Calculated Informality in Governing (Non)return: An Evolutionary Governance Perspective

ABSTRACT

Afghans’ protracted displacement is a geopolitical legacy from the Cold War. Although Pakistan’s return policymaking has foreseen the complete voluntary return of Afghans since the end of the Cold War, then as now, about three million Afghans reside in Pakistan. This article advances the notion of calculated informality to dissolve this seeming contradiction. Pakistan’s policymakers have excelled in calculated informality by successfully navigating the domestic and geopolitical arena over time based on practices of deregulation and ambiguity. Methodologically, the article applies Evolutionary Governance Theory (EGT) to reconstruct Pakistan’s return governance path based on the analysis of legal documents, previous research and secondary literature. EGT reveals the dependencies and layering at work in return governance and points out how the geopolitical positionality of Pakistan has determined its return policymaking. The structured interconnectedness of path-, goal- and interdependencies illustrates rigidities of the governance path and why opacity and ambiguity in return governance persist.

Introduction: Governing (Non)Return

As the European Union’s migration policymaking concerning countries of origin in the so-called Global South is re-designing the geopolitical space around the Mediterranean (Zardo 2020), the protracted displacement of Afghans – many of whom also migrate to the EU via this route – constitutes a geopolitical legacy from the Cold War. The neighbouring countries, Pakistan and Iran, bore the brunt of immigration from Afghanistan and still do so today. Currently, up to 2.3 million documented Afghans reside in Pakistan alone; the number of those undocumented is unclear but likely amounts to around one million, thus resulting in a total of more than three million Afghans.¹ This is against the background that Pakistan has already established a return governance regime from the early 1990s. From its inception, it focused on (the complete) return of Afghans to Afghanistan as the only viable option among the three so-called durable solutions (Zieck 2008, 257–258). In this article, I will endeavour to make sense of this seeming contradiction between policy goals – framed as full return – and outcomes manifest in the continued presence of an estimated three million Afghans in Pakistan. Given that the conflict in Afghanistan has entered its fifth decade and the international protection regime stresses the right to asylum and demands durable solutions, the observer is inclined to deduce that Pakistan is responding to the need of the day, which may be framed as a humanitarian act and necessity to tolerate and assist Afghan refugees. However, I argue that this interpretation applies only partially. Pakistan is not a signatory to the International Refugee Convention 1951 and its 1967 Protocol. The country’s refugee governance framework consists of few basic regional agreements, ad hoc and irregularly issued, mostly responsive national policies lacking a national refugee law, and complementary implementation mechanisms that manifest in heterogeneous practices such as Pakistani dominance and rejection, confrontation and empowerment of Afghans. Afghans have reacted to the return policies with efforts aiming at avoidance, exit, compliance, remigration and adaption over time. However, the agency of Afghans has been diminishing steadily (Mielke et al. 2021).

In a novel conceptual contribution to this special issue’s theme on return governance, I introduce the notion of calculated informality as governance mechanism that characterises Pakistan’s (non)return governance path. I argue that the seeming contradiction between the political rhetoric of return and de facto non-returns of Afghans from Pakistan does not constitute a ‘failure’ of governance or steering. Quite to the contrary, it is indicative of how Pakistan’s policymakers have been successfully navigating the geopolitical arena over time by excelling in calculated informality (Roy 2009, 2018) based on practices of deregulation and ambiguity. I employ the framework of Evolutionary Governance Theory (EGT) to draw out how calculated informality in Pakistan’s return policymaking is multi-layered and entangled with broader governance evolutions. These encompass the institutional and domestic political context; however, in addition, Pakistan’s return governance path reflects the government’s attempts to negotiate its geopolitical positionality in the international and regional political arena. EGT’s analytical focus on a governance path’s underlying path-, inter- and goal-dependencies allows concluding that neither Afghans nor the United Nations Refugee Agency (UNHCR) or other agents possess the ability to induce transformative change in the protracted displacement situation of Afghans in Pakistan. Moreover, the rigidity of the return governance path silences alternative voices from within the policymaking institutions in Pakistan who are protagonists of local integration and/or dual nationality. As a result, prospects for ‘durable solutions’ to Afghans’ displacement in Pakistan are dim, despite the regional governance framework and rhetoric commitment to protection.

It is important to note from the beginning that the very idea of return as a (durable) solution to (protracted) displacement is a highly charged geopolitical notion (Chimni 1998). It rests on several problematic assumptions, such as methodological nationalism (Wimmer and Glick Schiller 2003) and a sedentary bias which manifests in the idea that immobility is the normal state of being (Monsutti 2008). Scholars of (Forced) Migration and Refugee Studies (Horst 2018; Malkki 1995) and many other (trans-)disciplines have long pointed to the importance of translocal connectivities (Carling, Erdal, and Talleraas 2021). They make the case that mobility trajectories of displaced persons and migrants reflect transnational lives, even if the degree of mobility is not constant (Brun 2015) and transnational lives are exposed to various influences, including geopolitical ones. Incentivising, pushing, supporting or facilitating returns is thus in line with the geopolitical interests of host, origin and possibly third states. Moreover, return can be highly problematic in humanitarian terms (Fielden 1998, 480), especially when provisions or practices such as irregularization, dehumanisation, securitisation, and generating deportability violate the principles of asylum-seeking and non-refoulment (De Genova 2002; Gerver 2019). Pakistan stipulates repatriation of Afghans to Afghanistan as the preferred and de facto only durable solution (UNHCR 2012), even though the identity of second-, third- and fourth-generation Afghans as refugees is ambiguous at best. They do not share the displacement experience, and their notion of Afghanistan as ‘home’ they ought or want to return to is vague because they have never been there. Nevertheless, in the logic of the geopolitically informed return governance framework, Pakistan insists that return ought to be complete(d).

In the following, I hope to shed light on the field of tension between refugee protection (asylum and non-refoulment), geopolitically informed return policies and practices, and their underlying mechanisms. I start out with a clarification of the conceptual and methodological approach. Subsequently, two empirical sections serve to reconstruct Pakistan’s return governance path from a historical and practice angle of policy and policy-agency encounters. The subsequent discussion will reflect on the path- and interdependencies that structure the return governance path and shape ambiguities of calculated informality. I next reflect on the implications of the analysis for the situation of Afghans in Pakistan. Concluding, I summarise and contextualise the main findings and underline the analytical value of employing an EGT-lens for understanding return governance.

Calculated Informality as Governance Mechanism of Pakistan’s Return Governance Path

A decade ago, Ashutosh and Mountz (2012) discerned a research desideratum regarding the relevance of geopolitics in return and/or repatriation. A brief review of the literature on return governance indicates that this void is still valid. Scholars pointing out how biopolitics forms a central element of the geopolitics of mobility and how potentials for (embodied) ‘geopolitics from below’ stem from refugees’ agency (Hyndman 2012) have advanced Chimni’s (1998) important contribution on the geopolitics of asylum. Mobility and migration scholars have emphasised the political nature of repatriation as a process (Gerver 2019; Long 2013). De Genova (2002, 2020) pointed out that the law governing displaced and migrant populations should constitute a key object of study itself and appealed to scrutinise how ‘the law’ shapes migrants’ and refugees’ mobility and how it is ‘resisted from below’. Scholars focusing on the policymaking side of refugee governance have recognised the significance of uncertainty, informality, liminality, and exceptionalism in migration policymaking. Şahin Mencütek et al. (forthcoming) suggest the concept of strategic temporality to function as a governance mechanism in migration management. Stel (2020) elaborated the notion of institutional ambiguity as a refugee governance strategy pointing out the politics of uncertainty in Lebanon’s response to the Syrian refugee crisis (Nassar and Stel 2019). Accordingly, institutional ambiguity is shaped by refugeeness, defined as the agentic dimension of experiencing forced displacement, and governance as ‘the organization of public authority’ (Stel 2020, 3). Both constitute national-level derived determinants for ambiguous refugee governance. Tsourapas (2019) and Micinski (2021) developed the concept of refugee rentier states around the idea that hosting states construct and instrumentalise the deportability of refugees (De Genova 2002, 2020) to pursue their foreign policy goals, e.g., increasing their geopolitical leverage and importance and extracting aid (Micinski 2021, 5).

This article speaks to and enhances preceding strands of scholarship by putting forward the concept of calculated informality as an alternative notion to account for informality in refugee (return) governance. The term calculated informality was coined by Roy (2009), an urban study scholar, when analysing the planning crisis in Indian cities. Inspired by her findings on urban governance (Roy 2009, 86), I argue that it is exceptionality and the ability of the governing elites to suspend the law or use its absence to negotiate return governance outcomes in its own interest that explain the puzzle of decade-long return orthodoxy versus de facto non-return of Afghans living in Pakistan. Calculated informality thus constitutes a strategically and purposefully applied governance mechanism reflecting a state-sanctioned mode of deregulation. As the main interest of the Pakistani elite in foreign policymaking is to maintain Pakistan’s geopolitical positionality in relation to India, the following analysis adds a crucial geopolitical dimension to understanding return (non)governance that is neither included in Stel’s model of institutional ambiguity nor sufficiently covered by the notion of the refugee rentier state.

I am introducing Evolutionary Governance Theory (EGT) to draw out how ambiguity and deregulation in calculated informality has evolved layered and in close entanglement with the institutional and (geo)political context. According to EGT, governance is understood as a history of steering attempts, a management act that reflects how binding decisions are being coordinated through the mutual interplay of actor/institutional configurations and knowledge/power configurations. Works in critical geopolitics (O’Tuathail 2005), critical geopolitics of mobility (Chimni 1998; Hyndman 2012), and much of the scholarship in Forced Displacement and Migration studies all share the EGT assumption of complex entanglements between actor/institution and power/knowledge configurations. Actors are not distinct units of analysis; instead, it is assumed that actors and institutions mutually shape each other and co-evolve in close relation to power and knowledge co-evolution (van Assche, Beunen, and Duineveld 2014, 5). I use ‘Pakistan’ and ‘Government of Pakistan’ (GoP) or ‘the state’ synonymously below to represent this entangled actor/institution cum knowledge/power configuration as composite. Knowledge/power configurations serve, characterise, and even create actors (van Assche, Beunen, and Duineveld 2015, 24) as discourses create subjects and objects at the same time. Thus, even if we see Afghans in Pakistan as objects of return governance, EGT encourages us to reverse the gaze and scrutinise Afghans’ agency as subjects in governance. In this proposition, EGT shares a concern with scholars of critical geopolitics who conceptualise agency in the framework of geopolitics ‘from below’ (Hyndman 2012). Rooting in social-constructivist and poststructuralist thinking, the main idea of EGT is that governance is constantly evolving. As EGT scholars suggest,

This constructivist perspective is not a denial of an external reality, nor does it deny the agency of actors and institutions. It is the basic recognition that the ways in which humans understand and govern their environment are contingent structures. Contingent in the sense that these are the result of ongoing interactions between discourses and material realities […]. They are also contingent in the sense that the evolution of governance elements and their relations shows pathways in which past, present and future are linked (Beunen, van Assche, and Duineveld 2015, 138).

Institutional entanglements and underlying power dynamics and mechanisms characterise and structure any individual governance path, thereby constituting an inherent logic based on path-dependencies, inter- and goal-dependencies that enables us to account for the past, present and future of return governance trajectories (van Assche, Beunen, and Duineveld 2014, 29–32). The concept of path dependency expresses that certain (co-evolving) elements of governance might be subject to restrictions by the existing governance framework and its history of evolution. Return policymaking is thereby seen as both process and outcome of constant knowledge and power adjustments that underlie the revalorisation of policy provisions depending on the perceived ‘need of the day’. Interdependencies can be detected where relations between actors restrict actions, for example, when geopolitical considerations, e.g., a country’s geopolitical positionality (Sheppard 2002) shape return governance planning and decision-making, thereby influencing the institutional framework imposed ‘from above’, or when modes of agency as embodied ‘geopolitics from below’ (Hyndman 2012) cause policies to shift course. The notion of goal-dependencies captures how visions of actors’ shared future translate into plans and policies in the present that act upon the scenarios with a spectrum of possible strategies, ranging from facilitating to subtly undermining or reframing, to opposition. These dependency-relations provide for a structured connectedness and determine the degree of rigidity and flexibility of a governance path (van Assche, Beunen, and Duineveld 2015, 28).

Methodology

Methodologically, this research requires reconstructing the return governance path and contexts since 1980 through path mapping and scrutinising possible co-evolving goal-, path-, and interdependencies. Given the protractedness of Afghans’ displacement in Pakistan and co-evolved return governance manifest in policies, it is necessary to carry out a post-hoc categorisation of key choices and positions in governing returns and key shifts in the goals and implementation procedures. To identify possible path-creating and sustaining structures, I will focus on governance sites as historical junctures (Chimni 2009, 18) that reveal when and where decisions were taken or prepared that led to certain policy shifts (Beunen, van Assche, and Duineveld 2015, 340) within the overall return governance trajectory. For evidence, this paper relies on an analysis of the available primary sources, such as governmental and inter-agency programmes and policy documents concerning the governance of refugee returns and policies towards Afghans in Pakistan over the years wherever possible. However, the more informal the policy or policy shift, the less the likelihood that documentation about the policy has been available to this author. Thus, the developments and path trajectory had to be reconstructed through process tracing with the help of secondary sources. Interview material documented in previous research on Afghans in Pakistan (Grawert and Mielke 2018; Mielke 2016; Mielke et al. 2021) complements the database.

Reconstructing Pakistan’s Return Governance Path

The following two empirical sections serve to reconstruct Pakistan’s return governance path from two distinct epistemological angles. First, the diachronic tracing of the governance path along return governance sites (Beunen, van Assche, and Duineveld 2015, 340) distinguishes four periods 1980–1992, 1992–2001, 2002–2014, and 2015 to date. These governance sites reflect crucial phases and turning points of return policymaking in relation to geopolitical developments. Second, a practice perspective on policy–agency encounters draws out how policies and accompanying operational practices resonated with Afghans and were responded to.

Four Return Governance Sites: the Diachronic Perspective

1980–1992: 10 years after the onset of the armed conflict in Afghanistan, the 1988 Geneva Accords about Soviet troops’ withdrawal from Afghanistan constituted the raison d’être for governing the return of an estimated 3.2 million registered Afghan refugees from Pakistan (GoP 1997). After the 1979 intervention of the Soviet Army to support the Afghan Marxist government that had toppled the republican regime in a coup d’état in April 1978, a proxy war unfolded between the Soviet Union and the United States (US), causing an exodus of Afghans into Pakistan. This provided Pakistan’s military ruler with a welcome opportunity to bolster Pakistan’s significance as an ally in the US-framed war against communism. The respective Bilateral Agreement between the Republic of Afghanistan and the Republic of Pakistan on the Voluntary Return of Refugees holds that the Government of Pakistan (GoP) ‘shall facilitate the voluntary, orderly and peaceful repatriation of all Afghan refugees staying within its territories … ’ (UNSC, 1988, Art. III, 8) with the cooperation and assistance of UNHCR (Art. VI). However, although the security threat of Soviet presence was removed by February 1989, the Soviet-backed government in Afghanistan remained in place. Mass repatriation did not commence in 1989, which is partly explained by the active prevention of returns by the so-called Afghan (resistance) parties in Pakistan (UNHCR 1990, 1), e.g., through intimidation and roadblocks (Ghufran 2011, 948). Neither the United Nations Good Offices Mission in Afghanistan and Pakistan (UNGOMAP) – established as peacekeeping mission in May 1988 to assist the Accord’s implementation – nor UNHCR could facilitate the voluntary return of Afghans in the first 18 months (Baczko and Dorronsoro 2015, 273). The situation did not change substantially until the collapse of the Soviet-backed government in Kabul in April 1992. However, the refugee management in the early 1980s likewise provided several anchor points for path dependencies in the following decades. For one, the UNHCR was toothless because upon its entering Pakistan in April 1979, the GoP dictated that all aid for refugees would be channelled through government authorities (Schöch 2008, 5). The GoP delegated the registration of refugees to the Afghan parties who exercised close control over the Afghan exile population they claimed to represent (Schöch 2008, 12). Consequently, registration as a refugee was highly selective and not in line with UNHCR’s initial prima facie approach (Martin 2000, 74). Within 2 months after the Soviet invasion, Pakistan set up the Chief Commissionerate for Afghan Refugees (CCAR) in February 1980 to administer refugee aid. The CCAR and its provincial branches counted between 6,000 and 9,000 staff members in the subsequent years (Schöch 2008, 14).

1992–2001: The key governance framework of the second return governance site is the 1993 Tripartite Agreement with the post-communist Afghan government and UNHCR on the Repatriation of Afghan Refugees in Pakistan (Zieck 2008, 258). Following a policy-shift of UNHCR at the global level (Langenkamp 2003, 235) and the announcement of a ‘decade of return’ after the end of Cold War bipolarity, UNHCR aimed to repatriate and support the successful reintegration of 100,000 Afghans annually. Repatriation was aided by renewed registrations (verifications of previously distributed ration passbooks) which constituted the basis for assistance in voluntary returns. Despite a civil war unfolding among the former resistance parties inside Afghanistan from 1992 onwards and causing renewed displacement, the international community decreased its aid for Afghan refugees substantially because Afghanistan had lost its geostrategic importance for the previous sponsors of anti-Soviet warfare. UNHCR assistance for voluntary returns had to stop in late 1998 because of insufficient funding. The loss of attention in the international spotlight and of subsequent funds (Langenkamp 2003, 236–37) caused Pakistan to take harsh measures throughout the decade to create push factors for Afghans to return and block new mass immigration. On a discursive level, Pakistan justified this policy with the international community’s lack of burden-sharing via funding and resettlement. In 1997, the GoP diplomatically recognised the Taliban-led Islamic Emirate of Afghanistan and subsequently urged Afghans to return, employing a ‘safe country of origin’ narrative stressing that 95% of the territory was safe and peaceful. In 1998, the GoP abandoned its previous prima facie recognition of Afghan arrivals and denied entry without visas. This illegalisation of new arrivals resulted in deportations. However, forced returns were carried out selectively. They mainly targeted non-Pashtun ethnic minorities, such as Hazara in Quetta even though they were most at risk of persecution under the Taliban and subsequently made up the bulk of those 100,000 Afghans who entered Pakistan in 1999 versus the 91,834 individuals who repatriated voluntarily. Reportedly, Pakistani authorities handed illegal arrivals over to Taliban authorities until 2001 (Langenkamp 2003, 243).

2002–2014: The 9–11 terrorist attacks of 2001 marked a tectonic shift in the geopolitics of the Afghanistan–Pakistan region. For return governance, they rang in a ‘second decade of repatriation’ (again as such propagated by UNHCR) facilitated by UNHCR, IOM and other (I)NGOs after the transitional government for Afghanistan was put in place. A renewed high inflow of funds to assist refugee returns – as returns were seen as an indicator for the success of the international community’s state-building efforts in Afghanistan – was characteristic for this period from early 2002 until the end of the so-called transition period in Afghanistan 2014. As an ally in the Global War on Terror, Pakistan enjoyed renewed patronage from the US that bolstered its geostrategic position in the region, especially towards India. When the second Tripartite Agreement between the Governments of Pakistan and Afghanistan (GoA) and UNHCR was signed in 2003 (UNHCR 2003), between 1.5 and two million Afghans were documented in Pakistan (UNHCR n.d.), the number of undocumented Afghans being unknown. When far more Afghans were repatriated than were suggested to reside in Pakistan in early 2002, and refugee camps initially did not seem to empty, the GoP and UNHCR signed a Memorandum of Understanding (MoU) in 2004 for conducting a census in 2005. Another MoU followed in 2006 (UNHCR 2006), regulating the prima facie registration of Afghans (Zieck 2008, 265) and providing 3 years temporary protection against refoulment until the end of 2009. Of the enumerated three million Afghans in 2005, 2.1 million registered until spring 2007, rendering those who remained unregistered and had not meanwhile repatriated ‘illegal’ immigrants and subsequently not of concern to UNHCR.

Repatriations continued in high numbers until 2008 but stagnated thereafter (UNHCR n.d.), mainly due to deteriorating security conditions inside Afghanistan, which caused large-scale secondary displacement after return (Mielke 2016). An additional Tripartite Agreement (Zieck 2008, 264) referred to the GoA’s limited absorption capacity of returnees and introduced the notion of gradualism related to repatriation, which led to an extension of the repatriation strategy beyond 2009 (UNHCR 2010) and its linking with the GoA’s national development strategy (ANDS) 2009 to 2013 (Ahmad 2017, 642). In a move to adapt its return governance framework and shift from humanitarian to development measures (Ahmad 2017, 646), the GoP adopted a very ambitious Afghan Management Repatriation Strategy 2010–12 (AMRS), which, however, remained a paper tiger. Lack of internal political support prevented the enforcement of measures that aimed at regularising Afghans through the provision of visas and work permits. Only the regional Solutions Strategy for Afghan Refugees (SSAR) to Support Voluntary Repatriation, Sustainable Reintegration and Assistance for Host Countries (SSAR 2012) initiated in 2012 enabled development measures in the host countries (including Iran) and Afghanistan, based on the narrative of empowerment as a prerequisite to achieving Afghans’ self-reliance and resilience (UNHCR 2016). While the SSAR recognised that the sustainability of Afghans’ repatriation depends on a domestically conducive environment for reintegration, it also perpetuated the conventional durable solutions framework with a renewed focus on return as the only viable solution to Afghans’ protracted displacement.

2015 to date: Simultaneously with the conclusion of the security transition phase in Afghanistan, a domestic terrorist attack at a public school run by the Pakistan Army in Peshawar in December 2014 caused several casualties among the children. The media and the public blamed it immediately on Afghans, although later proof showed that Afghans were not involved. Nevertheless, this incident yielded severe consequences for the situation of Afghans in Pakistan. With domestic terrorism on the rise in Pakistan throughout the 2000s, the activities of the Pakistani Taliban Movement (Tehreek-e-Taliban Pakistan) had become discursively connected to Afghans’ presence in the country. The Peshawar Army school attack led to the adoption of a nationwide National Action Plan (NAP), a 20-point counter-terrorism package still in 2014. Of that, point 19 foresaw the ‘Formulation of a comprehensive policy to deal with the issue of Afghan refugees, beginning with registration of all refugees’ (NACTA 2014). As a result of the NAP, political pressure on registered Afghans increased, and unparalleled discrimination and harassment by police and security organs (HRW 2015) – combined with increasing rejection by the host community – caused 381,300 Afghans to return to Afghanistan in 2016 alone, more than had returned voluntarily between 2011 and 2016 (UNHCR n.d.). UNHCR was accused of aiding de facto forced repatriation and thus mass refoulment in 2016 after it had doubled its cash incentives for those willing to return in summer 2016 – under heavy pressure exerted by the GoP (Haider 2016; HRW 2017). The GoP’s Refugee Management Policy 2017 prescribed the introduction of a visa regime for non-registered Afghans in Pakistan, similar to the provisions included in the previous AMRS. In June 2016, border controls were introduced at Torkham border crossing, requiring all Afghans to have valid documentation (IOM 2017, 1) and thereby sealing the border for all those Afghans residing in Pakistan who regularly crossed as part of transnational living. After a large-scale registration exercise in 2018, 804,942 undocumented Afghans were issued Afghan Citizen (AC) cards that protected them as ‘illegal’ labour migrants or stateless against refoulment by the police until they had acquired a passport and visa regularising their presence in the country based on an MoU between the GoA and GoP. This schedule was not met, and the deadline was extended several times until 30 June 2020 and is since pending for extension, thus in effect irregularizing all AC cardholders in Pakistan as ‘illegal aliens’ and without protection, a practice De Genova (2002, 429) coined as the ‘legal production of (migrant) illegality’ in another context.

Regional and National Policymaking Meets Afghans’ Agency: the Practice Perspective

The four governance sites introduced the regional agreements that constitute the formal policy framework governing returns. Over time, these consisted of the mentioned Geneva Accord, the trilateral agreements of 1993, 2003 and 2007 and their respective extensions until 2021, and the regional 2012 SSAR with extensions. Complementing these, in a second layer of return governance, are national policies of an ad hoc, irregular, and mostly responsive nature, while a national refugee law is lacking. The policies originate from different authorities (e.g., the Ministry of the Interior and different courts, including the Supreme Court) besides the Ministry of States and Frontier Regions (SAFRON) and the CCAR, and the Chief of Army Staff (as in the case of the NAP). Besides UNHCR and (I)NGOs, they rely on registration authorities like the National Database and Registration Authority (NADRA) and security organs such as army personnel and the police for implementation. In this section, I will further distinguish a third layer consisting of operational processes that accompany regional and national policy implementation, marking, for example, practices such as dominance, rejection, confrontation, or empowerment towards Afghans. The fourth layer is constituted by the return policy-resonating space or sites of Afghans’ agency that incorporate different strategies in response to Pakistani state practices and government techniques – ranging from efforts at avoidance, avoidance and exit, avoidance followed by exit and remigration, adaption, and/or compliance.

A continuous pattern in policy–agency encounters within the wider refugee return governance framework is that of dominance (dominating policy) and compliance (complying objects). Dominance–compliance patterns constitute a path-dependent interaction mechanism that roots in the relationship between Pakistani authorities and Afghan refugees in the 1980s. With the institutional infrastructure (CCAR, CARs under SAFRON) in place to administer Afghan ‘refugees’, the latter’s dependence on the former was undeniable because the option to withhold or approve supplies to the Afghan parties constituted a powerful steering instrument during times of significant aid flows (Schöch 2008, 7; Martin 2000, 76). Seen from this angle, Afghan ‘refugees’ on Pakistani territory constituted an important political asset for the GoP and the base for refugee rentierism (Tsourapas 2019) in much of the 2000s and 2010s. In acts of confrontation, the GoP conducted the census of Afghans in 2005 and the subsequent registration exercise in 2006/07. Afghans reacted with compliance and evasion because the purpose of both was not transparent, and many Afghans feared that it would serve deportation rather than protection. Those who did not register and stayed undocumented were illegalised and henceforth at risk to be deported if caught – if they could not adapt by obtaining fake ID documents or luring the police with another strategy, such as fluency in the local idiom or simply paying money as a bribe (Grawert and Mielke 2018, 22, 37; Mielke et al. 2021).

The evolved documentary regime, comprising PoR cards, AC cards, passports and visas, constitutes a government technology of dominance and as such embodies a strong biopolitical dimension of geopolitics (Hyndman 2012) in return governance. Alimia (2019, 397) made the point that PoR cards (and I add AC cards) have been used as surveillance techniques to assist the profiling of Afghans. The cardholders, who had to provide bio-features like iris scanning, are registered in cross-checkable databases of NADRA, which required several rounds of re-issuing cards between 2009 and 2015. The ad hoc irregular extension of the cards’ validity has exposed cardholders to numerable offences and harassment by the police and security authorities (Mielke et al. 2021; Grawert and Mielke 2018, 20–23, 37). The long non-extension of PoR and AC cards that expired on June 30, 2020 constituted the latest manifestation of rejectionist policies because during the time of uncertainty over extension it deprived Afghans of basic protection against non-refoulment. PoR cards have been reissued as smart cards until end of December 2021 with a validity until 30 June 2023. AC card holders continue to live in legal limbo. They are left at the mercy of local police and security personnel on the one hand and their network of Pakistani friends, neighbours, and colleagues on the other. Informally, the omnipresent venality in Pakistani society enables Afghans to navigate adverse policy contexts, for example, by moving around with borrowed national ID-cards (CNICs) from Pakistani friends or by bribing police officers when stopped and checked for valid documents.

Besides these strategies of avoidance and adaptation, exit has been a common reaction to adverse and rejectionist Pakistani policies. Afghans were deported, returned (de facto) coerced or voluntarily, reminding the observer of what Gerver (2015) described when stating in another context, ‘each individual’s ultimate decision to return was not quite a choice to repatriate, but a choice to take particular actions for particular ends, with repatriation a means to achieve these ends’. In many cases, repatriation preceded the immediate or later remigration to Pakistan. Remigration is behind much of the confusion about repatriation numbers vs. the number of Afghans still residing in Pakistan. Many Afghans left the camps and deregistered in the 1990s and early 2000s but either did not return or returned and remigrated, even multiple times, for example, to cash in repatriation assistance more than once, or as part of circular migration patterns and transnational living. From early on, Afghans often chose seasonal return, with one part of the family remaining in the refugee camp. Alternatively, they repatriated to scope out local conditions inside Afghanistan and prepare their family’s return in the medium to long term (UNHCR 1990, 1).

These transnational mobility patterns of Afghans caused Pakistan to tighten its physical border in a counter-rejectionist move paired with militarisation (Safri 2011, 595). From the overview of the governance sites, it is obvious that the Afghanistan–Pakistan border has undergone a shift in border performativity from the late 1980s until today, resulting in an ever-increasing securitisation of the border. The physical border and its borderlands have seen a shift towards offensive border management from 2001 onwards, discursively legitimised with the risk of terrorist overspill from Afghanistan (Margesson 2007). This involved the successive installation of fences and ditches along the entire 2,430-kilometre border since 2005 (Oztig 2020, 216). Given that cross-border mobility could not be stopped while domestic terrorist groups like the Pakistani Taliban, who were blamed to have roots in Afghanistan, posed an increasing challenge, refugee camp removals and military operations in the tribal areas of the country’s northwest further militarised the border (Safri 2011, 595). This securitisation drive culminated in Pakistan’s introduction of the shoot-to-kill order at the border to prevent illegal flows in 2017 (Oztig 2020).

Discussion

Based on the empirical insights, I will now reflect on the discernable path- and interdependencies that have structured the evolution and presence of Pakistan’s return governance path. I show that Pakistan’s return governance path is characterised by a discursive and operational logic that is formal and informal at the same time and highly flexible in adapting to changing contexts such as geopolitical shifts at the global and regional scale. The informality observed does not suggest a failure of steering but is on purpose and as such calculated. It stems from the perceived necessity by Pakistani state agents to maintain a balanced geopolitical positionality in the region, especially in relation to India. The calculated informality in the governance of (non)return frames the continued presence of Afghans in Pakistan who are not granted asylum rights but are (mis)used as biopolitical mass exposed to omnipotent disposability and increasing void of agency.

Pakistan’s Return Governance Is Highly Informal

Pakistan did not sign the 1951 Refugee Convention and its 1967 Protocol, nor has it ever issued a national refugee law. The refusal to become a signatory of the Convention constitutes one anchor point of departure for a path dependency that grounded the state’s successively evolving approach to ‘managing’ Afghans with a temporary status. Because of the lack of these two types of essential legal protection frameworks, which are standard elsewhere, protection has always relied on alternative regimes ranging from ‘guesthood’ in the early 1980s to a seemingly arbitrary formulation of state policies aiming to regulate the presence of documented and/or undocumented Afghans in Pakistan. The previous sections highlighted the discrepancy between regional frameworks, national policies and the translation of both into practice, and how policy-agency encounters caused frictious processes (Björkdahl and Höglund 2013). Friction arose where rejection, domination and confrontation caused by return policy implementation met with Afghans’ practices of adaption, exit, and avoidance. These policy–agency encounters occur in a mutual interplay and constitute one main type of interdependence that structures the return governance path. National policy enforcement is full of ambiguities, not least because of large-scale non-enforcement of the formal return governance framework, in particular the Tripartite Agreement. The evidence of shifting policy measures at the national level between domination, empowerment, rejection, and confrontation points to underlying changing preferences and political interests due to interest fragmentation among Pakistani political actors, i.e., leaders of political parties, rights activists, representatives of the state administration, and the army.

A striking example of non-enforcement is that where the stated goal of repatriation is paired with repeated attempts to redefine refugees as migrants or illegal aliens (Mourad and Norman 2020; De Genova 2002, 2020), while mass deportations of the latter have never taken place in the last three decades. Even the status determination of Afghans was never wholeheartedly tackled and provisions for screening in the Tripartite Agreement of 2003, for example, were postponed until after the completed repatriation for an anticipated ‘residual caseload’ (UNHCR 2003). Neither the 2005 census nor the subsequent registration of Afghans sought to aid status determination but conferred prima facie status on every Afghan who registered, regardless of possible protection concerns in individual cases. Then again, even the temporary protection status has not consistently given protection because of irregular ad hoc policy formulation for extending the validity of PoR-cards after 2015. Thus, instead of haphazardness and steering failures, the potency of the state (Roy 2018) becomes apparent in several dimensions, most powerful in the ad hoc extension practices for PoR-cards from 2016 onwards. It constitutes a government technique that keeps Afghans deportable and at arm’s length ‘disposable’. The fact that large-scale coerced returns as a practice of power (Walter 2002, 266) were never officially enforced and carried out only selectively reflects a technology of power that characterises Pakistan’s return governance framework. This form of power is not the omnipresent power (Foucault 1990) which tends to be ‘totalizing’ but rests on ambiguity, open-endedness (Heyman and Smart 1999, 7, cited in: De Genova 2002, 429) and the production of uncertainties for Afghans. The de facto closure of Torkham border crossing in 2016, which repatriable Afghans relied on in their everyday transnational lives, while Chaman border crossing in Baluchistan remains open for unchecked passing, is another seeming contradiction, as this step complicates Afghans’ mobility (Mielke et al. 2021) but does not go as far as to stop it.

In another dimension of policy-agency dynamics, Pakistan has embraced the empowerment discourse targeting repatriable Afghans. While at first glance, it seems to starkly contrast with previously described defensive (rejection, dominance) and semi-defensive (confrontation) steering modes of Pakistan’s return governance framework, enabling measures like the permission to open bank accounts and register SIM-cards decreed in 2019 merely constitute an additional element of control. For the GoP, the empowerment discourse is a strategy to attract funding from the international community in the wider framework of the GoP’s burden-sharing appeal cum rent-seeking behaviour. The international community is implicated in ambiguity and informality here because it acquiesces with return as the only durable solution, as it fears having to shoulder more Afghan refugees coming to Europe through irregular migration channels or large increases in resettlement quotas. The example reveals how power relations between Pakistan and, for example, European governments underlie inter- and goal-dependencies and structure the (non)return governance path. However, while the GoP’s commitment to SSAR’s several empowerment programmes proves beneficial for the GoP because of the connected donor funding and possibilities for refugee-rentierism (Tsourapas 2019), it is not transformative for Afghans and has not brought any changes to their protracted displacement.

As Alsayyad and Roy (2006, 5) point out, ‘informality operates through the constant negotiability of value’, which I translate here as the potency of the state to flexibly assign and reassign value to certain laws via (non)enforcement or through categorisations of Afghans as status refugees versus undocumented migrants. This highlights how law and policies constitute social processes that are subject to (re)negotiation by the state. As emphasised earlier, ‘the state’ does not represent a homogeneous entity but includes politicians and party leaders outside the cabinet, bureaucrats, policymakers, the provinces, and representatives of the army as well as the refugee administration machine (SAFRON and the CCAR/CARs). They all use refugees for their own agendas at times (Borthakur 2017, 501). It deserves particular emphasis that while conventional scholarship on informality conceptualises it as ‘outside the law’, I follow Roy (2009, 2018) in her notion of ‘the state’ as a highly informalized entity (Roy 2009, 81) with the law ‘permeated by the logic of informality’ (Roy 2009, 82). Accordingly, the state utilises informality as an instrument of authority and accumulation – here as a return governance mechanism that allows large-scale non-return as long as this is opportune to Pakistani elites.

Calculated Informality to Negotiate Pakistan’s Geopolitical Positionality

The rigidity of the return governance path is, however, substantially co-shaped by Pakistan’s positionality (Sheppard 2002) in the geopolitical arena of South Asia – an additional interdependency that has its own path-dependency. From the very establishment of the state of Pakistan after the partition of former British India in 1947, India and Pakistan entered a relationship of enmity, with India as the large, potent neighbour. The fear that India could extend its power extraordinarily and thus threaten Pakistan’s existence has guided Pakistani foreign policy ever since. With every alliance and strategic partnership, Pakistan aims to outweigh the power asymmetry towards India and enhance its political and military position in the region. The Soviet intervention in Afghanistan and the role of Pakistan in the proxy war during the Cold War in the 1980s gave it a boost, including the US acquiescing with Pakistan obtaining the nuclear bomb. After the disengagement of the US in the region following the end of the Cold War and the fall of the Soviet/Russia-backed government in Afghanistan, Pakistan struggled to find its role and meaning without superpower patronage. The post 9–11 developments revalorised its geopolitical clout because the US made Pakistan its key ally in its Global War on Terror. Besides the symbolic recognition as US ally, this position brought immense flows of money, military equipment, and training (ICG 2012, 1–11), and US-support at the international level. It only waned again around 2014 when the US started questioning Pakistan’s role in supporting the Afghan Taliban. Subsequently, Pakistan’s geopolitical positionality towards India deteriorated when the US put pressure on the country and withheld symbolic and economic capital. With the Taliban takeover in Afghanistan in mid-August 2021, this could change once again should the US stop talking directly to the Taliban interim government or the political and humanitarian conditions worsen, triggering a mass exodus of ordinary Afghans towards the largely sealed border with Pakistan.

In a move to compensate and diversify the basis of its geopolitical positioning, Pakistan embraced China as its main economic and geopolitical partner in the region, which meant that the presence of Afghans as a bargaining chip lost value and the return governance path was reconfigured once again. Since then, the value of Afghans has only been relevant when international donor decisions were threatening not to include Pakistan among the main beneficiaries of aid earmarked to support the continued reconstruction in Afghanistan. In this reconfiguration, Afghans have turned into a residual issue that is being activated in times of need based on the logic of refugee-rent seeking. Due to its geopolitical dependencies, the GoP feels the need to constantly negotiate this geopolitical positionality to outbalance India’s geopolitical potency. In Pakistan’s own perception, to be inventive and find governmental techniques in every field of governance that influence its geopolitical positionality is a matter of survival. The purposeful employment of ambiguity and deregulation constitutes one such core strategy and technique. Pakistan has adopted calculated informality as a mechanism in the management of Afghans and their return. It allows Pakistan to negotiate its geopolitical positionality meaningfully as long as Afghans in Pakistan depend on state policies in their everyday exposure to insufficient protection (non-asylum and risk of refoulment) and are kept in limbo through the state’s ‘refugee governance’. In the logic of the governance path outlined in this article, calculated informality is producing exactly this lack of protection and subsequent uncertainties for Afghans.

Implications

Having uncovered the limitations and structure of return governance, I can now deduce indications concerning its future trajectory. Return governance operating through techniques of calculated informality yields implications in at least two dimensions of power: Most significantly, first, calculated informality in return governance subjects Afghans to – for them seemingly arbitrary – structural forces in the form of adhocist state practices and policies that generate uncertainty and deprive Afghans of any confidence in their lives’ future predictability and planning. The conditional awarding of rights of asylum and stay perpetuates this dependency, and the risk of renewed irregularization is looming at any time. The renewed temporary extension of PoR-cards and the de facto irregularization of all AC cardholders since June 30, 2020, in absentia of a deportation regime are cases in point. Adding to this, Afghans’ scope of agency has been constantly on the decline for two reasons. On the one hand, the increasing impediments to translocal (within Pakistan) and transnational mobility due to the tightening of the physical state border and the closure of crossing points for Afghans without a passport and visas pre-empted adaption practices like evasion and remigration. On the other hand, the everyday harassment and increasing rejection of Afghans by security organs and ordinary Pakistanis has narrowed spaces of cooperation and avoidance strategies. The informal protection regime of guesthood that existed in the 1980s and the early 1990s based on a joint historical, cultural, social and religious background of Afghans and Pakistanis, and especially Pashtuns on both sides of the border, is by now completely eroded and has been replaced by a rise in xenophobia and ‘Afghanophobia’ among ordinary Pakistanis. As a result, it appears that Afghans are less able to influence their situation in their interest today than they were ever before.² Given that the alternative option – coerced returns (which de facto took place throughout 2016) to insecurity and difficult-to-avail of livelihoods in Afghanistan – is no option, Afghans ‘prefer’ a tolerated continued presence in Pakistan and have so far refrained from any active mobilisation, resistance or protest against the lack of formal protection. The resulting situation constitutes a deadlock and non-solution because of the permanent threat of deportability being evoked and the principle of non-refoulment violated.

Second, the return governance path implies Pakistan’s rigidity in and continued insistence on return as the only acceptable durable solution for Afghans in the country. It thus rules out alternative and de facto lived options like local integration because the return governance path does not foresee decreeing a sound legal framework for such alternatives as they cannot be acknowledged by the state. It lies in the logic of the governance path that even dissident voices within the political establishment of Pakistan cannot escape the meta-narrative of a necessary, full return, which is the base for calculated informality and ensures that the country can continue to navigate its geopolitical positionality. When the newly elected Prime Minister Imran Khan announced in September 2018 that he would grant all Afghans born in Pakistan citizenship (Barker 2018), he faced strong opposition from political and civic forces and dropped the issue. During the 2020 International Refugee Summit in Islamabad, when the global and Pakistani public expected the hosting Pakistani government to announce a policy change towards Afghans, the Prime Minister used this occasion to talk about the military confrontation with India in Kashmir (Dawn 2020). This being quite in line with the path-inherent logic, the rhetoric commitment to finding durable solutions with return as a desired option must be evaluated as merely strategic and an expression of calculated ambiguity, which, as I have shown, is a core mechanism of Pakistan’s informal return governance. In such a configuration, it is highly unlikely that change can be expected from within the existing return governance framework in the future.

Concluding Remarks

In this article, I have elaborated on how calculated informality constitutes a key feature of Pakistan’s (non)return governance that shapes immobility patterns of transnational spaces of living for Afghans residing in Pakistan. The ontology of return governance has been shown to be formal and informal in terms of policy coherence with regional frameworks, national policy formulation and enactment, and fragmented in terms of initiatives and actors involved. I have demonstrated how Pakistan’s governing of return has excelled in calculated informality based on practices of deregulation and ambiguity. Through the lens of an evolutionary governance framework, the multiple dependencies and layering at work in return governance were pointed out. The rigidity of today’s return governance can be traced in historical path dependencies reaching back to the early 1980s’ Cold War- and refugee policies in Pakistan. Interdependencies have been shown to operate in policy-agency encounters involving various Pakistani actors, on the one hand, and undocumented as well as temporary documented Afghans, on the other hand. Pakistan’s geopolitical positioning is another expression of interdependencies in international politics that has a bearing on the rigidity of the governance path. The analysis also identified goal dependencies between the GoP and European donor states that share an interest in containing Afghans in the region in exchange for aid.

The described path- and interdependencies account for non-linear path creation conditional upon the governance context. In effect, Afghans’ humanitarian and protection needs and their preferences for long-term stay arrangements, including perspectives for transnational living and local integration, are overruled by the perceived need of the Pakistani government to constantly negotiate its geopolitical positionality in relation with India. Towards this goal, the GoP employs calculated informality as technology of power, which manifests in everyday structural violence affecting Afghans. The (non)return governance regime neither allows obtaining citizenship or permanent protection in Pakistan nor facilitates resettlement due to the implication of European donor states who purposefully employ strategies to escape burden sharing since the onset of the so-misnamed ‘refugee crisis’ in Europe in 2014.

The analysis of Afghans’ continued presence in Pakistan in light of repatriation as the only durable solution by the Pakistani state, examined the tension between principles of refugee protection and geopolitically informed return policies. The above findings speak to the scholarship on critical geopolitics of mobility and governance in that they confirm the role of geopolitical relations in structuring mobility (Ashutosh and Mountz 2012, 352). Calculated informality in governing (non)return constitutes a governmental technology that effectively deprives refugees and migrants of rights and uses deportability as bargaining leverage (Fakhoury 2021, 175). I conclude that the current situation of Afghans in Pakistan qualifies the potency of refugee agency as embodied geopolitics from below because the securitisation of the border, surveillance, bordering practices of the state, and increasing xenophobia of Pakistani citizens towards Afghans have narrowed their space of agency in several graduations from the 1980s until today. The future applicability of previously well-working practices of avoidance and adaption that enabled Afghans to lead relatively normal transnational lives, whereby they actively contested and negotiated the return governance framework (Ashutosh and Mountz 2012, 345), is highly uncertain. The uncovered ‘limited’ logic of the return governance path suggests that the future path trajectory will be contingent in adversity.

With its focus on the implementation history of plans, policies and laws, EGT is very well positioned to connect with several strands of scholarship related to refugee return governance and policymaking. The EGT lens proved useful in trying to make sense of governance outcomes while having to consider multiple actors and different levels of authority in sufficiently complex governance settings in longer-term perspective. It has provided historical depth and context-sensitivity in tracing the return governance path. Thereby, EGT required to go beyond a governance analysis of law and formal policies at the international and national levels, on the one hand, and the practice–implementation dimension and Afghans’ agency on the other. I thus conclude that there is a value in applying EGT in critical geopolitical studies of mobility. Follow-up research should indulge in more systematic historical and process analyses of governance paths and explore, for example, how return governance is embedded in broader governance evolutions and the policy-relevant consequences thereof, e.g., possibilities to alter ambiguity, deregulation and rigid (return) governance paths.

Acknowledgments

This article is the result of year-long research and intellectual work enabled by several project fundings since 2009: (1) “Sub-urban movements: Social inequality and dynamics of micro-mobilization” as part of the Research Network “Crossroads Asia—Conflict, Migration, Development” (Grant No. 01UC1103A), funded by the German Federal Ministry of Science and Research, (2) “Protected rather than Protracted - Strengthening Refugees and Peace” and (3) “Trajectories of reintegration” both funded by the German Federal Ministry of Economic Development and Cooperation, and (4) “Transnational Figurations of Displacement” (TRAFIG) funded by the European Union’s Horizon 2020 research and innovation programme (Grant No. 822453). I would like to thank the two anonymous reviewers and the thematic issue editors for their helpful reading of earlier drafts.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1. The figure of up to 2.3 million documented Afghans consists of those registered as status refugees (1.4 million, see UNHCR n.d.) plus 879,198 previously undocumented migrants that applied for Afghan Citizen (AC) cards with Pakistani authorities in 2017–18 (IOM 2019, 13). The figure of additional one million undocumented Afghans is a medium estimate established from a range of sources, e.g., IOM refers to half a million undocumented in the same source.

2. It is important to note that nuances exist, however, because Afghans’ protracted displacement is classed. See Mielke et al. (2021) for detailed explorations on this theme.