Bordering Through Exemption: Extracontinental Migration Flows in Mexico

This paper examines Mexico’s governmentality of extracontinental migration in transit to the United States. It argues that, in the context of transit control regimes, exemption is instrumentalised as a bordering mechanism and practice in which transit states assume, react and utilise their role as a ‘transit’ country. By drawing on statistical information about migrant populations from Asia and Africa intercepted by Mexican authorities from 2010 to 2019, four arrangements are identified: (1) sporadic expulsion, (2) regularisation façade, (3) guardianship and (4) self-deportation. The analysis sheds light on the transformative and adaptive dimension of the Mexican Transit Control Regime and how this is geared towards maintaining its focus on intercepting and deterring Central American migrants in transit to


Introduction
In August 2019, after two weeks of protest outside the Mexican migration detention facility, Siglo XXI, 812 African migrants stranded in Tapachula, Chiapas, obtained the nullity of the migratory resolution issued the previous month that ordered them to leave Mexico through the southern border, basically targeting and preventing their transit mobility to the United States (Pradilla 2019;Romero 2019). The Mexican migration enforcement agency, Instituto Nacional de Migración, was requested to issue a new resolution allowing migrants to leave Mexico by the border of their choice. This migratory resolution was, in fact, an oficio de salida del país, a definitive departure (DD) resolution; however, it is commonly termed as a free transit agreement (FT) since for migrants, it means the possibility of reaching the Unites States' (US) southern border without the risk of being intercepted again by Mexican authorities. This case shows the use and instrumentalisation of exemption in the governmentality of extracontinental migration in transit through Mexico. Yet, it is noteworthy that this is not the first time that the Mexican authorities have exempted an organised group of migrants in transit from deportation; in 2014, in response to what is known as the first migrant caravan, they also granted this type of migration arrangement to 1,500 migrants 1 (Vargas Carrasco 2018). In this paper, I claim that the use of this and other exceptional measures is crucial in understanding the operation of the Mexican Transit Control Regime (MTCR) and its multiscale, transformative and adaptive characteristics.
While encounters with border regimes shape migrants' journeys and their strategies of mobility (Andersson 2014; Brigden and Mainwaring 2016;Schapendonk et al. 2020), border regimes are also influenced by and react to migrants' mobility strategies, resilience and stubbornness (Fassin 2011). Although this influence often leads to reinforcing and militarising the policing of migration (Jones and Johnson 2016), it might also lead to the implementation of exceptional arrangements, as shown in the previous vignette. This paper considers that exemption is crucial for understanding the policing and enforcement of migration control in transit states. It argues that exemption is a distinctive bordering mechanism and practice to govern the migrant population that is not the target population of the control regime and, therefore, a way in which the transit states assume, react and use their role as countries of 'transit'. Thus, although multiscalar bordering lays at the very core of this type of control regime (Wonders 2017), the examination of exemption in transit regimes draws attention to the multiscalar dimension of the micropolitics of bordering and the role of discretion therein.
Exemption as a bordering mechanism and practice is examined through the issuing of exceptional migratory resolutions and, more particularly, its instrumentalisation at the convenience of the control regime. This approach is inspired by Nancy Wonders' invitation to examine what she terms 'border reconstruction projects', defined as 'state sponsored strategies designed to reinforce and/or reconstitute borders in response to challenges posed to nation-states and borders by globalization ' (2007: 34). Hence, although the MTCR fulfils the characteristics of a 'state of exception' (Agamben 2005), in this text, I do not aim to examine the sovereign power at the border and the politics of migration control (Salter 2008;Tazzioli and De Genova 2020), nor the multiple ways in which migrants experience and contest them (Varela Huerta and McLean 2019). Conversely, this paper examines exemption as a constitutive and yet disruptive aspect in the governmentality of transit migration control in Mexico.
The issuing of exceptional migratory resolutions in the MTCR is used to bordering and ordering extracontinental migration in transit for practical and political reasons. By avoiding the elevated costs and other logistical difficulties involved in the expulsion of this population, the regime can uphold its focus on intercepting and deterring Central American migrants. That is, by exempting extracontinental migrants from deportation, the regime responds and recalibrates to continue achieving its aims; in the words of Nancy Wonders (2007), it 'reconstructs' itself. Yet, this does not mean that the logic of punishment and deterrence embedded in the regime is absent for these migrants; they, like the rest of the intercepted migrants, experience the state 'bordered penality' (Aas 2014;Barker 2017). However, what is different is that the result of the encounter between these migrants and the control regime is not expulsion but exemption. Thus, these migration arrangements are as much a part of Mexico's 'politics of hostility' IJCJSD 10(3) 2021 (Domenech 2020) as the expulsion of migrants. By shedding light and problematising the use and instrumentalisation of exemption in the MTCR, this paper aims to contribute to the growing literature on bordering and bordering practices in so-called transit states (Coddington 2020;Frowd 2020). This paper continues as follows: after a brief methodologies section examining the use of freedom of information (FoI) requests as a method of enquiry and data gathering, I describe the main aspects of the MTCR as a facet of the externalisation of the US border control and map the most recent reverberation of US border and migration control agenda in Mexico's governmentality of transit migration. I outline the main aspects of its implementation during the past 20 years, emphasising how Mexican migration policing has changed and the disputes and struggles that have pushed for transforming the regime's practices. After discussing the regime's bordering through expulsion approach, I delve into the examination of what I call bordering through exemption. I analyse Mexico's governmentality of Asian and African migration in transit and identify four migration arrangements: (1) sporadic expulsion, (2) regularisation façade, (3) guardianship and (4) self-deportation. These arrangements respond to an instrumental rationality and show the intertwining of legal and extrajudicial procedures, the high degree of discretion, and, ultimately, underline the central importance of logistics in bordering.

Methodology
FoI requests are used as a method of enquiry and data gathering. FoI, Keen (1992: 46) writes, 'offers the possibility of unparalleled access to the working of one of our most pervasive and influential formal organisations, the federal government'. Further, in the light of the increasing official secrecy, FoI has proven to be a powerful tool in social science research and particularly in policing and security research (Belcher and Martin 2019;Walby and Luscombe 2020). This paper draws on a quantitative dataset of migrants apprehended by Mexican authorities and their migratory resolutions between 2010 to 2019. The information was obtained through several FoI requests to the Mexican National Migration Institute (INM), the Mexican border enforcement agency, about migration control enforcing actions. The analysis presented below is particularly based on apprehensions and migratory resolutions of African migrants (n = 31,558) and Asian migrants (n = 30,491).
Although the use of FoI can overcome the methodological obstacles to investigate secretive or guarded organisations (Walby and Larsen 2012), we must not disregard the multiple methodological challenges faced when trying to use it as a method of data gathering systematically, including slowness of release, partial and incomplete data, and even the 'malicious non-compliance' (Roberts 1998: 9) but also technical aspects that hamper the systematisation and analysis of the information, such as lack of uniformity in datasets and poor quality of digitalised files. For the information examined here, the lack of uniformity of the dataset was especially challenging, with elements included or excluded, the decision to register nationality over the country of origin, or the naming of the migratory resolutions. As many other researchers have experienced (Bows 2019), all these intentional and unintentional obfuscations made it more complicated to conduct the statistical analysis. This lack of uniformity responds to the fact that the information was requested during two federal administrations (2012-2018 and 2018-2023), and given that in the INM, as in many other Mexican government agencies, public servants are hired on a temporal basis, it is most likely that they are removed from their positions when a new administration starts, particularly if it involves a transition of power like the one in 2017.
The FoI requests focused on migration enforcement actions, particularly in the registry created for each migrant upon entry to the detention facility. Hence, the NoData input is noteworthy. This means that, even though a migrant was registered as being intercepted and detained in a migration detention facility, the INM does not have information on the outcome of the apprehension, usually reporting it as 'null' or in a blank space. Although this could be framed simply as incomplete data, in fact, it shows something more profound about the operation of the regime. Through a bureaucratic lens, NoData indicates that, although 'clerical control' is a crucial aspect of the regime when a migrant enters a detention facility (Campos-Delgado 2021), there is not enough attention to the development of migrant cases and migratory resolutions. 'Documents-the tangible evidence of bureaucratic inscription', Sarah Horton writes (2020: 4), 'constitute a particularly useful site at which to analyze the power relationship between migrants and the state'. Thus, NoData outlines the administrative disarray but, more profoundly, shows that the INM is failing to be responsible and accountable for all the migrants, who, from the moment they are intercepted, remain in its legal custody; this failure could lead to illegal practices that would put migrants in a situation of even greater vulnerability.
Despite the methodological drawbacks to access and systematise the data, the information received from the INM is taken as certified and official. Yet, it is essential to note the existence of discrepancies within the data obtained and the INM's public information in the Yearbooks of Statistics 2 (Unidad de Política Migratoria 2019). For example, on the number of African migrants intercepted from 2010 to 2019, the public information and that obtained through FoI differs by 10,314 cases. These discrepancies, as well as other 'official silences' identified, could be interpreted as a way to conceal the magnitude of Mexico's role as a migration control enforcer and the multiple irregularities associated with this role. This underscores the strength of FoI as a critical method of enquiry to challenge and unveil these silences and irregularities.

The Mexican Transit Control Regime
'In the last 30 years', Nancy Hiemstra (2019: 44) notes, 'the USA has constructed a complex architecture throughout Latin America aimed at stopping migrants in transit before they reach US borders'. She terms this process 'the elasticisation of the US southern border' (2019: 54) and identifies the development of security 'partnerships' with transit countries as one of the mechanisms through which the US reterritorialises and disperses its border policing activities. The MTCR implemented in Mexico is the epitome of this process.
The MTCR was initiated in June 2001 through the US-led bilateral Plan of Action for Cooperation on Border Safety between Mexico and the US (US Department of State 2001). From this moment, with funding, equipment, and training from the US, Mexico's migration management and bordering practices have been primarily focused on intercepting and deterring irregular migrants in transit through the US (Campos-Delgado 2018). It should be noted that, although the regime's blueprint includes all irregular migrants in transit, in practice, the regime has focused on the Central American population, to the point that, from 2001 to 2019, the citizens of Honduras, El Salvador and Guatemala represented 91 per cent of the migrants intercepted by Mexican authorities (Unidad de Política Migratoria 2019). The regime's targeting is determined by geopolitics in the Central America-Mexico-US migration corridor and the predominance of Central American migratory flows, which, as analysed below, leads to differential enforcement of the regime. As for the policing of the regime, although highly concentrated in the southern border region, the MTCR is spread throughout the entire Mexican territory; this dissemination has created what Vogt (2017: 12) conceptualises as an 'arterial border', 'a complex social arena where migration flows are policed, exploited and contested'.

Twenty Years of Enforcement
The first 10 years of the regime's operation had a securitising approach, attested by the rebranding in 2005 of the INM as a national security agency. Intercepting and funnelling irregular migrants towards dangerous paths of mobility were the main bordering mechanisms (Amnesty International 2010). These, together with the proliferation of criminal actors as a result of the so-called 'war against drugs', which began in 2006, 3 and the impunity and corruption in the country, increased migrants vulnerability by making them a target for criminals (Bustamante 2010). Theft, extortion, rape and kidnapping are some of the most common crimes that migrants face when travelling through Mexico, with transit through the southern border region being particularly risky for them, mainly through Chiapas, Oaxaca and Veracruz (REDODEM 2019). That is, there is consistency between the highest concentration of border enforcement and the vulnerability of migrants. Hence, the multiple displays of violence faced by migrants are created because of, and as part of, the regime (Tazzioli and De Genova 2020).
The enactment of the Mexican Migratory Act in 2011 changed the discourse of migration enforcement in Mexico, incorporating humanitarian rhetoric. This was the outcome of a gradual decriminalisation path in Mexico in response to lobbying and awareness-raising involvement from organised civil society and migrant advocates (Basok et al. 2015;Gonzalez-Murphy 2013). However, it was also an official reaction to the national and international consternation caused by the San Fernando massacre in August 2010, where Los Zetas, a criminal organisation, murdered 72 migrants. The Migratory Act was enacted in May 2011 and sixteen months after the Act's Rules of Procedure. Nevertheless, what we see in practice is that humanitarian discourse is used to justify and legitimate severer border controls (Campos-Delgado and Côté-Boucher 2021).
During the last decade, 2014 and 2019 are key moments to exemplify the reverberation of the US border and migration control agenda in Mexico's migration control policy and policing. In July 2014, after the socalled crisis of unaccompanied minor migrants at the US southern border (Rosenblum 2015), the Enrique Peña's administration (2012-2018) launched the program Frontera Sur, 'with the purpose', Daniel Villafuerte Solis and María Aguilar wrote (2017: 50), 'of sealing the border and preventing Central American migrants from reaching the United States'. In June 2019, a Mexico-US joint declaration stated that 'given the dramatic increase in migrants moving from Central America through Mexico to the United States', the Andres Manuel Lopez Obrador administration (2018-2023) agreed to 'take unprecedented steps to increase enforcement to curb irregular migration, to include the deployment of its National Guard throughout Mexico, giving priority to its southern border' (US Department of State 2019: 1). Both episodes, although they correspond to different administrations and political parties, answered to the US migration control agenda with militarisation and greater repression of irregular mobility, creating greater vulnerability for migrants and violations of their human rights as well as intimidation and persecution against migrant advocates (CEJIL 2019; REDODEM 2019). 'A politics of hostility', Eduardo Domenech emphasises (2020: 4), 'is not limited, nor does it necessarily have to coincide with a period of government; it can precede it, exceed it, or determine a moment of varying durability'; the analysis of these episodes of strengthening border control in direct response to the US border-control agenda confirms that.
The humanitarian rhetoric on the Migratory Act, together with the increasing securitisation and militarisation of the migrant trail, has turned the MTCR into a fully consolidated 'military-humanitarian border' (Pallister-Wilkins 2015; Tazzioli 2015), and, therefore, as will be examined below, bordering practices are guided by this framing.

Bordering through Expulsion
The MTCR focuses on stopping and deterring Central American migrants in transit. For this population, the regime operates through an intercept-detain-expel circuit: 97 per cent of migrants intercepted by Mexican authorities are expelled (Unidad de Política Migratoria 2019). It is noteworthy that 99 per cent of these expulsions are framed as 'assisted returns' (or voluntary deportations), a bordering practice that, as Stephan Dünnwald observes, hides the mandatory feature and the coercive state's role by focusing on the returnee (2013: 232). Consistent with the intercept-detain-expel circuit, Central American migrants have a low chance of being granted refugee status in Mexico. From 2013 to 2018, Mexican authorities received 43,717 requests for refugee status from migrants from Honduras, El Salvador and Guatemala, which represented 73 per cent of the total, with percentages of acceptance of 19, 38, and 21 for Honduras, El Salvador, and Guatemala, respectively (COMAR 2019). Also, like in other migration control regimes, waiting is used as a technology to control and dissuade migrants requesting refugee status (Jacobsen, Karlsen and Khosravi 2021); by April 2019, out of the 43,717 requests received, only 41 per cent had been resolved by the Mexican authorities (COMAR 2019).
Yet, in the 'temporal bordering' (Tazzioli 2018;Khosravi 2021) inflicted on Central American migrants, waiting and haste are synchronised as migration management technologies. Celerity is the characteristic of the intercept-detain-expel circuit: on average, migrants are held in detention between three and five days (CNDH 2019: 93). This celerity is possible, at a normative level, because of the international agreements signed (2003, 2006, 2007 and 2009) between Mexico and Honduras, Guatemala, and El Salvador to regulate the expulsion or repatriation of migrants and, at a practical level, because migrants are deported massively by land.
Detention infrastructures and dynamics are also organised following this circuit and its celerity characteristic. Detention facilities are conceived and operated to hold migrants for short periods; out of the 59 detention facilities spread throughout the country, 14 are allowed to hold migrants no longer than 48 hours and 12 no longer than seven days. Most of these temporary facilities, and even some removal centres, have limited infrastructure; for example, there are no dining rooms, patios or recreation areas. When this is the case, migrants must remain inside their usually overcrowded cells all day (CNDH 2019: 136). Besides, as migrants are usually quickly processed and expelled, there is no laundry service, and they are not provided with clean clothes (CCINM 2017: 93). However, this should not be examined as an inadvertent structural deficiency of the regime, as I argued somewhere else (Campos-Delgado 2021); the precarious and dire conditions within Mexico's detention centres are articulated to control, punish and deter migration. 'The political dream of assisted return schemes', William Walters argues (2016: 438), 'is to provide a sufficient level of material inducement such that the migrant places themselves on the plane, without the need for guards, restraints or any spectacle of enforcement'. In the case of the MTCR, the appalling conditions of the detention centres could be seen as the material inducement for the migrants to sign their voluntary deportation.
The regime's aim for celerity, directly and indirectly, increases the vulnerability of migrants. Detained migrants are not informed or superficially informed about their rights, such as their right to seek asylum in Mexico, and are frequently coerced to accept to sign their voluntary deportation (CCINM 2017: 105). Once deported, given that the conditions they are seeking to leave behind have not changed, migrants are likely to travel again shortly and be involved in situations of greater risk and vulnerability; for instance, during 2018, eight out of 10 migrants from Guatemala, Honduras and El Salvador declared their intention to travel irregularly through Mexico again (EMIF-SUR 2019). In other words, expedited deportations are also aligned with the reproduction of the 'cross, apprehend, deport, cross' border circuit (Nail 2015: 31), which forces migrants into even more precarious and dire conditions.

Bordering through Exemption
This section examines the use of exemption as a bordering mechanism and practice in the MTCR. Drawing on statistical information on the Asian and African migrant population intercepted by Mexican authorities from 2010 to 2019, four migration arrangements are identified: (1) sporadic expulsion, (2) regularisation façade, (3) guardianship and (4) self-deportation.
From 2010 to 2019, Mexican authorities expulsed one per cent of migrants from Africa and six per cent from Asia. This percentage is striking when compared to the 97 per cent deportation rate for Central American migrants mentioned above. Although Mexican authorities do enforce deportations to migrants from Asia and Africa, the numbers remain insignificant compared to other forms of bordering and other groups. In other words, although detention and expulsion are used as the main mechanisms in the governmentality of Central American migration, sporadic expulsion guides the bordering of extracontinental flows.
According to Article 133 of the Migratory Act, the conditions under which irregular adult migrants can regularise their immigration status are (i) having a direct familial relationship with a Mexican citizen or a non-citizen with a residence status, that is, a spouse, parent or child; (ii) having been a witness or victim of a serious crime, delito grave, committed on national territory; or (iii) 'if the foreigner's level of vulnerability hinders or makes impossible his or her deportation or assisted return' (LM 2011: 50). When the migrant meets any of these requirements, the INM issues an oficio de salida de la estación migratoria, 4 releasing the migrant from detention. From 2010 to 2019, 60 per cent of migrants from Africa and 41 per cent from Asia were released to initiate regularisation. Yet, although this could indicate 'hospitality' within the MTCR (DeBono 2019), the real possibilities of regularisation are limited because of the INM's scarce infrastructure and deficient bureaucracies (Bustamante, Martínez and Montañez 2020). Thus, as Martha Rojas Wiesner and Tanya Basok (2020) examine through the concept of 'legal illegality', these limited possibilities push migrants to engage in alternative and sometimes illegal practices in order to be legalised such as acquiring fraudulent documents. Further, the probability of migrants obtaining protection through refugee status is low, and those who receive it must endure a long wait. In 2018, only three (out of 185) migrants from Asia and two (out of 119) migrants from Africa obtained this status. However, what is striking is not only the low number of approvals but also the high abandonment rate of 74 and 58 per cent, respectively (Unidad de Política Migratoria 2018). Most likely, they abandoned their refugee application in Mexico to resume their journey towards the US; this could be because they never intended to seek refuge in Mexico or because the poor Mexican institutional response, perhaps along with other socioeconomic factors, discouraged them. Thus, there is a regularisation façade used similarly and disruptively by both migrants and authorities.
Guardianship, that is, the release of a migrant into someone's custody, is another exceptional arrangement to release intercepted migrants from detention. According to Article 214 of the Migratory Act's Rules of Procedure, this arrangement can be made when 'the immigration authority has a legal impediment to resolve the immigration situation of a foreign person'; that is, when the INM did not provide a resolution after 60 working days after the migrant was intercepted. In this case, the custody of the migrant will be granted to a person or institution of recognised solvency whose purpose is linked to the protection of human rights' (R-LM 2012: 60). The initiation of the process must be by the guardian, and the bond is established at the discretion of the migration authority; this means that migrants without social networks in the country are less likely to have the possibility of being released through this exceptional arrangement. This explains the low representation of migrants from Asia and Africa in this case: under this arrangement, between 2010 and 2019, Mexican migration authorities released from detention 233 and 201 migrants from Africa and Asia, respectively, representing 0.7 and 0.6 per cent of the total of intercepted migrants from those continents.
Self-deportation is the quintessential exceptional arrangement. The FT and DD are used recurrently as formal figures to explain that migrants are released from detention and must leave the country by their own means; that is, they can freely transit in the country as long as they are moving towards a border. Although a conventional figure, the FT is not stipulated in the Migratory Act and, nonetheless, in the INM records, it is distinguished from DD, a migratory resolution that it does appears in the law as oficio de salida del país. 5 Yet, since, in practice, these terms refer to the fact that the Mexican transit control apparatus will turn a blind eye to migrants' mobility to the US, both terms are used interchangeably or are replaced with the euphemism salvoconducto de salida. The exemption from state deportation through these arrangements is issued at the discretion of agents due to the slow response of home countries, the migrants' lack of consular or diplomatic representation in Mexico or, as shown in the vignette at the beginning of this paper, due to political pressure (Narváez Gutiérrez 2015). From 2010 to 2019, 38 per cent of migrants from Africa and 52 per cent from Asia were issued a self-deportation migratory resolution.
Bordering through exemption responses to instrumental rationality shows the intertwining of legal and extrajudicial procedures and the high degree of discretion but also underlines the central importance of logistics in bordering. While deportation by land is the key factor for enforcing the bordering through expulsion, the significant financial outlays and logistical challenges posed by the expulsion of extracontinental migrants are one of the critical reasons for the bordering through exemption, and, thus, by circumventing these outlays and challenges, the regime remains afloat and continues to focus on intercepting and deterring Central American migrants.

Conclusion
From 2001 to 2019, according to official public information, Mexican authorities intercepted 2,613,915 irregular migrants and expelled 2,398,419 (Unidad de Política Migratoria 2019); however, when closely examining the reports of migration resolutions, differential enforcement of the regime is revealed. In this paper, I argued that Mexico's 'politics of hostility' (Domenech 2020) towards irregular migrants in transit to the US could not be understood simply by examining the MTCR bordering through expulsion approach but in direct relation to its bordering through exemption strategy. Thus, I claimed that the issuing of exceptional migratory resolutions in the MTCR is used for bordering and ordering extracontinental migration in transit, but more importantly, to maintain the regime's focus on intercepting and deterring Central American migrants.
Through the analysis of the MTCR, I made a case on exemption as a distinctive bordering mechanism and practice to govern the migrant population that is not the target population of the control regime. I emphasised that by bordering through exemption, the transit states assume, react and use their role as a country of 'transit'; ultimately, this adaptive and transformative aspect of the control regimes sheds light on the multiscale objectives of border control practices and the micropolitics of bordering.
Further, although a naive reading might consider that the instrumentalisation of exemption works, to some extent, for extracontinental migrants' mobility objectives, it does not. The normalisation and routinisation of exemption highlight the deepening of the regime's 'abnormality' (Aas 2014;Campos-Delgado 2021). This 'reconstruction project' (Wonders 2007) is anchored in a migration securitisation framework and responds to the US border-control agenda. Therefore, it is also part of the politics of invisibility that seeks to obscure the moral and ethical implications of this type of border regime.