Liaison officers as influential ‘immigration risk’ brokers in visa policy implementation: intermediaries across institutional and national borders

ABSTRACT This article studies the hidden world of extraterritorial liaison officers, mid-level civil servants posted abroad whose agency influences UK visa implementation within a global framework. Specifically, we unpack their influential role in translating vague policy objectives into specific institutional justifications, norms, and practices, which bureaucrats apply when implementing visa decisions on location. ‘Risk’ knowledge production is crucial: they mobilise, broker and communicate so-called ‘immigration risks’ applied to specific foreign nationals across institutional levels and national boundaries. Liaison officers are intermediaries, ‘risk’ brokers, who: (a) interpret (and feedback on) the Home Office’s supposedly objective central ‘risk’ assessments; (b) construct ‘risk’ assessments based on local knowledge and intelligence to guide and legitimate street-level bureaucrats’ (consulate, airline) decisions; and (c) co-operate to a surprisingly high degree over ‘risk’ assessments with peers in Global North multi-state frameworks. Importantly, their interventions for the UK state effectively reinforces an unequal North–South global mobility regime. To examine how ally and target states are treated differently, we compare across France, USA, Thailand, Ghana, and Egypt. High state secrecy makes studying liaison officers difficult. Our original research applies document analysis of public policy statements, interventions via freedom of information requests, and interviews with twenty mid-level operational officers.


Introduction
Requiring travellers to hold visas authorised by a destination state's consulate prior to embarkation is classic 'remote control' (Zolberg 2006, 443). Issuing visas abroad is an important tool for states to remove bureaucratic decisions away from domestic territory, where foreigners gain access to judicial processes for appeals, and other constitutional, human, civil and personhood rights. Although visas are usually short-term, or for transit, states increasingly define the mobilities of specific foreign nationals as longterm 'immigration risks'. Today visa policies are the central instrument of mobility requests from UK Visas and Immigration (UKVI) managers and making judgements on whether to undertake the work, responding flexibly to changing priorities, risk assessing and disseminating intelligence, developing risk profiles and crime work with minimal input from the ILM (immigration liaison manager)' (HM Government 2016b). Official documents show they collect, analyse and circulate information on 'risk profiles', potential 'visa abuse' and 'emerging trends' for determining 'high risk applications' (Bolt 2015, 28;Toms and Thorpe 2012, 26;HM Government 2017). We also know they feedback 'risk' intelligence from their enforcement experiences to UK-based HO sections responsible for general policy on attributing 'risks' differentially to foreign nationals (Ostrand and Statham 2021).
Surprisingly, there is no research on officers' influence in the visa process, even in 'remote control' literature. Otherwise excellent studies by Scholten (2015) and FitzGerald (2019) mention liaison networks from policy documents and their work advising airlines and officials, but fail to investigate what this neutrally-termed 'advising' consists of substantively. There is also high-quality ethnographic research on consulate staff (Alpes and Spire 2014;Infantino and Rea 2012;Scheel 2018;Infantino 2021), but nothing on experts who provide 'risk' knowledge and guidance for their visa decisions. This article tries to fill this empirical gap. We study officers' influence in visa implementation, i.e. how their interpretive agency over 'risk' translates vague policy, as it appears on paper, into specific norms and practices that 'street-level bureaucrats' (Lipsky 2010) use to make decisions on-theground. We undertake the difficult task of unpacking this hidden world by analysing policy documents, intervening through FOI requests, and interviewing mid-level officers.
Next, we outline our perspective on officers as 'risk' brokers, building on constructionist perspectives that see policy implementation within governance as a social learning process that requires knowledge-production and agency (e.g. Burstein 1991). Then we discuss methods and how we used visa policy categories to select the state cases. The empirical study has four parts: first, we examine the source of 'risk' in HO policy statements; second, we evidence officers' role as 'immigration risk' brokers; third, we examine officers' own perceptions of how they apply 'risk' relative to the five states; and fourth, we examine their 'risk' knowledge exchanges with peers from ally and target states. The conclusion argues we need to know more about officers as hidden but influential brokers in the global mobility regime.

Liaison officers: 'risk' brokers across institutional and national boundaries
We advance a perspective on officers as 'immigration risk' brokers. They exercise interpretive agency through knowledge production, exchange and mediation of a state's perceived 'risk', while operating within the constraints of institutional culture and professional norms. Their scope for influence is increased by vague visa policy objectives, a distant management structure, and the on-the-ground nature of the job, which requires local knowledge and working closely with foreign officials, police, airlines, and consulates (Ostrand and Statham 2021). They draw authority as the HO's messengers and 'risk' experts abroad, translating and updating supposedly 'objective' quantitative central risk assessments for countries into specific implementation norms and practices.
Liaison officers sit within a unique framework of working relationships, acting as interlocutors between distinct communities, across HO institutional levels and national borders (Ostrand 2022). This intermediary role clearly makes them 'brokers' in the policy process, who are 'often critical to communication and learning between (policy) communities', and whose efficacy 'depends on the trust and complicity of others' (Freeman 2008, 378). Brokers realise this trust in their practical work by carrying negotiations back and forth across the distinct boundaries of policy communities and networks (Freeman 2008). Importantly, this sees policy implementation as a social learning process rather than carrying out instructions, where officials have scope for agency because they are continually confronted by 'new' problems requiring creative responses and knowledgeproduction (Pressman and Wildavsky 1984). The currency officers broker is 'risk' knowledge, which they interpret, frame 3 and communicate as interlocutors between distinct professional groups: HO and consulate and airline staff; HO and foreign officials; and with peers in multi-state frameworks. Ultimately, their actions cognitively package and frame the meaning of 'risk' applied to specific foreign nationals into practical guidelines for administrative staff to implement visa decisions.
Officials' understandings of policy 'problems' that resonate institutionally are strongly influenced by cultural theories about how society works, which importantly shapes policy decisions and implementation (March and Olsen 1989;Burstein 1991). Although policy approaches are often justified by rational scientific claims, '(i)n public policy, solving problems is an embedded, social process … We learn with others as much as from others' (Freeman 2008, 377-378). Existing research on immigration and border officials emphasises that administering policy decisions is primarily a social process driven by values, norms and practices embedded within the working environment. Eule, Loher, and Wyss (2018) find a wide gap between policy 'categories' on paper and how decisions are administered that allows significant scope for discretion, while Côté-Boucher, Infantino, and Salter (2014, 198) highlight the importance of socially embedded institutional practices, whereby 'they make decisions on the basis of both the formal and tacit understandings granted by their work routines and training'. Similarly, Jubany's (2011) study of the HO's asylum-screening, shows how bureaucrats apply (and rarely challenge) 'professional knowledge' or 'knowledge taken-for-granted' when making decisions, that is often based on experiences and social prejudices that resonate within the institution. Liaison officers are important providers of this institutional 'knowledge taken-forgranted' that consulate and airline staff use in administering visa decisions and deciding who can travel to the UK. Their authority is based on their expertise in 'risk' communication. Douglas and Wildavsky (1982) add to this social constructionist perspective on implementation, by focussing explicitly on how institutions use scientific notions of 'risk' to justify policies and their implementation into actions. Importantly, they argue that officials fetishise claims about scientific objectivity, but in fact institutions have a socially embedded 'cultural bias' for selecting the specific 'risk' assessment science that supports their already preferred stance. Similarly, immigration governance scholars find that expert knowledge is used not only to inform policies, but to substantiate policy choices already made, while legitimating the institutions making them (Boswell 2009). We argue that officers' agency in 'risk' knowledge production, framing, and communication, i.e. their expertise, is essential to understanding how visa implementation works as a social process across the HO's institutional culture. Later, we verify this perspective by putting empirical flesh on the bones.
Our first aim in conceptualising liaison officers as 'immigration risk' brokers is to unpack officers' influential role in providing justifications for 'risk' approaches to France, USA, Thailand, Ghana, and Egypt, empirically. By examining how they construct and use 'risk' to justify visa policy approaches for specific states, we gain access to the legitimating devices that resonate within an institutional culture, which determine whether a practically implemented approach is more liberal or restrictive. Our hypothesis is that officers' interpretations of 'risk' are strongly shaped by internalised social pressures and professional values, norms and practices embedded within the IEI of the HO, i.e. they use 'risk' science, local knowledge and expertise selectively to justify visa enforcement approaches that are predetermined by where a state stands relative to the UK in the global hierarchy of immigration control.
Our second aim is to examine the degree to which officers' 'risk' understandings are generated and shared through interdependent relationships with peers from foreign states. A key task for officers is to mobilise effective working collaborations with international partners, within an aim of preventing 'unwanted' and irregular immigration outside the juridical border (HM Government 2019). Given the clear power asymmetries between North and South states, we think it is important to place our study within a framework that recognises this multi-state hierarchy. Importantly, this requires moving away from 'remote control' perspectives that study bilateral relations between North and South states, and beyond those studying North-North co-operations exclusively within EU frameworks (Ostrand and Statham 2021). Specifically, we examine the degrees and forms of cooperation between officers and their professional peers from France, USA, Thailand, Ghana, and Egypt, over 'risk' knowledge production and exchanges as a legitimating basis for visa decisions and their enforcement. Here it is relevant that the UK participates in two multi-state cooperative Global North frameworks: with the USA in the Five Country Conference (FCC) 4 ; and with France in the EU/EEA. By studying this cooperation, we can see the degree to which there is shared knowledge, common understanding and usage of 'risk', as well as trusted professional relationships, operating across national boundaries.
A compelling argument supporting this approach is that, within advancing globalisation, increasing inter-state cooperation, especially among powerful states, has transformed state sovereignty over immigration controls (FitzGerald 2020), so that a state's migration management is increasingly dependent on inter-state agreements and working within multi-state systems. In this view, a 'global hierarchy of visa freedom' (Mau et al. 2015) is part of an interdependent world where there is a 'hierarchy of sovereignty' (Lake 2009) between powerful and less powerful states that influences which mobility pathways are available across geographical space, and for whom. It means the UK's scope for autonomy (including using 'risk' categorisations) is strongly constrained within multi-state norms, agreements and institutional co-operations. However, this loss of sovereignty is compensated because the state works together with powerful allies who share similar restrictive norms towards the mobility of certain nationals. In effect, the UK is a member of the powerful club of states who co-operate over enforcement through knowledge exchanges to such a degree that FitzGerald (2020, 5/6) calls it 'shared coercion of movement'.

Method
This research draws on twenty semi-structured interviews with current and former Home Office officials, between July 2016 and Oct 2017. Our interviews were primarily mid-level officials with operational experience working in one or more foreign jurisdictions; only three were based primarily in the UK (Table 1). All interviewees had significant knowledge on the implementation of UK extraterritorial migration control, especially by the overseas liaison network IEI, and had cooperated with officials from (but not limited to) France, USA, Thailand, Ghana and Egypt. Access was obtained informally by emailing individuals in the Home Office and British consulates and by referral. Given secrecy surrounding HO activities, and consent commitments made to interviewees, we are restricted in information we can report to preserve anonymity. Table 1 details limited information on the sample. In the text, we cite the letter corresponding to an interviewee to indicate a source.
The interviews are supported by original FOI requests and document research on primary and secondary visa legislation, explanatory memorandums, impact assessments, Independent Chief Inspector of Borders and Immigration (ICIBI) reports, policy papers and press releases (online appendix). Due to the limited information on the liaison network in UK legislation and official HO documents, alternative sources like job advertainments (posted online between 2016 and 2020) were useful for filling gaps on IEI's role in the visa process and the types of activities and responsibilities liaison personnel are expected to perform. Information from interviews and FOI requests are all original and previously unavailable in the public domain. This allows us to go beyond existing literature that relies mostly on policy documents and interviews with high-ranking UK-based civil servants. By contrast, we focus on operational mid-level actors. Importantly, this allows us to examine liaison officers' interpretative agency within the implementation of UK visa policies, which is otherwise publicly invisible, because states keep such operations and the rationales

Case selection: France, USA, Thailand, Ghana and Egypt
We used visa categorisations to select five foreign states where the UK has posted officers with different statuses in the 'hierarchy of sovereignty' over global mobility: France, USA, from the Global North, and Thailand, Ghana and Egypt, from the South. All countries are significantly large and populated, but vary by region, political interconnectedness and historical/cultural ties with the UK. The UK's differential treatment is demonstrated by its visa categorisations per state (see Table 2). The visa regime, before Brexit, divided foreign nationals into four categories for pre-entry controls, ranging from liberal to restrictive: free movement; nonvisa; visa; and transit visa. First come 'free movers' from states allowed to enter without a visa, including nationals from the EU/EEA and Switzerland. 5 France is our 'free mover' case, where the UK granted entry on the same basis as UK citizens. Here we discuss the pre-Brexit time-period of our interviews. New rules governing the EU/ UK relationship started in January 2021. Second, 'non-visa nationals' are people from countries who do not require a visa to visit, but require one for immigration purposes (e.g. work, study) or if they plan to stay longer than six months. 6 The USA is our 'non-visa nationals' case. Third, 'visa nationals' are from states where prior authorisation is required to enter for visits and immigration. 7 Thailand, Egypt and Ghana fit this category. Finally, 'transit visa nationals' are the most restricted category, who always need prior permission to enter, including a visa to transit through UK airports (Carriers Liaison Section 2015). Egypt and Ghana are states that face the UK's most restrictive visa controls.
The UK manipulates this categorisation by 'nationality' by adding facilitative schemes to ease border access for selected individuals. Effectively, they ease speed of decisions and entry. Officially, Registered Traveller and Super Priority services target 'high net-worth individuals and businesses in growth markets' (HO 2015a, 22). Again, their availability follows the general pattern of visa requirements; a greater availability in Thailand compared to Ghana and Egypt indicates Thailand's higher status in the international pecking order of a visa mobility regime. Facilitative schemes highlight the UK state's dual interests in encouraging certain types of mobilitylargely by those who are wealthier, perceived as contributing to the UK's economy, and from Global North ally stateswhile restricting it for others who do not fit this image. The idea of an international hierarchy dominated by Global North states is further underpinned by examining the degree of inter-state co-operations between the UK and the five states over extraterritorial visa controls. Table 3 shows that the UK has very strong institutionalised co-operations with the USA and France, including reciprocal data-sharing, knowledge-transfer and resource-pooling arrangements, and engages in joint training, investigation and enforcement activities. Importantly, these are coordinated through multi-state frameworks, specifically the FCC for USA, and EU/EEA for France. By contrast there is virtually no institutional co-operation with Egypt, not even a Police Referral Programme (PRP), which is an agreement that allows officers to share 'intelligence' with foreign police, giving the UK leverage on non-sovereign territory by formalising expectations that foreign authorities investigate and prosecute those implicated of 'abuses' by UK intelligence (Vine 2010b, 22-23). The UK has made some formal inroads into Ghana and Thailand, through bi-lateral agreements, including PRPs, and providing limited training and resources.
We study liaison officers' perceptions of their working relationships with peers over 'immigration risk' within these institutional co-operations from the five foreign states in the last empirical part. First, we examine how they draw authority from HO policy statements, before examining their 'risk' knowledge production and brokering role in visa implementation, and how they understand 'risk' relative to the five states.
Interpreting and implementing 'risk' from policy statements As civil servants posted abroad, liaison officers draw their legitimacy relative to other officials as key messengers on 'immigration risk' at the policy implementation stage. Here we examine how vague visa policies based on 'risk' set the agenda and institutional scope for officers' agency. Official documentation makes clear that centralised HO risk assessments are the original source of knowledge and expertise that liaison officers use to interpret and generate their own local understandings of specific visa 'risks': '(R)isk assessments are drawn from quantitative analysis of the known threats posed to the UK by visa nationals. The core data is produced by UK-based analyst teams and is supplemented by RALON (IEI's predecessor) overseas. Using local knowledge and intelligence, RALON refines and updates risk profiles for use in visa sections. Profiles are generally grouped by nationality and contain indicators around personal, domestic and employment circumstances, supporting documents and previous travel and immigration history' (Bolt 2015, 28).
At face value this suggests liaison officers' straightforwardly implement central risk assessments (verified by an apparently objective scientific method) into a set of relevant local practices and norms. However, closer scrutiny of how the HO constructs and applies 'risk' to foreign states raises important questions about the degree to which this is actually based on an objective scientific method. An alternative explanation from cultural 'risk' theory is that institutions have a 'cultural bias' for selecting and using the specific risk assessment science that supports their already preferred stance (Douglas and Wildavsky 1982). Indeed, this fits perspectives that see immigration governance as a field where expert knowledge is used largely to substantiate policy choices already made and legitimate institutions making them (Boswell 2009). The idea that 'risk' is socially constructed within an organisation's institutional norms to legitimate preferred practices (and that institutions selected their preferred objective science) matters, because liaison officers will have scope to interpret and define visa 'risks' on location, to the degree that they receive a clear steer to guide their actions from central risk assessments and policy statements.
The HO is highly secretive about how it defines 'immigration risks' and uses data to legitimate visa policies for specific foreign states. In response to our FOI request, it refused to release information on countries it classifies as 'high risk'a classification that authorises officials to apply greater scrutiny to people based on their nationality alone 8and the criteria/methodology it uses to make these decisions (Ostrand FOI 46078 2018). Official sources simply state they use 'data' and 'statistical and intelligence-based evidence to identify the nationalities that pose the greatest risk to immigration controls' (Brokenshire 2015;Ostrand FOI 46078 2018). They reveal no indication of data types, methods or evaluations for defining visa policy, nor access to 'risk' definitions. The official response justified this secrecy by claiming disclosure 'would substantially prejudice the operation of immigration controls and our ability to prevent and detect crime' (Ostrand FOI 46078 2018), a standard unverifiable national security argument. Unusually, a second justification claimed it would 'prejudice international relations' by eroding 'trust and confidences' with partner states (Ostrand FOI 46078 2018). If categorisation is a scientific method applied equally to all states, then it is highly questionable that transparency would erode rather than enhance relations with foreign states. Indeed the 'prejudice international relations' argument only makes sense if the HO thinks it better to be able to apply different criteria to states and values the 'trust and confidences' of some states more than others. That is a plausible realpolitik stance in international relations but is not based on scientific method.
The HO's review  of the visitor visa system provides rare public insight into the official language and logic of how the UK applies 'immigration risk' to different nationals.
It assessed all non-EEA states for the 'risk that their citizens potentially posed' to the UK regarding 'illegal immigration, crime and security' (HO 2009). Cited evaluation criteria for 'immigration risk' include: security of a foreign state's passport issuing procedure; its cooperation in deportations; frequency of 'immigration abuse' by nationals; perceived 'risk' of nationals regarding 'terrorism and criminality'; and 'steps taken' by a state 'to combat terrorism, crime and immigration abuse internally' (HO 2007, 9). Against these, the Home Office balanced 'likely economic, cultural and political consequences' of introducing or waiving visitor visas (HO 2007;HO and FCO 2010, 14). 9 This balancing of 'risk' against economic, cultural and political benefits follows a standard trade-off calculation for states' pre-entry visa regimes (Neumayer 2010). Again, this suggests a realpolitik rationale more than visa policy by 'risk' science.
On top of applying 'risk' in a blanket way to all nationals from specific foreign states, officials use 'risk profiles' to target individuals (based on personal information like employment, finances and travel history). This occurs regardless of nationality, though clearly people from 'risky' states receive higher scrutiny that subsequently increases chances of individual criteria being applied. A HO report effectively shuts the door on any possibility for understanding tangible scientifically meaningful criteria for such evaluations by stating, 'detailed discussion on risk profiling is restricted for security reasons' (HO 2015b, 15). Official statements on risk profiles are vague descriptive tautologies. For example, the ICIBI defines a risk profile as a document showing 'the relative potential harm (to the UK of a visa applicant/travelling passenger) based on characteristics of an individual when compared to existing evidence of adverse activity either in the UK or overseas' (Vine 2012, 31).
Policy documents use extremely vague definitions for 'risk', talking about 'immigration abuse', 'threats to the UK border' and 'criminality', but without specifying meanings (Bolt 2015;Vine 2014, 21). One exception is that perceived potential for asylum claims leads to higher visa requirements and 'risk profiles' (Bolt 2015, 4;HO 2011;NAO 2004, 4-5). Visa applications are 'held to a higher standard of proof on the basis of risk profiles' (Bolt 2015, 21) and additional scrutiny and verification checks are intrusive. Official documents show that in 2013-14, Iraqis, Palestinians, Syrians, Lebanese and Jordanians faced assessments of family members' immigration histories, interviews, scrutiny of financial documents, and employment history (Bolt 2015, 28-29). Again, such actions are justified in quasi-scientific terms, but without public information that allows verification against objective criteria: 'assessments are drawn from a quantitative analysis of the known threats posed to the UK border by visa nationals' (Bolt 2015, 28).
Another rare official document that comments on the state's method is an ICIBI report (Vine 2011, 25) that acknowledges risk assessments are partly based on the number of nationals refused visas for the UK. This insight, albeit limited, offers the intriguing prospect that a circular logic is at play. Nationals from specific foreign states are highly targeted because they are perceived as 'risks'; this leads to relatively more visa rejections. Subsequently, this provides de facto numerical evidence to justify applying higher scrutiny on people from that state on the basis of 'higher risk'. However, a higher visa refusal rate could result from higher scrutiny or perceived suspicion of an applicant's intentions, not from higher actual 'risk'however defined/operationalised relative to nationals from other states. 'Risk assessments' attached to specific states and profiles can become a self-fulfilling prophecy.
This ICIBI report (Vine 2011, 25), for example, even pointed out that 'accurate risk assessment' under this method requires that decisions, i.e. officials' interpretations, are 'correct' to be valid. However, an earlier report (Vine 2010a, 18) documented that decision-making on settlement visas for Pakistani nationals, 'was poor, to such an extent, that it was almost impossible in some cases to determine why visas had been issued, when others had been refused on identical or very similar evidence'.
Importantly, we see the HO uses hard scientific and legal language to attribute a pseudo-objective status to what it defines as 'immigration risks', but it refuses to make the science or criteria for risk assessments public to allow independent verification of their factual validity. State secrecy overrides any concerns for methodological transparency. Liaison officers are posted abroad with a duty to assess and act on 'risk'. However, there is a wide gap between how official statements use 'risk' to justify visa policies on paper and how officers interpret and define 'risks' to justify norms and practices at the implementation stage. Vague policy definitions allow officers significant autonomy for interpreting what specific 'risks' might be on location. In this way, their local 'risk' constructions transform abstract policy into justifications for specific institutional practices that are applied for visa decision-making. Officers' local 'risk' constructions will still be expected to fit HO general norms. So, while officers' scope for interpretive agency through implementing notions of 'risk' is substantial, it remains bounded by an institutional culture that strongly defines the overall direction of interpretation regarding specific foreign nationals.
Liaison officers as 'risk' knowledge-brokers: framing, communicating and legitimating 'risk' for practical use in visa decisions Turning to what liaison officers do, interviewees confirmed their institutional role is defined within an overall strategy, where visa enforcement is central to preventing irregular immigration: 'the UK is focused very much upstream on trying to manage irregular migration flows. So, (liaison personnel) are the key element of that because we provide the support and the led services in order to try to ensure the visa regime is as effective as possible' (T). This fits FitzGerald's (2020) perspective that extraterritorial efforts are central to contemporary immigration controls, but also underlines officers' importance in the visa component of that strategy.
Job advertisements confirm 'visa risk' responsibilities are one of their 'main duties' and that they work as intermediaries across different institutional levels: 'A main role of the (liaison officer) is to identify the level of fraud or abuse in all of the main categories of visa applications … (and) having identified the risks, to then produce information and intelligence based products that help (UKVI) Entry Clearance Officers make wellinformed, high quality decisions' (HM Government 2016c). Substantively, liaison officers inform visa implementation by providing training, advice, information, and intelligence to UK staff at overseas consulates. This includes creating and updating 'risk profiles' used by consulate staff to determine higher and lower 'risk' applications (Toms and Thorpe 2012, 26;Bolt 2015); delivering training and guidance on issues like document verification, forgeries and 'imposters' (HM Government 2016a; Vine 2011Vine , 2013; providing verification checks and advice on specific applications (Bolt 2015, 29;Bolt 2017, 13) and producing and distributing 'intelligence' on trends in 'visa abuse' to inform decisions (HM Government 2016a, 2016c).
The central HO does not design IEI's strategy for specific foreign states. Instead, liaison officers interpret 'risks' and use this as a knowledge basis for guiding their strategies on location. An interviewee explains: '(Activities are) pretty much decided by the (liaison manager) in the country because you know what your current threats are, what is the continuous problem you keep seeing' (U). Officers' interpretive agency for defining visa 'risk' on location extends far beyond a simple translation of HO institutional policy from the centre into implementation practices at the periphery. Officers make important judgments about 'risks' and 'need' for action. They take decisions that design and implement specific action strategies to address perceived 'needs' on location (D, E, M, O, T, U).
In this way, officers act relatively autonomously; their agency is akin to 'risk' knowledge production. Significantly, 'risk' knowledge production is emphasised in their official responsibilities for 'developing operational intelligence products through the collection, interpretation and evaluation of relevant quantitative and qualitative information sources' (HM Government 2020). Their constructions of perceived 'risks' are significantly removed from abstract general HO risk assessments and importantly become the knowledge basis that legitimates the practices and norms that consulate staff use to make visa decisions.
Officers use practical knowledge drawn from their experiences on-the-ground, including joint activities with foreign state officials, to define precisely what local 'risks' consist of. Interviewees were clear that data and intelligence inputs provided by foreign state officials influenced their assessments of local 'risk' and enforcement 'need' (A, C, F, M, T). In places where peer exchanges are constructive working relationships, officers train local police, immigration and border officers in investigative skills, profiling and document evaluation techniques to support the gathering and implementation of 'risk' intelligence (Ostrand and Statham 2021). Public job advertisements make clear that officers are expected to be the driving force, responsible for '(i)ntelligence exchange with local law enforcement partners … to develop intelligence on visa abuse' (HM Government 2016b), and '(l)ead(ing) on the exchange of actionable intelligence with local law enforcement partners and counterparts in other Diplomatic missions to identify and mitigate threats to the UK border' (HM Government 2020).
Importantly, officers are tasked institutionally with communicating this knowledge they produce on local 'known risks' to visa staff in consulates. They are important local conduits of 'risk' knowledge, which requires dissemination by 'regular liaison and close working with UKVI managers and (Visa) Officers regarding risk data provided … to assist UKVI in improving decision quality' (HM Government 2016b). Their 'risk' knowledge-provider role is hands on: extending to training visa staff on document verification, forgeries and child protection, issuing 'forgery alerts', responding to daily referrals and queries, and carrying out checks on specific cases (HM Government 2016a; Bolt 2015). An officer confirmed they provide direct guidance to staff on individual cases and an overall steer on visa implementation norms and practices by 'managing the issues they face on (specific) cases but also providing the reassurance that the decisions they are making are appropriate' (T).
Ethnographic research importantly documents how consulate staff are 'street-level bureaucrats' (Lipsky 2010) who exercise considerable discretion and apply 'local practical knowledge' when administering visa policies into actual decisions (Alpes and Spire 2014; Infantino and Rea 2012; Infantino 2021). Scheel finds that consulate staff codify applicants' profiles informally based on local office knowledge, while evaluating 'risk' under pressurised conditions of 'time constraint, incomplete information and uncertainty ' (2018, 2752). We think it is vital to add liaison officers to this story because their 'risk' knowledge provision provides influential support and legitimacy when visa staff make decisions under these constraints.
Liaison officers perform a similar 'risk' knowledge role for airline staff, as airlines face financial sanctions for transporting passengers with inadequate documentation 10 to the UK. Consequently, airline staff are responsible for checking passengers' documents (including visas) prior to boardingeffectively serving as a pre-departure immigration control. Liaison officers train airline personnel on visa and entry rules, 'passenger profiling', identifying forgeries, and general document security (HO 2019, 2). They also provide on-the-spot advice on cases, taking calls from the departure gate (D, O, T), and issuing alerts on emerging trends in irregular migration and forgeries (C). An officer underlined their authority within this relationship: 'we say to the airline … pick up the high-profile cases and concentrate on those three passengers … we'll help them with that. We will give them the profiles to look for, and we do some of this ourselves' (Q).
Overall, officers' have a relatively autonomous intermediary role brokering the meaning of 'risk' across different HO institutional sections and levels, and stakeholders, not least foreign state officials and private companies (airlines). They become especially influential as an important messenger informing UK visa enforcement on location, where their 'risk' knowledge production influences and legitimates the norms and practices that street-level bureaucrats apply. We see liaison officers as 'risk' brokers. Their knowledge production frames abstract notions of 'risk' into specific ready-to-use categories, i.e. cognitive packages, that inform 'local common knowledge' for bureaucrats to use in administering and enforcing visa decisions. In this way, consulate and airline bureaucrats' scope for discretion over decisions operates within limits set by liaison officers' 'risk' definitions. They provide a legitimating basis for local decisions abroad.

Liaison officers' perceptions, interpreting and defining 'risks' for foreign nationals
Given this influential role, how do liaison officers understand 'risk'? Interviewees discussed their visa actions very much within a language and framing of immigration 'risk' and 'criminality'. They included many references to 'risk', 'immigration crime' and 'organised immigration crime', though such terms are mostly presented as accepted common knowledge and not interrogated further. It seems that this unreflective talk about 'risk' is a social norm within the liaison network's professional institutional culture. Also, officers were clear that 'risk assessments' are increasingly the justificatory basis for enforcement: 'We work closely with risk analysis for visa operations' (M). 'There has been a shift towards a more intelligent-led, risk assessment and trends approach' (D). Interestingly, officers seemed to buy into the authority and objective validity of 'risk assessments', unquestioningly, sometimes showing an almost blind faith in 'risk' expertise: 'Technology is taking over. So, you know, risk is now managed by metrics' (O).
Perhaps interviewees' unwillingness to move beyond rather vague justifications based on 'risk' was due to secrecy, but we strongly suspect another process is at work. Earlier we discussed how institutions are 'culturally biased', so that internalised social norms within a bureaucratic professional culture can determine how and which 'risk assessments' are selected to justify institutionally preferred decisions, while making claims that decisions are based on scientific 'objectivity' (Douglas and Wildavsky 1982). We think that liaison officers' unreflective talk about 'risk', that almost fetishises its presumed objectivity, demonstrates that they work within a professional institutional culture, that is 'culturally biased' towards framing visa applications from specific nationalities (and with certain characteristics) as more 'suspicious' and 'risky', accepting expertise on 'risk' assessments which support this perspective uncritically. This is not to claim that officers as individuals are unreflective, insincere, or cultural dupes, but are professionals working within an institutionalised culture which offers no rewards for people who might try to interpret 'facts' differently than the general institutionalised trope of restrictive norms towards certain types of nationals (backed by 'immigration risk' science).
Officers have no incentives to challenge the dominant Home Office orthodoxy for 'immigration risk' that is socially embedded into standard practices. On the contrary, they receive performance 'targets' for meeting restrictive visa goals. Almost by necessity, their institutional culture works to reinforce and reproduce 'immigration risk' as an unquestioned mantra. We suspect this 'risk' bias is so deeply embedded within their institutional culture that it exhibits features of path dependency (Pierson 1993), where it reproduces itself over time, as the dominant self-legitimating organisational logic for decision-making. Importantly, this means that although officers have significant interpretive agency, this is only to the degree that their actions reinforce and provide increasing returns for the HO's visa policy objectives, which we argue are based on relatively preconceived and socially internalised ideas of 'immigration risk' posed by specific foreign states.
Secrecy and officers' inaccessibility makes it virtually impossible to conclusively demonstrate this point from interviews. However, we can examine how officers applied 'immigration risk' to nationals across the five foreign states, and whether approaches were consistently applied to all, or differed according to a state's position in the 'hierarchy of sovereignty'. If the way officers interpret 'risk' is shaped by social norms of their institutional culture, we would expect them to apply somewhat unreflective cultural understandings and national stereotypes to specific countries and nationalities. Similarly, research on immigration officers' asylum screening in the UK demonstrated the importance of an institutional culture of 'disbelief' and how officers applied categories derived from national stereotypes more than regulations and laws (Jubany 2011).
When talking about the USA, officers declared that 'immigration risk' was non-existent and Americans pose no 'risk' (O, Q, R, T). One explained, 'of course we don't need to secure the border from the States … there is no risk there' (O). This perspective reinforces that in policy statements. For example, in 2019, the UK allowed US visitors permission to use e-passport gates without the Registered Traveller scheme because Americans were a 'cohort of low-risk individuals'. 11 There is a sense that officers are deferential to US immigration control operations, technology and techniques for assessing and controlling 'risk': 'It is (liaison officers') remit to train airlines etc., but I doubt they would need to conduct forgery detection awareness in the USA' (R). They understand the UK often follows a US lead: 'The US has been one of the driving forces behind much of the collaborative work with the UK' (E). They see the UK behind the US in its visa control efforts: 'The US is not high risk, but is important for two reasons: one, they had quite advanced technology and two, invariably bad things that arrived in the US probably passed through the UK on its way there' (E). In sum, UK liaison officers understand their lower position in the international hierarchy and look up to the USA for inspiration.
The UK could not impose visa requirements on EU/EEA nationals prior to Brexit, which meant officials had no formal basis to contemplate 'immigration risk' for French nationals. Paradoxically, however, this pooled sovereignty with EU members made states in the Schengen free movement area a perceived 'immigration risk'. France is an important focus for liaison officers' implementation efforts: 'There are parts of Europe where we see more of what we would call 'harm cases' coming consistently … . France, now for that matter, lots of refugee, migrant traffic' (Q). Liaison officers see France as an ally, but also an important source of transit for what they call 'unwanted' immigration: '(T)he absence of internal border controls allows third-country nationals to travel through the EU without border checks. And some officials are not interested in identifying illegal migrants in their territory because they become responsible for them' (F). Interestingly, UK officials sometimes call into question the trustworthiness of French counterparts, something almost unthinkable towards US officials: 'Small provincial airports (in France) might not notice abuse or willingly ignore it, allowing onward travel to the UK' (F).
Regarding who liaison officers target in France as a 'risk', we find loose arbitrary criteria applied based on assumptions about a traveller's national origins and travel history. An officer explained that on a flight from France they would instruct airline workers to pick out names that do not 'sound' European or American, look at passengers' routes, and ask 'why is that person going here and there and there? That is suspicious' (U). While our evidence is not conclusive, this sounds like a proxy for picking out nonwhite travellers from Global South states.
Turning to target states, officers justified high visa requirements for Ghana claiming it has a medium to high potential as a source for 'unwanted' immigration (D, I, M, T) with the main risks 'document fraud for visas and settlement applications' (I). Ghanaians were viewed as an economic immigration risk: 'Ghana, still a risk, is generally more benign and less organised, people try to migrate illegally for economic reasons' (D). However, one officer inadvertently demonstrated that higher 'risk' perceptions are the starting point for decisions, rather than an evidence-based outcome: '(t)he actual number to the UK for illegal entry by Ghanaians are pretty low. But that doesn't mean people are not there irregularly, either by using deceit or fraud in the visa application process or overstaying a visa etc.' (I). It seems the officer does not let low numbers of irregular entry get in the way of a strongly applied institutional norm that Ghanaians pose a potential 'risk'.
Medium to higher 'risk' perceptions were also applied to Egyptians after the popular uprising against President Mubarak in 2011 (I, P), when economic instability and political and civil unrest were defined as an 'immigration risk' (HO 2011). A 2011 impact assessment justified a need for imposing transit visas by citing a rise in asylum applications: '(s)tatistics for Egypt show that nearly double the figure for asylum claims in 2010 have been made to date in 2011', deducing that, 'there is a high risk of further increases to (transit) abuse and the number of asylum applications from visa holders' (HO 2011, 4). Official figures actually show an increase in asylum applicants from 90 in 2010 to 154 in 2011 (ONS 2019), which is significant, but it errs on the side of exaggeration to call this 'nearly double'. Again, this suggests an interpretive frame reinforcing a restrictive lens of 'risk' perceptions toward Egyptians.
Interviewees viewed Thailand as significantly less of an 'immigration risk', notwithstanding concerns over 'human trafficking', visa overstaying and false documentation (A, U). The UK's approach is still cautious, and officers argued waiving visitor visas would mean too many people from Thailand travelling to the UK and staying without authorisation (A, B). Insight into the institutionalised perspective for Thailand is demonstrated when a liaison officer claimed, 'Thais are not a big issue … there are almost no returns. If someone has been staying in the UK without authorisation they will normally return on their own' (A). It is difficult to see how this assumption that irregular Thai migrants 'return on their own' could be verified by hard evidence or official statistics.
Overall, our findings support the idea that institutional culture matters a great deal in shaping the way that liaison officers understand and apply 'immigration risk'. Similar to Jubany's findings on asylum screening in the UK, officers seem to draw on stereotypes about the presumed behaviour of specific nationals, drawing on 'taken-for-granted' local knowledge more than objective claims about 'risk'. 'Risk' perceptions for people from Ghana, Egypt and Thailand were clearly described in this way. Interestingly, 'risks' were also perceived in France, a Northern ally, but because it was perceived as a conduit for potential immigration from the Global South. In this way, officers' 'risk' perceptions work institutionally to reinforce a global hierarchy of mobility.

Liaison officers working with allies and targets across borders in visa enforcement
The UK has generated extensive institutional and informal co-operations for implementing extraterritorial visa controls with Global North allies. Our FOI revealed the UK shared eighty joint visa application centres on foreign territories in 2017, with Australia, Belgium, Canada, France, Ireland, the Netherlands, New Zealand, Singapore and Switzerland (Ostrand FOI 42565 2016). This provides clear evidence for the significant depth of the state's institutionalised co-operations with selected wealthier states. Interestingly, the HO legitimated this institutional 'pooling' on cost efficiencies: that committing fewer resources to visa applications and gathering biometric data meant more resources for preventing 'immigration abuses' (Tai FOI 40510 2016;VFS.Global 2013). This operational logic apparently trumps any concerns over national sovereignty being compromised through institutional sharing with these foreign states.
Within inter-state co-operative frameworks liaison officers undertake a key role by engaging directly in building effective working relationships with peers from foreign states (Ostrand and Statham 2021). Their exchanges with peers within multi-national forums can be influential as 'significant sites for the circulation of knowledge across organisational and national borders' (Ostrand 2022, 3). Earlier we established the asymmetry between the UK's deep institutional co-operations with Global North states (USA, France) perceived as allies in visa enforcement, compared to weaker engagements with Southern states (Ghana, Egypt, Thailand) seen as targets. Here we focus on how officers' relationships with foreign peers can become conduits for shared perceptions of 'risk' and establish mutual trust, or not, and how this can reinforce a global visa regime of unequal mobility between states from the North and South.
Working relations with the USA and France are often embedded within multi-state co-operations rather than bilateral arrangements, namely FCC initiatives with USA, Canada, Australia and New Zealand and initiatives with EU/EEA states and Switzerland. Interviewees emphasised how their work is significantly embedded in co-operations within these frameworks: '(W)e work very closely with European partners and FCC partners … it is very rare that you see us do something just on our own. More often than not we are working with the Australians, the Canadians, the Germans, the French, the Americans. It is very much making sure that we are delivering as a whole to build up the capability (T).
Perceived common interests with allies even extend to visa decisions. For example, the UK's transit visa waiver scheme exempts certain individuals from transit visa countries if they have a valid permanent residence card, or entry visa, from Australia, Canada, New Zealand, USA, an EEA member state or Switzerland. 12 Effectively, the UK waives transit visa application processes and fees for people already granted permission to enter an EEA/Swiss or FCC state. An interviewee explained the rationale: '(W)e have confidence in these countries' vetting process, especially FCC states, and it would be unlikely for them to grant a visa to an individual who is an immigration or security risk … if they grant them visas then the individual should be fine to transit the UK' (F). Here we see the UK's own 'risk' assessment criteria are dropped and replaced by trust in an ally state's performance, presumably based on their 'risk' assessments and vetting processes.
The UK's European and FFC multi-state co-operations are underpinned by legal frameworks and sharing agreements. Receiving knowledge, information, intelligence and data through these networks contributes significantly to UK visa enforcement activities. Our FOI (Ostrand FOI 47192 2018) determined that officers checked all applications against the EU Second Generation Schengen Information System (SIS II) 13 a European-wide IT system on law enforcementand INTERPOL's Stolen and Lost Travel Documents Database. UK staff also accessed criminal records from European states. With FCC states, the UK has memorandums of understanding with USA, Canadian, Australian and New Zealand immigration and border agencies to obtain information on a case-by-case basis (Ostrand FOI 47192 2018;Blacque FOI 42284 2017). Officers also accessed the US Department of Homeland Security's biometric database to automatically check fingerprints of visa applications made in the USA and Jamaica. Here the willingness to use foreign state data systems and intelligence somewhat contradicts the myth that the UK acts as an autonomous sovereign state using its own 'risk' assessments to make decisions.
Liaison officers are key to these exchanges over 'risk'. The FCC and European blocs have initiated increasing operational and interpersonal engagements. An EU regulation on cooperation between liaison officers states that officers 'shall constitute local or regional cooperation networks among each other' and participate in regular meetings, exchanges of information and expertise and 'adopt common approaches as to methods of collecting and reporting strategically relevant information, including risk analyses'. 14 Similarly, a FCC working group on data-sharing 'meet(s) regularly to take forward work on progressing existing co-operation, and to identify areas in which co-operation on immigration exchanges could be enhanced or initiated'. 15 Interviewees confirmed input from foreign colleagues influenced their daily working practices by providing data and intelligence on migration trends, forgeries and information on individuals and groups (A; C; F; M; T). One recounted: '(P)redominantly, where you have day-today sort of stuff, you are dealing with passing information and intelligence and communications, saying we identified this false document being used trying to travel to the UK, etc.' (M). Officers also confirmed they work closely with officials from EU members' visa centres to verify information for visa decisions (C). Others clarified that FCC colleagues regularly provide data on trends in 'immigration abuses' and 'risk' profiling, and information for verifying the travel documents of their nationals (A; U). Specifically, regarding the USA, another claimed they work together daily, 'sharing information on passenger profiling, targeting and selection, verifying each other's nationals' documents, etc.' (R) Again, we see the UK is willing to forego autonomy to some degree by exchanging data, 'risk' assessments and techniques with allies.
Importantly, this systematic sharing, communication, and circulation of intelligence, data and 'risk' profiles has built mutual trust. Officers express a shared common sense of purpose with FCC and European colleagues. Talking about US colleagues, one recounted there were, 'shared goals and challenges, there is also a deep sense of respect' (E). This is underlined and institutionalised through FCC relationships: 'The UK knew it had these other four countries to rely on for advice and support. There was a sense of shared security among FCC countries. That was useful' (E). On French colleagues, the officer stated, 'There was a lot of respect from both sides in the capabilities of their counterparts' (E). It is clear officers share a professional institutional culture with colleagues from allies, based on trust and a perceived common purpose, that crosscuts national boundaries. This effectively means sharing assumptions about 'risk' and accepting foreign colleagues' data, analyses and opinions.
This multi-state routine co-operation leads de facto to a common standardisation of extraterritorial enforcement and 'risk' interpretations imposed on target states. For the UK, this deep sharing with powerful allies constitutes a significant pooling of sovereignty, which also makes it hard to deviate from the two blocs' norms. The only time officers thought UK objectives deviated concerned 'risk' for onward migration through Schengen states. Here they suspected that European allies might be 'burden shifting' their legal obligations to asylum claimants towards the UK by turning a blind eye to onward travel by irregular migrants (D, F, M, Q).
The UK clearly embraces collaboration with allies as the best way to achieve visa enforcement goals. Officers participate in an institutional culture with foreign peers, whom they trust professionally, and share common 'risk' perceptions and enforcement norms. This interpersonal trust in working together is reinforced by legal and institutional arrangements. Global North allies cohere around common restrictive visa objectives largely targeting the South. The UK's significant participation in this systematic multi-national pooling of sovereignty is largely hidden domestically, where governments continue to publicly claim visa controls are autonomous acts of national sovereignty.
Clearly, there is much less to say about Global South states targeted by the UK as sources of visa 'risks', regarding opportunities for implementing shared 'risk' definitions and building working relationships on mutual trust. Ghana is the case where the UK has established the deepest bi-lateral co-operations (Table 3) and is often used in UK publicity promoting the effectiveness of its extraterritorial interventions. Liaison officers trained Ghanaian officials in interviewing skills, 'risk' profiling techniques, document assessment, investigative skills, intelligence and IT. They donated equipment to support investigations and helped counterparts set up specialised units in 'human trafficking' and immigration crime (I, S). However, the asymmetric basis of this relationship and transfer of 'risk' intelligence norms from the UK to Ghana is revealed by officers talking about cooperation in paternalistic colonial overtones. The basis for agreement is not presented as mutual consensus over 'risk' between peers, but as a functional exchange for UK aid: '(I)f they cooperate then they are likely to get more assistance from us on building their capability, and aid and assistance on other fronts … it improves the business environment, improves the credibility … and probably development aid and trade' (L).
Egypt, another former colony, was less open to this basis for cooperation, which demonstrates Southern states' 'push back' and autonomy (Ostrand and Statham 2021;Zampagni 2016). Officers characterised how they saw Egypt's unwillingness to co-operate as creating feelings of mutual distrust and bad faith: 'Although IEI had good cooperation with airlines and individuals, generally due to lack of political will the authorities just did not cooperate. Also, they use all levers to pursue their own agenda, such as resumption of flights to Sharm el Sheikh, their pursuit of Muslim Brotherhood, foreign aid, a possible deal like with Turkey on migration … delaying visas for (UK) embassy officials etc.' (L). In a similar way, the UK's inability to generate effective cooperation with Thailand on terms suiting UK interests, was presented as resulting from the untrustworthiness of Thai authorities and police: 'We have to be careful in what (intelligence) we share to ensure the information won't be abused by Thai authorities' (A). '(T)he … thing in Thailand is about trust … that is why we are wary of what kind of cooperation we give' (U).

Conclusion
Here we tried to establish the influential, but largely hidden, role of liaison officers in shaping institutional understandings of 'immigration risk' that are implemented in visa decisions. This required assembling scant available public official information, making FOI requests, and interviews with operational officers. Without full HO disclosure, our data has limitations, and our interpretations regarding the internal workings of the liaison network remain somewhat speculative. Nonetheless, we advance an original evidenced account that unpacks a new perspective on liaison officers as influential interlocuters, who mobilise, frame, and communicate 'risk' definitions across institutional levels and national borders. While drawing a steer from the HO's centralised risk assessments, their day-to-day engagements with foreign law and border enforcement officials allows them to generate their own specific local 'risk' assessments. This important 'risk' knowledge is packaged into guidelines and disseminated directly to consulate and airline staff who make decisions on visas and check visa and travel documents at airports. They are effectively local experts-cum-messengers who deliver current thinking on visa 'risk' on location. Officers also feedback their own risk assessments through the network to inform the HO's general 'risk' approach. In a sense, their agency in 'risk' knowledge production and communication is vital in brokering understandings of 'immigration risk' between UK visa policy and implementation stages, because it works across institutional levels, discreet immigration policy community sectors, and national borders.
Notwithstanding their high autonomy for defining 'risk' in the field, we see that the perceptions that shape officers' decision-making still work within a set of professional norms and institutional culture that reinforces a HO perspective. We argue their 'risk' understandings are most likely driven by internalised institutional social pressures, where there are few incentives to challenge dominant orthodox norms for treating specific foreign nationals. Here the institution's 'cultural bias' (Douglas and Wildavsky 1982) in constructing and using 'immigration risk' is instructive. While the Home Office refuses to publicly acknowledge its supposedly scientific criteria and method for assessing 'immigration risk' for foreign nationals, we closely examined visa policy statements and liaison officers' perceptions. This revealed circular logics and inconsistencies in the way 'immigration risk' was applied to justify stances towards nationals from specific countries. Generally, there seems to be predetermined perceptions of people from target Global South states, which contrast to those from Global North allies. 'Risk' assessments applied largely support these institutionally preconceived stances. Officers use a pseudo-scientific mantra of objective 'risk' assessment, but usually make decisions based on personal interpretations of local situations, sometimes even falling back on national cultural stereotypes. Our claim is not that officers are insincere dupes, but professionals who work within a set of institutional norms that sets them 'targets' for preventing 'immigration risk', but few incentives or benefits for reflecting over the substantive basis of such 'risks'. We suspect specific biases with respect to foreign nationals are deeply embedded within the organisational culture, so that they exhibit 'path dependency' (Pierson 1993), and selfreproduce institutionally over time, as a dominant legitimating logic for decisions.
Another important contribution is that we show that officers also co-operate to a highly significant degree sharing 'risk' assessments with peers in Global North multistate frameworks. There is a very high degree of institutionalised multi-state co-operations. The UK participates strongly in alliances with Northern states through FCC and EU/EEA frameworks. Although some literature discusses institutional co-operations among EU member states, there have been few attempts to grasp the substantial scale, scope, and depth of multi-state frameworks at the operational level. Our study shows they are powerful, developed and extend well beyond Europe. Importantly, these alliances require pooling sovereignty and adopting common standards of 'risk'. Interestingly, officers recount that mutual professional trust is sufficiently high, that foreign ally states' visa decisions and 'risk' assessments are largely accepted at face value. Officers indicated that they share an institutional culture and norms with Northern peers, underpinned by intelligence exchanges, and with common shared perceptions of 'risk' and targets circulating among them. This places UK 'risk' norms and practices for visas within a broader international framework that overall works to underpin an inequality of global mobility between North and South.
However, another feature of our study relativised this simplistic North/South division to some degree. Our research explores liaison officers' agency in the 'opaque spaces between formal policies and outcomes' (Brodkin 2011, i199). By focussing on liaison officers' role in visa policy implementation, we see that although the clear distinctions between Northern allies (France and USA) and Southern targets (Ghana, Egypt, Thailand) evident from visa policy as it appears on paper are broadly sustained, there are also nuances that appear in the social world of implementation. Most clearly, France was a 'free mover' on paper, but was defined as a 'risk' in the field, where officers perceived 'risks' of onward migration by non-EU nationals, notwithstanding their overall good relations, trust and mutual respect with French officials. We also saw different degrees of perceived 'risk' applied towards nationals from Thailand, Egypt and Ghana. To understand how visa policies work as a form of immigration control, it is necessary to study implementation. While previously the focus has largely been on the street-level bureaucrats administering decisions in consulates, here we hope to have added liaison officers' 'risk' knowledge-production and communication role to the story.