The decision

UT Neutral citation number: [2022] UKUT 00335 (IAC)

EMAP (Gang violence – Convention Reason) El Salvador CG

Upper Tribunal
(Immigration and Asylum Chamber)

Heard at Manchester Civil Justice Centre


Heard on 27th April and 9th June 2022
Promulgated on 16 November 2022




(anonymity direction made)

Secretary of State for the Home Department

For the Appellant: Ms G. Patel and Mr C. Holmes of Counsel instructed by Hallmark Legal Solicitors
For the Respondent: Mr C. Thomann, Counsel instructed by the Government Legal Department


(i) The major gangs of El Salvador are agents of persecution.

(ii) Individuals who hold an opinion, thought or belief relating to the gangs, their policies or methods hold a political opinion about them.

(iii) Whether such an individual faces persecution for reasons of that political opinion will always be a question of fact. In the context of El Salvador it is an enquiry that should be informed by the following:

(a) The major gangs of El Salvador must now be regarded as political actors;

(b) Their criminal and political activities heavily overlap;

(c) The less immediately financial in nature the action, the more likely it is to be for reasons of the victim’s perceived opposition to the gangs.

(iv) As the law stands at present, so taking the disjunctive approach, those fearing gang violence in El Salvador may be considered to be members of a particular social group where they can demonstrate that they share an innate characteristic, a common background that cannot be changed, or a characteristic so fundamental to their identity or conscience that they should not be forced to renounce it.


El Salvador: Country Background
The Gangs: History and Context

The Relationship Between the Gangs and the State

The Third Generation Gang: Political and Social Demands

Size and Reach

Social Control and Influence

Infiltration of the State
The Refugee Convention: General Approach

Political Opinion

Membership of a Particular Social Group

Country Guidance: Discussion and Findings
Political Opinion

Membership of a Particular Social Group
The Appellant’s Case





1. The Appellant is a national of El Salvador born in 1988. The Respondent accepts that he cannot be returned to El Salvador because he faces a real risk of serious harm at the hands of the gangs there, from which the Salvadoran government is unable to protect him. The sole question to be determined in this appeal is whether that harm amounts to persecution for one of the five reasons set out in the Refugee Convention.

2. The Refugee Convention is not designed or intended to offer protection from any kind of harm. Its ambit does not extend to the terror caused by natural disaster, or to the inconvenience and discomfort of living in country A when you would rather enjoy the more liberal society offered by country B. It is long established that it is not engaged by the fear of generalised violence, or of criminality. As Baroness Hale puts it:

“Not all persecution gives rise to a valid asylum claim. Very bad things happen to a great many people but the international community has not committed itself to giving them all a safe haven.”

Fornah v Secretary of State for the Home Department [2007] UKHL 46 [at §97]

3. The Secretary of State contends before us that this is one such case. The First-tier Tribunal agreed, and dismissed the Appellant’s appeal. The Secretary of State submits that the Appellant fears violence and oppression from criminals who are hostile to him because he has resisted their attempts to exploit him for financial gain. This, she says, places his claim squarely outwith the protection of the Refugee Convention, and in this regard the Secretary of State finds support from domestic and comparative authority. In the Secretary of State’s view, the Appellant can succeed on neither of the grounds advanced on his behalf: he does not fear persecution ‘for reasons of’ his political opinion (imputed or otherwise), nor his membership of any particular social group.

4. The Appellant accepts that a fear of crime per se does not engage the Refugee Convention. He contends however that the situation in El Salvador is now such that the gangs must be considered to be actors of persecution, applying the ‘minimum standard’ definition found in the Qualification Directive 2004/83/EC. The harm he fears is “for reasons of” his stance against these actors of persecution, and so properly understood, his is a claim rooted in political opinion. The Appellant asks us to set the First-tier Tribunal’s rejection of this argument aside, and to find that he is a refugee. In the alternative the Appellant asks us to find that the persecution he fears is for reasons of his membership of a particular social group, defined as ‘those opposed to the gangs in El Salvador’ or ‘actual or perceived informants’.

5. We are grateful to the parties for their preparation and presentation of these competing arguments. This is a judgment to which we have both contributed. We begin by summarising the present situation in El Salvador, as it emerged in the detailed evidence before us.

El Salvador: Country Background

6. We have considered all of the country background evidence before us but have drawn on the following evidence in particular:

i) Country Policy and Information Note El Salvador: Actors of protection [Version 1.0 February 2021] (‘Protection CPIN’)

ii) Country Policy and Information Note El Salvador: Fear of gangs [Version 3.0 January 2021] (‘Gangs CPIN’)

iii) Response to Information Request Country: El Salvador [19th April 2022] (‘CPIN Response’)

iv) The UNHCR Eligibility Guidelines for Assessing the International Protection Needs of Asylum-Seekers from El Salvador [15th March 2016] (‘UNHCR Guidelines’)

v) The written and oral evidence of Dr Andrew Redden, Senior Lecturer in Latin American History at the University of Liverpool

vi) Recent reports on the current state of emergency, from Al-Jazeera El Salvador extends state of emergency amid gang crackdown [26th May 2022] and the Guardian El Salvador accused of massive human rights violations with 2% of the adult population in prison [2nd June 2022]

vii) Political Refugees from El Salvador: Gang Politics, the State and Asylum Claims by Professor Patrick J. McNamara, Department of History, University of Minnesota, published in the Refugee Survey Quarterly 2017, 36, 1-24.

viii) Resisting Criminal Organisations; Reconceptualising the ‘Political’ in International Refugee Law by Professor Amar Khoday, Department of Law, University of Manitoba, published in the McGill Law Journal 2016, 60:3, 461

ix) The Transformation of El Salvador’s Gangs Into Political Actors, by Douglas Farah, published online in June 2012 by the Center for Strategic and International Studies

x) A series of articles by ‘InSight Crime’, an investigative organisation widely cited in the CPINs, published online: Political Mafias Helped Empower Gangs, says El Salvador Expert (24th January 2019), 3 Dirty Secrets Revealed by the El Salvador Gang ‘Negotiations’ (4th September 2020) and US Blacklists El Salvador Officials, Bolstering Accusations of Gang Pacts (9th December 2021)

7. The evidence in this appeal is largely uncontentious. The expert witness Dr Redden has been writing about Latin America for 15 years and has since 2013 spent a considerable amount of time in El Salvador. When he is out of the country he maintains regular communication with his contacts there. We found Dr Redden’s evidence to be balanced, helpful and informative. Mr Thomann took no issue with his expertise or with the accuracy of his reporting about the general situation and we are satisfied that he is an expert, and that he understood his duty to the court. The Appellant has taken no issue with any of the information reproduced in the CPINs. Both parties agreed that the UNHCR is well placed to comment on the situation pertaining in El Salvador, and that the UNHCR Guidelines should be afforded a considerable degree of respect.

8. We begin by summarising the historical context and some general background about El Salvador, before going on to set out the particular factors which are said by the Appellant to indicate that the gangs can properly be considered to be political actors.

The Gangs: History and Context

9. El Salvador has by far the highest murder rate in the world, which is more than doubled when ‘disappearances’ are added to the calculation. The vast majority of the non-state violence is perpetrated by gangs. The evidence consistently reports that there are as many as 60,000 gang members in El Salvador, with up to 700,000 others – some 11% of the population of 6.5 million – in some way connected to their activities: this estimate, reported by the UNHCR in their 2016 Eligibility Guidelines [at page 10], comes from the government of El Salvador itself.

10. Most gang members operating in El Salvador today are affiliated to either Mara Salvatrucha (referred to hereinafter as MS-13) or Barrio-18 (B-18), described by UNHCR as “large transnational gang structures that have their origins in the Californian gang scene”. Research cited in the Gangs CPIN estimates that 84% of gang members owe allegiance to one of these groups: smaller gangs do exist but these are increasingly aligned with one of these major players. MS-13 is thought to be the one of the largest gangs in the world. B-18 is smaller and has split into two factions: the Sureños (Southerners) and Revolucionarios (Revolutionaries). At street level the gangs comprise local cells, and these are generally based in poor, or lower-middle class neighbourhoods. If affiliated with MS-13 these are referred to as clicas, and if B-18, canchas. These cells pursue a strategy of “exclusive control” over their neighbourhood, using force to repel challengers. Within each territory the gang tries to control all local criminal enterprises which sustain members’ livelihoods, such as extortion, prostitution, and selling drugs.

11. Above the clicas / canchas is the next strata of organisation. These regional level groups are known as programmas if MS-13, and tribus if B-18. These groupings comprise more senior gang members, sometimes referred to in English as ‘captains’, and ranfla or palabreros in Spanish. Where these leaders are imprisoned they continue to manage the gang’s affairs by telephone: UNHCR, for instance, reports that many of the extortion rackets targeting businesses or transport links in wealthier neighbourhoods are directed in this way. Above that strata are the ultimate, national leadership.

12. The modus operandi of the different gangs is very similar. Whilst UNCHR draws a distinction between B-18 which is less disciplined and more “trigger-happy” than the more organised and bureaucratic MS-13, they do emphasise that both are extremely violent. Persons who resist the authority of a local gang or who even just inadvertently cross it, who collaborate with the security forces or rival gangs are reportedly subject to swift and brutal retaliation. Not only are such persons killed, but UNHCR reports their families can be targeted as well. At times of heightened confrontation with the government the gangs have been known to impose collective punishment on entire neighbourhoods for perceived infractions.

13. Dr Redden explains how El Salvador found itself in this situation by setting out the historical context. Three themes emerge from his survey of the past century. The first is that El Salvador has persistently high levels of social inequality and poverty. The second is that it is a country which, by its close proximity to the United States, became a proxy host for a vicious cold-war era conflict. The third is that for both of those reasons, El Salvador’s modern history has been marred by the use of extreme violence as a means of political control.

14. In this deeply stratified society the ruling, landed class have thought little of suppressing the dissent of the poor with extreme violence: see for instance La Matanza (‘the slaughter’), a 1932 massacre of an estimated 30,000 indigenous farmers who protested against their conditions. The state’s recourse to violence finds expression later in the 20th century with the formation of organisations such as ORDEN, vehemently anti-communist death squads whose raison d’être was to identify, and exterminate, perceived opponents of the elite. Dr Redden sets out how these forces coalesced to create a society where violence, and particularly violence for political ends, is entirely normalised.

15. It was against this background that civil war erupted in 1980 between supporters of the left-wing Frente Farabundo Marti para la Liberación Nacional (FMLN) and the right-wing American-backed Alianza Republicana Nacionalista (ARENA). In the 12 years that followed approximately 75,000 people were killed and over 2 million people displaced. Dr Redden writes that from the outset of war, “the state began a ‘scorched earth’ campaign against its most marginalised rural populations. Entire communities were massacred. Rape was commonly used as a weapon of war by elite troops”. In 1993 the International Truth and Reconciliation Commission reported that 85% of the acts of violence committed during the war were attributable to the state, or to state-sponsored agents.

16. Unfortunately, writes Dr Redden, the underlying social problems which caused the war were completely unaffected by those 12 years of bloodshed. In addition to its historical social inequalities, the country now had to grapple with the legacy of war: two large standing armies had to be demobilised and found employment. Children who had fought were basically ignored because no one wanted to admit to having used child soldiers. In the absence of any gainful employment materialising, death squads linked to the extreme right resurfaced. Weapons were not taken out of circulation. There were no resources to provide therapy or psychological help for the brutalised population. All of this sowed the seeds for the social problems faced by El Salvador today. By 1995 the number of people killed by violence annually in the country was higher than at any time during the war itself.

17. Meanwhile in the United States the Salvadoran diaspora was subject to its own pressures. Dr Redden points to the marginalisation of migrant communities – Irish, Jewish, Italian - in the cities of the United States as the main driver for the evolution of urban gang culture from as early as the 1800s. By the time that Salvadorans started to arrive in America in large numbers during the 1970s, other Latino communities had already formed their own organisations. Salvadoran youth found themselves at the margin of the margins, rejected not only by mainstream American society, but by the other Latino gangs who were already well established in cities like Los Angeles. They therefore created their own. The two most prominent of these were MS-13 and B-18. Dr Redden writes “while the motives for joining a gang are often complex, in broad terms these gangs offered solidarity, protection and a group identity that was otherwise lacking among young people who had experienced and continue to experience significant trauma, dislocation, social marginalisation and violence”.

18. In 1996 the Clinton administration introduced automatic deportation provisions for migrants who were sentenced to more than 12 months in prison. This, and other immigration-control measures, saw tens of thousands deported from the United States to the countries of Central America. In El Salvador this wave of ‘returnees’ consisted primarily of young, male gang members who had little experience of anything other than crime. As the ICG, cited in the Gangs CPIN, put it: “thousands of adolescents were roaming the streets with no jobs and little else to do. The sense of belonging offered by the gangs was too much for many of them to resist” [at 4.1.2]. The Los Angeles gangs MS-13 and B-18 thus quickly assimilated and outgrew existing local outfits. Their rivalry caused them to pursue an aggressive - and hugely successful - recruitment strategy which as Dr Redden explains, “took advantage of the disaffection and social deprivation of demobilised, ex-combatant youths who had not been recognised in the peace process”.

19. Today the economic activity of the gangs has expanded from localised, small scale racketeering – primarily the extortion of small businesses in exchange for ‘protection’ – to larger-scale mafia-like control of specific industries such as transport, and to multinational trafficking in arms and drugs in co-operation with other crime groups such as the Mexican cartels.

The Relationship Between the State and the Gangs

20. From the early 2000s the response of the Salvadoran state to the challenge of the gangs was the ‘Iron Fist’ policy: mano dura. This consisted of tough, and often violent, repression of the gangs, with the military operating in tandem with the civilian security services. Dr Redden writes that this strategy was only partially successful. Although many thousands were imprisoned or killed, the gangs – at that stage consisting of small, largely independent street gangs or clicas – responded by becoming more unified and professional, developing an identifiable hierarchy and creating an organised crime structure with reach across El Salvador and links with other crime groups in the region.

21. In 2012 the then government tried a different approach. President Mauricio Funes of the FMLN entered direct negotiations with the gangs, desperate to do something about the violence ravaging the country. As FMLN officials signalled that they wanted to meet the gang leadership, MS-13 and the two B-18 factions came together to form what Professor McNamara calls a “criminal superstructure” which then entered into negotiations with the government, brokered by the Catholic Church and overseen by the Organisation of American States. In return for peace the gangs would get less restrictive prison conditions for the jailed, and fewer police raids into gang-held towns and neighbourhoods in the cities. The truce lasted almost two years, during which it is estimated 5500 lives were saved.

22. Dr Redden identifies two new problems that arose from the truce. First, the gangs used the space it afforded them to regroup, re-arm and reorganise. Second, the leadership came to realise the extent of the power that they held. When the truce collapsed the homicide rate immediately soared to record levels, and once again the government came under massive political pressure. The gangs then understood that in effectively being able to “put their foot on and off the gas”, they held enormous power and influence over the government.

23. By the 2014 Presidential elections the demands of the gangs had become more specific. Courted by both major parties, the gangs secured a promise from the FMLN of $10 million to be used by gang members as a source of micro-grants to create legitimate businesses. It is not known what they were promised by ARENA, although it is known that negotiations took place. Whatever it was, the FMLN offer apparently trumped it, because the gangs took to the streets to actively support their candidate. Gang members controlled access to voting stations and prevented ARENA members and supporters from voting. In several cases ARENA election monitors had their credentials removed from them, were told to get away from polling stations and were threatened with physical harm if they tried to get the help of the police. The FMLN were secured victory: “gang leaders learned from this experience that they could play a decisive role in electoral politics – a completely new possibility that could extend their power and influence throughout the country”.

24. Six months after that election, President Sanchez Ceren of the FMLN reneged on his promises. The result was an immediate and dramatic rise in the body count, beginning in March 2015. Police officers, both off and on duty, were targeted for assassination; police stations were attacked with grenades, high power machine guns, and improvised explosive devices. Gangs used safe houses, encrypted satellite phones and drones to monitor police movements. Car bombs were used for the first time since the war. Government attempts to damage the gangs by de-segregating the prison population – hoping that the incarcerated leadership would kill each other, thus weakening the organisations on the outside – backfired, as MS-13 and B-18 instead reached a ‘prison pact’. By the end of July 2015 their co-operation was in evidence on the outside as they called a national bus strike with the aim of disrupting the economy: it was enforced by the killing of non-striking drivers. This upsurge in violence eventually led to the Supreme Court of El Salvador designating the gangs as ‘terrorist organisations’: in his article Professor McNamara cites one of the Justices’ explanation that in that context the term ‘terrorism’ should be understood to mean “politically motivated violence designed to destabilise society”.

25. In the decade since 2012 El Salvador has seen this cycle repeat itself time after time. Repressive mano dura policies lead to a violent backlash from the gangs; the government, or parties hoping to be the government, enter into negotiations with the gangs to end the bloodshed; the agreement is broken; the repression starts again. In his oral evidence Dr Redden explained how it emerged that the gangs were courted by both ARENA and FMLN in the 2014 elections, and then again in 2019. On the table from the government side was improved prison conditions for gang members, while the gangs are always able to offer relative stability and low rates of violence. Leading up to and following both elections there was a lull in violence, followed by an obvious upswing as the gangs expressed their displeasure at terms not being met.

26. Today El Salvador is at the repression stage of the cycle. President Bukele, who came to power in 2019, famously uses the twitter profile of the “coolest dictator in the world”. He took office claiming to have a mano dura policy which would actually work: the ‘Plan for Territorial Control’ combined a renewed iron fist security strategy with promises to invest in the poorest neighbourhoods. He is however widely believed to have negotiated with the gangs in order to reward voters with a lull in violence following his election. In December 2021 the United States’ Treasury Department imposed sanctions on two members of his cabinet for negotiating with – and paying – MS-13 for precisely such an accommodation.

27. This tacit truce has this year, for reasons that remain unclear, dramatically collapsed. In March 2022 the gangs effectively declared war on the government. Between the 25th and 28th March the country saw a huge spike in killing, the like of which the country had not seen since the war: police officers, bus drivers and passengers, street vendors and their customers were gunned down in apparently indiscriminate violence. In those 72 hours 87 people were killed. Dr Redden’s view, that this rampage was a deliberate negotiation strategy, is strongly supported by the other country background evidence before us. The CPIN Response of April 2022 cites ‘Insight Crime’ [at 1.2.2]:

“According to analysts, these brief spikes typically occur when there is a rupture in negotiations between the government and the gangs, with the gangs using bodies as bargaining chips…the gangs use their ability to alter the levels of violence as leverage to press the government into meeting certain demands….’we shouldn’t see it the start of a war, but rather a cry for attention’ said Juan Martinez d’Aubiusson, a Salvadoran anthropologist and gang expert, adding that the gangs are likely using the killings to express discontent with secret government negotiations”.

28. Another source cited in the CPIN Response, Harvard academic Manuel Meléndez-Sánchez, describes the killing spree as “political leverage” over the government. An article from the Associated Press, which draws on interviews with members of Bukele’s Nuevas Ideas (New Ideas Party), suggests that the gangs were enraged by the government take-over of two key bus routes in San Salvador. The gangs use their control over the transport system to launder money, and the loss of some 300 buses would have slashed a revenue stream.

29. In response to the March violence President Bukele declared a state of emergency, for which he has the overwhelming support of parliament, and if polls are to be believed, the Salvadoran public. In the media articles supplied by the parties it is estimated that approximately 35,000 people have been rounded up and imprisoned. Although the President claims that these detainees are all gang members, sources cited by al-Jazeera and The Guardian allege that many are just young males from poor neighbourhoods; civil rights groups, both domestic and international, have protested about the apparent indiscriminate nature of the arrests, lack of due process and prison conditions that have seen as many as 20 young men die in a matter of weeks. The CPIN Response cites multiple sources (AP, The Guardian, the BBC and InSight Crime) to the effect that Bukele has reacted to such criticism by ‘doubling down’. Prisoners have been confined to their cells, windows have been boarded up to block out light and fresh air and food rations cut. The state of emergency has been extended again and four constitutional rights have been suspended: the right to free association, the right of defence, the 72 hour limit for administrative detention and the sanctity of private correspondence and communication.

The Third Generation Gang: Political and Social Demands

30. In his article Professor McNamara sets the recent history of the gangs in El Salvador in the context of a classification system used by sociologists and military scholars to describe the evolution of gangs. ‘First generation’ gangs are small, local, and interested in controlling turf and making money from petty crime in their immediate neighbourhood: in the context of Central America, the typical clica. ‘Second generation’ gangs evolve from these small cells into larger, organised groups with defined hierarchies: in the UK these would be termed ‘organised crime groups’. A gang becomes classified as ‘third generation’ when it develops strategies and goals that seek more specific social and political objectives.

31. Professor McNamara believes that the major gangs in El Salvador have reached this point. Having set out the history summarised above, and in particular the socio-economic demands made by the gangs during the 2014 elections, he pinpoints August 2015 as a pivotal moment:

“One month after the transportation strike, the gangs announced the formation of a new supra-organisation that represented a powerful opposition movement against the national government, local governments and civilians who challenged gang demands. Instead of fighting against rival gangs for control over territory, the new organisation sought recognition as an active political actor on the local national and international stage. Gang leaders named this new political and economic force ‘Mara 203’ which is El Salvador's international telephone code. Mara 203 spokespersons stated that they intended to seek greater political influence by participating in elections, controlling local elected officials, especially mayors, infiltrating police and military forces and taking over civilian associations created in the aftermath of the 1980 - 1992 civil war”.

32. Professor McNamara writes that in taking this step, the gangs were in effect taking a well-established path to power since both the major political parties at that point had gained legitimacy through the use of force: the FMLN had been a guerrilla insurgency, and ARENA had its roots in paramilitary death squads. Like those parties before them, gang leaders adopted the mantle of representing the unrepresented. Gang leaders argued that they were looking to protect the most vulnerable people in society from the arbitrary abuses of the security forces and corrupt politicians. Statements alluded to the economy benefitting only a small minority: they argued that the poor were tired of being ignored and frustrated by the lack of assistance to relieve extreme poverty.

33. By March 2016 the now united gangs – the “Co-ordinating Committee” - had announced an end to their previous hostility towards each other. The committee consisted of six leaders of MS-13, and three from each of the B-18 factions. They asked the government to begin formal negotiations to allow them to participate in the political process. The Co-ordinating Committee enlisted the assistance of the Lutheran Bishop of San Salvador to host meetings, and Heinrich Haupt, the German Ambassador, was a regular observer. In the 2017 elections the Committee flexed its political muscle by prohibiting FMLN flags and symbols in their territory and vowing to keep their candidates out of office. El Faro reported at the beginning of 2017 that the Committee was looking at disbanding the gangs altogether in exchange for immunity, improved prison conditions and the delivery of “better health care, jobs and education for their communities”. These statements have been accompanied by the diversification of gang economic activities: members now own legitimate businesses such as nightclubs, gas stations, bakeries and pharmacies. A new level of professionalisation has been encouraged, with the most promising members being sent to university to become lawyers and accountants. Professor McNamara concludes that:

“International observers have recognized the changes that have taken place among the gangs in El Salvador. Dr M. Gaborit, PhD, Chair of the Department of Psychology at the Universidad Centroamericana, argues that “the gangs have more power than any other political party – in terms of money, arms, vertical structure, national presence, and personnel. The gangs are the principal political actors for many and they have recognized their common interests”. The United Nations first reported these changes in March 2016, noting that the gangs, especially MS-13, have demonstrated a “growing military sophistication that is increasingly transforming the affiliated gangs into a force that can combat the State and hold territory”. The import of this transformation signals that “one of the most remarkable changes occasioned by the truce has been the dramatically increasing political sophistication with which the leaders of B-18 and MS have come to couch their grievances with the government and assert their increasingly overt political ambitions”. Gang members themselves speak more often about a “metamorphosis” that sounds at times like a self-serving way of acknowledging past violent crimes without taking responsibility for victimizing so many innocent civilians. But gang members themselves have been raised in a post-Civil War society where human rights violations, massacres, kidnappings, and extra-judicial killings that occurred during the war have gone unpunished. Organizations, if not individuals, who were responsible for those crimes have become legitimate political leaders. As long-time human rights activist Benjamın Cuellar acknowledges: “The gangs are sons of impunity”.

34. In his article Professor Amar Khoday argues, like Dr Redden, that the gangs have become emboldened by their success in negotiating with successive governments. He draws on the growing body of scholarly literature on gangs, in particular for our purposes, the work of national security analyst Douglas Farah. This led us to look at the Farah article cited by Professor Khoday ourselves. The parties were informed of this at the hearing and they were given a copy, with no objection being raised. Farah’s research suggests that the gangs are now assessing support for particular candidates in exchange for the ability to dictate policy. We cite it here because it is the primary source for the views expressed by Professor Khoday about the politicisation of the gangs. Mr Farah reports:

“Gang members I interviewed last month indicated they are interested in political power. Surprised and pleased with the results of the negotiations, their leaders are beginning to understand that territorial control and cohesion make it possible for them to wring concessions from the state while preserving their essence of their criminal character. They are already discussing backing certain candidates for local and national office in exchange for protection and the ability to dictate parts of the candidate’s agenda”.

35. This evidence is largely consistent with material cited in the various CPINs we have been given. Although Mr Thomann was able to identify one contrasting view - expressed by the United Nations Special Rapporteur (‘the UNSR’), who in her 2017 report found “no indication that gangs have an ideological basis or political programme” [cited at 9.14-15 of the Gangs CPIN] - the remainder of the evidence was consonant with the findings of Mr Farah and the academics we have cited.

36. In particular, the CPINs draw heavily on the work of a consultancy called InSight Crime, analysts who specialise in the study of organised crime in Latin America. InSight Crime is run by investigative journalists, academics and a former British Army officer. It notes on its website that “unlike many organizations, which rely on open source material to compile their analyses, InSight Crime goes into the field to speak with local sources, government entities, international law enforcement and the criminals themselves”.

37. InSight Crime has written extensively on MS-13, which it describes as “a social organization first, and a criminal organisation second. The MS-13 is a complex phenomenon. The gang is not about generating revenue as much as it is about creating a collective identity that is constructed and reinforced by shared, often criminal experiences, especially acts of violence and expressions of social control” [at 6.1.3 Gangs CPIN]. In respect of the gangs’ involvement in the political structures of El Salvador, Insight Crime concur with the other sources before us in reporting that all major political parties – including Bukele’s Nueves Ideas alliance – have repeatedly negotiated with, and worked together with, the gang leadership to deliver overt political results for the former, and more subtle political results in the form of greater power and increased social and economic control for the latter. For instance, they report that jailhouse intelligence reports, prison logbooks and interviews with key players reveal that since at least October 2019, Bukele engaged in talks with senior gang leadership – mostly MS-13 but also B-18 Sureños. The talks were conducted by Osiris Luna, the Director of Prisons, and Carlos Marroquin, the head of the government’s Unidad de Reconstruccion del Tejido Social (Social Fabric Reconstruction Unit), a body designed to implement social, educational and economic programs in marginalized neighbourhoods. In its investigations into these talks Salvadoran newspaper El Faro reported that the gang leaders agreed to keep violence to a minimum, and to support New Ideas in the congressional and local elections; the quid pro quo was more concessions for prisoners and “the promise of social and economic programs inside and outside prisons”.

38. InSight Crime, and again El Faro, have also produced evidence that Bukele – and his close ally Carlos Marroquin - have a long-standing history of collaborating with the gangs. When he was candidate for mayor of San Salvador in 2015 he sent Marroquin and others into gang held neighbourhoods to “open space” for his campaign. Once he won, he created the Tejido social unit referred to above, and appointed Marroquin its head. The unit then spearheaded ongoing talks with the gangs aimed at reforming and clearing the capital’s historic city centre. Their success in managing to do this then fuelled Bukele’s presidential run. It was success due in no small measure to arrangements with the gangs, who under his leadership “steadily tightened their grip on all aspects of life” in that historic city centre [see Gangs CPIN at 7.2.2 and 7.3.8].

39. All of this leads InSight Crime to the view that “the gangs are a political force”. They are cited in the CPIN Response [at 7.3.2] explaining that “The relationship between the gangs and politicians is mutually beneficial by nature; the gang expands or maintains its influence over a certain area, while the politician gains or maintains local power. This partly explains why this dynamic is so frequently observed in El Salvador….this type of mutually beneficial relationship has also been observed in national politics”. In their September 2020 article they put it like this:

“Try as the government might to diminish their importance via forceful rhetoric and nationwide plans to incarcerate them en masse, the gangs continue to illustrate that they have political and social capital. Part of this, of course, comes from their keen understanding that homicides, public opinion polls and foreign direct investment are intimately linked”.

(emphasis in original).

40. Asked to explain the notion that a criminal enterprise should be seen as political, Dr Redden pointed out that the gangs have to be assessed in the social milieu in which they exist. Their origins lie in the need to express social solidarity, group identity and achieving protection. The same can be said for the FMLN and ARENA. He argues that at a grassroots level where you have a group of young people coming together for these reasons that is itself a political act; today the gangs and the official political parties are all looking to advance their constituents’ interests. Another factor that had to be considered, and one that Dr Redden was at pains to emphasise, was the long-standing normalisation of violence in El Salvador, as a political and social tool. The extreme nature and extent of violence perpetrated by the gangs might be thought to preclude them from any political legitimacy. However, in practice and when considered in the context of El Salvador, where all sides have perpetrated horrifying violence, Dr Redden argues that the criminal activities and violence of the gangs are simply a reflection of the status quo. The present government has overseen the resurgence of political death squads. Violence is perceived to be “completely normal” in El Salvador – it permeates all aspects of life. There is for instance an extremely high rate of domestic violence. Ultimately Dr Redden cautions against looking for a particular ideological motivation in the assessment of whether the gangs are engaging in ‘politics’ (if they have one, it is an “extreme form of nihilism”). In the context of El Salvador they do not need to express any particular or formalised belief system to be engaged in the exercise of power, and to see those who cross them as acting contrary to that.

41. On this latter point, Dr Redden finds support in the conclusions of Professor McNamara:

“Gang members themselves view a person’s refusal to comply with their demands as being anti-gang. From the perspective of the gangs, civilians commit acts of political resistance when they report crimes to the police, when they sponsor or participate in anti-gang youth programmes, when they refuse to pay extortion demands, and when they abandon their homes to try to find safety rather than submit to gang authority…At issue here is a more complete understanding of “political”. The gangs are now committed to more than simply earning money. They seek power and influence to protect allies and to improve the lives of people who support them. Poor people, young people, and women have political opinions even if they do not use obvious or partisan political language to express those views….”

Size and Reach

42. All of the evidence before us indicates that the gangs are now in effective control of a good deal of the country. The UNHCR Guidelines, for instance, state that the gangs “operate country wide”, and the Gangs CPIN cites the International Crisis Group estimation that they are active in 94% of all municipalities. Dr Redden argues that this “vast territorial control” puts them at least on a par with the officially elected administration. This control is organised from the top down. In his report Dr Redden had cautioned that due to the secretive nature of the gangs, scholars could not be certain about just how organised and hierarchical they are, but in his oral evidence he resiled from this to say that having conducted further research he now agreed with the US Department of State and Homeland Security which describes the major gangs in El Salvador as “highly organised, hierarchical, transnational criminal organisations”. The scholarship, and his own observations, indicate that when orders are given they are followed. That is how they have come to control so much territory. Dr Redden concurs with Professor Khoday that this is an important factor in understanding the extent of the power that the gangs now hold.

Social Control and Influence

43. In those areas where gangs hold a total or significant degree of power, they can be seen to exercise clear and comprehensive social control of the population. Dr Redden points out that during the pandemic it was gangs who enforced lockdown, issuing threats against civilians for flouting the government imposed curfew. We note that evidence recorded in the Gangs CPIN states that in certain areas gang members also reduced renta ‘taxation’ for the duration of lockdown, and even handed out food parcels. Dr Redden explained that in his experience populations under gang control may look to the gangs, rather than the government, to get things done. For instance community leaders would approach the local clica to ask for permission to paint a mural on a certain wall; small business people would look to them for help. The gang can dictate whether or not local children go to school. Civilians cannot simply cross from one territory into another without permission. Dr Redden is also aware of examples where the gangs have negotiated with local government to ensure that certain services are provided to the population.

44. Dr Redden also cited the practice of renta as a signifier of social control, and stressed that it is not perceived simply as a criminal act. Certainly for many victims it is plainly extortion under threat of violence: in these cases it is nothing more than a ‘protection’ racket. Dr Redden knows of examples where entire villages have been cut off because of the suffocating effect of extortion checkpoints on all roads in and out. In the communities where the gangs are most entrenched, however, renta can represent something more nuanced – a relationship between the local clica and the population that they see themselves as representing. The way that money is collected will accordingly vary. Businesses seen as thriving are more likely to be targeted regularly because the gang members – often themselves poor and marginalised - think that the owners can afford it. These businesses can often be squeezed into bankruptcy. But for the poor, the gangs may take a very different approach. Dr Redden cites the example seen on camera in the 2008 documentary La Campanera in which gang members went door to door in their community to collect money to support the family of a fallen comrade: “the youths asked nicely and explained what the money was for and did give the option to refuse (on camera): ‘if you can’t don’t worry’”. Although Dr Redden qualifies this evidence by pointing out that people could be “willingly” giving because there is always the implied threat of violence, he also referred in his oral evidence to one area that he is familiar with where the local MS-13 clica refused to extort the local population because the members had themselves grown up there- they saw themselves as representing that community and did not wish to exploit them (we note that in his oral evidence Dr Redden acknowledged that this benevolence was subsequently overridden by orders from above).

45. As he explained, the behaviour of different clicas in different areas is very variable, but it is often the perception of the gang members themselves that they are part of the communities in which they ‘work’. They share a sense of solidarity with the local population who see politicians are corrupt and simply there to serve the interests of the elite. Often this sense of solidarity is shared by the local civilians who are prepared to look past the means of the gangs: these are “people who are very marginalised and used to normalised violence”. It is in this context that Dr Redden feels able to draw an analogy between renta and taxation. Although he acknowledges that local people are not getting ‘services’ as such for the payment of renta - they are not getting their bins collected – they are getting relative peace in their neighbourhood. The gangs offer a “(para)military policing function”. The people don’t see much distinction between renta and official taxation: one goes in the pocket of the gangs, the other the politicians.

46. The Protection CPIN cites the ICG who report that “in some areas gangs have accumulated so much power that they have become de facto custodians of these localities, setting up road blocks, supervising everyday life and imposing their own law” [at 9.2.3]; a University of London study states that “strict compliance” is required of civilians living in gang held territory [at 10.10.1]. The same paper suggests that “the lives of gang members and residents are affected by anything from curfews to rules determining clothing and haircuts and any infraction – real or suspected – is punished with a severity the gang deems commensurate with the ‘offence’”. Professor McNamara points to a correlation between heightened control by the gangs, and the withdrawal of the state. He cites the view of gang expert Dr M Gaborit that “the government has simply abandoned parts of the country and the duties of governing; the gangs have taken over that space”.

Infiltration of the State

47. Another feature of the gangs’ power is the fact that they have successfully compromised all government institutions so that lines have become blurred. There is in Dr Redden’s estimation “chronic and critical” infiltration of all branches of the security services, including elite divisions in the army. He observes that this is particularly concerning since it suggests that gang-members are “receiving formal weapons and counterinsurgency training, and subsequently go on to have access to arms, equipment and intelligence”.

48. The infiltration of civilian state structures is reported to be similarly widespread. The police force is the institution most affected, particularly in rural areas. Professor McNamara writes that “many civilians know it would be dangerous to report crimes committed by gang members directly to the police, because police often forward that information to gang leaders”. There is also evidence of gang infiltration of the prison system, community organisations and municipal government. As to the latter, Professor McNamara states that there are town mayors who are believed to have negotiated with the gangs, with some, such as Mayor Canales of San Francisco Gotera, known to work directly for them. Similarly, the gangs have involved themselves in local civic organisations such as community development organisations, youth sports committees, and school boards. Gangs control access to public schools, with the teachers’ union estimating that as many as 100,000 children had dropped out as a result.

49. The judiciary is weakened by corruption and intimidation: in one case personally reported to Dr Redden prisoners on remand secured their release after the trial judge’s daughter was “given a note as she left school”. The judge and his family left the country. Local councils are subject to endemic corruption and gang infiltration. In his oral evidence Dr Redden gave the example of a mayor known to him who is openly in the gang’s employ. Even where the connection is more subtle, gangs are able to exercise control over, for instance, which way traffic flows down certain streets in order to protect their territory: all of this is arranged with the co-operation of local councils. As we have set out above, this liaison with politicians and elected officials is also documented to go well beyond the local municipality to the national level.

The Refugee Convention: General Approach

50. Article 1A(2) of the Refugee Convention 1951 provides that a refugee is:

"a person who owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable, or owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return to it”.

51. Article 1F is the only other place in the Convention where the term ‘political’ appears, but here in a different context:
F. The provisions of this Convention shall not apply to any person with respect to whom there are serious reasons for considering that:
(a) He has committed a crime against peace, a war crime, or a crime against humanity, as defined in the international instruments drawn up to make provision in respect of such crimes;
(b) He has committed a serious non-political crime outside the country of refuge prior to his admission to that country as a refugee;
(c) He has been guilty of acts contrary to the purposes and principles of the United Nations.
52. Article 31 of the Vienna Convention on the Law of Treaties lays down a ‘general rule of interpretation’ to be applied by parties to an international convention. Article 31(1) provides: a treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.

53. The domestic authorities have applied Article 31 of the Vienna Convention to the text and preamble of the Refugee Convention, such that the following propositions are now clearly established.

54. First, the Refugee Convention has a broad humanitarian purpose in that it exists to protect fundamental rights and freedoms: see R v IAT ex p Shah; Islam v SSHD [1999] UKHL 20, [1999] Imm AR 283; Fornah and K v SSHD [2006] UKHL 46, [2007] Imm AR 247. Linked to this, as noted by Lord Bingham in Sepet and Bulbul v SSHD [2003] UKHL 203, the Refugee Convention “must be seen as a living instrument in the sense that while its meaning does not change over time its application will”. To do otherwise would be to risk it becoming an anachronism. Lord Bingham cited with approval the following passage from the judgment of Laws LJ in R v SSHD ex parte Adan [1999] EWCA Civ 1948:

"It is clear that the signatory states intended that the Convention should afford continuing protection for refugees in the changing circumstances of the present and future world”.

55. Second, the need to adopt a broad purposive construction, consistent with humanitarian aims, applies not only to the Refugee Convention grounds [see Shah and Islam at 651F, 656], but also to the question of nexus [see Gomez (Non-state actors: Acero-Garces disapproved) Colombia * [2000] UKIAT 00007 at 22].

56. Third, the Refugee Convention does not protect everyone who fears serious harm: this would place an unduly heavy burden on its signatories. The key that opens the door to protection is discrimination, which is reflected in the ‘for reasons of’ nexus and the five ‘Convention grounds’: see Shah and Islam [at 650G-651B, 656E, 659F].

57. Fourth, in determining the causal nexus what matters is the “real reason”. As Lord Bingham said in Fornah and K (supra) [at 17]:

“The ground on which the claimant relies need not be the only or even the primary reason for the apprehended persecution. It is enough that the ground relied on is an effective reason. The persecutory treatment need not be motivated by enmity, malignity or animus on the part of the persecutor, whose professed or apparent motives may or may not be the real reason for the persecution. What matters is the real reason.”

58. It is against this background that decision makers are tasked with evaluating whether feared persecution is ‘for reasons of’ one of the five Convention reasons.

Political Opinion

59. In what circumstances might serious harm inflicted by a gang amount to persecution for reasons of political opinion?

60. In order to answer that question we were referred to a wealth of comparative jurisprudence in which courts in other jurisdictions have reached their own conclusions. Having read it all we now find, with some regret at having put the parties to the trouble of identifying it, that most of these cases were not of any great assistance to us. Many turn on their facts. Some claims were lost because they could not establish that the gangs in question had any political intent or motivation: see for instance the US case Santos-Lemus v Mukasey, 542 F. 3d 738 (9th Circuit, 2008), the Canadian appeal of Re XXX 2020 CanLII 1186590 or the more recent Belgian case of XX v the Commissioner for Refugees and Stateless Persons NR 243 676 (5th November 2020). Some failed because the evidence did not establish a causal nexus: Alvirurez-Gomez v Lynch, Attorney General No 15- 2181 (1st Circuit, 2016), Marroquin-Ochoma v Holder 574 F 3d 574 (8th Circuit, 2009). As is to be expected when it comes to factual assessment, we observe that there are also cases which go the other way: Re Orozco-Polanco No A75-244-012 US EOIR, Bolanos-Henandez v Immigration and Naturalisation Service 767 F 2d 1277. Others, like the lead American case of Immigration and Naturalisation Service v Elias-Zacarias 502 US 478 (1992) turn instead on the specifics of the applicable statute in the jurisdiction in which they were heard. In that matter the American statute required, inter alia, that the claimant establish that the ‘one central reason’ for persecution was a Convention ground, and further that the persecution feared had to be on account of the victim’s political opinion, rather than any views held by the actor of persecution. Neither of those requirements will feature in our analysis, which must be framed in accordance with the applicable law in the UK. Perhaps unsurprisingly we have in the end derived the most assistance from our own domestic authorities, in particular the starred Presidential decision of Gomez (Non-state actors: Acero-Garces disapproved) Colombia * [2000] UKIAT 00007.

61. We are conscious that Gomez is now 22 years old, and that it and related decisions pre-date the Qualification Directive by which, the parties agree, we are today bound. We have nevertheless found it helpful to begin at the beginning and to consider the approach that the courts there took to the question of political opinion before going on to consider the effect of the Directive.

62. Ms Gomez was a young lawyer who found herself in fear of armed men who had been threatening a client of hers. She believed that these men were in fact FARC guerrillas, who would see her defence of their target as political opposition to their cause. She relied on the decision of Mr RG Care in Acero-Garces [1999] INLR 460. This had been a similar claim. Mr Care had held that the appellant in that case would be “seen by her persecutors to be on the side of law, order and justice” and that since the protection of the Colombian authorities was only selectively applied, her claim succeeded on political opinion grounds. Gomez argued that she, too, was on the side of law order and justice, and that the persecution she feared was for reasons of this alignment.

63. As the headnote suggests, the Presidential panel in Gomez rejected that analysis. They were not satisfied that a wholesale approach of ‘good v bad’, or as they memorably put it “Star Wars generalisations” about “the side of law and order” v “dark forces” could be made about the situation in late 1990s Colombia. The analysis had to be far more nuanced.

64. The Tribunal begins by noting that the term ‘political opinion’ is not defined in the Convention itself but that taking a broad, purposive approach the following principles had developed. One did not have to be a member of a group or party to succeed. Protection should not be limited to cases where the individual had demonstrated political activity: a genuinely held opinion would suffice. The individual concerned does not even have to hold such an opinion himself; it is enough that the actor of persecution imputes that belief to him. An appeal could succeed where there was a mixture of motives for the persecution.

65. The decision then considers what kind of opinions might attract international protection. It notes that in conventional political theory the term “political” is confined to matters pertaining to government or governmental policy, and finds it hard to quarrel with Hathaway’s proposition that “essentially any action which is perceived to be a challenge to governmental authority is therefore appropriately considered to be the expression of a political opinion” [Law of Refugee Status, page 154]. On the other hand the term ‘political’ has a much wider application in other disciplines and contexts. One might for instance speak of “sexual politics” at the micro, family level, but this would be extremely unlikely, absent some other particular feature, to engage the Convention.

66. The Tribunal identifies that persecution by non-state actors is liable to fall somewhere between these two positions. In Canada (Attorney-General) v Ward (1993) 2 SCR 689, 746 the Supreme Court of Canada had to decide whether a former fighter in the Irish National Liberation Army, who had incurred the wrath of his former comrades by freeing civilian hostages, faced persecution for reasons of his political opinion. The Court rejected the AG’s submission, based on a proposition by Atle Grahl-Madsen, that the Convention only protected “opinions that are contrary to or critical of the policies of the government or ruling party”. Deeming this too narrow an approach, the Court instead adopted the definition suggested by Guy Goodwin-Gill:

“any opinion on any matter in which the machinery of the state, government, and policy may be engaged”.

Ward’s actions could not sensibly be construed as being contrary to the British government, but they were political nonetheless.

67. This approach was, shortly before Gomez, endorsed by the Federal Court of Appeal in Klinko v Canada (Minister of Citizenship and Immigration), a case involving an anti-corruption whistleblower in which the court emphasised that Goodwin-Gill’s definition did not depend on the machinery of government actually being engaged: it was enough that it “may be” so. Even with that reassurance, the Tribunal in Gomez expressed doubts that this definition was “in fact broad enough to encompass every type of situation relating to non-state actors of persecution”. It noted from the American caselaw that individuals could face persecution for reasons of their political opinion in situations far removed from the machinery of government. At [40] it said this:

40. As well as the need to adopt a broad definition of the term “political” there is also a need to recognise that the term is a malleable one. In the nature of politics, the boundaries between the political and the non-political shift in historical time and place. In Shah and Islam the point was made by Lord Hoffman that although women in contemporary Pakistan could constitute a particular social group, that did not mean that women anywhere or at any time could. It seems to us that the parameters of time and historical place are even more present in relation to the political opinion ground. That the definition of the adjective “political” must always be to some extent malleable flows from the fact that the nature of the power relationships and transactions that compose what is political vary from society to society. Sometimes political opinion may be located in a particular type of expression or activity, e.g. wearing western clothes in a highly fundamentalist Muslim country with strict social mores; sometimes not. In society A where trade unions adopt a combative posture towards the government, membership of a trade union may be tantamount to holding a political opinion; in society B it may not be so. The risk of extortion threats from a criminal gang will not normally be on account of political opinion, but in some societies where criminal and political activities heavily overlap, the picture may be different. Persons who hold posts in governmental agencies of the state at central or local level will not normally be capable of having political opinions attributed to them by groups opposed to the government. But if for example there is a major armed conflict going on between the authorities and guerrilla groups (e.g. Islamic fundamentalists in Algeria in the 1990s) then it may be that they will have attributed to them the political opinion of being on the government`s side rather than the fundamentalist Islamic side (Doufani (14798); see also Woldemichael (17663)).

68. In recognition of that malleability, the Tribunal cautions against drawing sharp distinctions: between criminal and political, economic and political or personal and political. Each case must be examined on its facts. At [§38] the Tribunal says this:

“…Even in the case of non-state actors therefore one cannot easily see how differences they may have with someone they persecute could be described as political unless they themselves have or express a political ideology or set of political objectives, i.e. views which have a bearing on the major power transactions relating to government taking place in a particular society. That is to say, the Tribunal doubts that the Refugee Convention ground of political opinion was meant to cover power-relationships at all levels of society…. Cases where an individual has been accepted as a non-state actor capable of imputing political opinion appear to be ones where that individual is effectively implementing the political views of either the state or some other body with political aims and objectives”.

69. In the executive summary at the end of the decision this is expressed more succinctly [at 73 (vii)]:

“To qualify as political the opinion in question must relate to the major power transactions taking place in that particular society. It is difficult to see how a political opinion can be imputed by a non-state actor who (or which) is not itself a political entity”

70. In the Colombian context, the Tribunal noted the expressly political aims of organisations such as FARC, and found that where the threat emanates from such a powerful guerrilla group “there will be less difficulty than otherwise in establishing that a possible opinion which such a group will impute to those who stand in their way will be a political one”. It also recognised the evidence that narcotics traffickers such as the Cali cartel had achieved such a size and influence over the state that their aims could not “easily be described as wholly non-political”: they had for instance funded the Presidency. There was considerable overlap between governmental activities and those of the cartels: there may therefore be a political reason for their persecutory acts. This does not however mean that in every case of persecution a political opinion will be attributed to the victim. In its summary of its findings [at §73 (x)] the Tribunal puts it like this:

“Even in a case where an appellant can make out a Convention ground of political opinion, he or she must still also establish that the persecution is on account of that political opinion. It is common sense under this nexus test that even where persecutors have political views about those they target, it may not always be the political opinion that motivates their actions. As was said in Jeah, the mere existence of a generalised political motive does not lead to the conclusion that the persecutor perceives what the claimant has said or done as political”

71. The position in Gomez may therefore be summarised as follows. The paradigm case of persecution for reasons of political opinion involves an oppressive state suppressing dissidents. The need for the Refugee Convention to be interpreted in light of its purpose has given rise to many variants of that classical situation. Individuals who have no opinion at all may be persecuted by the state which wrongly imputes dissenting views to them. Actors of persecution other than the state might persecute others on behalf of the state, or for their own benefit. It is not axiomatic that any persecution by a political entity will constitute persecution for a Convention reason, but the more overtly political in nature its objectives, the more readily it can be inferred that the persecution inflicted would be “for reasons of” political opinion. Where an agenda is not explicitly articulated it may nevertheless be gleaned where the persecutor has “views which have a bearing on the major power transactions relating to government taking place in a particular society” [at §38]. Each case must be determined on its facts.

72. In the course of submissions we have been referred to a clutch of domestic authorities in which Gomez was either expressly approved, or its logic otherwise applied, by the Court of Appeal. In Storozhenko v Secretary of State [2001] EWCA Civ 895; [2002] Imm AR 329 the court considered the claim of a Ukrainian man who had been assaulted by a police officer. He had pursued a complaint against this officer, and then fled, fearing the consequences of having done so. Brooke LJ dismissed the claim, finding it “manifestly artificial” to frame Storozhenko’s feared harm as persecution for reasons of his imputed political opinion: if the police in Ukraine sought to harm him it was from the desire to silence him as a witness and protect their colleague and themselves. In Montoya v Secretary of State [2002] EWCA Civ 620 a rich landowner sought sanctuary from Marxist rebels in Colombia who were demanding of him extortion payments that he could not meet; the Court upheld the analysis of the Tribunal that “there is no satisfactory evidence to show that the guerrillas’ motives were anything other than the purely criminal ones of desiring to extract extortion money”. Finally, in Jon Hairo Ortiz Suarez v Secretary of State [2002] EWCA Civ 722 the appellant deserted from the Colombian Army upon discovering his captain’s complicity in human rights abuses and corruption. Endorsing the approach in Gomez the court found no causal nexus to any Convention reason, observing that the persecutory intent of the claimant’s captain was to silence him as the witness to his crimes.

73. As is evident from this summary of the Court of Appeal authorities, and indeed Gomez itself, none of them turned on whether the actor of persecution in question was a political entity. Although that was certainly a relevant question, it could not be determinative. In each of these cases the courts found it conceptually possible that there was a Convention reason, but on the facts there were actually alternative reasons for the harm feared. See for instance the judgment of Keene LJ in Suarez [at §46]:

“I would only add that I accept that there can be cases where the risk of persecution arises from a mixture of political and criminal reasons, particularly in a society such as Colombia currently is, where criminal economic activity may support political structures. But it is wrong to assume that all actions aimed at preventing the exposure of criminal activities in such a society can be characterised as imputing a political opinion to the witness. These matters need to be looked at on a case by case basis”

74. Although the reasoning in Gomez was obiter dicta, we are bound by the Court of Appeal decision in Suarez. We therefore decline Mr Holmes’ invitation to widen the scope of what is understood by ‘political opinion’, as it is set out in Gomez, an invitation supported by submissions which, although attractive, sounded very much like those accepted by Mr Care in 1999 but subsequently rejected by the higher courts. The Tribunal in Gomez was correct to express doubts about whether the Ward test would always be wide enough, even when given the emphasis required by the Federal Court of Appeal decision in Klinko:

“any opinion on any matter in which the machinery of the state, government, and policy may be engaged”.

75. The Tribunal rightly emphasised that the Convention must be malleable, and that “the boundaries between the political and the non-political shift in historical time and place”. To that end we are satisfied that the Gomez test remains a helpful and accurate touchstone:

“To qualify as political the opinion in question must relate to the major power transactions taking place in that particular society. It is difficult to see how a political opinion can be imputed by a non-state actor who (or which) is not itself a political entity”

76. That is in truth an already wide definition. It does not confine itself to opinions contrary to the government, nor even opinions about the government.

77. Mr Thomann does not take issue with any of that. He submits, however, that the Ward/Gomez formulation must be applied cautiously, because to do otherwise would be to ignore three elephants in the room.

78. The first is the uncontroversial proposition that the Refugee Convention is not designed to protect victims of crime. We certainly do not resile from that. That is not however a principle so rigid that all victims of crime must be denied from protection: most acts of persecution are, after all, crimes.

79. The second is the comparative caselaw to which we have already referred. It is right to say that most of these claims failed for the very reasons advanced by Mr Thomann, and insofar as they underline the general principle that the Convention cannot be read to extend protection to those who fear crime simpliciter, or a general breakdown in law and order, that is a principle that as we say, we accept. Beyond that, however, we cannot be bound by findings of fact made in other jurisdictions, in other cases, sometimes about countries other than El Salvador. As all of those authorities underline, whether the Convention is engaged is always a fact-sensitive assessment.

80. The third of Mr Thomann’s elephants is Article 1F. The word ‘political’ appears twice in Article 1 of the Refugee Convention. First at 1A(2) where protection is extended to include those fearing persecution for reasons of their “political opinion”, and then again at 1F(b) where protection is denied to those who commit “non-political” crimes. Mr Thomann relied upon the definition of ‘political’ in T v Secretary of State for the Home Department [1996] AC 742; [1996] Imm AR 443 wherein Lord Berwick observed [at 786-7]:

“A crime is a political crime for the purposes of article 1F(b) of the Geneva Convention if, and only if (1) it is committed for a political purpose, that is to say, with the object of overthrowing or subverting or changing the government of a state or inducing it to change its policy; and (2) there is a sufficiently close and direct link between the crime and the alleged political purpose. In determining whether such a link exists, the court will bear in mind the means used to achieve the political end, and will have particular regard to whether the crime was aimed at a military or governmental target, on the one hand, or a civilian target on the other, and in either event whether it was likely to involve the indiscriminate killing or injuring of members of the public.”

81. Mr Thomann submits that to interpret the ‘political’ in Article 1A(2) in a materially different way would be contrary to the ordinary rules of treaty interpretation. Worse still, he submits, it would expose state parties to the risk that they could no longer exclude these gangsters from protection, since they could argue their crimes to be ‘political’.

82. We do not accept the premise of the former or that the anticipated outcome postulated by the latter is justified.

83. Mr Thomann’s point about consistency is, at first blush, a powerful one. If the ‘political’ in Article 1F is read to require some direct nexus to the state, why should the ‘political’ in Article 1A(2) be interpreted differently? It is on this basis that he urges us to take a restrictive reading of Article 1A(2). When we step back, however, we see that what that might mean in practice is hard to say. The Secretary of State does not ask us to depart from the Gomez principles; nor does she urge us to roll back on decades of caselaw to reject claims based on a fear of non-state agents; it is not her case that to engage the Convention the persecutory relationship must always be between an individual and the repressive state. In fact the debate between the parties is not really legal at all, it is simply an argument about the facts on the ground in El Salvador. On that basis we find it difficult to understand the utility of this point. There is already an uncontroversial fundamental difference in emphasis and approach between the two ‘politicals’ found in Article 1. The broad humanitarian purpose of the Convention has compelled decision makers to take, on the one hand, a flexible and evolutionary approach to who should qualify for protection, and on the other a restrictive, narrow approach to who can properly be excluded. Any differences in approach to a “non-political crime” and “political opinion” can therefore be properly explained by giving the word “political” its ordinary meaning in the light of the relevant context, object and purpose. It is noteworthy that the nouns qualified by the adjective “political” are very different: see the contrast between a political act and a political offence in R v Governor Of Pentonville Prison, Ex p Cheng [1973] AC 931 at 944 as discussed in T (supra) [at 767]. In any event as T makes clear (see by way of example [772]), any politically motivated crime would not inevitably be cast as a political crime coming outside of Article 1F, particularly where it involves terrorism.

84. This leads us to the Secretary of State’s second concern: that gang members who persecute others may be able to resist exclusion on the grounds that their crimes were politically motivated.

85. Whether an individual should be excluded from the benefit of the Convention is a complex question involving numerous considerations, but it is ultimately an exercise in assessing the nexus between the harm inflicted and the reasons for it. It is always open to decision makers to find that a crime committed by a political actor is not a political crime at all. Furthermore, even where crimes are undeniably political in nature that does not operate as a total bar on exclusion: perpetrators may still be excluded under the other clauses of Article 1F, or in circumstances where the offence is particularly egregious or disproportionate still be denied protection under Article 1F(b). We note in this regard that s36 of the Nationality and Borders Act 2022 specifically provides for such a scenario: “Article 1(F)(b), the reference to a serious non-political crime includes a particularly cruel action, even if it is committed with an allegedly political objective”.

86. We are therefore not satisfied that either the comparative caselaw, or Mr Thomann’s submission about Article 1F, can make any significant impact on our reading of the domestic authorities. As to the point about criminality in general, it is a point that we keep at the forefront of our minds as we analyse the facts in respect of El Salvador.

87. Before we get to that however, we must address a significant development since Gomez, and since the Court of Appeal last had cause to consider this question. Council Directive 2004/83/EC of 29 April 2004 on minimum standards for the qualification and status of third country nationals or stateless persons as refugees or as persons who otherwise need international protection and the content of the protection granted (‘the Directive’), adopted in its original form by the UK and subsequently transposed in material terms by the Refugee of Person in Need of International Protection (Qualification) Regulations 2006 (‘the Regulations’), provides a straightforward framework for the consideration of such claims. Article 6 of the Directive defines ‘actors of persecution’ as including:
(a) The state;
(b) Parties or organisations controlling the State or a substantial part of the territory of the State;
(c) Non-state actors, if it can be demonstrated that the actors mentioned in (a) and (b), including international organisations, are unable or unwilling to provide protection against persecution or serious harm as defined in Article 7

This text is replicated at Regulation 3 of the Regulations1.

88. And Article 10(1)(e) of the Directive explains:

“the concept of political opinion shall in particular include the holding of an opinion, thought or belief on a matter related to the potential actors of persecution mentioned in Article 6 and to their policies or methods, whether or not that opinion, thought or belief has been acted upon by the applicant”2.

(See Regulation 6(1)(f) of the Regulations).

89. That the gangs of El Salvador meet the definition at Article 6 (c) of the Directive is implicitly accepted in the Secretary of State’s grant of humanitarian protection. They therefore constitute ‘actors of persecution’ for the purpose of Article 6. By virtue of Article 10(1)(e) of the Directive anyone holding an opinion, thought or belief about them, their policies or methods, holds a ‘political opinion’. As simplistic as this analysis appears, we do not regard it as being inconsistent with the approach taken in Gomez. We say this because we regard Article 6 as an acknowledgment by member states that any organisation capable of meeting the tests under Article 6(b) or (c) is very likely to be engaged in “major power transactions” so that an opinion about those transactions would count as “political”. The focus in Gomez, and in the submissions before us, about whether the gangs see themselves as political, then becomes relevant to the question of causation.

Particular Social Group

90. We now consider the circumstances in which claimants from El Salvador might be found to be members of a particular social group.

91. Article 10 of the Qualification Directive reads:
Article 10
Reasons for persecution
1.   Member States shall take the following elements into account when assessing the reasons for persecution:

a group shall be considered to form a particular social group where in particular:

(i) members of that group share an innate characteristic, or a common background that cannot be changed, or share a characteristic or belief that is so fundamental to identity or conscience that a person should not be forced to renounce it, and
(ii) that group has a distinct identity in the relevant country, because it is perceived as being different by the surrounding society

(roman numerals and emphasis added)

92. Article 10(1)(d)(i) is concerned with the protected characteristic of the individual, and Article 10(1)(d)(ii) with the social perception of the group to which he belongs. The Upper Tribunal has held both limbs must be satisfied in order for a group to be found, because the ‘and’ joining sub-sections (i) and (ii) should be given its natural meaning: SB (PSG – Protection Regulations – Reg 6) Moldova CG [2008] UKAIT 00002 [at §74]. The CJEU too has indicated that the requirements in Article 10 are to be read conjunctively, for instance in Ahmedbekova and Ahmedbekov v Zamestnik-predsedatel na Darzhavna agentsia za bezhantsite (Case No C-651/16) [at §89]:

“For it to be found that there is a ‘social group’, within the meaning of that provision, two cumulative conditions must be satisfied. First, members of that group must share an ‘innate characteristic’, or a ‘common background that cannot be changed’, or share a characteristic or belief that is ‘so fundamental to identity or conscience that a person should not be forced to renounce it’. Second, that group must have a distinct identity in the relevant country, because it is perceived as being different by the surrounding society (judgment of 7 November 2013, X and Others, C‑199/12 to C‑201/12, EU:C:2013:720, paragraph 45)”. 

93. Article 10 is however a ‘minimum standard’, and it remains open to state parties to take a more generous approach, reading the two requirements disjunctively, so that the ‘and’ at the end of (i) should be read as ‘or’. The rationale for such an approach is set out fully in DH (Particular Social Group: Mental Health) Afghanistan [2020] UKUT 00223 (IAC) [at §46-75], and is supported by the obiter dicta of four members of the Appellate Committee in Fornah (FC) v Secretary of State for the Home Department [2006] UKHL 46, who found the proposition that both elements must be satisfied to be inconsistent with international authority.

94. For the Appellant Ms Patel and Mr Holmes urge us to follow the latter line of authority. The Secretary of State prefers the former.

95. We are satisfied that for many claimants from El Salvador, either approach will do. The Gangs CPIN [at 2.3.5] sets out the Respondent’s position that both women and LGBT people constitute particular social groups in El Salvador, meeting both the innate characteristic and social perception tests. Curiously, and somewhat perversely from our perspective, gang members themselves could also qualify under either limb, since they cannot change their history, and are likely to be highly visible to society at large. Witnesses brave enough to publicly take the stand against the gangs could well qualify since the fact of having given evidence is immutable, and in close-knit communities, or in a public trial, they are likely to be visible within society: this was the finding in Henriques-Rivas v Holder (US 9th Circuit 09-71571) where a witness had appeared in San Salvador court to testify against MS-13.

96. There are however those for whom the conjunctive approach – requiring both limbs of Article 10(1)(d) to be met - will be more challenging. It might be said that people in El Salvador who believe in law and order, and who consequently oppose the gangs, share a belief so fundamental to their identity or conscience that they should not be forced to renounce it. Thus someone who makes a principled refusal to pay renta, or who secretly passes information to the police would thereby satisfy the ‘innate characteristic’ test. But in what circumstances might such an individual be able to demonstrate that such actions, or the belief that led to them, would also create a distinct identity in El Salvador: how would they be perceived as different by the surrounding society? As the US 11th Circuit point out in Castillo-Arias v US Attorney General (US 11th Circuit Court of Appeals 04-14662), confidential informants, or those who privately oppose the gangs, are not identifiable by anyone else in society because these are, by their nature, covert actions and beliefs. For the Appellant Mr Holmes was prepared to accept that an individual whose opposition was wholly discrete would face a difficulty in founding a claim should the conjunctive approach be taken.

97. We are conscious that whatever our findings on this matter, they will soon be academic since the Nationality and Borders Act 2022 (the 2022 Act) now mandates that the more restrictive, conjunctive approach is taken (emphasis added):

33 Article 1(A)(2): reasons for persecution

(2)A group forms a particular social group for the purposes of Article 1(A)(2) of the Refugee Convention only if it meets both of the following conditions.
(3) The first condition is that members of the group share—
(a) an innate characteristic,
(b) a common background that cannot be changed, or
(c) a characteristic or belief that is so fundamental to identity or conscience that a person should not be forced to renounce it.
(4) The second condition is that the group has a distinct identity in the relevant country because it is perceived as being different by the surrounding society.

98. Here the drafter has adopted the wording of Article 10, with one key difference. Article 10 provides that a group shall be considered to form a particular social group where “in particular” both tests are met. In section 33 of the 2022 Act parliament has supplanted that term with the words “only if it meets both of the following conditions”. Tribunals to come will of course be bound by that statutory requirement. It is therefore important to understand what it means, and in particular how the ‘social perception’ limb should be understood.

99. It is difficult to trace precisely the origin of the ‘social perception’ approach. In many of the early and leading cases it was self-evident that a group that could be defined with reference to its protected characteristics would also be cognisable in society as such. UNHCR certainly have, since at least the 2001 San Remo roundtable, regarded that convergence as a matter of fact rather than law, but if it is to be understood to be a mandatory requirement, its parameters should be clearly delineated.

100. The first point to note is that any proposed particular social group must be assessed in the context of the society in which it is said to exist. Groups regarded as having social visibility in one country may not do so in another: in his dissenting judgment in Shah and Islam Lord Millett posits ‘westernised women’ as an example of that. We would suggest that it is this central tenet of the jurisprudence which has perhaps given rise to the fact of social visibility morphing into a requirement3.

101. The second point is that a group that makes out its claim to ‘social visibility’ by pointing to the discrimination it faces will not necessarily fall foul of the principle that the group cannot be defined by reference to the feared persecution. The two are not the same thing: see for instance the defining features of life as a Pakistani woman in Shah and Islam.

102. It can even, with some care, be said that the persecution itself may give rise to the social visibility of a certain group within society. To this end in Shah and Islam Lord Steyn cites with approval this well known passage from A v MIMA:
In A. v. Minister for Immigration and Ethnic Affairs 142 A.L.R. 331, 359 McHugh J. explained the limits of the principle. He said:
 "Nevertheless, while persecutory conduct cannot define the social group, the actions of the persecutors may serve to identify or even cause the creation of a particular social group in society. Left-handed men are not a particular social group. But, if they were persecuted because they were left-handed, they would no doubt quickly become recognisable in their society as a particular social group. Their persecution for being left-handed would create a public perception that they were a particular social group. But it would be the attribute of being left-handed and not the persecutory acts that would identify them as a particular social group."
The same view is articulated by Goodwin-Gill, The Refugee in International Law, 2nd ed., (1996) at p. 362. I am in respectful agreement with this qualification of the general principle. I would hold that the general principle does not defeat the argument of counsel for the appellants.
103. The fourth point is that members of the group need not be identifiable ‘on sight’ by members of the society in which they live. When we consider some of the most vulnerable groups that have benefitted from this provision of the Convention – for instance homosexuals or former victims of trafficking – the logic of this approach becomes apparent. It is the perception of the group in general that counts. An individual gay man need not therefore establish that he has personally been perceived or identified as gay in the hostile environment from which he comes, it is enough that he can show that gay men in general are perceived as different by the surrounding society.

104. This leads to the final, and for the purpose of such cases, perhaps the most important point about social visibility. The group does not need to be perceived as different by society as a whole. It certainly can be – eg women in Pakistan – but in practice the perception need only be held by some members of the society. It is today uncontroversial that members of a family can constitute a particular social group, and that they would (absent special notoriety or fame) only be perceived as being part of that group by the immediate community in which they live. In Henriques-Rivas v Holder the majority of the 9th Circuit, sitting en banc, posited that visibility to the persecutor, or potential persecutor, could be enough. In that case the claimant was a Salvadoran girl who testified against MS-13 at a trial which saw the lead defendant sentenced to up to 30 years in prison for the murder of her father. Her face was visible to all in court, and so the clica in her locality was well aware of her actions. Her bravery in speaking against them marked her out as different in that community. The Court goes on to find [at its page 24] that the Board of Immigration Appeals:

“…failed to consider significant evidence that Salvadoran society recognizes the unique vulnerability of people who testify against gang members in criminal proceedings, because gang members are likely to target these individuals as a group….Notably, as Henriquez-Rivas cited in her briefing before the BIA as well as in her opening brief on petition for review, the Salvadoran legislature enacted a special witness protection law in 2006 to protect people who testify against violent criminal elements, such as MS, in Salvadoran court.…. It is difficult to imagine better evidence that a society recognizes a particular class of individuals as uniquely vulnerable, because of their group perception by gang members, than that a special witness protection law has been tailored to its characteristics.

When a particular social group is not visible to society in general (as with a characteristic that is geographically limited, or that individuals may make efforts to hide), social visibility may be demonstrated by looking to the perceptions of persecutors. Such perceptions may be highly relevant to, or even potentially dispositive of, the question of social visibility. Cf. Sanchez-Trujullo, 801 F.2d at 810 & n.7”.

105. Those are the principles which guide the social perception approach. Returning to the present appeal, for which the 2022 Act has no immediate consequence, we find good reason to interpret Article 10(1)(d) in the same way that it was interpreted by the House of Lords in Fornah. In asking us to do otherwise Mr Thomann simply adopted the submissions made by the Secretary of State in DH: we have read those submissions, but with respect reject them for the following reasons.

106. First, their Lordships’ judgments in Fornah, although obiter, are highly persuasive. They are made in the context of an appeal which was wholly concerned with the definition of ‘particular social group’, and in which their Lordships were referred to all of the relevant international authorities. UNHCR was an intervener in the case, and as Lord Bingham notes, their opinion in favour of a disjunctive reading amounted to a “very helpful distillation” of the effect of those authorities. At [§46], for instance, Lord Hope of Craighead cites with approval the speech of McHugh J in Applicant S v Minister for Immigration and Multicultural Affairs (2004) 217 CLR 387:

“46. In Applicant S v Minister for Immigration and Multicultural Affairs, paras 67-69 McHugh J was at pains to emphasise that it was a mistake to say that a particular social group does not exist unless it is always perceived as such by the society in which it exists. He said that it was not necessary that society itself must recognise the particular social group as a group that is set apart from the rest of that society, or that the persecutor or persecutors must actually perceive the group as constituting a particular social group. As he put it in para 69:

"It is enough that the persecutor or persecutors single out the asylum-seeker for being a member of a class whose members possess a 'uniting' feature or attribute, and the persons in that class are cognisable objectively as a particular social group."

In their judgment in paras 17-18 Gleeson CJ, Gummow and Kirby JJ appear to disagree with McHugh J in requiring recognition within the society subjectively that the collection of individuals is a group that is set apart from the rest of the community. My own preference, with respect, is for the more cautious approach of McHugh J that it would be a mistake to insist that such recognition is always necessary. I agree with him that it is sufficient that the asylum-seeker can be seen objectively to have been singled out by the persecutor or persecutors for reasons of his or her membership of a particular social group whose defining characteristics exist independently of the words or actions of the persecutor. That is as true in cases where the family is identified as the particular social group, as it was in that case where it was contended that the particular social group comprised young, able-bodied Afghan men”.

107. Lord Bingham of Cornhill takes the same approach [at §16]:

“… If, however, this article were interpreted as meaning that a social group should only be recognised as a particular social group for purposes of the Convention if it satisfies the criteria in both of subparagraphs (i) and (ii), then in my opinion it propounds a test more stringent than is warranted by international authority. In its published Comments on this Directive (January 2005) the UNHCR adheres to its view that the criteria in sub-paragraphs (i) and (ii) should be treated as alternatives, providing for recognition of a particular social group where either criterion is met and not requiring that both be met".

As do Lady Hale [at §111] and Lord Brown [at §118].

108. The careful analysis in Fornah is to be contrasted with the approach taken by the CJEU in Ahmedbekova and X and Ors (supra), where no submissions on the point were made, and no consideration given to the competing arguments that we have considered.

109. Second, we would respectfully suggest that in the application of Article 10(1)(d) the focus on the word “and” has been misplaced: the true key to the provision lies in the words “in particular”, which are quite properly transposed into the Regulations as “for example”. Thus Regulation 6(1)(d) accurately reflects the Fornah understanding of Article 10, that the two requirements simply serve as illustrations:

(d) a group shall be considered to form a particular social group where, for example:
(i) members of that group share an innate characteristic, or a common background that cannot be changed, or share a characteristic or belief that is so fundamental to identity or conscience that a person should not be forced to renounce it, and
(ii) that group has a distinct identity in the relevant country, because it is perceived as being different by the surrounding society;
(emphasis added)

110. We agree with UNHCR’s analysis [cited at §15 of Fornah] that where both tests are satisfied there certainly will be a particular social group, but that the point of the ‘social perception’ approach is as follows:

“If a claimant alleges a social group that is based on a characteristic determined to be neither unalterable or fundamental, further analysis should be undertaken to determine whether the group is nonetheless perceived as a cognizable group in that society. So, for example, if it were determined that owning a shop or participating in a certain occupation in a particular society is neither unchangeable nor a fundamental aspect of human identity, a shopkeeper or members of a particular profession might nonetheless constitute a particular social group if in the society they are recognized as a group which sets them apart.”

111. Whichever approach is taken, it is of course only one part of the process. Decision makers must first evaluate whether the individual claimant is a member of a particular social group, but then go on to determine whether the persecution he or she faces is for reasons of their membership of that group.

Country Guidance: Discussion and Findings

Political Opinion

112. The major gangs of El Salvador – by which we mean MS-13 and B-18, to whom the vast majority of gang members owe their allegiances – are agents of persecution as defined by Article 6(c) of the Qualification Directive. They are non-state actors, from whom the state is, the Secretary of State expressly accepts, unable or unwilling to provide protection. Although much of the evidence before us indicated that these gangs are, collectively, in control of large parts of the country we have therefore not found it necessary to make a finding on whether they could also qualify as actors of persecution under Article 6(b) (“parties controlling…a substantial part of the state").

113. Article 10 of the Qualification Directive provides that the concept of ‘political opinion’ shall include the holding of an opinion, thought or belief related to an agent of persecution, and to their policies or methods. We think it unlikely that there are many victims of the gangs who do not hold an opinion about them, their policies or methods.

114. The question remains whether, and in what circumstances, such individuals might be facing persecution ‘for reasons of’ that political opinion. As the many authorities to which we have been referred make clear, whether the causal nexus is established is always a question of fact. In the context of El Salvador it is an enquiry that can today be informed by the following.

115. We are wholly satisfied that MS-13 and B-18 must today be regarded as political actors in El Salvador. These gangs, whose leadership now work in tandem against the government, are now estimated by the ICG to have a presence in 94% of municipalities. They are in control, or have a significant degree of control, across “vast” areas of the country, where they subject the resident population to “an extraordinary level of social control”. This may not involve the provision of ‘services’ as we would understand it, but they do not have to be acting as a proxy government in order to be exercising power. The Supreme Court of El Salvador has declared gang violence to be “politically motivated” in its designation of the gangs as ‘terrorists’. The evidence consistently indicates that they have infiltrated all major branches of government and the security services, at both national and local level: to borrow the phrase used in Gomez [at 40], here “criminal and political activities heavily overlap”.

116. More significantly it is clear that in the past decade, the government of El Salvador, as well as the official opposition of the day, have engaged with the gang leadership in dialogue and negotiations aimed at delivering to them identifiable political aims. In 2012 the demands related primarily to servicing the needs of the imprisoned; by 2014 they had come to understand that they could play a “decisive role in electoral politics” and so micro-grants to benefit the grassroots membership was added to the list; in 2015 the broadened horizons of the leadership saw the unveiling of ‘Mara 203’, and in 2016 the “criminal superstructure” of a united MS-13 and B-18 coalesced into the formation of the ‘Co-ordinating Committee’; in the 2017 elections the Committee demanded “better healthcare, jobs and education for their communities” whilst orchestrating a targeted campaign against FMLN candidates; the rise of President Bukele is now documented to have involved long term strategic negotiations with the gangs, who for a time at least delivered relative peace and stability in exchange for social and economic programmes aimed at benefitting the poor.

117. Mr Thomann is of course right to say that this foray into politics has not been particularly successful. Mara 203 has come to nothing, and any relationship once enjoyed with Bukele has certainly now soured. This is not however the point. With the exception of the 2017 statement of the UNSR, all of the analysts to whom we have been referred found evidence of the gangs pursuing specific social and political objectives. That these objectives may also benefit the gangs economically, by opening up space for further criminal enterprise, does not negate that intention: in that regard we must respectfully disagree with the conclusion to the contrary reached by the Belgian Council for Immigration Disputes in NR 243 676. As the Tribunal make clear in Gomez, reality is rarely binary. The reality in El Salvador is that these violent criminals have come to understand their own political power, and how it can be leveraged to benefit not only themselves, but the communities which they perceive themselves to represent. Nor is any of this particularly surprising, given the history of organisations such as MS-13. All of the sources before us explain that at their inception these street gangs were formed as expressions of community solidarity in the face of adversity. It is that powerful sense of collective identity that lead Insight Crime to describe MS-13 as “a social organization first, and a criminal organisation second”. The fact that the policies they support or espouse benefit their own members cannot logically preclude them from being legitimately considered to be political demands: that is very often the nature of politics.

118. What we glean from this history is that it is not simply the aspirations of the Coordinating Committee or other leaders that we are concerned with. The notion that the gang represents ‘them’ is one deeply ingrained in its rank-and-file membership. The evolution to ‘third generation’ gangs appears to be something that individual members are aware of, and invested in: see for example those interviewed by Douglas Farah, who expressed their “surprise and pleasure” at how well negotiations were going, or the research cited by McNamara that members speak of the gangs as undergoing a “metamorphosis”. It is in this context that Redden, Khoday, and McNamara all argue that resistance is now framed by gang members themselves as political opposition.

119. We accept that in many cases this is likely to be true, particularly given the existential fight for survival that the gangs are currently facing in El Salvador. It is not however always going to be the case.

120. There will be cases at one end of the spectrum where the motive for persecution is purely political. Professor McNamara gives the example of an individual involved in anti-gang youth programmes. Another example would be the targeting of an individual who speaks out against a gang-selected candidate, or a local politician who refuses to advance the policies they urge upon him.

121. There will be cases at the other end of the spectrum where the motive for persecution is purely criminal. The most obvious example of that would be the shopkeeper subject to extortion by his local clica. The act of extortion itself may be crippling for the shopkeeper, and he may be living in terror of what might happen should he refuse to pay, but absent other features the motive is wholly financial, and criminal in nature. We doubt the gang has given any thought at all to what the shopkeeper thinks about their policies or methods.

122. In between those two poles is the area of overlap where the criminal and the political motivations of the gangs are harder to separate. It is true that punishment for resistance will often be inflicted in pursuit of criminal, economic objectives, but in the context of El Salvador that is not all it is. The subject of extortion who takes a stand and refuses to pay, the victim of violence who turns to the state for assistance, the youth who resists the pressure to join a gang are all in our view likely to be able to establish that an effective cause of the persecution they fear is the opinion or belief that they hold about the gang. The less immediately financial in nature the point of the adverse attention, the more likely it is going to fall towards the political end of the spectrum.

Particular Social Group

123. Individuals who resist the gangs in El Salvador will be able to establish that they are members of a particular social group if they can demonstrate that they share an innate characteristic, a common background that cannot be changed or that they share a characteristic or belief that is so fundamental to identity or conscience that they should not be forced to renounce it. We are satisfied that this could include women, LBGT individuals, former gang members and those who for reasons of conscience take a stand against the gangs.

124. Applying Fornah we do not regard it as necessary to make findings on whether any of these groups have a distinct identity in El Salvador, although we note that many of them do: women, LGBT individuals, former gang members would all meet the ‘social perception’ test should it be applied4. We are also satisfied that those who make a public or visible stand against the gangs would qualify since they are likely to be perceived as different by the surrounding society. Those who privately, discreetly oppose the gangs are not, and taking the conjunctive approach their claims would fail.

The Appellant’s Case

125. The Appellant is married with two children, and these family members are dependent upon his claim. Prior to coming to the United Kingdom to seek protection they all lived together in La Libertad, on the south-west coast of El Salvador. The Appellant was employed as a graphic designer.

126. His problems began in December 2018. He was kidnapped at gunpoint by members of MS-13 in what appeared to have been a case of mistaken identity. Having managed to talk his way out of that incident he was then targeted, in March 2019, by members of B-18 who were in control of his neighbourhood. Over period of some months they demanded that he pay them renta for their “homies in jail”. The Appellant complied with their demands, albeit unwillingly.

127. In October 2019 a change in the law in El Salvador relating to property registration resulted in a visit to the Appellant from the police. Members of the local gang observed the police attending his home, and wrongly assumed that it related to a complaint about their extortion demands. On the 30th October 2019 members of the gang surrounded him in the park outside his house. He was held at gunpoint and assaulted. He was told that if he had any more contact with the police he, and his daughters, would be killed. The Appellant decided that despite that threat, he would have to try and seek protection from the police. He did not know what else to do. He travelled into San Salvador, so that he could go to a police station where he would not be seen. He reported the threats. He was told that the complaint could take up to five months to be investigated. Fearful of what could happen if he waited that long, the Appellant and his family gathered some money together and left the country.

128. Fifteen days after they left El Salvador, the Appellant’s brother-in-law was killed by members of B-18. People from the same gang have threatened his sister and his father-in-law.

129. The Appellant claimed asylum on arrival in the UK. In her refusal letter dated the 7th August 2020 the Respondent confirmed that the Appellant and his family would be granted 5 years Humanitarian Protection. The Secretary of State was not however prepared to grant refugee status. In her view this was not a claim which fell within the ambit of the Refugee Convention, because neither MS-13 nor B-18 sought to cause the Appellant harm for any reason other than the fact that they are criminals.

130. On appeal to the First-tier Tribunal (Judge Lodato) the Appellant argued the case that he has maintained before us: he placed particular reliance upon the UNHCR Guidelines. The First-tier Tribunal upheld the Respondent’s reasons for refusal. Although the Tribunal accepted that the gangs in El Salvador exercised “extraordinary levels of control”, it did not accept that they were yet akin to a pseudo-government, or that they could be said to be wielding the kind of “major power” identified in Gomez. As to whether the Appellant could be said to hold the kind of characteristic that might permit him to claim membership of a particular social group, the Tribunal accepted that prior complaints to the police could be said to give rise to an immutable characteristic – the past cannot, after all, be changed – but it was not satisfied that this action, by its nature undertaken in private, could possibly give rise to any wider public perception of the group existing. In reality, the Tribunal held, the harm feared was for reasons of the gangs’ criminal intent.

131. Permission to appeal to the Upper Tribunal was granted by Upper Tribunal Judge Plimmer on the 21st September 2021. Judge Plimmer considered it arguable that in reaching its decision the First-tier Tribunal had failed to have regard to, or failed to give reasons for rejecting, the evidence and views set out in the UNHCR Guidelines.

132. It cannot be said that the First-tier Tribunal ignored the UNHCR guidelines entirely. The Appellant’s reliance upon them is recorded at paragraph 19, and at paragraph 26 the Tribunal alludes to the UNHCR position when it says: “I have noted the ‘extraordinary levels of control’ exercised by the criminal gangs in El Salvador…”: we read this as a reference to UNHCR’s evidence that the gangs exercise “extraordinary levels of social control”. It nevertheless goes on to conclude “the evidence did not establish that the gangs the Appellant feared in this case were akin to pseudo governments in his home area such as they might be regarded as wielding the kind of ‘major power’ identified in Gomez, that might, exceptionally, qualify as a non-state political actor”. We are satisfied that this reasoning reveals several errors in approach.

133. No authority, Gomez included, required the Appellant to demonstrate that the gangs he feared were operating as “pseudo governments in his home area”. The starting position should have been the ‘minimum standards’ set out in the Qualification Directive. Nowhere does the Tribunal consider whether the gangs were actors of persecution under Article 6, or whether the Appellant held an opinion about them which would amount to a political opinion under Article 10. This was the framework against which the causal nexus should have been assessed.

134. Further, we are satisfied that the Tribunal failed to engage with the evidence presented by the UNHCR, and upon which the Appellant had placed such reliance before it. The 2016 Guidelines detailed not just the reach and extent of the gangs’ control in El Salvador, but set out many of the features of their behaviour that we have outlined above: had the Tribunal engaged with this evidence it may have seen it capable of meeting the ‘major power’ test which it derived from Gomez. The guidelines set out not only the “extraordinary levels of social control” exercised by the gangs over the civilian population, but their origins, the political negotiations with the major political parties, their hierarchical, even bureaucratic structures, and the:

“dramatically increasing political sophistication with which the leaders of B-18 and MS have come to couch their grievances to the government and assert their increasingly overt political ambitions. The gangs reportedly continue to decide which political parties can campaign on their territories and they are also reported to have control of several local churches. Indeed, the MS and B-18 gangs claim to be able to influence the elections in El Salvador”

135. None of this evidence is considered. There is no presumptive weight to be attached to the views or evidence reported by UNCHR, but their reports will “typically command very considerable respect” because of their intrinsic quality: see HF (Iraq) and Ors [2013] EWCA Civ 1276 [at §44], AS (Afghanistan) [2021] EWCA Civ 195 [at §21]. In fact in this case the matters reported in the UNHCR Guidelines were supported by almost all of the remaining evidence before the Tribunal, including much of the material in the Gangs CPIN. The failure to consider that evidence was a material error of law.

136. At its paragraph 27 the Tribunal finds no causal nexus to be established because there was “no direct evidence” that the gangs could know that he reported their behaviour to the police. This appears to overlook the Appellant’s evidence that the gang expressly threatened him because they knew he had been talking to the police. It also demonstrates a failure to consider the multiple references made in the UNHCR Guidelines to gang infiltration of, or ability to compromise, government and security structures in El Salvador: “public officials linked to sophisticated drug-smuggling structures”, “victims do not report these crimes to the police for a fear of retribution and a lack of confidence in the authorities”, “gang members who are sheltered under State-run protection programmes often do not escape punishment”, “corruption in the Salvadorian security forces and judiciary reportedly contribute to creating high levels of impunity”, “gangs had reportedly penetrated the State through the police force”, “gangs reportedly have their own infiltrators in the police”. The Appellant was not required to show “direct evidence” that the gang knew he had gone to the police. He was required to prove that it was reasonably likely. In its assessment of that question, it was incumbent on the Tribunal to assess all of the evidence before it, and that included multiple references to the police being compromised by the gangs. That was the context in which the gang members’ threats to the Appellant should have been assessed. We would also observe that this was all accepted by the Respondent: see paragraphs 57-58 of the reasons for refusal letter: “your claim that you were threatened with violence after it was perceived by the gang that you had spoken to authorities about gang activities is consistent with external evidence”.

137. We are satisfied that these are errors such that the decision of the First-tier Tribunal must be set aside. We emphasise that in reaching this decision we have only had regard to the material before the First-tier Tribunal at the date that it took its decision.

138. We remake the decision as follows.

139. When the Appellant was kidnapped by members of MS-13 in December 2018 he was certainly subjected to serious harm by actors of persecution. We accept that as he lay blindfolded in the back of the vehicle he would certainly have had an opinion or belief about their methods and policies. We are however unable to make a finding about the motivation for that attack, since it is apparent from the evidence that it was a case of mistaken identity: we have no means of knowing why the intended target was being hunted down. The second instance of harm was the imposition of renta by the local B-18 cancha. We accept that the implied threat of violence that accompanied this ‘taxation’ amounted to serious harm, but again we cannot be satisfied that it was harm which engaged the Convention. At this stage the Appellant was just another civilian contributing to the maintenance of B-18 “homies” from the neighbourhood. The motivation for those extortion demands was wholly financial in nature: we doubt that the men involved cared less whether the Appellant had an opinion about them, and indeed they had no reason to think that he did, since he obliged and made the payments. That harm – an oppressive, terrifying harm visited on Salvadoran citizens on a daily basis – is economic, and criminal in nature.

140. The Appellant does not however seek protection because he no longer wishes to pay renta. He seeks protection because the gang is aware that he had contact with the police on two occasions. On the first occasion the gang was unaware of why the police had visited his home, only that they had. Perhaps it is for this reason that he was only issued with a warning. On the second occasion, however, it would appear that the reasons for the Appellant’s actions in visiting the police station in El Salvador did become known to the gang. He took a deliberate action to inform on them; he sought to evade their control by travelling to the capital. We have read the evidence on the widespread infiltration of the police force by the gangs. The accepted evidence is that after the Appellant took the action that he did his brother-in-law was killed and B-18 issued threats against other family members. We are satisfied, having regard to the overall context of this claim, and the lower standard of proof, that it is reasonably likely that the Appellant’s relatives were targeted because of his actions in taking a stand against the gangs. In doing so he revealed that he held a thought, belief or opinion about their policies and methods. We are satisfied that the threat he faces today is for reasons of that fact. Accordingly. we find that the Appellant has established a well founded fear of persecution for reasons of his political opinion.

141. In the alternative, and taking the disjunctive approach, we are satisfied that he faces persecution for reasons of his membership of a particular social group. He is a police informer. His past is an immutable characteristic that he cannot change. If we had applied the conjunctive approach the appeal would have been dismissed on this ground.


142. It is the accepted evidence in this case that at least one member of the Appellant’s family has been killed after he left El Salvador and others subject to threats of serious violence. Having had regard to paragraph 28 of the Guidance Note 2022 No 2: Anonymity Orders and Hearings in Private we are concerned that identifying him could lead to the identification of members of his family who may still be in El Salvador and for that reason we find it necessary to make an order for anonymity in the following terms:

“Unless and until a tribunal or court directs otherwise, the Appellant is granted anonymity. No report of these proceedings shall directly or indirectly identify him, any of his witnesses or any member of his family. This direction applies to, amongst others, both the Appellant and the Respondent. Failure to comply with this direction could lead to contempt of court proceedings”


143. The decision of the First-tier Tribunal is set aside.

144. The decision in the appeal is remade as follows: the appeal is allowed on protection (asylum) grounds.

145. There is an order for anonymity.

Upper Tribunal Judge Bruce
16th October 2022