 percent of transnational repression targets are political activists, former insiders/government elites, journalists, human rights advocates, and national, ethnic, religious, and linguistic minorities), this robustness check increases our confidence that the results are not being driven by nonpolitical cases (e.g., ordinary extraditions) or incidents where the origin country may not have been responsible—and the study’s main findings hold.

Second, the online appendix, table A7, excludes instances where the targeted individuals are accused of terrorism and/or extremism. While many of these origin state accusations are not supported by evidence (Freedom House 2021), this test allows us to rule out that the results are being driven by more extreme cases of transnational repression where host countries may have thought they were simply assisting the origin country in combating terrorism rather than political dissent. Again, our results remain the same.

Third, we test an alternative explanation that focuses on the role of Interpol red notices and requests on international cooperation in transnational repression, as Interpol notices legitimize extradition requests from origin countries and reduce the perception among host governments that the request is political in nature (i.e., that the transfer of an individual to the origin country would be a potential human rights violation). Online appendix, table A8 includes an Interpol dummy variable where directed dyad-years are coded 1 if an Interpol red notice, diffusion request, or extradition request contributed to the transnational repression incident (detention, rendition, or unlawful deportation) and 0 otherwise. As expected, the results show that Interpol notices and requests increase the likelihood that countries will participate in transnational repression. Importantly, the inclusion of this new variable does not affect the study’s main finding on the influence of bilateral trade flows and the rule of law on cooperation in transnational repression.

Finally, to ensure that the link between trade and transnational repression is not being driven by an outlier, namely cooperation between Thailand and China (the dyad with the largest volume of transnational repression incidences), we test our hypotheses using a sample that excludes Thailand as a host country and China as an origin country in online appendix, table A9—and the results still hold. Together, these robustness tests provide further empirical support for our hypotheses.

Conclusion

While transnational repression is not a new phenomenon, the rate at which states target their citizens living abroad using repressive measures has increased in recent years due to the globalization of activism, technological advances in government surveillance, and international cooperation on migration (Freedom House 2021). Information and communication technologies have provided diaspora with new opportunities to engage in activism against their country of origin as well as increased the capacity of states to monitor, locate, and silence extraterritorial threats (Weidmann and Rød 2019; Dukalskis et al. 2022). Additionally, an increase in human movement, hardening of international borders, and perception of migration as a security issue have enabled abusive governments to expand their reach outside of their territory and target diaspora by co-opting international and regional institutions and securing assistance from foreign governments by framing cooperation as an effort to fight crime and terrorism (Cooley and Heathershaw 2017; Simmons 2019).

Although recent research has improved our understanding of this topic, most studies have focused on why origin countries engage in this repressive tactic rather than why host countries agree to cooperate (Shain 2010; Cooley Heathershaw 2017; Glasius 2018; Dukalskis 2021; Furstenberg, Lemon, and Heathershaw 2021; Tsourapas 2021; Moss, Michaelsen, and Kennedy 2022; Dukalskis et al. 2023). This is important since many instances of transnational repression would not be possible without international cooperation (Michaelsen and Ruijgrok 2023). To fill this gap, our article provides one of the first quantitative cross-national accounts of the political and economic factors that drive foreign complicity in transnational repression. Why do some states assist other countries to reach across national borders and repress their diaspora, while others do not? As expected, we find evidence that international cooperation on transnational repression relies on a country’s economic ties to the origin country and the quality of the rule of law. On the one hand, countries with a stronger rule of law provide fewer opportunities to repress because of their unwillingness to trade off normative and legal commitments to due process and human rights. On the other hand, origin countries are better able to leverage cooperation on human rights violations from states with shared economic interests that cooperate with one another in other areas of international politics. We also find support for an interaction effect whereby the most desirable transnational repression partners are those countries that are more willing to trade off civil liberties for economic cooperation (e.g., states with both a weak rule of law and greater financial dependence on the origin country).

These findings contribute to emerging research on extraterritorial repression and existing debates in international relations in several ways. First, the results suggest that domestic institutions and commitments to upholding the rule of law can be an effective tool for deterring international cooperation in silencing dissent. This finding complements existing research on the role robust legal and liberal democratic institutions, domestic norms, and public opinion play in constraining the government’s behavior on human rights (Davenport 2007; Kelley 2007; Lupu and Wallace 2019). Second, the results show that rather than improving human rights, economic interconnectedness can incentivize states to trade off human rights for material gains. This indicates that although the outcome is starkly different, illiberal states can use their financial linkages to pressure and buy cooperation on human rights violations in the same manner that liberal governments can leverage their economic position to induce state compliance with human rights (Hafner-Burton 2005; Dreher, Nunnenkamp, and Thiele 2008; Cao, Greenhill, and Prakash 2013; Peterson, Murdie, and Asal 2016). Third, although many studies suggest that violating human rights is politically costly for states (especially democracies), this study reconfirms that governments likely anticipate less public backlash for complicity in extraterritorial violations of foreign nationals than direct abuses of their own citizens within their jurisdiction. This finding speaks to research on the limitations of extraterritorial human rights obligations, the dynamics of refugee rights violations, and the potential challenges of attributing “derivative responsibility” to foreign governments for assisting another state in the commission of an internally wrongful act (Cerone 2008, 527; Heupel 2018; Abdelaaty 2021). Finally, studying the causes and dynamics of international cooperation on transnational repression has implications for research on whether globalization, the internet, migration, and institutions help or hinder the promotion of human rights and provide a cautionary tale on the cross-border security apparatus of authoritarian regimes (Brysk 2002; Simmons 2019; Gohdes 2020; Weidmann and Rød 2019).

The results also have important implications for policymakers and civil society activists who seek to prevent autocratic governments from reaching across national borders to repress their diaspora. If countries are more willing to assist trade partners in violating human rights, lawmakers should pay closer attention to renditions and unlawful deportations of diaspora to states with strong economic ties. Especially if they concern minority groups, political activists, journalists, former insiders/government elites, or human rights advocates (the most common targets of transnational repression). Additionally, efforts to strengthen the domestic rule of law in all countries (not just the most repressive states) appear to be a robust way to ensure that foreign governments comply with international human rights norms, due process, and principles on refugee protections. More generally, this study highlights the extraterritorial nature of repression and the interconnectedness of state practices with human rights.

However, it is important to keep in mind some potential limitations of this study. First, any dataset measuring transnational repression is only able to capture the “tip of the iceberg” of this global phenomenon, as many instances are likely to go unobserved due to the secret nature of human rights violations and government attempts to cover up, censor, and deny abuses (Freedom House 2021; Dukalskis et al. 2022, 2). There is also likely to be an underreporting bias in the Freedom House (2021) data that fails to accurately capture the involvement of countries with lower levels of press freedom and access to information since the database is based on public information. However, we prefer to use a conservative estimate of transnational repression to avoid including false positives in our data and do not find a country’s media freedom to be a statistically significant predictor of participation in our empirical models. Future research on this topic could use quantitative measurement models (e.g., latent variables) and qualitative methods (e.g., case studies and interviews) to corroborate and examine the dynamics of international cooperation in transnational repression in countries where it is least likely to be observed, as recommended by Dukalskis et al. (2022).

Second, our study only considers the theoretical mechanisms that explain why foreign governments decide to cooperate in transnational repression. While outside the remit of this article, examining cooperation between origin countries and non-state actors (e.g., cultural, educational, and religious institutions, diaspora groups, and criminal gangs) at the cross-national level is a crucial step in understanding how to prevent future acts of transnational repression. Future studies could build on our research to see whether certain host country characteristics (like the rule of rule), or relations between the host country and origin country (like trade) cause origin states to outsource repression to non-state actors or make it more logistically difficult or politically costly to co-opt non-state actors in their territory, even if the host country itself is not directly involved.

Finally, we have only examined physical instances of transnational repression in our study. New studies may wish to examine the factors that drive origin countries to engage in nonphysical instances of transnational repression (e.g., harassment, intimidation, and threats) in some host countries but not others. Additionally, future research on this topic may seek to disaggregate participation in transnational repression into the types of activities that a host country is involved in, as international cooperation in some repressive activities may be more politically costly than others (e.g., detentions versus renditions versus unlawful deportations).

More generally, while we have tried to triangulate our results and theoretical logic with reliable qualitative evidence, the covert nature of cooperation in transnational repression makes it difficult to know for certain how and why origin countries approach some host countries and not others. In addition to conducting in-depth qualitative case studies, examining global patterns and trends among Interpol requests is a particularly fruitful area for research on this topic as it provides us with an observable measure of when we know an origin country approach has taken place, whether the request was successful, and may allow us to trace the host country mechanisms through which cooperation was achieved.

Author Biography

Rebecca Cordell is an Assistant Professor of political science at the University of Pittsburgh. She received her PhD in 2017 from the University of Essex. Her research interests include state repression, human rights, political violence, and measurement.

Kashmiri Medhi is a Graduate Student of political science at the University of Texas at Dallas. She received her Master’s in political science from the University of Delhi. Her research interests include mass protests, political violence, and human rights.

Notes

Author’s note: The authors thank the editors of International Studies Quarterly, two anonymous reviewers, Paul F. Diehl, Allison Koh, Connor Nicholas Kopchick, Cyanne E. Loyle, Anton Sobolev, and Ezgi Yildiz for their helpful comments and suggestions. Prior versions of this manuscript were presented at the American Political Science Association Annual Meeting 2023, the International Studies Association Annual Meeting 2023, and the Political Economy, Public Policy, and Political Science Seminar at the University of Texas at Dallas.

Footnotes

References